malaria in kenya case study

CDC Activities in Kenya

A pregnant woman receives a treated mosquito net during the Intermittent Preventive Treatment of Malaria in Pregnancy Trial in Kenya’s Siaya District Hospital.

In Kenya, there are an estimated 3.5 million new clinical cases  and 10,700 deaths each year, and those living in western Kenya have an especially high risk of malaria. As it does in many countries around the world, CDC has worked closely with the Kenya Ministry of Health to fight malaria. CDC’s efforts in Kenya are also supported by the U.S. President’s Malaria Initiative External .

Over three decades ago, CDC established a collaboration with the Kenya Medical Research Institute External (KEMRI), which is part of Kenya’s Ministry of Health. Led by the Division of Parasitic Diseases and Malaria , CDC’s malaria work in Kenya includes

Capacity building and technical support

• Surveillance, monitoring and evaluation
• Case management
• Transmission reduction research

Throughout the world, CDC works with Ministries of Health to build capacity and to offer technical support to protect the public’s health. In Kenya, CDC provides onsite technical assistance and funding for malaria research. Over the past three decades, CDC’s investment in Kenya has resulted in a well-trained staff of Kenyan scientists, clinicians, laboratory technicians, and field workers. This achievement is due to CDC’s subject matter experts who serve as advisors for degree-seeking Kenyan students, providing regular trainings and seminars. Technology transfer from CDC to KEMRI has also played an important role. CDC has provided KEMRI with key technological resources that now allow work to be performed locally, which was not possible in the past.

Surveillance, monitoring, and evaluation

CDC and KEMRI work together to conduct health facility surveillance. This surveillance activity involves documenting infections discovered during hospital or health center admission. As many people infected with malaria in western Kenya do not feel ill and therefore do not go to health facilities for malaria treatment, and because malaria transmission fluctuates throughout the year, CDC and KEMRI have developed a novel surveillance platform where we test community members in their households every working day of the year.. Both of these surveillance activities allow the Kenya Ministry of Health to monitor progress in their implementation of malaria control strategies. Vector and insecticide resistance surveillance is also conducted to better understand how mosquitoes interact with malaria parasites and to measure effectiveness of indoor residual spray and long lasting insecticide-treated nets. Lastly, surveillance of deaths using the verbal autopsy and minimally invasive tissue sampling (MITS) is supported through our health and demographic surveillance system (HDSS) and the Child Health and Mortality Prevention Surveillance (CHAMPS) Network study in order to assess the impact of malaria intervention scale-up on malaria-associated mortality over time.

More Information

• About Malaria
• Malaria and travelers
• Malaria worldwide

Vaccines are critical to reducing malaria morbidity and mortality. CDC works closely with KEMRI in both areas. CDC, KEMRI, and other partners conducted a phase III trial on an experimental malaria vaccine (RTS,S/AS01 vaccine candidate). Findings from this study resulted in a positive opinion of the vaccine from the European Medicines Agency (EMA), and a decision from the World Health Organization to sponsor the evaluation of it’s feasibility, impact and safety in 720,000 children in Ghana, Malawi and Kenya beginning in 2018. CDC and KEMRI are also collaborating on a phase II evaluation of a Plasmodium falciparum Sporozoite Vaccine, and a phase II study of a fractionated dose of the RTS,S/AS01 E vaccine.

Vector Control

Vector control is one of the interventions CDC and other global partners use to control malaria. Current activities include:

• Evaluating the durability of different types of long-lasting insecticide-treated nets against malaria vectors (anopheline mosquitoes)
• Evaluating the efficacy of a new attractive targeted sugar bait (ATSB)
• Evaluating the efficacy of a novel host decoy trap for vector surveillance and for transmission reduction
• Evaluating the efficacy of spatial repellents for malaria transmission reduction

Prevention in Pregnancy

Collaborating with the Global Malaria in Pregnancy Consortium External , CDC works with KEMRI to control malaria by

• Assessment of the effectiveness of sulfadoxine-pyrimethamine for intermittent preventive treatment of malaria in pregnancy
• Evaluation of new drugs and strategies for prevention of malaria in pregnancy, including women who are living with HIV
• Evaluation of health systems and sociocultural obstacles to improve uptake of intermittent preventive treatment of malaria in pregnancy

Case Management

Artemisinin combination therapies (ACTs) are the standard treatments for malaria across Africa. It is important to monitor the efficacy of ACTs and to test new therapies. Continued monitoring will provide information that can be used to make decisions about changes to national policy if drug resistance develops. CDC, in partnership with KEMRI, has been performing drug efficacy studies for this purpose since 2007.

ACTs are not recommended for use in the first trimester of pregnancy due to lack of safety data. Thus, CDC and KEMRI have developed and validated a pharmacovigilance system for monitoring the safety of antimalarial drugs during pregnancy. A study was conducted to assess health-care workers’ and drug vendors’ knowledge of and adherence to treatment guidelines for malaria in pregnancy. Data are being collated to evaluate the safety of these drugs in pregnant women.

In 2014, with partners from the Malaria Elimination Consortium of Western Kenya, CDC and KEMRI conducted drug dosing and safety studies for low-dose primaquine and high-dose ivermectin as novel tools for malaria transmission reduction strategies (e.g., mass screen and treat or mass drug administration) when provided in combination with ACTs.

Transmission reduction

CDC is a member of the Global Malaria Transmission Consortium. The Consortium’s initial focus was identifying the best way of measuring malaria transmission. This research has included investigating indoor residual spraying and durable wall liners. In 2015, CDC and KEMRI completed a large-scale intermittent mass test and treat (MTaT) study targeting nearly 30,000 people 3 times a year for 2 years. Results of this study are being analyzed for dissemination in mid-2018. Additional transmission reduction activities include evaluating ATSBs, spatial repellents, and the use of surveillance platforms to identify transmission foci for targeted interventions.

KEMRI has state-of-the-art malaria laboratories that can support epidemiologic studies and conduct research on

• The immunology of malaria in children and pregnant women
• Parasite resistance to antimalarial drugs
• Host genetic risk factors for severe malaria

KEMRI’s and CDC’s Atlanta-based laboratory staff collaborate regularly to evaluate substandard antimalarial drug use in the community and conduct pharmacokinetic studies of antimalarials.

To learn more about activities in Kenya, follow CDC Kenya on Twitter: @CDCKenya or visit the CDC Kenya Web page . For more information about malaria, visit: http://www.cdc.gov/malaria/index.html .

To receive email updates about this page, enter your email address:

New! Locally Acquired Cases of Malaria in Florida, Texas, Maryland, and Arkansas

New! Update to Guidance for use of Artemether-Lumefantrine (Coartem®) in Pregnancy for Uncomplicated Malaria New! Discontinuation of CDC’s Distribution of Intravenous Artesunate as Commercial Drug Guidance for Malaria Diagnosis in Patients Suspected of Ebola Infection in the United States -->

See all Malaria Notices

• New! Malaria is a Serious Disease
• New!   La malaria (paludismo) es una enfermedad grave
• How to Report a Case of Malaria
• CDC Yellow Book
• Red Pages: Malaria-endemic areas by country
• Drugs for Prevention
• Choosing a Drug to Prevent Malaria
• Drugs for Treatment in the U.S.
• Frequently Asked Questions (FAQs)
• Blood Banks

Click here for contact information

Exit Notification / Disclaimer Policy

• The Centers for Disease Control and Prevention (CDC) cannot attest to the accuracy of a non-federal website.
• Linking to a non-federal website does not constitute an endorsement by CDC or any of its employees of the sponsors or the information and products presented on the website.
• You will be subject to the destination website's privacy policy when you follow the link.
• CDC is not responsible for Section 508 compliance (accessibility) on other federal or private website.

• Open access
• Published: 03 October 2022

Clinical malaria incidence and health seeking pattern in geographically heterogeneous landscape of western Kenya

• Wilfred Ouma Otambo 1 , 2 ,
• Patrick O. Onyango 1 ,
• Kevin Ochwedo 2 ,
• Julius Olumeh 3 ,
• Shirley A. Onyango 2 ,
• Pauline Orondo 2 ,
• Harrysone Atieli 2 ,
• Ming-Chieh Lee 4 ,
• Chloe Wang 4 ,
• Daibin Zhong 4 ,
• Andrew Githeko 5 ,
• Guofa Zhou 4 ,
• John Githure 2 ,
• Collins Ouma 6 ,
• Guiyun Yan 4 &
• James Kazura 7  

BMC Infectious Diseases volume  22 , Article number:  768 ( 2022 ) Cite this article

2152 Accesses

3 Citations

1 Altmetric

Metrics details

Malaria remains a public health problem in Kenya despite sustained interventions deployed by the government. One of the major impediments to effective malaria control is a lack of accurate diagnosis and effective treatment. This study was conducted to assess clinical malaria incidence and treatment seeking profiles of febrile cases in western Kenya.

Active case detection of malaria was carried out in three eco-epidemiologically distinct zones topologically characterized as lakeshore, hillside, and highland plateau in Kisumu County, western Kenya, from March 2020 to March 2021. Community Health Volunteers (CHVs) conducted biweekly visits to residents in their households to interview and examine for febrile illness. A febrile case was defined as an individual having fever (axillary temperature ≥ 37.5 °C) during examination or complaints of fever and other nonspecific malaria related symptoms 1–2 days before examination. Prior to the biweekly malaria testing by the CHVs, the participants' treatment seeking methods were based on their behaviors in response to febrile illness. In suspected malaria cases, finger-prick blood samples were taken and tested for malaria parasites with ultra-sensitive Alere ® malaria rapid diagnostic tests (RDT) and subjected to real-time polymerase chain reaction (RT-PCR) for quality control examination.

Of the total 5838 residents interviewed, 2205 residents had high temperature or reported febrile illness in the previous two days before the visit. Clinical malaria incidence (cases/1000people/month) was highest in the lakeshore zone (24.3), followed by the hillside (18.7) and the highland plateau zone (10.3). Clinical malaria incidence showed significant difference across gender (χ 2  = 7.57; df = 2, p  = 0.0227) and age group (χ 2  = 58.34; df = 4, p  < 0.0001). Treatment seeking patterns of malaria febrile cases showed significant difference with doing nothing (48.7%) and purchasing antimalarials from drug shops (38.1%) being the most common health-seeking pattern among the 2205 febrile residents (χ 2  = 21.875; df = 4, p  < 0.0001). Caregivers of 802 school-aged children aged 5–14 years with fever primarily sought treatment from drug shops (28.9%) and public hospitals (14.0%), with significant lower proportions of children receiving treatment from traditional medication (2.9%) and private hospital (4.4%) ( p  < 0.0001). There was no significant difference in care givers' treatment seeking patterns for feverish children under the age of five ( p  = 0.086). Residents with clinical malaria cases in the lakeshore and hillside zones sought treatment primarily from public hospitals (61.9%, 60/97) traditional medication (51.1%, 23/45) respectively ( p  < 0.0001). However, there was no significant difference in the treatment seeking patterns of highland plateau residents with clinical malaria ( p  = 0.431).The main factors associated with the decision to seek treatment were the travel distance to the health facility, the severity of the disease, confidence in the treatment, and affordability.

Clinical malaria incidence remains highest in the Lakeshore (24.3cases/1000 people/month) despite high LLINs coverage (90%). The travel distance to the health facility, severity of disease and affordability were mainly associated with 80% of residents either self-medicating or doing nothing to alleviate their illness. The findings of this study suggest that the Ministry of Health should strengthen community case management of malaria by providing supportive supervision of community health volunteers to advocate for community awareness, early diagnosis, and treatment of malaria.

Peer Review reports

Malaria remains a major public health problem in Kenya, despite increased efforts by the Ministry of Health to scale up intervention strategies [ 1 , 2 , 3 ]. Approximately 70% of the country's 47 million inhabitants are at risk of the disease, with the western Kenya region having the highest burden of infection [ 4 ]. The disease accounts for about 30% out-patients attendance in both public and private health care facilities [ 1 , 2 , 5 , 6 ]. The most common symptom of clinical malaria is fever which drives people to seek treatment when it becomes severe [ 7 ]. Access to health care, housing type, proximity of human settlements to vector breeding sites [ 8 ], socioeconomic status, and bednet use [ 9 , 10 ] may all influence clinical malaria risk in the community. Children and pregnant women are the most vulnerable groups to infection [ 11 , 12 ]. School-aged children serve as a reservoir for malaria parasites, and the prevalence of infection among this age group is high [ 13 ]. Variation in the ecological landscape may result in differential risk exposures to malaria contributing to variation in febrile incidences in the community [ 14 ] driving residents to seek alternative treatment routes. Understanding the health-seeking behavior of clinical malaria cases across different topographical zones may aid in addressing the year-round infection in the community.

The Kenya National Guidelines for the Diagnosis, Treatment, and Prevention of Malaria call for seeking medical attention within 24 h of becoming ill [ 15 ]. Adherence to the guidelines, on the other hand, is becoming a problem as people suffering from malaria-related symptoms either self-medicate with over-the-counter medications or do nothing and only seek treatment from health facilities when their symptoms become severe. The COVID-19 pandemic has had a negative impact on fever care, with outpatient clinic attendance significantly lower since the virus was first detected in Kenya [ 16 ]. Prior to COVID-19, the majority of caregivers of children with fever sought treatment from public health facilities [ 15 ]. While the majority of parents seek medical attention for their feverish children, many do not do so right away. Some parents wait until their child's symptoms become severe before seeking medical help. Doing nothing about febrile symptoms, resulting in a delay in seeking treatment until the illness worsens, may lead to complicated malaria [ 17 ]. Although treatment in public health facilities is free, there is vast majority of underreported malaria cases at the community [ 18 ]. The availability and affordability of the local herbs is easier as they can be obtained from the fields or traditional healers as these traditional healers have good knowledge of symptoms of malaria [ 19 , 20 , 21 ]. The reasons for informing the decision to use the various treatment seeking rotes are unknown.

Self-medication for fever relief is a common practice, particularly in the early stages of illness when symptoms are mild [ 19 ]. Self-prescription and the use of antimalarial drugs without a confirmed diagnosis may result in antimalarial misuse, which may contributes to selection pressure [ 22 ]. In the absence of a confirmed diagnosis, alternative methods of treating malaria symptoms may result in disease complications. Monitoring the impact of topography on treatment seeking profiles in rural communities, as well as adherence to MOH-recommended prompt diagnosis and treatment guidelines, is critical for effective malaria control. As a result, more efforts must be made to encourage prompt fever treatment, as well as the adoption of more sensitive and accurate diagnostic tools to aid in community case management of malaria. The current study aims at assessing clinical malaria incidences and treatment seeking patterns across topography in a rural community in Kisumu County, western Kenya.

This study was carried out in Nyakach Sub-County of Kisumu in Western Kenya. Based on malaria prevalence and topographical features, the study area was divided into three eco-epidemiological zones based on a previous study [ 18 ]: Lakeshore, Hillside, and Highland plateau (Fig.  1 ). Landscapes in the three zones are very different from each other based on altitude and topography [ 18 ]. The altitude of the three eco-epidemiological zones varies, with the lakeshore zone located on the lakeside of the Lake Victoria region at an altitude ranging from 1100 to 1200 m above sea level and prone to flooding during the rainy season. The Highland plateau zone has an altitude range of 1500–1700 m and has more stable larval habitats, whereas the Hillside zone has an altitude range of 1300–1450 m and is located between the Lakeshore and Highland plateau zones. Permanent aquatic habitats are uncommon and larval habitats unstable in this zone. The study area is approximately 327 square kilometers in size, with a population of 168,140 people living in 35,553 households at a population density of 460 people per square km [ 23 ]. This region's economic activities are primarily fishing, subsistence farming, rock mining, and small-scale trading.

Map of Nyakach Sub-County in Kisumu County showing the study eco-epidemiological zones: Lakeshore zone (highlighted in blue), hillside zone (highlighted in brown), and highland plateau zone (purple highlighted)

Study design and data collection

A longitudinal study was conducted in three eco-epidemiologically distinct zones: Lakeshore, Hillside, and Highland plateau zones, between March 2020 and March 2021.Community health volunteers (CHVs) were trained on how to record febrile cases in each household, as well as how to take blood sample for ultra-sensitive malaria RDT and RT-PCR analysis. The survey was conducted during the Covid-19 era, and all infection prevention and control protocols were followed in accordance with Ministry of Health guidelines [ 24 ]. A febrile suspected malaria case was defined as an individual with fever (axillary temperature ≥ 37.5 °C) at the time of examination or complaints of fever and other nonspecific symptoms within 48 h prior to examination, according to the WHO definition [ 25 ]. The survey sought to ascertain the community's health, demographic, and socioeconomic characteristics. The study questionnaire collected self-reported information on age, gender, and active fever, treatment seeking method prior to the CHVs testing, primary occupation, travel history, and ITN use. Participants were divided into three age groups (< 5 years old, 5–14 years old, and ≥ 15 years old). Active fever was defined as an individual axillary temperature ≥ 37.5 °C at the time of examination. The participants' treatment seeking methods were based on their behaviors in response to the illness prior to the biweekly malaria testing by the CHVs. There are five types of treatment seeking methods (public hospital, private hospital, drug shops, traditional medication, and do nothing) Occupation was divided into four categories (farmer, small scale business, office worker, unemployed, student, non-school child, and others. Travel history was defined as having traveled outside the study zones within the previous two weeks. ITN use was defined as sleeping under an ITN the night before the survey. The study sought to identify the socioeconomic and demographic factors associated with the decision to seek treatment. Furthermore, the study questionnaire gathered self-reported data on wall material type, health insurance, and malaria information and awareness, marital status, and distance to the health facility, severity of the disease, confidence in treatment choice, affordability, and medication availability. The wall material type was used to assess house structure categorized into the following groups: Brick/Block, Mud & Wood, and Mud & Cement. The health insurance was classified as the mode of payment at the health facility, as either the hospital payment was by health insurance or cash. Malaria information was defined as the medium through which participants learned about malaria prevention, symptoms, and control. Malaria information was divided into three categories: no information received, information received from the media, and information received from the CHV. Participants who knew the symptoms and severity of malaria were classified as "aware," while those who didn't know were classified as "not aware." The marital status was divided into four categories: under age, single, married, and widowed/divorced. The travel distance to the health facility, the severity of the disease, confidence in the choice of treatment, affordability, and medication availability were all evaluated to see if they played a role in treatment selection. These information was reviewed daily by team supervisors for quality assurance. A total of 2205 finger-prick blood samples were taken from febrile cases for parasite examination with ultra-sensitive Alere ® malaria RDT (Reference number: 05FK140, Republic of Korea) and RT-PCR on dry blood spots [ 26 ]. The samples were then transported to the International Centre of Excellence for Malaria Research (ICEMR) at Tom Mboya University College in Homa Bay, Kenya, for further analysis. The CHVs administered AL to all RDT positive febrile residents. Residents who tested negative were referred to the nearest health facility for follow-up care.

DNA extraction and screening for Plasmodium falciparum infection

975 of the 2205 dried blood spots were randomly selected for DNA extraction to determine the sensitivity and specificity of the ultrasensitive malaria RDT. Chelex resin (Chelex-100) saponin method was used with slight modifications [ 26 ]. Plasmodium species-specific primers and probes targeting 18S ribosomal RNA were used [ 27 ]. PCR reaction volume was constituted as follows; 6 µL of PerfeCTa® qPCR ToughMix™, Low ROX™ Master mix (2X), 0.4 µL each of the forward and reverse species-specific primers (10 µM), 0.5 µL of the species-specific probe, 0.1 µL of double-distilled water and 2 µL of parasite DNA. Thermocycler conditions were set as follows, 50 °C for 2 min, (95 °C for 2 min, 95 °C for 3 s and 58 °C for 30 s) for 45 cycles (QuantStudio™ 3 Real-Time PCR System).

Data analysis

Data were analyzed using SPSS Version 21 software. The demographic profiles of the study participants were described using descriptive statistics. The Chi-square test, odds ratio, incidence ratio, and risk ratio were used to identify the factors associated with clinical malaria incidences and the treatment seeking patterns. Multiple regression was used to predict malaria febrile incidence across topography. Artificial neural network model was used to identify the variables importance associated with the decision to seek treatment. Frequency tables were used to describe categorical variables (counts and percentages). For all analyses, p  ≤ 0.05 was considered statistically significant.

Demographic information of the study participants

A total of 1,599 households were surveyed, with 5,838 residents participating in the study. The three zones' residents’ age structure and gender were all similar. Farming was the most important source of income (21.7%). Individuals aged > 15 years made up approximately 56.6% of the study population and literacy rates were high, with 54.7% completing primary school and 26.2% completing secondary school education (Table 1 ).

Clinical malaria incidence across topography by age and gender

In the study zone, 2205 residents reported febrile illness out of a total of 5838 residents. The Lakeshore zone had the highest clinical malaria incidence, with 24.3 cases/1000 people/month, followed by the hillside zone (18.7 cases/1000 people/month) and the highland plateau zone (10.3 cases/1000 people/ month).

A further Chi square test revealed a statistically significant difference in clinical malaria incidence by gender across topographic zones (χ 2  = 7.57; df = 2, p  = 0.0227). Males had a higher incidence of 26.7 in the lakeshore zone than females, who had an incidence of 22.3. In the hillside and the highland plateau, the females had the higher incidence of infection at 21.7 and 12.7, respectively (Table 2 ).

The chi square test revealed a significant difference in the incidence of clinical malaria by age group across topographical zones (χ 2  = 58.34; df = 4, p  < 0.0001). In the Lakeshore zone, hillside and the highland plateau the school going children aged between 5 and 14 years old had the highest incidence of infection at 38.8, 28.1, and 13.9 cases/1000 people/month, respectively (Table 2 ).

Among the females in the lakeshore zone, the risk of clinical malaria incidences was 1.72 times higher among the 5–14 years old school going children (IR:1.72, 95% CI = 1.24–2.18) and 0.51 times lower among individuals ≥ 15 years old (IR:0.51, 95% CI = 0.25–0.77) compared to children under 5 years old (Table 2 ). Among the males, the incidence risk of infection was 0.18 times lower among individuals ≥ 15 years old (IR: 0.18, 95% CI = 0.04–0.31) compared to children under five years old in the lake zone. In the hillside zone, the school going children had the highest incidence risk of infection compared to the children under five years old (IR: 2.66, 95% CI = 1.53–3.62) (Table 2 ).

Risk factors associated with clinical malaria incidences

Multivariate analysis found that residency in the lakeshore and hillside zone, being male, being between the ages < 5 years and 5–14 years, having a subjective fever, and an elevated body temperature at the time of the visit were all associated with an increased risk of clinical malaria incidences ( p  < 0.05) (Table 3 ). When compared to the highland plateau, the odds of clinical malaria cases were 2.01(95% CI = 1.63–2.49, p  < 0.0001) and 1.47 times higher in the lakeshore and hillside zones, respectively. Females were less likely than males to suffer from clinical malaria (OR: 0.83, 95% CI = 0.70–0.99, p  = 0.042). When compared to individuals over the age of 15, school-aged children aged 5–14 years and under 5 years were 2.00 times (95% CI = 1.66–2.43, p  < 0.0001) and 1.98 (95% CI = 1.54–2.54, p  < 0.0001) more likely to suffer from clinical malaria, respectively. Residents who did not have active fever at the time of testing by the CHVs were less likely to test positive for malaria than those who did (OR: 0.27 95% CI = 0.21–0.34, p  < 0.0001). However, seasonality, recent travel history, and bed net use were not associated with the risk of clinical malaria incidences ( p  < 0.05) (Table 3 ).

Symptoms presented by ultrasensitive malaria RDT positive and negative residents

The Chi square test revealed a significant difference in symptoms between residents who tested positive for malaria by ultrasensitive malaria RDT and those who tested negative (χ 2  = 20.273, df = 7, p  = 0.005). Fever, headache, chills, and vomiting were the most common symptoms among ultrasensitive malaria RDT positive residents, while fatigue, muscle and joint pain were common among ultrasensitive malaria RDT negative residents (Fig.  2 ).

Percentage of symptoms presented by ultrasensitive malaria RDT positive and negative cases. Significance levels in difference between slide positive and negative groups for the same symptom: *P < 0.01, **P < 0.01, ***P < 0.001, and n.s. not significant at level of 0.05

The Chi square test revealed no significant differences in the symptoms reported by ultrasensitive malaria RDT positive residents across age groups (χ 2  = 16.537, df = 14, p  = 0.282). Furthermore, the test revealed no significant difference in symptoms among those who tested negative for ultrasensitive malaria RDT across age groups (χ 2  = 6.577, df = 14, p  = 0.950) (Fig.  3 ).

A Percentage of symptoms of ultrasensitive malaria RDT positive within the age groups. B Percentage of symptoms of ultrasensitive malaria RDT negative cases within the age groups

Treatment seeking patterns of clinical malaria cases

The identified treatment seeking patterns of the clinical malaria cases were doing nothing, buying medicine from drug stores/chemists, and seeking treatment in public, private, and traditional medication which was mainly herbal remedies. There was significant difference in the treatment seeking profiles of the clinical malaria cases (χ 2  = 21.875; df = 4, p  < 0.0001). The most common health seeking behavior among the total 2205 febrile cases assessed was doing nothing (48.7%), buying medicine from drug shops/chemists (38.1%), and seeking treatment in public (12.5%), private hospitals (4.1%), and traditional medication (3.5%) (Table 4 ). Treatment seeking patterns for the clinical malaria cases differed significantly by the lakeshore (χ 2  = 22.471, df = 4, p  < 0.0001), hillside zones (χ 2  = 27.813, df = 4, p  < 0.0001), female sex (χ 2  = 19.447, df = 4, p  = 0.001), school going children (χ 2  = 21.717, df = 4, p  < 0.0001), residents with active fever (temperature ≥ 37.5 °C) at the time of visit (χ 2  = 11.943, df = 4, p  = 0.018), bednet users(χ 2  = 16.355, df = 4, p  = 0.003), and bednet non users (χ 2  = 15.945, df = 4, p  = 0.003) (Table 4 ). 28.9% (232/802) of caregivers of school-aged children aged 5 to 14 years old with fever sought treatment from drug shops, while 14.0% (112/802) sought treatment from public health facilities, with much lower proportions of children receiving fever treatment from traditional medication (2.9%, 23/802) and private health facility (4.4%, 35/802) ( p  < 0.0001). However, There was no significant difference in care givers' treatment seeking patterns for feverish children under the age of five ( p  = 0.086). Although the majority of children receive fever treatment at government facilities, the proportion of children seeking treatment varies by topography ( p  < 0.0001).

Residents with clinical malaria cases in the lakeshore zones sought treatment primarily from public hospitals (61.9%, 60/97) and purchased medication from drug shops (51.4%, 75/146) ( p  < 0.0001). Residents in the hillside zone with clinical malaria cases sought treatment primarily through traditional medication (51.1%, 23/45) ( p  < 0.0001). There was, however, no significant difference in treatment seeking patterns between Highland plateau residents with clinical malaria cases ( p  = 0.431) (Table 4 ).

Females with clinical malaria sought treatment primarily from traditional medications (50.9%, 27/53) and at public hospitals (43.8%, 64/146). Children, on the other hand, are unable to make treatment decisions on their own, and their treatment seeking pattern is determined by their parents/guardians (Table 4 ).

Majority of the clinical malaria cases were from school going children under traditional medication (60.9%, 14/23) and at the public hospital (51.8%, 58/112). Of the residents with active fever at the time of visits, clinical malaria cases were mostly from those who sought prior treatment from: public hospital (82.0%, 50/61) and those who did nothing (66.5%, 127/191) (Table 4 ).

Factors associated with decision to seek treatment

Artificial neural network model was used to identify the main factors associated with the decision to seek treatment. Independent variable importance analysis showed that distance (Importance = 0.184, normalized importance = 100%), severity of disease (Importance = 0.163, normalized importance = 88.7%), confidence in the treatment (Importance = 0.108, normalized importance = 58.5%), affordability (Importance = 0.100, normalized importance = 54.4%), availability of medication (Importance = 0.090, normalized importance = 49.1%), marital status (Importance = 0.072, normalized importance = 39.2%), health insurance (Importance = 0.057, normalized importance = 30.8%), malaria awareness (Importance = 0.054, normalized importance = 29.50%), socio-economic status (wall type: Importance = 0.054, normalized importance = 29.4% and floor type: Importance = 0.038, normalized importance = 20.70%), Knowledge of malaria (Importance = 0.034, normalized importance = 18.8%), net usage (Importance = 0.028, normalized importance = 15.0%) and gender (Importance = 0.018, normalized importance = 10.0%) were the main factors associated with decision to seek treatment. (Training: cross entropy error = 167.712, incorrect prediction = 25.1%; Testing: cross entropy error = 110.675; incorrect prediction = 35.4%) (Additional file 1 : Table S1).

A subsequent analysis revealed a significant relationship between treatment seeking pattern and distance to the health facility (χ 2  = 98.816, df = 4, p  < 0.0001). Residents who reported distance to the health facility as a factor in their decision to seek treatment did nothing (34.4%, 65/188). Residents who reported that distance was not a factor in their decision to seek treatment, on the other hand, primarily sought treatment from private (31.6%, 67/212) and public hospitals (27.4%, 58/212) (Additional file 1 : Table S2).

The severity of the diseases was significantly related to the treatment seeking preference (χ 2  = 121.246, df = 4, p  < 0.0001). Residents who said the severity of the malaria disease influenced their decision to seek treatment did so primarily through traditional medicine (26.2%, 67/256) and private hospitals (25.8%, 66/256). Residents who reported that the severity of the diseases was not a factor in their decision to seek treatment, on the other hand, largely did nothing (48.6%, 70/144) (Additional file 1 : Table S2).

Residents' confidence in their treatment options was significantly related to their decision to seek treatment (χ 2  = 33.442, df = 4, p  < 0.0001). The majority of residents were confident in seeking treatment in a public hospital (25.1%, 55/219), traditional medication (24.7, 54/219), and private hospital (20.1%, 44/219). Residents who did nothing (31.5%, 57/181) and bought drugs from drug shops (20.4%, 37/181), on the other hand, reported having little confidence in their treatment option (Additional file 1 : Table S2).

The cost of the medication had a significant impact on the decision to seek treatment. (χ 2  = 80.640, df = 4, p  < 0.0001). The majority of residents could easily afford treatment from traditional medications (27.2%, 68/250) and public hospitals (25.6%, 64/250). On the other hand, most residents could not afford medication from private hospitals (38.7%, 58/150) (Additional file 1 : Table S2).

Furthermore, the availability of medication was strongly related to the decision to seek treatment (χ 2  = 93.594, df = 4, p  < 0.0001). The majority of residents stated that medication was easily accessible from traditional medication (29.4%, 67/228), drug shops (27.6%, 63/228), and private hospitals (20.6%, 47/228). Those who did nothing (38.4%, 66/172) and those who sought medication from public hospitals (25.0%, 43/172) both reported that medication was not readily available (Additional file 1 : Table S2).

The current study looked at clinical malaria incidences, treatment seeking profiles of febrile cases and factors associated with the decision to seek treatment in western Kenya. Clinical malaria incidence (cases/1000people/month) was highest in the lakeshore zone (24.3), followed by the hillside (18.7) and the highland plateau zone (10.3). The most common health-seeking behaviors among the 2205 febrile residents were doing nothing (48.7%) and purchasing antimalarial from local drug shops (38.1%). Malaria was present in the residents who did nothing, sought traditional medication, and purchased antimalarial from drug shops, and from public hospitals at the time of the visit. The decision to seek treatment was heavily influenced by the distance to the health facility, the severity of the disease, confidence in the treatment, affordability, and socioeconomic status. Furthermore, the current study observed that ultrasensitive malaria RDT diagnosis is highly specific (98.7%) and has good sensitivity (65.5%).

The high clinical malaria incidences and the positivity rates in the Lakeshore zone may be attributed to the area's flat plains and frequent flooding during the rainy seasons, resulting in water stagnation and the presence of permanent mosquito breeding habitats, as well as households' proximity to open water sources, which are stable larval habitats and potential mosquito breeding grounds. The findings corroborate previous research from western Kenya that found a high prevalence of malaria along the lake basin [ 28 , 29 , 30 , 31 , 32 , 33 ]. The primary economic activities in the current study region are subsistence farming and small-scale businesses such as fishing and rock mining. Residents' economic activities, such as night fishing and dusk small-scale businesses, may cause them to remain outside without protective measures, exposing themselves to mosquito bites. However, the current study did not investigate the relationship between economic activity and malaria burden.

In the current study, male in the lakeshore zones were more likely to contract malaria than females. This could be attributed to socioeconomic differences, with the majority of adult males engaging in nighttime outdoor activities that expose them to mosquito bites if no protective measures are taken [ 34 , 35 ]. Females, on the other hand, were more likely to contract malaria in the hillside and highland plateau zones, most likely as a result of dusk activities such as selling vegetables and outdoor cooking at night, which exposes them to mosquito bites [ 36 ]. Females have pre-natal clinic appointments during pregnancy and frequently take their children to seek treatment, which may explain their high hospital seeking behavior and, as a result, their lower clinical malaria incidences when compared to males [ 37 , 38 ]. Clinical malaria incidences were high among school-aged children aged 5–14 years in all study zones, according to the findings. Lower bednet usage among school-aged children exposes them to high mosquito bites at night, which may explain why clinical malaria incidences are higher in this age group [ 34 , 39 , 40 ]. The low infection rate among children under the age of five compared to school-age children could be attributed to the children being cared for by their parents and sleeping under mosquito nets at night [ 12 , 41 , 42 , 43 ]. Children who sleep under insecticide-treated mosquito nets were less likely to contract malaria than those who did not sleep under bednets [ 41 ]. A similar study in Mozambique discovered that self-reported symptomatic malaria is extremely common among children, and that factors facilitating access to health care are associated with symptomatic malaria diagnosis [ 7 ]. Individuals in malaria-endemic areas develop adaptive immunity to the P. falciparum parasite, resulting in a decreasing rate of infection with age [ 44 ]. Similarly, a study in Burkina Faso linked increased fever cases among children to malaria infection (27). The current study, on the other hand, found that health seeking profiles did not differ by age group. Children, unlike adults, are unable to make treatment decisions on their own because their parents or guardians determine the treatment pattern [ 37 , 38 ].

Individuals suspected of having malaria often start by doing nothing, then self-medicate with drugs from drug stores or traditional medications, and when the condition worsens, they seek treatment at health facilities. Doing nothing was most commonly reported among febrile residents, but when their febrile condition worsened, these residents were more likely to seek other alternative treatment options. In the current study, more than 80% of residents either self-medicate or do nothing when they have febrile illness, with only less than 20% seeking treatment in a health facility. In the current study, less than 20% of residents were reported to seek malaria treatment in health facilities, with an estimated 80% of febrile cases being underreported, with a proportion of whom could be malaria cases not being recorded in health facilities. A large proportion of the community does not seek treatment at health care facilities for a variety of reasons, including a lack of antimalarial in health facilities, the affordability of malaria diagnosis and distance to health facilities, confidence in the treatment, and socioeconomic status. As a result, approximately 31% of febrile cases self-diagnose and self-treat with drugs obtained from local drug shops located in nearly every shopping center. A Nigerian study found that approximately 88% of residents prefer to manage malaria at home, with only about 12% visiting health facilities [ 17 ]. The use of antimalarial drugs in the absence of a confirmed test is a major source of concern. Despite seeking treatment from drug stores and traditional medication, inappropriate treatment may have contributed to the observed higher clinical malaria cases in the current study.

Traditional medicine is commonly used to treat fever in African communities, especially during the early stages of illness or when the symptoms are mild [ 17 , 21 , 45 ]. According to the current study, the hillside zone, which is mostly hilly and has a lot of herbs and shrub plantation, explains why the majority of the residents are more likely to seek traditional medicine. Local herbs are more accessible and affordable because they can be obtained from the fields or traditional healers. According to studies in the Democratic Republic of the Congo, Guinea, and Kenya, traditional healers have a good understanding of malaria symptoms and causes, resulting in consistent knowledge of antimalarial plants [ 19 , 20 , 21 ]. It has been reported that herbal medications are involved in parasite clearance [ 17 , 21 , 45 ]. Furthermore, healthcare facilities in the hillside zone were scarce, which could explain the decision to seek traditional medication form of treatment. The current study, however, did correlate the availability of health facilities across topographies and treatment seeking profiles. The current study, however, did not follow up on the parasite clearance by traditional herbs. The study showed socio-economic status such as that the type of housing wall and the floor type, distance to medication access, and hospital payment method all influenced the decision to seek treatment. Residents from the lake zone, for example, were more likely to seek treatment in a public hospital and purchase antimalarial drugs from local drug stores. This was greatly influenced by the distance and ease of access. The severity of fever as a result of P. falciparum infection drives people to seek treatment [ 7 ] which is heavily influenced by accessibility, availability, and affordability of treatment services [ 22 ]. The current study residents reported taking analgesics to relieve pain before taking antimalarials, which may explain why there were fewer active fever cases in the study zone.

The rapid emergence and spread of the COVID-19 has resulted in massive global disruptions that are affecting people's lives and well-being. The devastation caused by the pandemic could be greatly exacerbated if the response jeopardizes the provision of life-saving malaria services [ 46 ]. COVID-19-related challenges have contributed to an increase in antimalarial and RDT stockout rates, resulting in a drop in test-and-treat policy adherence [ 16 ]. Reduced funding for vector interventions, combined with competing public health challenges such as the ongoing COVID-19 pandemic, may result in a rollback of malaria control gains, leading to increased morbidity and mortality from malaria [ 47 , 48 , 49 ]. Furthermore, fear and stigma were generated as a result of the COVID-19 situation. Fear of contracting COVID-19 in a health facility, for example, as well as the stigma of being tested for COVID-19 infection, influenced facility attendance. The number of people visiting health-care facilities decreased as a result of such concerns. The Kenya malaria indicator survey has also reported similar findings [ 50 ].

According to the current study, ultrasensitive malaria RDT diagnosis had a higher specificity (99%) and a good sensitivity (66%) in detecting malaria febrile cases. The study's findings confirm the high sensitivity of the ultrasensitive malaria RDT when compared to RT-PCR, as previously reported [ 51 , 52 , 53 ]. Malaria intervention strategies are dependent on whether malaria patients can easily access and afford appropriate diagnosis and treatment. To reduce the complication of malaria cases, the government should invest in supportive supervision of CHVs as well as the provision of more sensitive RDTs and antimalarial to strengthen community malaria case management.

Malaria case treatment-seeking habit is critical in determining malaria infection at the community level. Despite high bednet coverage, the current study found that the community has a high rate of clinical malaria incidences and positivity rates with the lakeshore zones bearing the greatest burden. The number of febrile cases is high because only about 20% of residents seek diagnosis and treatment in health care facilities, while the other 80% self-medicate or do nothing. These health seeking behavior suggests that a portion of the community's reported 80% of febrile cases may be infected with malaria but not reported in the Kisumu's monthly DHIS-2 reporting system. More research should be done to determine the true number of malaria-infected people who aren't reported in the DHIS-2. This information will help the Ministry of Health strengthen its community case management strategy for malaria.

Availability of data and materials

The dataset used in this study is available from the corresponding author upon request.

Abbreviations

Confidence Interval

Community Health Volunteer

Dried blood spots

Deoxyribonucleic acid

International Center of Excellence for Malaria Research

Rapid diagnosis test

Real-time polymerase chain reaction

National Malaria Control Programme Ministry of Health. The Kenya Malaria Communication Strategy 2016–2021. 2016.

Kenya Malaria Control Programme. Kenya Malaria Programme Review 2018. Nairobi, Kenya, and Rockville, Maryland, USA: NMCP, KNBS, and ICF International; 2019.

Kenya Malaria Control Programme. The Kenya Malaria Strategy 2019–2023. 2019.

National Malaria Control Programme. Kenya Malaria Indicator Survey 2015. Nairobi, Kenya, and Rockville, Maryland, USA: NMCP, KNBS, and ICF International; 2016.

Githinji S, Oyando R, Malinga J, Ejersa W, Soti D, Rono J, et al. Completeness of malaria indicator data reporting via the District Health Information Software 2 in Kenya, 2011–2015. Malar J. 2017;16:344.

Article   PubMed   PubMed Central   Google Scholar  

Maina JK, Macharia PM, Ouma PO. Coverage of routine reporting on malaria parasitological testing in Kenya, 2015–2016. Glob Health Action. 2017;10.

Carlucci JG, Blevins Peratikos M, Cherry CB, Lopez ML, Green AF, González-Calvo L, et al. Prevalence and determinants of malaria among children in Zambézia Province, Mozambique. Malar J. 2017;16:1–13.

Article   Google Scholar  

WHO. Guidelines for Malaria Vector Control [Internet]. World Health Organization. 2019.

Guerra M, De SB, Mabale NN, Berzosa P, Arez AP. Malaria determining risk factors at the household level in two rural villages of mainland Equatorial Guinea. Malar J. 2018;17:203.

Bannister-tyrrell M, Srun S, Sluydts V, Gryseels C, Mean V, Kim S, et al. Importance of household-level risk factors in explaining micro- epidemiology of asymptomatic malaria infections in Ratanakiri Province , Cambodia. Sci Rep. 2018;1–15.

Kwenti TE, TayongDizzleBitaKwenti AL, Njunda LA, Nkuo-Akenji T. Epidemiological and clinical profile of paediatric malaria: a cross sectional study performed on febrile children in five epidemiological strata of malaria in Cameroon. BMC Infect Dis. 2017;17:1–13.

Hajison PL, Feresu SA, Mwakikunga BW. Malaria in children under-five: a comparison of risk factors in lakeshore and highland areas, Zomba district, Malawi. PLoS ONE. 2018;13: e0207207.

Cohee L, Laufer M. Tackling malaria transmission in sub-Saharan Africa. Lancet Glob Heal. 2018;6:e598–9.

Mwakalinga VM, Sartorius BKD, Limwagu AJ, Mlacha YP, Msellemu DF, Chaki PP, et al. Topographic mapping of the interfaces between human and aquatic mosquito habitats to enable barrier targeting of interventions against malaria vectors. R Soc Open Sci. 2018;5: 161055.

Kenya Malaria Control Programme. National Guidelines for the diagnosis, treatment and prevention of malaria in Kenya. 2016.

U.S. President’s Malaria Initiative. Kenya Malaria Operational Plan FY 2022 [Internet]. 2022.

Dave-agboola IO, Raji J. Health-Seeking Behaviour of Malaria Patients in Lagos, Nigeria. Int J Heal Sci Res. 2018;8:259–64.

Google Scholar  

Otambo WO, Omondi CJ, Ochwedo KO, Onyango O, Atieli H, Ming-chiehLee CW, et al. Risk associations of submicroscopic malaria infection in lakeshore, plateau and highland areas of Kisumu County in western Kenya. PLoS ONE. 2022;17:e0268463.

Article   CAS   PubMed   PubMed Central   Google Scholar  

Manya MH, Keymeulen F, Ngezahayo J, Bakari AS, Kalonda ME, Kahumba BJ, et al. Antimalarial herbal remedies of Bukavu and Uvira areas in DR Congo: an ethnobotanical survey. J Ethnopharmacol. 2020;249: 112422.

Article   CAS   PubMed   Google Scholar  

Traore M, Baldé M, Diallo M, Baldé E, Camara A, et al. Ethnobotanical survey on medicinal plants used by Guinean traditional healers in the treatment of malaria. J Ethnopharmacol. 2013;150:2013.

Gathwira JW, Rukunga GM, Mwitari PG, Mwikwabe MN, Kimani CW, Muthaura CN, et al. Traditional herbal antimalarial therapy in Kilifi district, Kenya. J Ethnopharmacol. 2010;134:434–42.

Karyana M, Devine A, Kenangalem E, Burdarm L, Poespoprodjo JR, Vemuri R, et al. Treatment—seeking behaviour and associated costs for malaria in Papua, Indonesia. Malar J. 2016;15:536.

Kisumu County. Kisumu County Integrated Development Plan II, 2018–2022 [Internet]. 2018.

Ministry of Health K. Interim guidelines on management of COVID-19 in Kenya. Repub Kenya. 2020;1–76.

WHO. WHO informal consultation on fever management in peripheral health care settings: a global review of evidence and practice [Internet]. Vol. 148. 2013.

Plowe CV, Djimde A, Bouare M, Doumbo O, Wellems TE. Pyrimethamine and proguanil resistance-conferring mutations in Plasmodium falciparum dihydrofolate reductase: polymerase chain reaction methods for surveillance in Africa. Am J Trop Med Hyg. 1995;52:565–8.

Veron V, Stephane S, Bernard C. Experimental parasitology multiplex real-time PCR detection of P. falciparum , P. vivax and P. malariae in human blood samples. Exp Parasitol. 2009;121:346–51.

Matsushita N, Kim Y, Ng CFS, Moriyama M, Igarashi T, Yamamoto K, et al. Differences of rainfall-Malaria associations in lowland and highland in Western Kenya. Int J Environ Res Public Health. 2019;16:19.

Amek N, Bayoh N, Hamel M, Lindblade KA, Gimnig JE, Odhiambo F, et al. Spatial and temporal dynamics of malaria transmission in rural Western Kenya. Parasit Vectors. 2012;5:86.

Matsushita N, Kim Y, Ng CFS, Moriyama M, Igarashi T, Yamamoto K, et al. Differences of rainfall–malaria associations in lowland and highland in western Kenya. Int J Environ Res Public Health. 2019;16:3693.

Article   PubMed Central   Google Scholar  

Dellicour S, Hill J, Bruce J, Ouma P, Marwanga D, Otieno P, et al. Effectiveness of the delivery of interventions to prevent malaria in pregnancy in Kenya. Malar J. 2016;15:1–13.

Ng’ang’a PN, Aduogo P, Mutero CM. Long lasting insecticidal mosquito nets (LLINs) ownership, use and coverage following mass distribution campaign in Lake Victoria basin, Western Kenya. BMC Public Health. 2021;21:1046.

Kapesa A, Kweka EJ, Zhou G, Atieli HE, Kamugisha E, Mazigo HD, et al. Utility of passive malaria surveillance in hospitals as a surrogate to community infection transmission dynamics in western Kenya. Arch Public Heal. 2018;76:1–11.

Ochwedo KO, Omondi CJ, Magomere EO, Olumeh JO, Debrah I, Onyango SA, et al. Hyper-prevalence of submicroscopic Plasmodium falciparum infections in a rural area of western Kenya with declining malaria cases. Malar J. 2021;20:427.

Diiro GM, Affognon HD, Muriithi BW, Wanja SK, Mbogo C, Mutero C. The role of gender on malaria preventive behaviour among rural households in Kenya. Malar J. 2016;15:14.

Jenkins R, Omollo R, Ongecha M, Sifuna P, Othieno C, Ongeri L, et al. Prevalence of malaria parasites in adults and its determinants in malaria endemic area of Kisumu County, Kenya. Malar J. 2015;14:263.

Franckel A, Lalou R. Health-seeking behaviour for childhood malaria: household dynamics in rural Senegal. J Biosoc Sci. 2009;41:1–19.

Article   PubMed   Google Scholar  

Budu E, Seidu AA, Armah-Ansah EK, Sambah F, Baatiema L, Ahinkorah BO. Women’s autonomy in healthcare decisionmaking and healthcare seeking behaviour for childhood illness in Ghana: analysis of data from the 2014 Ghana Demographic and Health Survey. PLoS ONE. 2020;15: e0241488.

Kamau A, Nyaga V, Bauni E, Tsofa B, Noor AM, Bejon P, et al. Trends in bednet ownership and usage, and the effect of bednets on malaria hospitalization in the Kilifi Health and Demographic Surveillance System (KHDSS): 2008–2015. BMC Infect Dis. 2017;17:720.

Minakawa N, Kongere JO, Sonye GO, Lutiali PA, Awuor B, Kawada H, et al. Long-lasting insecticidal nets incorporating piperonyl butoxide reduce the risk of malaria in children in Western Kenya: a cluster randomized controlled trial. Am J Trop Med Hyg. 2021;105:461–71.

Abossie A, Yohanes T, Nedu A, Tafesse W, Damitie M. Prevalence of malaria and associated risk factors among febrile children under five years: a cross-sectional study in arba minch zuria district, south Ethiopia. Infect Drug Resist. 2020;13:363–72.

Maketa V, Mavoko HM, Inocêncio R, Zanga J, Lubiba J, Kalonji A, et al. The relationship between Plasmodium infection, anaemia and nutritional status in asymptomatic children aged under five years living in stable transmission zones in Kinshasa, Democratic Republic of Congo. Malar J. 2015;14:83.

Roberts D, Matthews G. Risk factors of malaria in children under the age of five years old in Uganda. Malar J. 2016;15:246.

Agwu E, Ihongbe JC, Okogun GRA, Inyang NJ. High incidence of co-infection with malaria and typhoid in febrile HIV infected and AIDS patients in Ekpoma, Edo State, Nigeria Brazilian. J Microbiol. 2009;40:329–32.

CAS   Google Scholar  

Heng S, Durnez L, Mao S, Siv S, Tho S, Mean V, et al. Passive case detection of malaria in Ratanakiri Province (Cambodia) to detect villages at higher risk for malaria. Malar J. 2017;16:1–11.

WHO. Tailoring malaria interventions in the COVID-19 response [Internet]. 2020.

Ajayi IO, Ajumobi OO, Falade C. Malaria and COVID-19: commonalities, intersections and implications for sustaining malaria control. Pan Afr Med J. 2020;37:1–10.

Weiss DJ, Bertozzi-villa A, Rumisha SF, Amratia P, Arambepola R, Battle KE, et al. Indirect effects of the COVID-19 pandemic on malaria intervention coverage, morbidity, and mortality in Africa: a geospatial modelling analysis. Lancet Infect Dis. 2021;21:59–69.

Heuschen AK, Lu G, Razum O, Mumin AA, Sankoh O, Von SL, et al. Public health—relevant consequences of the COVID-19 pandemic on malaria in sub-Saharan Africa: a scoping review. Malar J. 2021;20:339.

Division of National Malaria Program. Kenya Malaria Indicator Survey 2020 [Internet]. Nairobi, Kenya and Rockville, Maryland, USA: DNMP and ICF; 2021.

Reichert EN, Hume JCC, Sagara I, Healy SA, Assadou MH, Guindo MA, et al. Ultra-sensitive RDT performance and antigen dynamics in a high-transmission Plasmodium falciparum setting in Mali. Malar J. 2020;19:323.

Danwang C, Kirakoya-Samadoulougou F, Samadoulougou S. Assessing field performance of ultrasensitive rapid diagnostic tests for malaria: a systematic review and meta-analysis. Malar J. 2021;20:245.

Acquah FK, Donu D, Obboh EK, Bredu D, Mawuli B, Amponsah JA, et al. Diagnostic performance of an ultrasensitive HRP2-based malaria rapid diagnostic test kit used in surveys of afebrile people living in Southern Ghana. Malar J. 2021;20:125.

Download references

Acknowledgements

We would like to thank the Nyakach Sub-County study participants for their participation in the study. We would like to express our gratitude to all the CHVs, and field and lab team members lead by Charles Omboko who worked tirelessly to complete this project.

This research is supported by grants from the National Institutes of Health (U19 AI129326 and D43 TW001505).

Author information

Authors and affiliations.

Department of Zoology, Maseno University, Kisumu, Kenya

Wilfred Ouma Otambo & Patrick O. Onyango

International Centre of Excellence for Malaria Research, Tom Mboya University College-University of California Irvine Joint Lab, Homa Bay, Kenya

Wilfred Ouma Otambo, Kevin Ochwedo, Shirley A. Onyango, Pauline Orondo, Harrysone Atieli & John Githure

School of Natural and Environmental Science, Newcastle University, Newcastle Upon Tyne, UK

Julius Olumeh

Program in Public Health, University of California Irvine, Irvine, CA, USA

Ming-Chieh Lee, Chloe Wang, Daibin Zhong, Guofa Zhou & Guiyun Yan

Centre for Global Health Research, Kenya Medical Research Institute, Kisumu, Kenya

Andrew Githeko

Department of Biomedical Sciences and Technology, Maseno University, Kisumu, Kenya

Collins Ouma

Department of Pathology, School of Medicine, Case Western Reserve University, Cleveland, OH, USA

James Kazura

You can also search for this author in PubMed   Google Scholar

Contributions

WOO Conceptualization, designed the study, oversaw its implementation, performed laboratory assays, interpretations, analyses, drafted the original manuscript and edited and reviewed the final manuscript. KO, JO, SO, PO aided in the coordination of sample collection, conducted laboratory analysis and reviewing the manuscript. HA provided administrative support. MCL helped in designing the figure, CW, DZ and GZ contributed to study design, data analysis, editing and reviewing the manuscript, AG contributed to study design, editing and reviewing the manuscript. JG conceived the study design, reviewed and revised the manuscript. CO and PO provided input in data analysis, supervision, editing and reviewed the manuscript. GY contributed to study design, editing and review of the manuscript, and funded the project. JK contributed to study design and editing and reviewed the manuscript. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Wilfred Ouma Otambo .

Ethics declarations

Ethics approval and consent to participate.

Maseno University's Ethics Review Committee (MSU/DRPI/MUERC/00778/19) and the University of California, Irvine's Institutional Review Board (HS# 2017-3512) provided ethical approval for this study. The Kisumu County Director of Health (GN133VOLIX-413) and the Deputy County Commissioner (NYK/PH/13/1-200) gave their approval for the study to be conducted in the villages. The survey was open to all residents who were willing to participate in the study, regardless of their demographics. Residents who refused to participate or changed their willingness to participate at any time were excluded from participating. Before the study began, adults provided signed informed consent, and minors provided assents through their parents/guardians.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1: table s1.

. Independent variable importance associated with decision to seek treatment. Table S2 . Factors associated with decision to seek treatment.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

Otambo, W.O., Onyango, P.O., Ochwedo, K. et al. Clinical malaria incidence and health seeking pattern in geographically heterogeneous landscape of western Kenya. BMC Infect Dis 22 , 768 (2022). https://doi.org/10.1186/s12879-022-07757-w

Download citation

Received : 12 April 2022

Accepted : 27 September 2022

Published : 03 October 2022

DOI : https://doi.org/10.1186/s12879-022-07757-w

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Active case detection
• Malaria incidence
• Ultrasensitive RDT
• Health seeking behavior
• Self-medication
• Traditional medication

BMC Infectious Diseases

ISSN: 1471-2334

• Submission enquiries: [email protected]
• General enquiries: [email protected]

Welcome to the United Nations

The real impact of the new Malaria vaccine in Kenya

Get monthly e-newsletter.

Africa’s Lake Victoria is the world’s second largest freshwater lake and its largest tropical lake. This awe-inspiring ecosystem supports a stunning variety of bird, aquatic and animal species. Unfortunately, the same ecosystem is also an ideal breeding ground for mosquitoes that carry malaria.

Malaria experts refer to the areas around the lake’s shore as Kenya’s “lake-endemic” region. Most people there have endured malaria infections, and many have suffered terrible losses, including children, malaria’s most common and vulnerable victims.

“When I was growing up, I suffered from several attacks of malaria, and I’ve seen children suffer from permanent disabilities,” says Vivienne, a mother of 5 from Chemelil market, a rural village outside Kisumu town. 

She added: “My 3 oldest children suffer from frequent attacks of malaria. When that happens, they lose their appetite, suffer from fever, diarrhea, and vomiting, and become very weak.”

This life-long experience with malaria is the reason Vivienne and other mothers throughout this part of Kenya were enthusiastic about the arrival in 2019 of the world’s first malaria vaccine, RTS,S/AS01 (or RTS,S).

The malaria vaccine has been a game-changer and a breakthrough. We’ve seen mortality go down in under-1 and under-5 children. People didn’t think we could have a vaccine for malaria, and now everyone’s excited.

That year, the vaccine became available in parts of Kenya, as well as Ghana and Malawi, in a pilot introduction through the national immunization programme, under the WHO-coordinated Malaria Vaccine Implementation Programme (MVIP). 

The purpose of the pilots was to evaluate the public health use of the vaccine, including whether caregivers would bring their children to clinics for the 4-dose regimen and the vaccine’s impact on reducing childhood illness and death from malaria in routine use.

Nearly 4 years on, more than 1.4 million children have received the vaccine across the 3 pilot countries, of which, 400,000 children in Kenya have received at least their first dose.

The impact of malaria vaccine implementation

“Whatever we were using before, we reached a point where the burden [of malaria] plateaued, and we needed an additional tool,” explained Dr. Gregory Ganda, Kisumu County Executive for Health. 

More than 3 years on, the vaccine has become an important and life-saving additional tool alongside other malaria interventions, such as insecticide treated bed nets, indoor residual spraying, preventive treatment for pregnant women, and effective malaria medications.

Since the vaccine was introduced in parts of Kenya, hospitalizations for children under 5 for severe malaria have fallen substantially, and there is a drop in child deaths.

“Over the past 3 years, we’ve witnessed a significant reduction in pediatric admissions from malaria,” adds Dr. Ganda. “It’s a great feeling as a doctor when you are considering closing a ward because of lack of patients.”

Meeting children of the malaria vaccine generation

Vivienne first heard about the malaria vaccine from Rose Akinyi, the community health volunteer who came to her home shortly after she gave birth to her 4th child, Isaac. 

In Kenya community health volunteers act as critical links between caregivers and the health care system. In addition to checking on Vivienne and Isaac’s health, Rose reminded her about the various vaccines routinely given in Kenya. 

Having the vaccine is important because it is protecting my child. I even wish you had another one for the bigger children so that they could be protected the same way Stella is protected.

Because Vivienne lives in one of the 26 sub-counties that took part in the pilot introduction, that included the malaria vaccine.

Isaac has now received all 4 doses of the malaria vaccine, while his younger brother, Moses, age 13 months, has received the first 3 doses.

“Isaac and Moses have suffered much less from malaria than my other children. The kids are stronger, and when they did get malaria, it was a lot less severe,” adds Vivienne. 

In neighboring Homa Bay, Margaret Atieno’s daughter, Stella, recently received the 3rd dose of the vaccine. She explained how she wished her older children could have benefited from the added prevention.

“Having the vaccine is important because it is protecting my child,” she said. “I even wish you had another one for the bigger children so that they could be protected the same way Stella is protected.”

Overall, the verdict is resounding: the vaccine saves children’s lives, caregivers want this vaccine for their children, and more children at risk are being reached with this additional malaria prevention.

“The malaria vaccine has been a game-changer and a breakthrough. We’ve seen mortality go down in under-1 and under-5 children. People didn’t think we could have a vaccine for malaria, and now everyone’s excited,” said Dr. Gordon Okomo, County Director of Health in Homa Bay County.

Kenya recently expanded delivery of the vaccine to more communities in the pilot areas, more than doubling access to the malaria vaccine and the Ministry of Health is committed to further phased introduction.

Beyond Kenya, demand for the malaria vaccine is unprecedented. At least 28 countries in Africa plan to apply for Gavi support to deploy the vaccine.

Leveraging new opportunities

Malaria vaccination visits are also creating new opportunities for healthcare workers to see children who might not otherwise come to health centers or hospitals—and screen them for any other missed vaccinations.

“[The] malaria vaccine has been an opportunity to help us follow clients and…to improve the uptake of other vaccines,” explains Maureen Atieno, nurse in charge of the maternal and child health clinic, at Homa Bay teaching and referral hospital. “When you follow-up for malaria vaccine we can identify people who have defaulted for other vaccines, including measles-rubella.”

Looking forward

Beyond Kenya, demand for the malaria vaccine is unprecedented. At least 28 countries in Africa plan to apply for Gavi support to deploy the vaccine. 

Initial supply is limited and will be allocated according to a  framework  that prioritizes initial doses to children living in areas of greatest need. 

As supply increases to meet demand the vaccine will reach more children, within and across endemic countries. Increasing supply to reap the full benefits of the vaccine is a priority for WHO, Gavi, UNICEF and partners.

Also in this issue

Guterres calls for phasing out fossil fuels to avoid climate ‘catastrophe’

Achieving debt sustainability and economic resilience in Africa

Ghana goes organic to build resilient food systems

Somalia’s first all-women media team marks first anniversary

New UN report: Risk of prolonged period of low growth looms large amid multiple global crises

Escaping Sudan: Home alone, with warplanes overhead

Horn of Africa: Over 7 million children under the age of 5 remain malnourished across Ethiopia, Kenya and Somalia

Africa Day 2023: UN Secretary-General’s message

Benin triples number of companies opened by youth

WHO congratulates Benin and Mali for eliminating trachoma as a public health problem

Africa Dialogue Series 2023 underway

Africa's largest polio vaccination drive since 2020 targets 21 million children

Award wining Ghanaian peacekeeper urges girls to choose career in military

Road Safety Week: African nations steer towards reducing deaths

African Finance Ministers call for an IMF that is “fit for the 21st Century”

Understanding plastic pollution and its impact on lives

AfCFTA: Seizing opportunities for a prosperous Africa

Meet the honey business owner creating a buzz in Uganda

As a UN Security Council member, Sierra Leone will promote peace, security and good governance in Africa

Mozambique embraces dialogue for lasting peace

Empowering women through bee project in Angola

Ghanaian peacekeeper wins UN Military Gender Advocate 2022

Activists encouraged to apply for UN SDG Action Awards

Togo plans investment reforms to drive sustainable growth

AfCFTA: African free trade area will spur sustainable growth, says UN chief

‘Silent emergency’: Premature births claim a million lives yearly

World Diabetes Day 2023: Need for equitable access to care for people with TB and diabetes

Nigeria to vaccinate 7.7 million girls against leading cause of cervical cancer

More from africa renewal.

Tackling emerging anti-malarial drug resistance in Africa

Are antimalarial drugs still effective?

The allies protecting pregnant women from malaria in Africa

Over 1 million African children protected by first malaria vaccine

Strengthening community and stakeholder participation in the implementation of integrated vector management for malaria control in western Kenya: a case study

Affiliations.

• 1 International Centre of Insect Physiology and Ecology (ICIPE), PO Box 30772, Nairobi, Kenya. [email protected].
• 2 School of Public Health, Jomo Kenyatta University of Agriculture and Technology, PO Box 62000, Nairobi, Kenya. [email protected].
• 3 International Centre of Insect Physiology and Ecology (ICIPE), PO Box 30772, Nairobi, Kenya.
• 4 University of Pretoria Institute for Sustainable Malaria Control (UP ISMC), School of Health Systems and Public Health, University of Pretoria, Private Bag X363, Pretoria, 0001, South Africa.
• PMID: 33740983
• PMCID: PMC7977174
• DOI: 10.1186/s12936-021-03692-4

Background: Integrated vector management (IVM) is defined as a rational decision-making process for the optimal use of resources for vector control. The IVM approach is based on the premise that effective control of vectors and the diseases they transmit is not the sole preserve of the health sector. It requires the collaboration and participation of communities and other stakeholders in public and private sectors. Community participation is key to the success of IVM implementation at the local level.

Case description: The study was conducted in Nyabondo, a rural area of Kenya where malaria is endemic. The objective of the project was to promote adoption and sustainability of IVM and scale up IVM-related activities as well as increase community participation and partnership in malaria control through outreach, capacity-building and collaboration with other stakeholders in the area. Collaboration was pursued through forging partnership with various government departments and ministries, particularly the fisheries department, ministry of education, ministry of health, forestry department and the social services. In total, 33 community-based organizations working within the area were identified and their role documented. Through distribution of information, education and communication (IEC) materials alone, the project was able to reach 10,670 people using various social mobilization methods, such as convening of sensitization meetings-dubbed 'mosquito days'-mainly spearheaded by primary school pupils. A total of 23 local primary schools participated in creating awareness on malaria prevention and control during the project phase. The collaboration with other departments like fisheries led to stocking of more than 20 fishponds with a total of 18,000 fingerlings in the years 2017 and 2018. Fish ponds provided an opportunity for income generation to the community. In partnership with the county government health department, the project was able to re-train 40 CHVs on IVM and malaria case management in the area. Additionally, 40 fish farmers were re-trained on fish farming as part of income generating activity (IGA) while 10 CBOs made up of 509 members received both eucalyptus and Ocimum kilimandscharicum seedlings that were distributed to four CBOs composed of 152 members. Four primary schools made up of 113 health club members also received eucalyptus seedlings as part of IGA in addition to fish farming. In total, around 20,000 eucalyptus seedlings were distributed to the community as part of IGA initiatives. By the end of 2018, the project was able to reach 25,322 people in the community during its two-year advocacy and social mobilization initiatives.

Conclusion: Through advocacy and social mobilization, the IVM strategy improved inter-sectoral collaboration, enhanced capacity building and community participation. However, more IVM related activities are needed to effectively mobilize available resources and increase community participation in malaria control.

Keywords: Advocacy; Capacity building; Community participation; IVM; Malaria; Social mobilization.

• Community Participation / statistics & numerical data*
• Malaria / prevention & control*
• Mosquito Control / organization & administration*
• Mosquito Vectors*
• Stakeholder Participation*

Grants and funding

• BV HH-07 / 2016-18/Biovision (BV) Foundation

• Open access
• Published: 09 November 2022

A prospective cohort study of Plasmodium falciparum malaria in three sites of Western Kenya

• Benyl M. Ondeto 1 , 3 ,
• Xiaoming Wang 2 ,
• Harrysone Atieli 3 ,
• Daibin Zhong 2 ,
• Guofa Zhou 2 ,
• Ming-Chieh Lee 2 ,
• Pauline Winnie Orondo 3 , 4 ,
• Kevin O. Ochwedo 1 , 3 ,
• Collince J. Omondi 1 , 3 ,
• Simon M. Muriu 5 ,
• David O. Odongo 1 ,
• Horace Ochanda 1 ,
• James Kazura 6 ,
• Andrew K. Githeko 3 , 7 &
• Guiyun Yan 2  

Parasites & Vectors volume  15 , Article number:  416 ( 2022 ) Cite this article

1938 Accesses

2 Citations

1 Altmetric

Metrics details

Malaria in western Kenya is currently characterized by sustained high Plasmodial transmission and infection resurgence, despite positive responses in some areas following intensified malaria control interventions since 2006. This study aimed to evaluate long-term changes in malaria transmission profiles and to assess patterns of asymptomatic malaria infections in school children aged 5–15 years at three sites in western Kenya with heterogeneous malaria transmission and simultaneous malaria control interventions.

The study was conducted from 2018 to 2019 and is based on data taken every third year from 2005 to 2014 during a longitudinal parasitological and mosquito adult surveillance and malaria control programme that was initiated in 2002 in the villages of Kombewa, Iguhu, and Marani. Plasmodium spp. infections were determined using microscopy. Mosquito samples were identified to species and host blood meal source and sporozoite infections were assayed using polymerase chain reaction.

Plasmodium falciparum was the only malaria parasite evaluated during this study (2018–2019). Asymptomatic malaria parasite prevalence in school children decreased in all sites from 2005 to 2008. However, since 2011, parasite prevalence has resurged by > 40% in Kombewa and Marani. Malaria vector densities showed similar reductions from 2005 to 2008 in all sites, rose steadily until 2014, and decreased again. Overall, Kombewa had a higher risk of infection compared to Iguhu ( χ 2  = 552.52, df  = 1, P  < 0.0001) and Marani ( χ 2  = 1127.99, df  = 1, P  < 0.0001). There was a significant difference in probability of non-infection during malaria episodes (log-rank test, χ 2  = 617.59, df  = 2, P  < 0.0001) in the study sites, with Kombewa having the least median time of non-infection during malaria episodes. Gender bias toward males in infection was observed ( χ 2  = 27.17, df  = 1, P  < 0.0001). The annual entomological inoculation rates were 5.12, 3.65, and 0.50 infective bites/person/year at Kombewa, Iguhu, and Marani, respectively, during 2018 to 2019.

Conclusions

Malaria prevalence in western Kenya remains high and has resurged in some sites despite continuous intervention efforts. Targeting malaria interventions to those with asymptomatic infections who serve as human reservoirs might decrease malaria transmission and prevent resurgences. Longitudinal monitoring enables detection of changes in parasitological and entomological profiles and provides core baseline data for the evaluation of vector interventions and guidance for future planning of malaria control.

Graphical abstract

Globally, an estimated 241 million cases of malaria were reported in 2020 resulting in approximately 627,000 deaths; 96% of these deaths occurred in Africa and children aged < 5 years accounted for 77% of these deaths [ 1 ]. In Kenya, an estimated 27 million malaria cases and 12,600 deaths attributed to malaria were reported in 2020 [ 1 ]. Since 2000, malaria mortality and morbidity have declined significantly in African countries, including Kenya, and have been attributed chiefly to the scale-up of insecticide-treated net (ITN) distributions, indoor residual spraying (IRS), and artemisinin-based combination therapies (ACTs) [ 2 ]. Nonetheless, malaria remains a major public health concern in Africa.

Kenya’s Ministry of Health began the country’s first free mass long-lasting insecticidal net (LLIN) distribution in 2006 to children under 5 years and pregnant women, followed by a second distribution in 2011 aiming for universal coverage in targeted areas [ 3 ]. Thereafter, there have been three successive rounds of distribution in 2014, 2017, and 2021 to boost LLIN coverage and replace worn nets [ 3 , 4 , 5 ]. Indoor residual spraying applications began in 2005 to prevent epidemics in malaria epidemic-prone areas in the highlands [ 6 ]. To reduce the malaria burden in the Lake Victoria endemic zone, IRS was implemented in targeted districts from 2008 to 2012 [ 6 , 7 , 8 ]. However, IRS was not applied from 2012 to 2016 because of the detection of widespread pyrethroid resistance in malaria vector populations and lack of a registered non-pyrethroid insecticide in the country [ 9 , 10 ]. After 5 years of no treatments, IRS was restarted in 2017 with the micro-encapsulated organophosphate insecticide pirimiphos-methyl (Actellic® 300CS) and applied during successive rounds from 2018 to 2021 in two targeted counties (Migori and Homa Bay) located in the Lake Victoria endemic zone, where intense malaria transmission occurs throughout the year [ 11 ]. Artemisinin-based combination therapies began in 2004 after several years of sulfadoxine-pyrimethamine treatments (1998–2003) and earlier recognition of widespread antimalarial drug failures (e.g. chloroquine) [ 12 , 13 ]. Malaria control programmes face numerous challenges, among them development of pyrethroid resistance in malaria vectors [ 14 ], changes in vector dominance and behaviour [ 15 , 16 , 17 ], and the emergence of antimalarial drug resistance [ 18 ]. In an effort to mitigate insecticide resistance, the World Health Organization (WHO) has recommended conducting IRS with organophosphate and neonicotinoid insecticides and using pyrethroid-piperonyl butoxide (PBO) synergized treated nets [ 1 ], which have been distributed in targeted counties in Kenya from 2020 to 2021.

Despite these malaria control efforts, areas in western Kenya are experiencing heterogeneity in malaria transmission after interventions, with some areas indicating a decline in transmission, while in others, transmission has remained unchanged or has resurged [ 5 , 19 , 20 , 21 ]. A study in western Kenya linked these contrasting outcomes to malaria vector species composition shifts, insecticide resistance, and climatic warming [ 21 ]. Similar observations of varying outcomes in malaria control have been observed elsewhere in Africa [ 22 ].

This study aimed to evaluate long-term changes in malaria transmission profiles and patterns of asymptomatic malaria infection in three sites with different transmission intensities in western Kenya after distributions of new pyrethroid-PBO treated LLINs and applications of new IRS formulations. Hopefully, the results presented here will help in assessing vector interventions, serve as a baseline for the evaluation of new interventions, and guide future control planning by the Kenya National Malaria Control Programme.

The study was conducted in three sites in western Kenya, each with different malaria transmission intensity: two highland sites, Iguhu (0°08′53′′N; 34°47′16′′E, 1430–1580 m elevation) (mesoendemic) in Kakamega County and Marani (0°35′13′′S; 34°48′11′′E, 1540–1740 m elevation) (hypoendemic) in Kisii County, and one lowland site in Kombewa (0°07′10′′S; 34°29′04′′E, 1150–1300 m elevation) (holoendemic) in Kisumu County (Fig.  1 ).

Map of the study sites in western Kenya

The climate in western Kenya consists mainly of a bimodal pattern of rainfall, a long rainy season between April and June, and a short rainy season between October and November [ 19 ]. The hot and dry season is from January to February while the cool and dry season from July to September [ 19 ]. All sites have shown variations in monthly cumulative precipitation and monthly mean maximum and minimum temperatures, ranging from 29.1 °C to 14.5 °C, respectively [ 19 , 21 , 23 ].

Plasmodium falciparum is the primary malaria parasite species in the three sites [ 19 ]. The first mass distribution of LLINS in 2006 in western Kenya led to a decline of both asymptomatic malaria and clinical cases [ 21 ]. The second mass distribution in 2011 was characterized by a positive response at Iguhu but Kombewa and Marani experienced sustained high P . falciparum transmission and infection resurgences, respectively, despite a third round of LLIN distributions in 2015 [ 21 ].

The predominant malaria vector species in the study sites are Anopheles gambiae s.s., An . arabiensis , and An . funestus [ 19 , 24 ]. In the lowland site, An . funestus is the most abundant and infectious malaria vector, while in the highland sites An . gambiae s.s. is the main vector responsible for Plasmodial transmission. Recent studies in this region have observed an increase in the proportion of An . arabiensis in the highlands because of vector interventions using LLINs and IRS; these measures may be suppressing the more anthropophilic and endophilic An . gambiae s.s. and killing fewer of the more zoophilic An . arabiensis [ 25 ]. Hence, high bednet coverage in western Kenya may explain decreases in vector densities of An. gambiae s.s. in the three sites, reductions of An. funestus in Iguhu and Kombewa, and temporal alterations in feeding behaviour of An. gambiae to earlier host seeking [ 20 ].

Study design

Historic plasmodium falciparum parasite prevalence and vector densities.

This study was based on longitudinal parasitological and adult vector surveillance that commenced in 2002 (Iguhu) and 2003 (Kombewa and Marani) [ 19 ] to date. Snapshots of these data were taken every 3 years from 2005 to 2014 [ 5 , 20 , 21 ]. Data (years 2005, 2008, 2011, and 2014) from this period form the basis for the current study conducted between 2018 and 2019.

Parasitological surveys

A cohort of 514 volunteer school-aged children aged 5–15 years were enrolled (January–March 2018) for monthly Plasmodium spp. surveys between 1 January 2018 and 31 October 2019 in Kombewa, Iguhu, and Marani (Fig.  2 ). The sample size was calculated based on the size of the study population and parasite prevalence from a previous study [ 5 ]. Consent was obtained from parents or guardians before children could participate in the study. Children with no reported chronic or acute illness, except malaria, were allowed to participate in the study. At the sampling time, children who were found to have fever were referred to the nearest government health facility for diagnosis and treatment according to Kenyan government malaria treatment guidelines [ 26 ].

Study design flow chart of the cohort study

Blood samples were collected using the finger-prick method and thick and thin smears prepared on labeled slides for malaria parasite species identification and parasite counts using microscopy. Malaria parasite counts were scored against 200 leukocytes. A second microscopist carried out random checks on the slide counts to ensure microscopy quality. Parasite density was expressed as parasites per μl, assuming a count of 8000 white blood cells per μl of blood [ 27 ]. Plasmodium spp. infection data collected from all participants were subjected to prevalence analyses; however, only participants with at least 6 months of follow-up were included in the Plasmodium spp. infection pattern analyses, including duration and probability of non-infections (Fig.  2 ).

Entomological surveys

Collections of indoor resting vector populations were conducted monthly by the pyrethrum spray catch (PSC) method [ 28 ] in 30 randomly selected houses in each study site between 1 January 2018 and 31 October 2019. Mosquitoes were identified morphologically as either Anopheles gambiae s.l. or An . funestus [ 29 ]. DNA was extracted [ 30 ] from the legs and wings of each mosquito specimen to speciate sibling species in An . gambiae s.l. and An . funestus using conventional polymerase chain reaction (PCR), as described by Scott et al. [ 31 ] and Koekemoer et al. [ 32 ], respectively. The DNA extracted from the abdomen of each freshly fed female mosquito was used to identify host blood meal sources using a multiplexed PCR assay [ 33 ]. The DNA extracted from the head and thorax of each mosquito specimen was used to determine sporozoite infections of Plasmodium spp. by using a multiplexed real-time quantitative PCR (qPCR) assay [ 34 , 35 ].

Climatic data

Mean monthly rainfall and maximum and minimum temperature from 2018 to 2019 were obtained from the Kenya Meteorological Department for meteorological stations in Kakamega (for Iguhu), Kisii (for Marani), and Kisumu (for Kombewa).

Data management and analysis

The variations in parasite prevalence between different time periods at Kombewa, Iguhu, and Marani were compared using Tukey-Kramer HSD test of analysis of variance (ANOVA) with repeated measures. In addition, the differences of vector densities between different time periods at each site were compared using non-parametric Wilcoxon rank-sum tests. Means (95% confidence interval, CI) and proportions were calculated for vector and parasite populations. For the primary malaria species, Plasmodium falciparum , parasite/gametocyte prevalence for each site, each month, was expressed as the percentage of microscopically positive samples over the total number of samples tested. The Chi-square test was used to determine statistical differences in the parasite/gametocyte prevalence among the study sites and parasite prevalence by age and gender category in each study site. Geometric mean parasite density and variations in proportion by month infected in the age and gender in each site were compared using Wilcoxon/Kruskal-Wallis tests. The variations in the distribution of the proportion of surveys being infected among the study sites were determined using Tukey-Kramer HSD test of ANOVA. Multiple Imputation by Chained Equations (MICE) simulation was done to impute the missing data in the time-to-event analysis. A Kaplan-Meier curve was built to analyze the probability of non-infection during malaria episodes in each study site. The log-rank test was applied to compare the probability of non-infection during malaria episodes in the three study sites adjusted for multiple comparisons with Bonferroni corrections. Wald approximations were used for hazard ratio 95% confidence interval limit effects. Hazard ratios for the asymptomatic malaria infections were compared with proportional hazards fit by study sites, gender, and age groups.

The monthly density of adult anopheline mosquitoes in each study site was calculated as the average number of females per house per night (f/h/n) based on monthly surveys. Vector density variation among study sites was compared using Wilcoxon/Kruskal-Wallis tests. The human blood index (HBI) was calculated as the proportion of blood-fed Anopheles mosquito samples that had fed on humans to the total tested [ 36 ]. The sporozoite rates for each site and vector species were calculated as the proportion of Anopheles mosquito samples positive for Plasmodium spp. to the total number tested. The annual entomological inoculation rates (EIRs) for each site and vector species were calculated as the product of the sporozoite rate and human biting rates [ 37 ]. Differences in the mean annual rainfall and mean annual maximum and minimum temperatures between the study sites were computed using the Tukey-Kramer HSD test of ANOVA with repeated measures. These analyses were done using JMP Pro 16 (SAS Institute, Inc.) and R statistical software (version 4.0.3; R foundation for statistical computing, Vienna, Austria).

Changes in parasite prevalence and vector densities in Kombewa, Iguhu, and Marani are shown in Table 1 from 2005 to 2014. Similar trends in parasite prevalence were observed in the three sites, i.e., declining parasite prevalence from 2005 to 2008 in all sites, and a rebounding trend in prevalence from 2008 in Iguhu and 2011 in Kombewa and Marani (Table 1 ). In Kombewa, parasite prevalence decreased slightly from 2005 (51.16%, 95% CI 46.79–55.54) to 2008 (48.06%, 95% CI 41.61–54.51) (Tukey-Kramer HSD test, P  > 0.05) and then declined sharply from 2008 to 2011 (29.80%, 95% CI 19.50–40.10) (Tukey-Kramer HSD test, P  = 0.006). After that, it rose significantly to 45.86% (95% CI 39.34–52.38) in 2014 (Tukey-Kramer HSD test, P  = 0.02). In Iguhu, a sharp decline in parasite prevalence was observed from 2005 (26.61%, 95% CI 21.88–31.34) to 2008 (6.45%, 95% CI 4.58–8.32) (Tukey-Kramer HSD test, P  < 0.0001) and rose steadily to 16.82% (95% CI 13.52–20.12) in 2014 (Tukey-Kramer HSD test, P  = 0.002). In Marani, a steady decline of parasite prevalence was observed from 2005 (1.95%, 95% CI 0.82–3.09) to 2011 (0.35%, 95% CI 0.05–0.66) (Tukey-Kramer HSD test, P  = 0.04), after which there was a sharp rise in 2014 (4.44%, 95% CI 3.37–5.51) (Tukey-Kramer HSD test, P  < 0.0001).

The indoor resting densities of An . gambiae s.l. and An . funestus varied significantly in all sites. The vector densities showed reductions from 2005 to 2008 in all sites and thereafter rose steadily until 2014 (Table 1 ). Studies from 2005 and 2008 indicate that the indoor resting densities of malaria vectors decreased sharply in Kombewa from 1.04 (95% CI 0.14–1.93) to 0.31 (95% CI 0.15–0.47) f/h/n for An . gambiae s.l. (Wilcoxon test, Z  = 1.24, P  = 0.21) and from 2.14 (95% CI 1.16–3.12) to 0.52 (95% CI 0.21–0.83) f/h/n for An . funestus (Wilcoxon test, Z  = 3.38, P  = 0.0007). Similarly, a decline was observed in Iguhu with a reduction from 2.56 (95% CI 0.21–4.91) to 0.36 (95% CI 0.24–0.48) f/h/n for An . gambiae s.l. (Wilcoxon test, Z  = 1.62, P  = 0.11) and that of An . funestus changed significantly from 0.29 (95% CI 0.09–0.49) to 0.02 (95% CI 0.01–0.04) f/h/n (Wilcoxon test, Z  = 4.03, P  < 0.0001). In Marani, An . gambiae s.l. densities decreased from 0.03 (95% CI 0.00–0.05) to 0.01 (95% CI 0.00–0.02) f/h/n (Wilcoxon test, Z  = 1.07, P  = 0.29) between 2005 and 2008, while no An . funestus were found during the 2 years. Between 2008 and 2014, the population of indoor resting vectors rose steadily in Kombewa ( An . gambiae s.l., Wilcoxon test, Z  = 3.23, P  = 0.001; An . funestus , Wilcoxon test, Z  = 2.51, P  = 0.01), Iguhu ( An . gambiae s.l., Wilcoxon test, Z  = 1.47, P  = 0.14; An . funestus , Wilcoxon test, Z  = 4.17, P  < 0.0001) and Marani ( An . gambiae s.l., Wilcoxon test, Z  = 3.00, P  = 0.003; An . funestus , Wilcoxon test, Z  = 4.41, P  < 0.0001).

Plasmodium falciparum parasite prevalence, gametocyte prevalence, and parasite density

In the 2018–2019 survey, only P . falciparum was found and evaluated. The P . falciparum prevalence in Kombewa was significantly higher compared to Iguhu ( χ 2  = 552.52, df  = 1, P  < 0.0001) and Marani ( χ 2  = 1127.99, df  = 1, P  < 0.0001) (Fig.  3 ). Compared to 2011, parasite prevalence in 2018–2019 has resurged by > 40% in Kombewa and Marani, whereas in Iguhu, it has decreased by 7.3%. There were no significant differences in P. falciparum prevalence between males and females in all sites except Kombewa and no significant differences in P. falciparum prevalence between age groups at all sites (Additional file 3 : Table S1).

Plasmodium parasite prevalence ( a ) and gametocyte prevalence ( b ) in Kombewa, Iguhu, and Marani in western Kenya. Differences in the parasite/gametocyte prevalence among study sites were determined using Chi-square test

The P . falciparum gametocyte prevalence was significantly higher in Kombewa compared to Iguhu and Marani ( χ 2  = 7.69, df  = 2, P  = 0.02) (Fig.  3 ).

In Kombewa, there was a significant difference in the geometric means of P . falciparum density between the two age groups, with higher parasite density in the 5–10 years old group. Similarly, males had higher parasite density compared to females (Additional file 3 : Table S1).

Plasmodium falciparum infection patterns

The proportion of months infected varied greatly in Kombewa (35.9%), Iguhu (14.9%), and Marani (5.8%) (Tukey-Kramer HSD test, P  < 0.0001) (Additional file 1 : Fig S1). No significant age and gender variations were found in the proportion of months infected in the study sites (Additional file 3 : Table S1). Additional file 2 : Fig S2 indicates the distribution of malaria infection patterns in the age and gender groups in the study sites.

As shown in Fig.  4 , the median time of non-infection during malaria first episode was 1.90 [interquartile range (IQR): 1.61–2.19] months, 5.46 (IQR: 4.30–6.62) months, and 10.86 (IQR: 9.03–12.69) months in Kombewa, Iguhu, and Marani, respectively. Median time from first to second malaria episodes was 1.95 (IQR: 1.64–2.32) months, 10.37 (IQR: 8.18–12.57) months, and 65.96 (IQR: 35.38–122.98) months in Kombewa, Iguhu, and Marani, respectively. When exploring time intervals from second to third malaria episodes, the median time was 3.24 (IQR: 2.56–4.10) months, 29.10 (IQR: 16.20–52.25) months, and 491.07 (IQR: 60.23–4003.94) months in Kombewa, Iguhu, and Marani, respectively. The median time of non-infection for all malaria episodes was 17.30 (IQR: 16.77–17.75) months, 23.26 (IQR: 22.40–24.13) months, and 33.40 (IQR: 30.14–34.65) months in Kombewa, Iguhu, and Marani, respectively. There was a significant difference in probability of non-infection during malaria first episode (log-rank test, χ 2  = 171.78, df = 2, P  < 0.0001), first–second episodes (log-rank test, χ 2  = 179.33, df  = 2, P  < 0.0001), second–third episodes (log-rank test, χ 2  = 245.77, df  = 2, P  < 0.0001), and all episodes (log-rank test, χ 2  = 617.59, df  = 2, P  < 0.0001) in the study sites.

Kaplan-Meier probability of non-infection during ( a ) p.f. malaria first episode, ( b ) first–second episodes, ( c ) second–third episodes, and ( d ) all p . f . episodes in Kombewa, Iguhu, and Marani in western Kenya. Abbreviations: p . f ., Plasmodium falciparum . The probability of non-infection during malaria episodes in the study sites were compared using log-rank test

For male gender, Kombewa and Iguhu sites were statistically significant risk factors associated with asymptomatic malaria infection. (Additional file 4 : Table S2). Unadjusted hazard ratios for the infection were significantly higher in Kombewa and Iguhu compared to Marani, with similar results after adjustment for gender and age. Females had a significantly lower unadjusted hazard ratio for the infection than males, but was insignificant after adjustment for sites and ages.

Vector species composition and densities

A total of 583 female anophelines were collected between 1 January 2018 and 31 October 2019, comprising 458 (78.6%) An . gambiae s.l. and 125 (21.4%) An . funestus . Of these, 479 specimens (391 An . gambiae s.l. and 88 An . funestus ) were analyzed for sibling species. For the An . gambiae s.l. specimens, PCR results indicated that 77.8% were An . gambiae s.s. and 22.2% An . arabiensis in Kombewa, 85.7% An . gambiae s.s. and 14.3% An . arabiensis in Iguhu, and 33.3% An . gambiae s.s. and 66.7% An . arabiensis in Marani. All the An . funestus subjected to species identification from the study sites were confirmed as An . funestus s.s.

The mean indoor resting densities of An . gambia e s.l. were significantly different among the study sites (Wilcoxon test, χ 2  = 253.44, df  = 2, P  < 0.0001), with Iguhu having the highest densities and Marani the lowest densities (Fig.  5 ; Table 2 ). Also, the mean densities of An . funestus were significantly different among study sites (Wilcoxon test, χ 2  = 26.03, df  = 2, P  < 0.0001), with Kombewa having the highest densities and Marani the lowest (Fig.  5 ; Table 2 ). Compared to 2014, vector density has decreased by > 60% in all sites except in Iguhu, where An . gambiae s.l. density decreased slightly by 9%.

Indoor resting densities of An. gambiae s.l. and An. funestus in Kombewa ( a ), Iguhu ( b ), and Marani ( c ) in western Kenya. Differences in vector density among study sites was compared using Wilcoxon/Kruskal-Wallis tests

Blood meal indices and annual entomological inoculation rate

The blood meals of An . gambiae s.l. and An . funestus were mostly of bovine (55.3%) and human (90.4%) origin, respectively, in both Kombewa and Iguhu (Additional file 5 : Table S3). Due to the small number of mosquito collections in Marani, the HBI was not analyzed. Overall, the HBI of An . gambiae s.l. and An . funestus was 41.10% and 88.00%, respectively.

The annual EIR of An . funestus was threefold higher in Kombewa compared to Iguhu (Table 3 ). In Iguhu, the annual EIR of An . gambiae s.l. was threefold higher than the corresponding value of An . funestus (Table 3 ). Due to the small number of mosquito collections in Marani, the annual EIR was not analyzed. The overall total annual EIRs were 5.12, 3.65, and 0.50 infective bites/person/year (ib/p/yr) at Kombewa, Iguhu, and Marani, respectively.

Rainfall among the three study sites was not statistically different (ANOVA, F (2, 69)  = 1.24, P  > 0.05). The mean annual maximum (ANOVA, F (2, 69)  = 29.72, P  < 0.0001) and minimum (ANOVA, F (2, 69)  = 77.19, P  < 0.0001) temperatures were significantly different among the sites (Additional file 6 : Fig S3). The mean annual temperature between Iguhu and Marani was not significantly different (Tukey-Kramer HSD test, P  = 0.0004), whereas it was significantly different between Iguhu and Kombewa and between Marani and Kombewa (Tukey-Kramer HSD test, all P  < 0.0001) (Additional file 6 : Figure S3).

This study evaluated long-term changes in malaria transmission profiles in three sites in western Kenya with heterogeneous malaria transmission and high coverage with malaria control interventions [ 10 , 38 , 39 ]. The study also described the pattern of asymptomatic malaria infection in the study sites. Findings of the study demonstrated that malaria prevalence remains high or has resurged in some sites despite continuous intervention efforts. Results also showed that Kombewa had a higher risk of asymptomatic infection than Iguhu and Marani and further reported a gender bias towards males in infection.

Parasite prevalence has been decreasing since 2005 in the three sites and is likely associated with a reduction in vector abundance after free mass LLIN distributions after 2006, application of IRS, and increased use of ACT treatment [ 8 , 19 ]. However, there has been an observed resurgence of parasite prevalence since 2008 (Iguhu) and 2011 (Kombewa and Marani) and malaria vector densities since 2008 in all sites. These changes may be attributed to worn-out bednets and irregular use of nets; reduced optimum efficacy of LLINs over time; development of pyrethroid resistance in malaria vectors and less coverage of IRS in epidemic-prone areas [ 6 , 19 ]. Additionally, in 2014 the resurgence in malaria transmission observed in Marani may also be explained by the increase in ambient temperatures between 2012 and 2015 and high rainfall in 2014 [ 21 ]. The sharp decrease in indoor resting vector densities since 2014 is likely due to continuous scaling up of LLINs in the study area. Nevertheless, despite the decrease in vector densities, persistent malaria transmission in the context of extensive malaria vector control has been observed, and this could be attributed to outdoor vector biting and resting behaviour to avoid physical contact with insecticide-treated materials, changes in vector behaviour to early evening biting and early exiting from houses, as reported in western Kenya and other parts of Africa [ 20 , 40 , 41 ].

The 2018–2019 study observed a higher prevalence of gametocytes in Kombewa and Iguhu than in Marani and shows that the populations living in Kombewa and Iguhu maintain a large reservoir of infectious gametocytes, thus leading to stable and continuous malaria transmission. In contrast, the population living in highland village of Marani consists of a high proportion of susceptible individuals and consequently, under suitable climatic conditions, may experience malaria resurgences [ 42 ]. Hence, monitoring air temperature and precipitation data is crucial in predicting vector and parasite dynamics, particularly in the highlands where slight changes in these parameters could lead to malaria epidemics [ 21 ].

Many factors have been associated with heterogeneity in malaria risk and include biotic, abiotic, and socio-economic factors [ 43 ]. Kombewa had the highest risk and hazard ratio of asymptomatic malaria infections in the study. Furthermore, the median time interval and probability of non-infection during malaria episodes were least in Kombewa compared to other study sites, indicating increased malaria exposure. The study further reported a gender bias towards males in asymptomatic malaria infection. Briggs et al. (2020) [ 44 ] observed that the sex-based difference might be elucidated by a slower clearance of infection in males than females due to differences in immune responses [ 44 , 45 , 46 ]. In other studies, this sex-based difference has been postulated to socio-behavioural factors that place men at a higher risk [ 47 , 48 ]. Higher risk of malaria in male children and adolescents is likely linked to an array of physiological and behavioural changes that could contribute to the observed gender bias in this study. The possible explanations put forward for the gender difference in malaria infection include roles of sex hormones in the functioning of the immune system, immunological factors, cultural factors, and vector exposure, such as not sleeping under a net [ 45 , 46 , 47 , 49 ]. Therefore, research studies on sex-based differences in infectious diseases such as malaria are essential for providing optimum disease management for both genders [ 46 ]. In Kombewa, young children had a higher parasite density than older individuals. The declining risk of parasitaemia as age increases has been documented in other parts of Africa with stable malaria transmission, since individuals develop semi-immunity after continued exposure to infectious mosquito bites [ 50 , 51 ].

Studies conducted over 2 decades ago showed that the HBI of indoor resting An . gambiae s.s. in western Kenya was 96–97%, indicating that they had fed exclusively on humans [ 52 , 53 ]. However, in this investigation, the overall HBI of An . gambiae s.l. in all study sites was only 41.1%. This behavioural plasticity in host seeking suggests that there has been a shift in blood meal sources, which could be attributed to extensive bednet coverage in the region [ 54 ]. Conversely, An . funestus was highly anthropophilic, an observation previously made in Kenya and other parts of Africa [ 52 , 55 , 56 ]. Furthermore, in studies conducted in Kombewa, the highly anthropophilic An . funestus has been reported to have high resistance to pyrethroids, and changes in their biting behaviour could be a major factor sustaining high transmission in the area amidst extensive malaria vector control [ 20 , 21 ].

The EIRs obtained in previous studies by Githeko et al. [ 57 ] and Beier et al. [ 58 ] were exceedingly high (91–416 ib/p/yr) in western Kenya. Since then, there has been a decline in the annual P . falciparum inoculation rates, as observed by Ndenga et al. (2016), who reported the total annual EIRs as 31.1, 16.6, and 0.4 ib/p/yr at Kombewa, Iguhu, and Marani, respectively [ 23 ]. In the current study, the lower inoculation rates recorded could be attributed to reduced vector densities and, to some extent, a shift to non-human feeding by the malaria vectors due to high bednet coverage in the study areas [ 54 ]. Nevertheless, An . funestus and An . gambiae s.l. played major roles in malaria transmission in Kombewa and Iguhu, respectively, despite the comparatively low vector densities, indicating high vectorial efficiency of these anophelines in transmitting malaria in the region.

One limitation of our study was that parasitological surveys were based on microscopy only, which may not detect light plasmodial infections compared to highly sensitive PCR-based techniques. Hence, the P . falciparum prevalence and infection pattern may have been underestimated. A second limitation was the lack of long-term information on outdoor malaria transmission dynamics, which may have provided insight to the resurgence in P . falciparum transmission despite continuous intervention efforts.

Malaria prevalence remains high and has resurged in some sites in western Kenya despite continuous intervention efforts. Hence, long-time monitoring of malaria transmission profiles is essential in evaluating the success of current interventions, accurately measuring changing malaria epidemiology, and directing strategies for future control and elimination efforts. Residing in malaria-endemic villages and male gender were significant risk factors associated with asymptomatic malaria infection, with these individuals serving as human reservoirs for sustained malaria transmission. Consequently, targeted control might effectively reduce those with asymptomatic infections and potentially decrease malaria transmission and prevent resurgences.

Availability of data and materials

All data generated or analyzed during this study are included in this published article and its additional files.

Abbreviations

Artemisinin-based combination therapies

Deoxyribonucleic acid

Entomological inoculation rates

Human biting index

International Center of Excellence for Malaria Research

Indoor residual spraying

Insecticide-treated nets

Long-lasting insecticidal nets

Piperonyl butoxide

Polymerase chain reaction

Pyrethrum spray catches

Quantitative polymerase chain reaction

sensu stricto

World Health Organization

WHO. World Malaria Report 2021. Geneva, World Health Organization; 2021 1–263. https://www.who.int/publications/i/item/9789240040496 .

WHO. World Malaria Report 2020: 20 years of global progress and challenges. Geneva, World Health Organization; 2020. 1–247. https://www.who.int/publications/i/item/9789240015791 .

Ministry of Health. The epidemiology and control profile of malaria in Kenya: reviewing the evidence to guide the future vector control. National Malaria Control Programme, Ministry of Health. Technical support provided by the LINK Project (London School of Hygiene and Tropical Medicine and the Information for Malaria (INFORM) Project, KEMRI-Wellcome Trust Research Programme), Nairobi, Ken; 2016. https://dhsprogram.com/pubs/pdf/MIS36/MIS36.pdf .

National Malaria Control Programme-Kenya. https://web.facebook.com/nmcpkenya?_rdc=1&_rdr .

Zhou G, Lee M-C, Githeko AK, Atieli HE, Yan G. Insecticide-treated net campaign and malaria transmission in western Kenya: 2003–2015. Front Public Heal. 2016;4:153.

CAS   Google Scholar  

Division of Malaria Control [Ministry of Public Health and Sanitation], Kenya national bureau of statistics and IM. 2010 Kenya malaria indicator survey Nairobi, Kenya: DOMC, KNBS and ICF Macro; 2011. https://dhsprogram.com/publications/publication-mis7-mis-final-reports.cfm .

PMI. Indoor residual spraying 2 (IRS 2) Kenya: task order 2 final report. https://pdf.usaid.gov/pdf_docs/pdacy423.pdf .

Gimnig JE, Otieno P, Were V, Marwanga D, Abong’o D, Wiegand R, et al. The effect of indoor residual spraying on the prevalence of malaria parasite infection, clinical malaria and anemia in an area of perennial transmission and moderate coverage of insecticide treated nets in western Kenya. PLoS ONE. 2016;11(1):e0145282.

Article   PubMed   PubMed Central   Google Scholar  

AIRS Kenya entomological monitoring annual report Rockville, MD. PMI|Africa IRS (AIRS) project indoor residual spraying (IRS 2) task order six, Abt Associates Inc; 2018. https://d1u4sg1s9ptc4z.cloudfront.net/uploads/2021/03/kenya-2017-entomological-monitoring-final-report.pdf .

National Malaria Control Programme (NMCP), Kenya national bureau of statistics (KNBS) and II. Kenya malaria indicator survey 2015 Nairobi, Kenya, and Rockville, Maryland, USA; 2016. https://dhsprogram.com/publications/publication-DM81-Other-Dissemination-Materials.cfm .

The PMI VectorLink Kenya project 2021. https://pmivectorlink.org/where-we-work/kenya/ .

Okech BA, Mwobobia IK, Kamau A, Muiruri S, Mutiso N, Nyambura J, et al. Use of integrated malaria management reduces malaria in Kenya. PLoS ONE. 2008;3(12):e4050.

Okiro EA, Hay SI, Gikandi PW, Sharif SK, Noor AM, Peshu N, et al. The decline in paediatric malaria admissions on the coast of Kenya. Malar J. 2007;6:151.

Ondeto BM, Nyundo C, Kamau L, Muriu SM, Mwangangi JM, Njagi K, et al. Current status of insecticide resistance among malaria vectors in Kenya. Parasit Vectors. 2017;10:429.

Takken W. Do insecticide-treated bednets have an effect on malaria vectors? Trop Med Int Heal. 2002;7(12):1022–30.

Article   CAS   Google Scholar  

Ojuka P, Ii YB, Denoeud-Ndam L, Nabasumba C, Muller Y, Okia M, et al. Early biting and insecticide resistance in the malaria vector Anopheles might compromise the effectiveness of vector control intervention in Southwestern Uganda. Malar J. 2015;14:148.

Bayoh NM, Mathias DK, Odiere MR, Mutuku FM, Kamau L, Gimnig JE, et al. Anopheles gambiae : historical population decline associated with regional distribution of insecticide-treated bed nets in western Nyanza Province, Kenya. Malar J. 2010;9:62.

Kozlov M. Resistance to key malaria drugs confirmed in Africa. Nature. 2021;597:604.

Zhou G, Afrane YA, Vardo-Zalik AM, Atieli H, Zhong D, Wamae P, et al. Changing patterns of malaria epidemiology between 2002 and 2010 in western Kenya: the fall and rise of malaria. PLoS ONE. 2011;6(5):e20318.

Article   CAS   PubMed   PubMed Central   Google Scholar  

Ototo EN, Mbugi JP, Wanjala CL, Zhou G, Githeko AK, Yan G. Surveillance of malaria vector population density and biting behaviour in western Kenya. Malar J. 2015;14:244.

Kapesa A, Kweka EJ, Atieli H, Kamugisha E, Zhou G, Githeko AK, et al. Why some sites are responding better to anti-malarial interventions? A case study from western Kenya. Malar J. 2017;16:498.

Nkumama IN, O’Meara WP, Osier FHA. Changes in malaria epidemiology in Africa and new challenges for elimination. Trends Parasitol. 2017;33:128–40.

Article   PubMed   Google Scholar  

Ndenga B, Githeko A, Omukunda E, Munyekenye G, Atieli H, Wamai P, et al. Population dynamics of malaria vectors in western Kenya highlands. J Med Entomol. 2006;43:200–6.

Degefa T, Yewhalaw D, Zhou G, Lee MC, Atieli H, Githeko AK, et al. Indoor and outdoor malaria vector surveillance in western Kenya: implications for better understanding of residual transmission. Malar J BioMed Central. 2017;16:443.

Google Scholar  

Ototo EN, Githeko AK, Wanjala CL, Scott TW. Surveillance of vector populations and malaria transmission during the 2009/10 El Niño event in the western Kenya highlands: opportunities for early detection of malaria hyper-transmission. Parasit Vectors. 2011;4:144.

Ministry of Health. National guidelines for the diagnosis, treatment and prevention of malaria in Kenya 2016. https://pdf.usaid.gov/pdf_docs/PA00N1G9.pdf

Slutsker L, Taylor T, Wirima J, Steketee R. In-hospital morbidity and mortality due to malaria-associated severe anaemia in two areas of Malawi with different patterns of malaria infection. Trans R Soc Trop Med Hyg. 1994;88:548–51.

Article   CAS   PubMed   Google Scholar  

WHO. Manual on practical entomology in malaria. Part II methods and techniques Geneva, World Health Organization 1975. https://apps.who.int/iris/handle/10665/42481 .

Gillies M, Coetzee M. A supplement to the Anophelinae of Africa south of the Sahara. Sth Afr Inst Med Res. 1987;55:1–143.

Musapa M, Kumwenda T, Mkulama M, Chishimba S, Norris DE, Thuma PE, et al. A simple chelex protocol for DNA extraction from Anopheles spp. J Vis Exp. 2013;71:e3281.

Scott JA, Brogdon WG, Collins FH. Identification of single specimens of the Anopheles gambiae complex by the polymerase chain reaction. Am J Trop Med Hyg. 1993;49(4):520–9.

Koekemoer LL, Kamau L, Hunt RH, Coetzee M. A cocktail polymerase chain reaction assay to identify members of the Anopheles funestus (Diptera: Culicidae) group. Am J Trop Med Hyg. 2002;6(6):804–11.

Article   Google Scholar  

Kent RJ, Norris DE. Identification of mammalian blood meals in mosquitoes by a multiplexed polymerase chain reaction targeting cytochrome B. Am J Trop Med Hyg. 2005;73(2):336–42.

Shokoples SE, Ndao M, Kowalewska-Grochowska K, Yanow SK. Multiplexed real-time PCR assay for discrimination of Plasmodium species with improved sensitivity for mixed infections. J Clin Microbiol. 2009;47(4):975–80.

Veron V, Simon S, Carme B. Multiplex real-time PCR detection of P falciparum, P . vivax and P . malariae in human blood samples. Exp Parasitol. 2009;121(4):346–51.

Garrett-Jones C. The human blood index of malaria vectors in relation to epidemiological assessment. Bull World Heal Organ. 1964;30:241–61.

Macdonald G. The epidemiology and control of malaria. London: Oxford University Press; 1957.

Division of National Malaria Programme (DNMP) [Kenya] and ICF 2021 Kenya Malaria Indicator Survey 2020 Nairobi, Kenya, and Rockville Maryland USA

Dulacha D, Were V, Oyugi E, Kiptui R, Owiny M, Boru W, et al. Reduction in malaria burden following the introduction of indoor residual spraying in areas protected by long-lasting insecticidal nets in western Kenya, 2016–2018. PLoS ONE. 2022;17(4):e0266736.

Bamou R, Rono M, Degefa T, Midega J, Mbogo C, Ingosi P, et al. Entomological and anthropological factors contributing to persistent malaria transmission in Kenya, Ethiopia, and Cameroon. J infect Dis. 2021;223(S2):S155–70.

Wanjala CL, Mbugi JP, Ototo E, Gesuge M, Afrane YA, Atieli HE, et al. Pyrethroid and DDT resistance and organophosphate susceptibility among Anopheles spp. mosquitoes, western Kenya. Emerg Infect Dis. 2015;21(12):2193–6.

Wanjala CL, Waitumbi J, Zhou G, Githeko AK. Identification of malaria transmission and epidemic hotspots in the western Kenya highlands: its application to malaria epidemic prediction. Parasit Vectors. 2011;4:81.

Essendi WM, Vardo-Zalik AM, Lo E, Machani MG, Zhou G, Githeko AK, et al. Epidemiological risk factors for clinical malaria infection in the highlands of western Kenya. Malar J. 2019;18:211.

Briggs J, Teyssier N, Nankabirwa JI, Rek J, Jagannathan P, Arinaitwe E, et al. Sex-based differences in clearance of chronic Plasmodium falciparum infection. Elife. 2020;9:e59872.

Fish EN. The X-files in immunity: sex-based differences predispose immune responses. Nat Rev Immunol. 2008;8(9):737–44.

Bernin H, Lotter H. Sex bias in the outcome of human tropical infectious diseases: influence of steroid hormones. J infect Dis. 2014;209(S3):S107-113.

Pathak S, Rege M, Gogtay NJ, Aigal U, Sharma SK, Valecha N, et al. Age-dependent sex bias in clinical malarial disease in hypoendemic regions. PLoS ONE. 2012;7(4):e35592.

Camargo LMA, dal Colletto GMD, Ferreira MU, de Gurgel SM, Escobar AL, Marques A, et al. Hypoendemic malaria in Rondonia (Brazil, western Amazon region): seasonal variation and risk groups in an urban locality. Am J Trop Med Hyg. 1996;55(1):32–8.

Finda MF, Moshi IR, Monroe A, Limwagu AJ, Nyoni AP, Swai JK, et al. Linking human behaviours and malaria vector biting risk in south-eastern Tanzania. PLoS ONE. 2019;14(6):e0217414.

Rogier C, Tall A, Diagne N, Fontenille D, Spiegel A, Trape JF. Plasmodium falciparum clinical malaria: lessons from longitudinal studies in Senegal. Parassitologia. 1999;41:255–9.

CAS   PubMed   Google Scholar  

Sarpong N, Owusu-Dabo E, Kreuels B, Fobil JN, Segbaya S, Amoyaw F, et al. Prevalence of malaria parasitaemia in school children from two districts of Ghana earmarked for indoor residual spraying: a cross-sectional study. Malar J. 2015;14:260. https://doi.org/10.1186/s12936-015-0772-6 .

Githeko AK, Service MW, Mbogo CM, Atieli FK, Juma FO. Origin of blood meals in indoor and outdoor resting malaria vectors in western Kenya. Acta Trop. 1994;58(3–4):307–16.

Shililu JI, Maier WA, Seitz HM, Orago AS. Seasonal density, sporozoite rates and entomological inoculation rates of Anopheles gambiae and Anopheles funestus in a high-altitude sugarcane growing zone in western Kenya. Trop Med Int Heal. 1998;3(9):706–10.

Ndenga BA, Mulaya NL, Musaki SK, Shiroko JN, Dongus S, Fillinger U. Malaria vectors and their blood-meal sources in an area of high bed net ownership in the western Kenya highlands. Malar J. 2016;15:76. https://doi.org/10.1186/s12936-016-1115-y .

Mzilahowa T, Hastings IM, Molyneux ME, McCall PJ. Entomological indices of malaria transmission in Chikhwawa District, southern Malawi. Malar J. 2012;11:380. https://doi.org/10.1186/1475-2875-11-380 .

Tanga MC, Ngundu WI, Tchouassi PD. Daily survival and human blood index of major malaria vectors associated with oil palm cultivation in Cameroon and their role in malaria transmission. Trop Med Int Heal. 2011;16(4):447–57.

Githeko AK, Service MW, Mbogo CM, Atieli FK, Juma FO. Plasmodium falciparum sporozoite and entomological inoculation rates at the ahero rice irrigation scheme and the Miwani sugar-belt in western Kenya. Ann Trop Med Parasitol. 1993;87(4):379–91.

Beier JC, Perkins PV, Onyango FK, Gargan TP, Oster CN, Whitmire RE, et al. Characterization of malaria transmission by Anopheles (Diptera: Culicidae) in western Kenya in preparation for malaria vaccine trials. J Med Entomol. 1990;27(4):570–7.

Download references

Acknowledgements

We thank the community in Kombewa, Iguhu, and Marani for their support and willingness to participate in this research. In addition, we thank the field and laboratory team of the ICEMR project in Kenya for providing technical support during the study.

This research is supported by grants from the National Institutes of Health (U19 AI129326 and D43 TW001505).

Author information

Authors and affiliations.

Department of Biology, University of Nairobi, Nairobi, 00100, Kenya

Benyl M. Ondeto, Kevin O. Ochwedo, Collince J. Omondi, David O. Odongo & Horace Ochanda

Program in Public Health, College of Health Sciences, University of California at Irvine, Irvine, CA, 92697, USA

Xiaoming Wang, Daibin Zhong, Guofa Zhou, Ming-Chieh Lee & Guiyun Yan

Sub-Saharan Africa International Center of Excellence for Malaria Research, Tom Mboya University, Homa Bay, 40300, Kenya

Benyl M. Ondeto, Harrysone Atieli, Pauline Winnie Orondo, Kevin O. Ochwedo, Collince J. Omondi & Andrew K. Githeko

Department of Biochemistry, Jomo Kenyatta University of Agriculture and Technology, Nairobi, 00200, Kenya

Pauline Winnie Orondo

Department of Biological Sciences, Pwani University, Kilifi, 80108, Kenya

Simon M. Muriu

Center for Global Health and Disease, Case Western Reserve University, Cleveland, OH, 44106, USA

James Kazura

Centre for Global Health Research, Kenya Medical Research Institute, Kisumu, 40100, Kenya

Andrew K. Githeko

You can also search for this author in PubMed   Google Scholar

Contributions

BMO, HA, GZ, JK, AKG, and GY conceived and developed the study. BMO, HA, M-CL, DZ, XW, PWO, KOO, CJO, SMM, DOO, and HO participated in the design and implementation of parasitological and entomological studies. M-CL generated the map. DZ, BMO, PWO, KOO, and CJO carried out the laboratory analysis. XW, BMO, and GZ did data analysis and interpretation. BMO wrote the first draft of the manuscript. The final manuscript was edited by GY and AKG. All authors read and approved the final manuscript.

Corresponding authors

Correspondence to Benyl M. Ondeto or Guiyun Yan .

Ethics declarations

Ethics approval and consent to participate.

The study obtained ethical approval from the Maseno University Ethics Review Committee (MUERC Protocol No. 00660) and the University of California, Irvine Institutional Review Board (UCI IRB). Written informed consent was obtained from children’s parents/guardians before they were enrolled in the study and household heads

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1: figure s1..

Distribution of the proportion of surveys being infected in Kombewa ( a ), Iguhu ( b ), and Marani ( c ) in western Kenya.

Additional file 2: Figure S2.

Heat map showing the  Plasmodium falciparum  infection patterns in Kombewa, Iguhu, and Marani in western Kenya.

Additional file 3: Table S1.

Malaria infection in Kombewa, Iguhu, and Marani in western Kenya [Mean (95%CI)].

Additional file 4: Table S2.

Hazard ratios for the infection in Kombewa, Iguhu, and Marani in western Kenya.

Additional file 5: Table S3.

Host feeding preference of Anopheles mosquitoes in Kombewa, Iguhu, and Marani in western Kenya.

Additional file 6: Figure S3.

Variations in monthly maximum temperature, minimum temperature, mean temperature and monthly rainfalls in Kombewa ( a ), Iguhu ( b ), and Marani ( c ) in western Kenya.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

Ondeto, B.M., Wang, X., Atieli, H. et al. A prospective cohort study of Plasmodium falciparum malaria in three sites of Western Kenya. Parasites Vectors 15 , 416 (2022). https://doi.org/10.1186/s13071-022-05503-4

Download citation

Received : 12 April 2022

Accepted : 14 September 2022

Published : 09 November 2022

DOI : https://doi.org/10.1186/s13071-022-05503-4

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Transmission
• Plasmodium falciparum

Parasites & Vectors

ISSN: 1756-3305

• Submission enquiries: [email protected]

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• My Account Login
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 14 September 2023

Estimating the burden of severe malarial anaemia and access to hospital care in East Africa

• Peter Winskill   ORCID: orcid.org/0000-0003-3001-4959 1 ,
• Aggrey Dhabangi 2 ,
• Titus K. Kwambai 3 ,
• Amani Thomas Mori   ORCID: orcid.org/0000-0002-9968-3635 4 , 5 ,
• Andria Mousa 6 &
• Lucy C. Okell   ORCID: orcid.org/0000-0001-7202-6873 1  

Nature Communications volume  14 , Article number:  5691 ( 2023 ) Cite this article

1070 Accesses

1 Altmetric

Metrics details

• Clinical microbiology
• Health services
• Parasitology

Severe malarial anaemia can be fatal if not promptly treated. Hospital studies may under-represent the true burden because cases often occur in settings with poor access to healthcare. We estimate the relationship of community prevalence of malaria infection and severe malarial anaemia with the incidence of severe malarial anaemia cases in hospital, using survey data from 21 countries and hospital data from Kenya, Tanzania and Uganda. The estimated percentage of severe malarial anaemia cases that were hospitalised is low and consistent for Kenya (21% (95% CrI: 7%, 47%)), Tanzania (18% (95% CrI: 5%, 52%)) and Uganda (23% (95% CrI: 9%, 48%)). The majority of severe malarial anaemia cases remain in the community, with the consequent public health burden being contingent upon the severity of these cases. Alongside health system strengthening, research to better understand the spectrum of disease associated with severe malarial anaemia cases in the community is a priority.

Similar content being viewed by others

Projected health impact of post-discharge malaria chemoprevention among children with severe malarial anaemia in Africa

Lucy C. Okell, Titus K. Kwambai, … Amani Thomas Mori

Malaria surveillance, outbreak investigation, response and its determinant factors in Waghemra Zone, Northeast Ethiopia: unmatched case–control study

Habtu Debash, Marye Nigatie, … Alemu Gedefie

Malaria seroepidemiology in very low transmission settings in the Peruvian Amazon

Bryan Fernandez-Camacho, Brian Peña-Calero, … Gabriel Carrasco-Escobar

Introduction

Malaria remains one of the most deadly infectious diseases in the world, being responsible for an estimated 619,000 deaths in 2021, mostly in young children 1 . Severe malaria caused by Plasmodium falciparum presents as a range of clinical manifestations including, but not limited to, cerebral malaria (CM), respiratory distress syndrome (RDS) and severe malarial anaemia (SMA), all of which can quickly lead to clinical deterioration and death in the absence of prompt treatment 2 , 3 .

SMA is an acute clinical manifestation of severe malaria, defined in children as haemoglobin (Hb) concentration <5 g/dl (or haematocrit <15%) combined with the presence of malaria parasites (with the additional criterion of parasite density threshold of >10,000/µl in the WHO definition although this is not always used in clinical settings) 4 . The pathogenesis of SMA is complex, with the destruction of both non-parasitised and parasitised red blood cells (RBC) and decreased RBC production from the bone marrow contributing to severity 5 . SMA pathogenesis also likely represents a spectrum, being influenced by previous malaria exposure, parasite density, existing anaemia and severe anaemia (both chronic and acute) and its related co-occurrence with other drivers such as helminths, sickle cell and malnutrition 6 . However, the diagnostic definition is broad and differential diagnosis is often difficult or not possible. The progression from uncomplicated malaria to SMA is strongly associated with delayed treatment, highlighting the importance of access to prompt diagnosis and treatment whenever symptoms occur 7 . The burden of SMA is disproportionately concentrated in children under 5 years of age 8 , and SMA is associated with high in-hospital mortality rates in excess of 5% 9 , with over half of the deaths occurring within the first 24 h of hospital admission 10 . Even short delays to effective treatment can have fatal consequences in the context of severe malaria 7 . In addition, surviving children continue to be at high risk of increased morbidity and mortality even after effective in-hospital treatment of an acute episode 11 .

The burden of SMA in malaria-endemic countries remains high despite efforts to scale up case management and preventive interventions that have led to a decline in malaria and severe malaria (including SMA) over the last 20 years. A recent study using hospital data estimated that there were in excess of 1.5 annual hospital admissions per 1000 children aged 3 months to 9 years old in areas of high malaria prevalence in East Africa 12 . In highly endemic areas, SMA can be present in 20–40% of paediatric hospital admissions and contribute to >50% of in-hospital malaria deaths 10 . In many malaria-endemic settings, it is possible that the observed burden only represents a proportion of the total burden of SMA, with further cases occurring in the community without being reported by hospital surveillance systems 13 , 14 .

Despite its high case fatality rate in children, the relative rarity of SMA at any one time in whole population surveys has made it challenging to assess the true burden of the disease. However, one growing source of additional information about the burden of SMA in the community is large-scale, representative household surveys, which collect the same diagnostic metrics used to define SMA in hospital. An analysis of over 180,000 children aged under 5 years from Demographic Health Surveys (DHS) and Malaria Indicator Surveys (MIS) across 19 countries found that nearly 1% of children testing positive for malaria also had severe anaemia (haemoglobin <5 g/dl) 15 , providing clues for burden analysis. Although surveys ask questions about treatment seeking, the limited responses obtained from the small numbers of severe malaria cases make conclusions drawn from these data highly uncertain. Current estimates of the global severe malaria morbidity and mortality have a high degree of uncertainty and rely on case incidence, hospitalisation, and verbal autopsy data alongside expert opinion combined within a framework that includes multiple causes of death 16 . Understanding the true burden of SMA is important to improve the current estimates as well as to inform investments in better access to hospital care and prevention interventions.

In this study, we correlate both household survey and in-hospital data sources using a statistical model of disease incidence and hospital access to estimate the proportion of SMA cases that access hospital care and quantify for the first time the burden of SMA in the community.

Overall, of the 209,542 children in the household survey data, 18% had malaria, 0.30% had severe anaemia, and 0.15% had both. After adjusting for background levels of severe anaemia in children without malaria in each country, an estimated 0.12% of children had SMA. Our model-fitted observed prevalence of severe malarial anaemia in 0.5- to 5-year-olds ( \({{SMA}}_{0.5-5}\) ) in the community, capturing trends with respect to Plasmodium falciparum parasite prevalence in 2- to 10-year-olds ( \({{Pf}\Pr }_{2-10}\) ) as well as between country differences (Fig.  1 ). Although there was substantial uncertainty in the trends within each country due to low numbers, pooling the data across all countries allowed estimation of a more precise relationship between \({{Pf}\Pr }_{2-10}\) and SMA. Observed SMA prevalence was estimated to have an increasing relationship with \({{Pf}\Pr }_{2-10}\) , that plateaued or decreased slightly at higher values of \({{Pf}\Pr }_{2-10}\) (>40%).

The main plot (top left) shows the overall fitted relationship across all countries (with country-level effects set to zero) and subplots show fits to country data. Orange lines show the posterior median fit (summarised across 120,000 posterior samples), teal lines show 100 draws from the joint posterior and black points, and lines are the DHS data and associated 95% confidence intervals. \({{SMA}}_{0.5-5}\) is estimated by adjusting the survey prevalence of malaria and severe anaemia in 0.5- to 5-year-olds by the background prevalence of severe anaemia. Source data are provided as a Source Data file.

After allowing for chance overlap between background severe anaemia and malaria, as well as imperfect hospital access, our model could also capture the observed trends in hospitalised SMA incidence with respect to \({{Pf}\Pr }_{2-10}\) , producing a similar relationship as found by Paton et al. 12 (Fig.  2 ). We estimate that hospitalised SMA incidence rises with increasing \({{Pf}\Pr }_{2-10}\) , before plateauing or falling slightly at higher values of \({{Pf}\Pr }_{2-10}\,\) (>40%). Estimates of annual hospitalised incidence per 1000 children of 0.5, 1, and 1.5 corresponded to median estimates of \({{Pf}\Pr }_{2-10}\) of 16%, 27% and 40%, respectively. There was substantial uncertainty associated with the estimates at higher levels of \({{Pf}\Pr }_{2-10}\) .

The orange line shows the posterior median fit (summarised across 120,000 posterior samples), teal lines show 100 draws from the joint posterior, black points are the observed hospitalisation data 12 , and the black dashed line is the previously fitted model from Paton et al. Source data are provided as a Source Data file.

Information in the surveys to inform estimates of the proportion of children with SMA who sought treatment was sparse. In Kenya, Tanzania and Uganda, 0% (0 of 2), 0% (0 of 2) and 40% (2 of 5) were reported to have sought treatment at a government hospital, respectively. Our estimates of the percentage of SMA cases hospitalised were low and consistent for Kenya (21% (95% CrI: 7%, 47%)), Tanzania (18% (95% CrI: 5%, 52%)) and Uganda (23% (95% CrI: 9%, 48%)) (Fig.  3A , Table  1 ). These estimates imply the presence of a large community burden of SMA approximately 5 times higher than cases observed in hospital (Fig.  3B ).

A The estimated posterior distribution of % hospitalised for a child aged 0.25–9 years with SMA in Kenya, Tanzania, and Uganda, box midline indicates the median, shoulders the 25% and 75% percentiles and whiskers the range from the main analysis. Dark points represent samples from the main analysis (100 per country) and faded points samples across all sensitivity analyses (100 per country). B A comparison of the estimated community incidence of SMA with respect to malaria prevalence in 2- to 10-year-olds ( \({{Pf}\Pr }_{2-10}\) ) compared to the incidence observed through hospital admissions. Dark points and lines represent 1000 samples and the median estimate, respectively, from the main analysis. Faded points show 1000 samples for each sensitivity analysis. Source data are provided as a Source Data file.

Estimates of total community burden were somewhat sensitive to the definition of SMA used, decreasing when fever (caregiver reported child as having fever in the 2 weeks preceding the survey) was included in the definition, which correspondingly increased the estimated probability of hospitalisation. Results were less sensitive to the malaria diagnostic used and the adjustment for non-malarial severe anaemia. Estimates were most sensitive to the assumed duration of SMA in the community, with a longer duration of SMA decreasing community incidence estimates and therefore increasing the estimate of the proportion of cases hospitalised (Table  1 ). The median posterior estimate of the percentage of cases hospitalised did not exceed 40% in any of the analyses.

We saw weak evidence for a decrease in the percentage of cases hospitalised with respect to the distance to hospital (Supplementary Fig.  S1 ); however, the range and resolution of distance to hospital data were not enough to identify a strong trend.

We compared our modelled estimates of SMA incidence to those observed during the RTS,S trial as a means of validation. Estimates of the observed SMA incidence in the trial fell between the median estimates of SMA hospitalised incidence and SMA total incidence predicted by our model, overlapping with our uncertainty ranges (Supplementary Fig.  S2 ).

Our analysis relates severe anaemia and malaria measured in representative household surveys with hospitalised SMA incidence and estimates that the percentage of SMA cases that are hospitalised is consistently low at 21% (95% CrI: 7–47%), 18% (5–52%), and 23% (9–48%) for Kenya, Tanzania, and Uganda, respectively (Fig.  3 , Table  1 ). These results indicate that the community burden of SMA may be around 5 times higher than inferred from the hospital data, although considerable uncertainties remain. Our results were robust over a range of sensitivity analyses that explored some key assumptions underlying the work (Table  1 ).

What does a community SMA case look like? This is a critical question key to understanding the full extent of the community burden, and one that is hard to fully answer given current data. Our estimate of the percentage of cases hospitalised represents a complex set of drivers influencing the likelihood that a child with SMA received hospitalised care. First, it is likely that there is a wide spectrum of syndromes and associated outcomes for children with malaria and severe anaemia. We acknowledge as a limitation that the household surveys we used to estimate community prevalence of SMA did not contain parasite density data, which may have improved the specificity of our case definition. It is likely, for example, that SMA as a result of mild malaria occurring together with chronic anaemia may follow a different course to a more acute anaemia as a result of a malaria infection in a previously healthy child 17 , 18 . Potentially, not all cases need hospitalisation or transfusion 9 , and some will spontaneously recover following treatment from an outpatient clinic or other non-hospital provider 19 . Even when children are seriously ill, SMA appears to be a difficult health condition for caregivers to recognise, which can lead to delays or a complete lack of treatment seeking 20 . This proportion can be significant; for example, in a study in Tanzania, 67% of mothers of symptomatic children with severe anaemia reported that they did not think the illness was severe 13 . Even when symptoms are recognised, there remain significant physical and financial barriers to accessing hospital care. Whilst our adjustment for non-malarial severe anaemia is included to account for other drivers of anaemia, such as helminths, sickle cell and malnutrition, significant heterogeneity in the distribution of these co-occurring variables may influence the estimates derived 21 . Malaria is commonly associated with poverty, and in this context, a host of health inequities create barriers that prevent children with SMA from accessing timely and appropriate treatment 13 . Moreover, the referral systems and networks for healthcare seeking for SMA vary widely across countries, for example, being influenced by community health workers or the availability of blood for transfusion. As a result, the generalisability of our findings outside of the settings in the study is uncertain. SMA often occurs alongside other severe malaria syndromes, such as respiratory distress or cerebral malaria, which can increase the probability of hospitalisation and case fatality rate 7 , 22 . These other types of severe symptoms may be more easily recognised as needing urgent care, although conversely, these cases may also be more likely to die before reaching a hospital. The case fatality of SMA in hospital is in the range of 2–20% 23 , 24 ; this value could be higher in the community due to lack of hospital care or lower if those who do not seek care have less severe disease. As a result, it remains unclear how the burden of community SMA translates to child health outcomes.

Our findings indicate that a large proportion of SMA is occurring in the community. In line with these findings, and faced with the same data limitations, other studies have also pointed towards significant gaps in hospital access and a large community burden for severe malaria. Camponovo et al . attempted to triangulate World Malaria report estimates of cases, admissions, and deaths at the national level with model predictions to produce estimates of the proportion of cases admitted in hospitals. They estimated that 9% (range: 7–31%), 48% (37–100%), and 63% (51–100%) of patients are admitted in Kenya, Tanzania, and Uganda, respectively, although they stress the need for further studies given that part of their analysis relied on expert opinion 14 . Whilst these ranges overlap with our estimates, the high degree of uncertainty in both makes direct comparisons challenging. The CARAMAL study, which tracked children with suspected severe malaria in the community, found that 58%, 67%, and 48% attended hospital following referral in Uganda, DRC, and Nigeria, respectively, which are higher than our estimates for SMA. However, these were cases that had sought initial treatment and that included all types of severe malaria manifestations in addition to severe anaemia, e.g., cerebral malaria or respiratory distress, which may be more easily recognised as severe, and also progress more rapidly 25 .

More widely, other studies have attempted to harness information from verbal autopsies to address questions about hospital access and the location of death in low-resource settings. These studies provide clues about the community burden of disease. Using verbal autopsy, Fraser et al . estimated that around one-third of 1324 deaths from time-critical conditions screened in South Africa did not seek care before dying 26 . A study by Gill et al . in Lusaka, Zambia, showed that the majority of respiratory syncytial virus (RSV) deaths in children occurred in the community 27 , similarly pointing towards a significant community burden of disease. A study of deaths occurring during pregnancy or childbirth in Burkina Faso and Indonesia showed that 41% (72/174) of deaths occurred at home, with 22% and 8% in each country, respectively, being associated with malaria 28 . Data from the million deaths study in India show that in excess of 80% of deaths in children under 15 years old occur outside of the healthcare facility, with 75% of all deaths occurring at home 29 . Whilst the context differs between these studies, large numbers of deaths occurring at home highlight a significant burden of severe disease in the community, as the results of the current study have also indicated. Overall, our results are consistent with the malaria mortality estimated from verbal autopsies in the community. Efforts to infer total malaria mortality rates 30 in the community and hospital combined have led to estimates that are similar in magnitude to the incidence of hospitalised severe malaria 12 . Since severe malaria case fatality in hospital is in the range of 2–60%, these numbers suggest a significant “hidden” community burden.

Our results are sensitive to the assumed duration of SMA in the community. The relationship between duration and hospitalisation probability is positively correlated (Supplementary Fig.  S3 ). Re-running the analysis with a less informative prior on the duration of SMA resulted in the estimates of the probability of hospitalisation and duration both increasing. However, it is not clear how plausible the posterior estimate of duration (Supplementary Table  S2 , 74 days (95% CrI: 21, 230 days) versus 41 days (95% CrI: 15, 100) in the main analysis) is in this instance. Although information on the duration of SMA in the community is scant, we do know that SMA cases admitted to hospital usually report to have had symptoms for less than 7 days (e.g., refs. 7 , 31 ), and other data from animal models and studies in adult populations indicate similar timescales (Supplementary Information: “Infection timescale”). This may indicate that our original assumption of a short duration is more plausible or could point towards a spectrum of severity associated with those with SMA in the community where duration might be longer than those with the most acute, severe presentation who may immediately seek hospital care. It is likely that the simple duration estimated is an aggregate across more than one duration distribution representing different clinical manifestations of SMA, however, there are currently neither data available nor resolution in the clinical definition widely used to disentangle any such mixtures if they are present.

Including fever in our definition of survey prevalence reduces our estimate of prevalence and consequently increases our estimates of the percentage of cases hospitalised (Fig.  4 ). However, this definition may be too restrictive as we know that a proportion of those presenting at hospital with SMA are not febrile 7 . Results were not as sensitive to the malaria diagnostic used or the lack of adjustment for non-malarial severe anaemia (Fig.  4 , Supplementary Information: “Sensitivity”).

Blue boxes at the top highlight the key adjustments and transformations as each step to move from malaria and severe anaemia prevalence in the survey data to severe malarial anaemia incidence in the hospitalisation data.

We fitted a hierarchical model to data on the prevalence of malaria and severe anaemia, which combined multiple survey-years for a number of countries. This approach was taken due to the low number of malaria and severe anaemia cases observed in the survey data (306 total across surveys from 21 countries). The upper estimate of the malaria and severe anaemia prevalence with respect to \({{Pf}\Pr }_{2-10}\) therefore borrows information from data from all 21 countries included, whilst the growth rate and shift parameters were assumed to be the same for all countries. The extent to which these assumptions are true across settings will influence estimates as the trend in a country is not only informed by data from that country but also, to a lesser extent, data from all countries included. We can see this influence in countries included in the hospitalisation fitting where we observe median model estimates at higher \({{Pf}\Pr }_{2-10}\) exceeds (e.g., Fig.  1 . Kenya) or is lower (e.g., Fig.  1 . Tanzania) than survey point estimates, albeit with a large degree of uncertainty due to the small amount of data from single countries. The extra complexity in our model compared to that of Paton et al. 12 results in more flexibility in the functional form relating hospitalised incidence to \({{Pf}\Pr }_{2-10}\) . Whilst the resulting fitted functional forms are very similar, the Paton et al. function monotonically increases whilst the fit presented here shows a slight fall above \({{Pf}\Pr }_{2-10}\) of 40% (Fig.  2 ) as a result of increasing non-malarial SMA incidental with a malaria infection. This does not strongly impact our estimates of the probability of a child with SMA accessing a hospital but is an area requiring future research to understand more fully. Our analysis may be confounded by the distribution of other factors associated with the prevalence of malaria, anaemia and severe disease that we cannot individually control. These include other infectious diseases (e.g., HIV), parasitic infections (e.g., hookworm), nutritional issues (e.g., Vitamin A or B12 deficiencies), enzymopathies (e.g., G6PD), haemoglobinopathies (e.g., sickle cell anaemia) 5 , age (further complicated by difference age-ranges in the population survey and hospitalisation data sets that we adjust for), socio-economic variables and malaria treatment. Furthermore, the hospitalisation data is not representative of the whole country, sampling populations from within 30 km of hospitals and excluding urban areas 12 . Differential case fatality rates in the community and hospital may also introduce bias via death censoring. Given a child has SMA, we assume a constant probability of hospitalisation. More complex models could be proposed, for example, assuming that the probability that SMA leads to severe disease needing hospitalisation is a function of \({{Pf}\Pr }_{2-10}\) . However, we saw a good fit to hospitalisation data using the simple model.

Our findings lead to two key conclusions. First, more work is needed to understand the spectrum of disease associated with SMA cases in the community and how Hb responds throughout the time course of an infection and recovery. This is vital to be able to enumerate the child health impact of community SMA. Second, the results point towards a substantial and largely unobserved burden of SMA occurring in the community and the importance of removing or reducing barriers that erode the probability that severely ill children receive the hospital care that they need.

Inclusion and ethics

Ethical approval for this secondary analysis of data study was granted by the Imperial College Research Ethics Committee (ICREC), ref: 22IC7782.

Our authorship team includes researchers from both malaria-endemic and non-malaria-endemic settings. Together we designed the study to be relevant to the study context and to maximise the contribution and impact of the combined skills, knowledge and experience of all authors.

We extracted data on malaria status, as determined by microscopy or rapid diagnostic test (RDT), haemoglobin level, reported fever status in the last 2 weeks, and reported hospital (public) attendance from DHS and MIS survey data among children aged 6 to 59 months from 21 countries between 2011 and 2020 32 . Children with haemoglobin levels of <5 g/dl were classified as severely anaemic. The main analysis used microscopy to determine malaria infection as it is indicative of current infection and has a higher parasite density threshold of detection than rapid diagnostic tests (RDTs) 33 and may therefore be a better indicator of severe anaemia caused by malaria. We aimed to relate the prevalence of SMA to the local transmission intensity and therefore estimated the cluster-level prevalence of malaria infection as the average Malaria Atlas Project predicted prevalence among 2- to 10-year-olds ( Pf PR 2-10 ) 34 within a 5 km radius of the cluster geolocations, to account for the random offset of geolocations in the data. Hospitalisation data were extracted from Paton et al. 12 using the measurement-adjusted hospital SMA (defined as Hb <5 g/dl on admission in combination with a positive malaria diagnosis) admission rates (see Paton et al ., Supplement Fig.  S3A ). We also extracted Pf PR 2-10 and distance to the nearest hospital (centre of range, km) estimates for each site.

We designed a statistical model to correlate the DHS and MIS survey estimates of the prevalence of malaria and severe anaemia in the community with the hospitalisation records capturing the incidence of SMA hospital admissions. Paton et al. found a sigmoidal relationship between community parasite prevalence and incidence of SMA in hospital 12 . We similarly allowed the prevalence of children with both malaria infection and severe anaemia in the community (MASA) to be determined by community parasite prevalence. Additionally, we allowed for background rates of severe anaemia with non-malarial causes, which may co-exist with malaria infection. First, we fitted a hierarchical model to the individual (Hb and malaria status by microscopy) and DHS cluster-level (malaria transmission intensity) data to capture the change in the prevalence of MASA with respect to malaria transmission. We included a country-level random effect to capture systematic differences in this relationship between countries. Following this, we adjusted for a constant country-level background prevalence of non-malarial severe anaemia (Hb <5 g/dl with an accompanying negative malaria test) (Fig.  4 , step 1). This accounts for the background rate of severe anaemia, including both acute and chronic causes and adjusts for co-endemic underlying drivers of non-malarial anaemia such as helminths, sickle cell and malnutrition. Next, we age-standardised to match the age range recorded in Paton et al . (3 months–9 years) using a previously defined relationship between age and SMA 12 (Fig.  4 , Step 2). We then converted this adjusted prevalence to an incidence rate with a fitted estimate of the duration of an SMA case in the community (Fig.  4 , step 3). This incidence was then further adjusted by a distance-dependent hospitalisation probability to fit the hospitalisation data (Fig.  4 , step 4). The hospitalisation probability was informed by the proportion of MASA cases reporting to have attended a public hospital in the survey data. We jointly fitted the models of MASA prevalence and hospitalised SMA incidence in a Bayesian framework using the DrJacoby R package 35 . Full mathematical details and descriptions of the parameters fitted are given in the Supplementary Information.

Sensitivity analyses and validation

We refitted the model four times to determine the influence of four key assumptions on our estimate of the percentage of SMA cases that are hospitalised; (1) severity associated with the definition of SMA in the community, (2) the diagnostic used in the definition of malaria (microscopy or RDT), (3) the adjustment for non-malarial anaemia and (4) our prior specification on the duration of SMA (Supplementary Information: “Sensitivity”). Although few estimates of population incidence of SMA have been published, we were able to compare our model predictions against findings from the RTS,S vaccine trial (Supplementary Information: “Model validation”).

Reporting summary

Further information on research design is available in the  Nature Portfolio Reporting Summary linked to this article.

Data availability

Data used in this analysis are either publicly available in the referenced publication, online ( https://malariaatlas.org/ ) or available online upon registration ( https://dhsprogram.com/data/available-datasets.cfm ). Source data for figures and tables are provided with this paper.  Source data are provided with this paper.

Code availability

Code for analysis is hosted at https://github.com/mrc-ide/commal ( https://doi.org/10.5281/zenodo.8238352 ).

World Health Organization. World Malaria Report 2022 (2022).

Trampuz, A., Jereb, M., Muzlovic, I. & Prabhu, R. M. Clinical review: severe malaria. Crit. Care 7 , 315–323 (2003).

Article   PubMed   PubMed Central   Google Scholar  

MacKintosh, C. L., Beeson, J. G. & Marsh, K. Clinical features and pathogenesis of severe malaria. Trends Parasitol. 20 , 597–603 (2004).

Article   CAS   PubMed   Google Scholar  

Severe malaria. Trop. Med. Int. Health 19 , 7–131 (2014).

White, N. J. Anaemia and malaria. Malar. J. 17 , 1–17 (2018).

Article   Google Scholar  

Soares Magalhães, R. J. & Clements, A. C. A. Mapping the risk of anaemia in preschool-age children: the contribution of malnutrition, malaria, and helminth infections in West Africa. PLoS Med. 8 , e1000438 (2011).

Article   PubMed Central   Google Scholar  

Mousa, A. et al. The impact of delayed treatment of uncomplicated P. falciparum malaria on progression to severe malaria: a systematic review and a pooled multicentre individual-patient meta-analysis. PLoS Med . 17 , e1003359 (2020).

Roca-Feltrer, A. et al. The age patterns of severe malaria syndromes in sub-Saharan Africa across a range of transmission intensities and seasonality settings. Malar. J. 9 , 1–9 (2010).

Akech, S. O. et al. Survival and haematological recovery of children with severe malaria transfused in accordance to WHO guidelines in Kilifi, Kenya. Malar. J. 7 , 1–9 (2008).

Obonyo, C. O., Vulule, J., Akhwale, W. S. & Grobbee, D. E. In-hospital morbidity and mortality due to severe malarial anemia in western Kenya. Am. J. Trop. Med. Hyg. 77 , 23–28 (2007).

Article   PubMed   Google Scholar  

Kwambai, T. K. et al. Articles Post-discharge morbidity and mortality in children admitted with severe anaemia and other health conditions in malaria-endemic settings in Africa: a systematic review and meta-analysis. Lancet Child Adolesc. Health 4642 , 1–10 (2022).

Google Scholar  

Paton, R. S. et al. Malaria infection and severe disease risks in Africa. Science 373 , 926–931 (2021).

Article   CAS   PubMed   PubMed Central   ADS   Google Scholar  

Schellenberg, D. et al. The silent burden of anaemia in Tanzanian children: a community-based study. Bull. World Health Organ. 81 , 581–590 (2003).

CAS   PubMed   PubMed Central   Google Scholar  

Camponovo, F., Bever, C. A., Galactionova, K., Smith, T. & Penny, M. A. Incidence and admission rates for severe malaria and their impact on mortality in Africa. Malar. J. 16 , 1–12 (2017).

Taylor, C., Namaste, S. M. L., Lowell, J., Useem, J. & Yé, Y. Estimating the fraction of severe malaria among malaria-positive children: analysis of household surveys in 19 malaria-endemic countries in Africa. Am. J. Trop. Med. Hyg. 104 , 1375–1382 (2021).

World Health Organization. World Malaria Report 2021 . (2021).

White, N. J. Severe malaria. Malar. J. 21 , 284 (2022).

Watson, J. A. et al. Improving the diagnosis of severe malaria in African children using platelet counts and plasma Pf HRP2 concentrations. Sci. Transl. Med. 14 , eabn5040 (2022).

Article   CAS   PubMed   PubMed Central   Google Scholar  

Newton, C. R. J. C. et al. Severe anaemia in children living in a malaria endemic area of Kenya. Trop. Med. Int. Health 2 , 165–178 (1997).

Dhabangi, A. et al. Community perceptions of paediatric severe anaemia in Uganda. PLoS ONE 14 , 1–13 (2019).

GBD 2021 Anaemia Collaborators. Prevalence, years lived with disability, and trends in anaemia burden by severity and cause, 1990–2021: findings from the Global Burden of Disease Study 2021. Lancet Haematol . 10 , e713–e734 (2023).

Cserti-Gazdewich, C. M. et al. Inter-relationships of cardinal features and outcomes of symptomatic pediatric Plasmodium falciparum malaria in 1,933 children in Kampala, Uganda. Am. J. Trop. Med. Hyg. 88 , 747–756 (2013).

Dondorp, A. M. et al. The relationship between age and the manifestations of and mortality associated with severe malaria. Clin. Infect. Dis. 47 , 151–157 (2008).

Akech, S. et al. The clinical profile of severe pediatric malaria in an area targeted for routine RTS,S/AS01 malaria vaccination in Western Kenya. Clin. Infect. Dis. 71 , 372–380 (2020).

Brunner, N. C. et al. Pre-referral rectal artesunate and referral completion among children with suspected severe malaria in the Democratic Republic of the Congo, Nigeria and Uganda. BMJ Glob. Health 7 , e008346 (2022).

Fraser, A. et al. Time-critical conditions: assessment of burden and access to care using verbal autopsy in Agincourt, South Africa. BMJ Glob . Health 5 , 1–10 (2020).

Gill, C. J. et al. Infant deaths from respiratory syncytial virus in Lusaka, Zambia from the ZPRIME study: a 3-year, systematic, post-mortem surveillance project. Lancet Glob . Health 10 , e269–e277 (2022).

CAS   Google Scholar  

D’Ambruoso, L., Byass, P., Qomariyah, S. N. & Ouédraogo, M. A lost cause? Extending verbal autopsy to investigate biomedical and socio-cultural causes of maternal death in Burkina Faso and Indonesia. Soc. Sci. Med. 71 , 1728–1738 (2010).

Jha, P. et al. Prospective study of one million deaths in India: rationale, design, and validation results. PLoS Med. 3 , 0191–0200 (2006).

Rowe, A. K. et al. The burden of malaria mortality among African children in the year 2000. Int. J. Epidemiol. 35 , 691–704 (2006).

Kurtzhals, J. A. L. et al. Low plasma concentrations of interleukin 10 in severe malarial anaemia compared with cerebral and uncomplicated malaria. Lancet 351 , 1768–1772 (1998).

USAID. The DHS Program . http://www.dhsprogram.com/ .

Wu, L. et al. Comparison of diagnostics for the detection of asymptomatic Plasmodium falciparum infections to inform control and elimination strategies. Nature 528 , S86–S93 (2015).

The Malaria Atlas Project. https://malariaatlas.org/ (2022).

Verity, R. & Winskill, P. Dr. Jacoby . https://mrc-ide.github.io/drjacoby/ (2023).

Download references

Acknowledgements

P.W. acknowledges the Imperial College Research Fellowship and the Bill & Melinda Gates Foundation (INV-043624). L.C.O. acknowledges the UK Royal Society Fellowship. The funders of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. All authors had full access to all the data in the study and had final responsibility for the decision to submit for publication. The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the U.S. Centers for Disease Control and Prevention.

Author information

Authors and affiliations.

MRC Centre for Global Infectious Disease Analysis, Department of Infectious Disease Epidemiology, Imperial College, London, W2 1PG, UK

Peter Winskill & Lucy C. Okell

Child Health and Development Centre, Makerere University College of Health Sciences, Kampala, Uganda

Aggrey Dhabangi

Division of Parasitic Diseases and Malaria, Global Health Center, Centers for Disease Control and Prevention, Kisumu, Kenya

Titus K. Kwambai

Health Economics Leadership and Translational Ethics Research Group (HELTER), Department of Global Public Health and Primary Care, University of Bergen, Arstadveien 17, 5009, Bergen, Norway

Amani Thomas Mori

Muhimbili University of Health and Allied Sciences, P.O. Box, 65001, Dar es-Salaam, Tanzania

Department of Infection Biology, London School of Hygiene and Tropical Medicine, London, UK

Andria Mousa

You can also search for this author in PubMed   Google Scholar

Contributions

P.W.: Conceptualisation, Methodology, Software, Formal analysis, Writing—Original Draft, Visualisation, Writing—Review & Editing, Data curation, Validation. L.C.O.: Conceptualisation, Methodology, Writing—Review & Editing, Data curation, Validation. A.D., T.K.K., A.T.M., A.M.: Methodology, Writing—Review & Editing.

Corresponding author

Correspondence to Peter Winskill .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Peer review

Peer review information.

Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Peer review file, reporting summary, source data, source data, rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Winskill, P., Dhabangi, A., Kwambai, T.K. et al. Estimating the burden of severe malarial anaemia and access to hospital care in East Africa. Nat Commun 14 , 5691 (2023). https://doi.org/10.1038/s41467-023-41275-w

Download citation

Received : 03 February 2023

Accepted : 29 August 2023

Published : 14 September 2023

DOI : https://doi.org/10.1038/s41467-023-41275-w

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing: Microbiology newsletter — what matters in microbiology research, free to your inbox weekly.

Loading metrics

Open Access

Peer-reviewed

Research Article

Facilitators and barriers to integrated malaria prevention in Wakiso district, Uganda: A photovoice study

Roles Conceptualization, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Validation, Writing – review & editing

* E-mail: [email protected]

Affiliation Department of Disease Control and Environmental Health, School of Public Health, College of Health Sciences, Makerere University, Kampala, Uganda

Roles Data curation, Formal analysis, Methodology, Project administration, Supervision, Writing – original draft, Writing – review & editing

Roles Formal analysis, Methodology, Project administration, Supervision, Writing – review & editing

Roles Methodology, Project administration, Supervision, Writing – review & editing

Roles Formal analysis, Methodology, Supervision, Validation, Writing – review & editing

Affiliation School of Interdisciplinary Health Programs, Western Michigan University, Kalamazoo, Michigan, United States of America

Roles Funding acquisition, Investigation, Supervision, Validation, Writing – review & editing

Affiliation Department of Medicine, School of Medicine, College of Health Sciences, Makerere University, Kampala, Uganda

• David Musoke, 
• Grace B. Lubega, 
• Filimin Niyongabo, 
• Suzan Nakalawa, 
• Shannon McMorrow, 
• Rhoda K. Wanyenze, 
• Moses R. Kamya

• Published: April 16, 2024
• https://doi.org/10.1371/journal.pgph.0002469
• Reader Comments

Malaria continues to cause significant morbidity and mortality globally, particularly in sub-Saharan Africa. Appropriate combinations of non-chemical and chemical methods of malaria vector control in the context of integrated vector management have been recommended by the World Health Organization. The aim of the study was to explore facilitators and barriers to using integrated malaria prevention in Wakiso district, Uganda. This qualitative study employed photovoice among 20 community members in Kasanje Town Council, Wakiso District. The photos taken by participants for 5 months using smartphones were discussed during monthly meetings with the researchers. The discussions were audio-recorded, and resulting data analysed using thematic analysis with the support of NVivo (2020) QSR International. Findings indicated that various conventional and non-conventional measures were being used for preventing malaria such as: insecticide treated nets; clearing overgrown vegetation; draining stagnant water; mosquito coils; smouldering of cow dung; spraying insecticides; plant repellents near houses; eating of prophylactic herbs; as well as closing doors and windows on houses early in the evening. Facilitators supporting the use of several malaria prevention methods holistically included: low cost and accessibility of some methods such as slashing overgrown vegetation; and support provided for certain methods such as receiving free mosquito nets from the government. Barriers to using several malaria prevention methods holistically included: inadequate knowledge of some methods such as housing improvement; allergic reactions to chemical-based methods such as insecticide treated nets; unaffordability of some methods such as insecticide sprays; and inaccessibility of certain methods such as body repellents. These barriers to integrated malaria prevention need to be addressed to achieve greater impact from the combination of methods in endemic communities.

Citation: Musoke D, Lubega GB, Niyongabo F, Nakalawa S, McMorrow S, Wanyenze RK, et al. (2024) Facilitators and barriers to integrated malaria prevention in Wakiso district, Uganda: A photovoice study. PLOS Glob Public Health 4(4): e0002469. https://doi.org/10.1371/journal.pgph.0002469

Editor: Khadime Sylla, Cheikh Anta Diop University: Universite Cheikh Anta Diop, SENEGAL

Received: October 4, 2023; Accepted: March 9, 2024; Published: April 16, 2024

Copyright: © 2024 Musoke et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All data are in the paper and/or Supporting Information files.

Funding: The study received funding from the EDCTP2 programme, supported by the European Union (Grant Number TMA2020CDF-3189) and the Fondation Botnar to DM. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: I have read the journal’s policy and the authors of this manuscript have the following competing interests: David Musoke is an Editorial Board member of PLOS Global Public Health.

Introduction

Malaria continues to be a major public health challenge in many parts of the world particularly in sub-Saharan Africa. The disease disproportionately affects children under the age of 5 years and pregnant women, who are at higher risk of severe illness and death [ 1 – 4 ]. Approximately, 90% of cases and 92% of deaths due to malaria occur in sub-Saharan Africa [ 5 ]. The disease has significant economic impacts, including reduced productivity and increased healthcare costs, with an estimated total economic burden of US$ 12 billion annually in sub-Saharan Africa [ 6 ]. In Uganda, malaria remains one of the leading causes of morbidity and mortality, with an estimated 8 million cases and over 15,000 deaths reported annually [ 7 ]. A national survey estimated that malaria accounts for over 40% of outpatient visits and approximately 15–20% of hospital admissions in the country [ 8 ]. In addition, the disease is known to lead to the loss of several school and workdays and time caring for the sick, as well as vast economic burden to households [ 9 ]. Despite the implementation of various prevention strategies with a key focus on the use of long-lasting insecticidal nets (LLIN) and indoor residual spraying (IRS), the disease remains a persistent problem in the country [ 10 ]. Therefore, innovative approaches that could contribute to the reduction in the occurrence of malaria are needed in Uganda and other endemic countries.

To combat malaria, an integrated preventive approach has been recommended by leading global public health entities including the World Health Organization (WHO), which has been a key advocate for this strategy [ 11 ]. Integrated malaria prevention refers to the use of multiple, complementary strategies at households and in the community to prevent the disease. This approach recognizes that no single intervention is enough to effectively prevent malaria and aims to maximize the synergistic effect of combined methods. By combining multiple strategies, integrated malaria prevention aims to reduce transmission and ultimately reduce the burden of the disease [ 12 – 15 ]. The approach is effective in reducing malaria incidence and mortality, particularly when implemented in a targeted and sustained manner [ 16 – 18 ]. In recent years, the importance of integrated malaria prevention has been reinforced by evidence demonstrating the effectiveness of this approach in reducing the occurrence of the disease. As a result, integrated malaria prevention has become a cornerstone of global efforts to control and eliminate the disease. Indeed, many countries, including Uganda, are implementing integrated malaria prevention programmes such as combining chemical based methods such as LLINs with reducing mosquito breeding sites [ 19 , 20 ]. However, evidence on the use of multiple malaria prevention methods holistically and their contribution to reducing the burden of the disease is limited.

There are significant barriers to the implementation of integrated malaria prevention programmes such as limited resources and inadequate community engagement. In addition, the burden of implementing several methods and the health concerns related to chemical-based methods such as IRS and body repellents have been found to affect the integrated approach to malaria prevention [ 21 ]. Furthermore, a systematic review of global malaria prevention strategies highlighted the need to understand contextual factors such as the role of cultural and social norms, access to health services, and health literacy when implementing integrated malaria approaches in rural communities [ 22 , 23 ]. Consideration of these factors can increase community ownership, promote the uptake and use of multiple malaria prevention measures, and help to address barriers that may exist. This, in turn, can help to enhance the facilitators for integrated malaria preventive measures in communities [ 23 , 24 ]. For example, increased awareness on the various potential methods that could be used by individuals and households could enhance utilisation of the integrated approach to prevent malaria. This therefore necessitates more literature on the use of integrated malaria prevention to contribute to efforts of controlling the disease.

Most of the existing evidence on integrated malaria prevention used traditional research methodologies such as questionnaire surveys and interviews [ 25 , 26 ]. In these methodologies, there is little power with the participants as the researchers lead the entire processes [ 27 ]. A community-based participatory research method such as photovoice in which participants have more power in the conduct of studies was therefore needed. Photovoice puts cameras in the hands of participants which gives them the opportunity to take an active role and have influence on the study through the photographs they take [ 28 , 29 ]. In addition, photovoice has the potential to unveil aspects of a community that may be overlooked in traditional research methods. By allowing participants to visually express what matters most to them, researchers gain access to rich and contextually relevant data including for diseases such as malaria that might not be fully captured through other methods such as surveys or interviews [ 30 ]. Use of photovoice in malaria research would also add to evidence on use of participatory approaches in communities to reduce the burden of the disease. We therefore used photovoice to explore facilitators and barriers to using integrated malaria prevention in Wakiso district, Uganda.

Methodology

Study design.

The study employed photovoice to explore issues related to implementation of the various methods in integrated malaria prevention in a rural, malaria endemic community in Wakiso district, Uganda. This community-based participatory research approach was carried out by 20 community members over a period of six months between 23 rd May and 01 st December 2022. The 20 participants were sufficient for the study based on existing photovoice literature [ 31 ]. The photos taken were discussed by the participants together with the researchers to explore facilitators and barriers regarding use of the integrated approach to prevent malaria.

Study area and participants

The study was carried out in Kasanje town council, Wakiso district, a largely rural area in the central region of Uganda. The town council has one government health facility (Kasanje Health Centre III) and several private facilities including clinics, with the communities having limited access to malaria prevention and control services. The population is engaged in various economic and social activities such as brick making, crop farming, animal husbandry, petty trading, stone quarrying, and sand mining. Brickmaking and sand mining generate large pools of stagnating water and facilitate mosquito breeding in the area, hence the occurrence of malaria. Kasanje town council has a population of 29,008, with 14,597 males and 14,411 females [ 32 ]. Rural villages in the town council were purposively selected for inclusion in the study due to their higher prevalence of malaria than urban settings. These villages (cells) were from 6 wards of Jjungo, Kasanje, Makko, Bulumbu, Zziba, and Ssazi.

The 20 participants who were involved in the study were aged 18 years and above. They were selected purposively with support of local leaders in the study area based on the villages they served. The criteria for selecting the participants was provided to the local leaders by the researchers. This criteria included identifying individuals situated in various localities in the area such as those in hilly and valley settings. Other criteria used in the selection to ensure diversity included: wealth of households, gender, and occupation. Once the local leaders had identified potential study participants, the researchers reviewed and confirmed that they were suitable for the study. A diverse group of participants was selected to ensure a wide variety of perspectives and experiences on using integrated malaria prevention by different community members as was the case in our earlier photovoice studies [ 33 – 35 ]. Only 1 participant per village was selected for their involvement in the study.

Training workshop and photography assignment

After recruiting the participants, a training workshop was conducted to provide the required knowledge and skills for the research such as the use and care of smartphones, as well as ethics in photography. The training also discussed various malaria prevention methods in integrated malaria prevention, and how to approach people and get consent before taking their pictures. Participants were asked to use the smart phones provided to them by the researchers to capture aspects and situations related to integrated malaria prevention in the community, including within their households. These photographs were used to facilitate the research by identifying facilitators of and barriers to using the different malaria prevention methods in their own local setting both individually and holistically. The participants were given 5 months for taking the photographs which was found to be an adequate duration for such photovoice studies [ 33 , 34 ]. Indeed, the duration of the study was sufficient to reach saturation as no photos on new themes emerged in the final 2 months of photography. A follow-up visit 2 weeks after commencement of photography was carried out by the researchers to ensure the assignment was being undertaken as planned. During this visit, the challenges faced by the participants with using the smart phones, among others, were addressed. Monthly meetings were held between participants and researchers to discuss the photos taken during the previous month. During the entire 5 months of photography, regular supervision of the participants by the researchers was also carried out through site visits to ensure fidelity.

Discussing photos and data analysis

After the photographs were taken, each participant was asked to talk about all the photographs they felt were relevant to the study aim of malaria prevention in the community. Emphasis was given to what facilitated use of any of the methods, and challenges faced while attempting to use other practices at their households and in the community. Participatory analysis involved participants themselves discussing and analysing the photos taken during the monthly meetings. The discussions were conducted in Luganda , the local language most used in the study area. The discussion of photos over the 5-month period provided information related to facilitators and barriers to use of the various malaria prevention methods. The discussions, which were facilitated by the researchers, were audio recorded and later transcribed verbatim in the local language by one of the researchers (FN) who had experience in qualitative methods. Once the transcripts were verified, they were translated into English for analysis.

Analysis was done by the researchers (GBL, FN and DM) with vast experience in qualitative research. Thematic analysis was done with the support of NVivo (2020) QSR International. The analysis initially involved the researchers reading the transcripts several times to familiarize themselves with the data. Thereafter, words and related phrases from the transcripts were grouped to form codes, with the initial coding done by GBL and FN. Any disagreements in the generated codes were resolved by another researcher (DM). This involved a meeting where the 3 investigators met and discussed contentious codes in detail, moderated by the third researcher. Related codes were then grouped to form sub-themes, and related sub-themes grouped together to form themes. During consolidation of themes, several sub-themes were fit within the respective main themes that emerged from the data. The themes generated from the analysis were used in writing the manuscript. During manuscript writing, selected photos and quotes were used as part of the findings.

Dissemination

On completion of the 5 rounds of photography, monthly discussions of photos and participatory analysis between participants and researchers, a dissemination workshop was held for various local stakeholders. These stakeholders included community members, local leaders, community health workers (CHWs), health practitioners, and researchers. During the workshop, a selection of the most relevant photos was showcased by the participants to share findings, experiences, and recommendations for integrated malaria prevention to the wider community.

Ethical considerations

Ethical approval to conduct the study was obtained from Makerere University School of Public Health Research and Ethics Committee (SPH-2022-250). The research also received approval from the Uganda National Council for Science and Technology (HS2270ES). Participation in the research was voluntary, and participants provided written informed consent after an explanation of the proposed research including the anticipated risks and potential benefits before taking part. Participants were informed about ethics and safety in research including concerns in the use of photography such as getting people’s consent before taking their photos. It was also stressed during the training of participants that acceptance of a community member to take their photo would be voluntary, and refusal would not affect them in any way. A notebook was provided to each participant to record any scenario that would not be captured as a photo. For example, when someone refused to consent for a photo to be taken or when a participant did not have their smart phone at the time. Use of such a notebook ensured participants did not feel obliged to take photos of every scenario as was the case in our previous photovoice studies [ 36 , 37 ]. No photograph identifying individuals was used for any form of dissemination including in the publication without the consent of both the photographer and identified people. All data emanating from the study was confidentially stored by the researchers, with restricted access given to other members of the research team whenever the need arose.

The study involved 12 female participants, and the majority were farmers (14) and married (13) ( Table 1 ). Over the 5 months, 912 photos were taken by the participants. From the ensuing discussion and data analysis emerged 3 major themes: malaria prevention methods used; facilitators of integrated malaria prevention (low cost and accessibility of methods; and support provided for certain methods); and barriers to integrated malaria prevention (inadequate knowledge and inaccessibility of some methods; allergic reactions to chemical-based methods; and unaffordability of some methods).

• PPT PowerPoint slide
• PNG larger image
• TIFF original image

https://doi.org/10.1371/journal.pgph.0002469.t001

Malaria prevention methods used

Findings revealed that various conventional and non-conventional measures were used for preventing malaria in the study area. Among the conventional methods, sleeping under LLINs was the most prominent. Other well-known methods being used included: slashing of overgrown vegetation surrounding houses; draining of stagnant water; screening of windows; closure of windows and doors before 6:00 pm; use of mosquito coils; spraying of insecticides; and IRS. Participants also mentioned that prophylaxis with antimalarials was being used among pregnant women to prevent malaria. According to the participants, this medication was predominantly given to pregnant women as part of their antenatal care visits to health facilities. Mosquito coils were particularly favoured by many households that were not using other methods.

“ A mosquito coil is one of those materials that release a scent that repels mosquitoes . It is lit and the smoke repels mosquitoes . It is always appropriate to light it before people plan to go to sleep . It helps so that by the time people go to sleep , the house is free from mosquitoes . ” Male, Participant 14, Meeting 2

Regarding the non-conventional methods, participants mentioned the use of plant repellents around houses and eating of herbs they said were prophylactic. Examples of common plant repellents included neem tree, rosemary, and lemon eucalyptus. Participants stated that they grew these plants near their homes which repelled mosquitoes through their scent, while others mentioned that they burnt plant repellents inside their houses before going to bed. According to the participants, the smoke from the burning repelled mosquitoes. In addition, smouldering of cow dung inside and outside houses was used to prevent mosquitoes from biting household members. Another non-conventional method was the eating of herbs that were believed to have malaria-inhibiting properties. Examples of such herbs included: Vernonia amygydalina , Aristolochia littoralis , Gynandropsis gynandra , and Cleme gynandra . The herbs were cooked and eaten as vegetables during meals or added to sauces.

“ Whenever my parent cooks vegetables , she adds leaves of a local medicinal herb . When she also cooks groundnut paste , she adds in another herb . They [herbs] are very bitter and in most cases , we never wanted to eat the food but were forced to . She used to explain to us that she was protecting us from malaria and indeed we did not fall sick from the disease .” Male, Participant 16, Meeting 4

Facilitators of integrated malaria prevention

Facilitators to using several malaria prevention methods holistically were captured through photos and accompanying narratives. These included low cost and accessibility of methods, and support provided for certain methods.

Low cost and accessibility of methods

According to the participants, closure of windows and doors before 6:00pm, and slashing overgrown vegetation could easily be used with other methods. This was because both methods could be done daily without any costs involved. Participants also agreed that the repairing of LLINs that had holes could be done easily once community members had a needle and thread. Having a sewing kit was therefore considered a necessity for many homes. Other participants mentioned that if their mosquito nets got torn, they would tie the part with the hole at no cost ( Fig 1 ).

https://doi.org/10.1371/journal.pgph.0002469.g001

The use of plant repellents was also found to be cheap because the repellent seeds were usually bought at once and planted around compounds ( Fig 2 ). Some participants mentioned that at times these plants grew on their own, while others were found in neighbouring shrubs hence participants could easily use them at no cost. Other participants highlighted that the plant repellents were also versatile and multi-purposeful. For example, having plant repellents around homes beautified the compound and conserved the environment.

https://doi.org/10.1371/journal.pgph.0002469.g002

Participants also mentioned that some methods such as the use of mosquito coils were readily available in local shops at a low cost. Among participants who reared animals, the smouldering of cow dung was one method that they easily used as the material was available most of the time. Given that most participants preferred to keep animals near their houses which they mentioned attracted mosquitoes, implementing the strategy of burning dry cow dung in addition to using LLINs was considered a viable option for holistic malaria prevention.

Support provided for certain methods

According to the participants, several methods could be used holistically if community members were regularly educated and reminded of how to integrate different malaria prevention methods. One participant reported that after showing their community members the importance of slashing the compound regularly in addition to using LLINs, they started practicing both methods. Participants also mentioned that the Government of Uganda through the Ministry of Health provided free LLINs regularly which could be used together with other methods such as spraying insecticides, clearing overgrown vegetation, and draining stagnant water.

“ There was a lady in that home . The interesting thing is that where you see the little green now was a bush . But of recent , the lady had cleared the bush after our conversation . She had a good compound but due to the long grass previously , mosquitoes were very many at her home at night . She also informed me that she didn’t know mosquitoes were coming from her compound because it was bushy . She had thought they come from some other places . ” Male, Participant 15, Meeting 4

Barriers to integrated malaria prevention

Barriers to using several malaria prevention methods holistically included: inadequate knowledge and inaccessibility of some methods; allergic reactions to chemical-based methods; and affordability of certain methods.

Inadequate knowledge and inaccessibility of some methods

Participants highlighted that some community members were not aware of certain malaria prevention methods such as the use of body repellents. In addition, participants mentioned that they did not know about some malaria prevention methods that were presented by their colleagues during the monthly meetings such as electric mosquito traps. Other participants stated that many community members did not know the importance of using multiple methods. For example, many participants reported that some of the people they interacted with during the study did not know that housing improvements such as properly fitting windows and doors, filling of holes in walls, placement of glass panes in windows and doors, and the use of screens in windows reduced entry of mosquitoes in houses. Participants also revealed that body repellents and electric traps were not readily accessible within their communities thus prevented them from using these methods alongside other measures. For example, many participants stated that they had not seen body repellents in their communities, despite being aware of the possible use of vaseline-based repellents especially while outdoors at night.

“ When I was moving around , I realised that people have no knowledge and need to be taught about other methods apart from the mosquito net usage and closing doors early . Therefore , people need to be taught to understand the different ways that can be used to prevent malaria especially housing improvement . ” Male, Participant 5, Meeting 3

Allergic reactions to chemical-based methods

Participants mentioned that some of the malaria prevention methods such as the use of LLINs, IRS, and insecticide sprays caused allergic reactions such as skin irritation and discomfort among users. Therefore, some people could not integrate them with other prevention methods. Other participants stated that smouldering of plant repellents and cow dung sometimes caused discomforts such as respiratory irritation among some people, thereby preventing their holistic use. According to the participants, it was very common for pregnant women to have shortness of breath when inside houses that had been sprayed with insecticides.

“There was a lady who was seated in the bedroom at night without the mosquito net fitted on the bed. When I asked her why she was not using the net, she told me that it disturbs her and gets breathless when she sleeps under it.” Female, Participant 19, Meeting 5

Unaffordability of some methods

Participants stated that some malaria prevention methods were expensive, especially those that necessitated regular buying. Some participants said that members of their community did not have the money to buy electric mosquito traps and body repellents. Furthermore, the use of electric mosquito traps was linked to increased electricity costs due to routine charging. It was therefore very common for people including children to eat dinner and do homework outside the houses at night due to adequate lighting indoors which exposed them to mosquito bites. This was because household heads could not always afford to purchase kerosene for lamps and therefore used natural light from the moon ( Fig 3 ). Indeed, all the participants agreed that poverty was a major barrier to the integration of the different malaria prevention methods in the community.

https://doi.org/10.1371/journal.pgph.0002469.g003

According to the participants, money was a necessity for implementing different methods such as supporting housing improvement with the right quality of building materials and replacing broken glass window and door panes. Participants established that many people were sleeping in incomplete and poorly built houses because they lacked funds. Indeed, many participants captured photos of houses with poorly fitted doors and windows, those with missing window panes, those with holes in their walls, those with screening in vents, and others with windows and doors having spaces in them that allowed mosquito entry. The use of cloth materials, iron sheets, and rods to cover the window and door spaces as a way of improvising for the lack of standard panes on houses was common ( Fig 4 ).

Iron sheets and pieces of wood had been placed in the window which left gaps that permitted mosquito entry.

https://doi.org/10.1371/journal.pgph.0002469.g004

This qualitative photovoice study presented community members with the opportunity to assess facilitators and barriers to holistic malaria prevention. It also explored realistic and feasible ways for integration of malaria prevention practices, both conventional and non-conventional, thereby contributing to reducing the burden of the disease. The prominence of non-conventional methods in the study such as plant repellents, smouldering animal dung, and prophylactic herbs are important findings that should be explored when discussing holistic malaria prevention in rural settings. Indeed, these methods may require further investigation to better appreciate their contribution to malaria control in Uganda and elsewhere. If these methods are proven to be effective, they could complement existing malaria control strategies locally and globally. Facilitators of integrated malaria prevention such as those related to low cost and availability of various methods could be harnessed to enhance integrated control of mosquitoes in communities. Strategies to address the barriers to using multiple methods such as inadequate knowledge and inaccessibility of some methods need to be addressed as part of national malaria control efforts. These findings provide evidence on integrated malaria prevention interventions in the context of a rural setting in Uganda which could be used to support future research as well as policy and practice related to integrated vector management as recommended by the WHO [ 37 ].

Many non-conventional methods established in our study such as smouldering animal dung and use of prophylactic herbs were considered cheap and easily accessible by the community which facilitated integration. Rural communities are characterised by vast vegetation cover which facilitates growth of plant repellents and rearing of animals which provide cow dung. Herbs that were used for malaria prophylaxis can easily be collected from gardens, bushes, as well as nearby shrubs and neighbourhoods [ 38 ]. Previous qualitative and quantitative studies in Uganda [ 38 , 39 ], Kenya [ 40 – 42 ], West Africa [ 43 ], Tanzania [ 44 ], and India [ 45 ] have all reported burning of cow dung and plant repellents to prevent mosquito bites and entry into houses. Eating herbal plants in groundnut paste to prevent malaria has also been documented elsewhere [ 38 ]. Research on the effectiveness, efficacy, and safety of non-conventional methods such as prophylaxis with herbs and burning of cow dung, alongside conventional ones should be conducted to potentially contribute to national and global malaria control efforts.

Support from the government and health workers was seen as another facilitator to integrated malaria prevention in our study. Other quantitative and qualitative studies we have recently conducted documented the importance of government support in malaria prevention especially in rural settings [ 46 , 47 ]. Future provision of free LLINs to the population by MOH is likely to further facilitate use of multiple malaria prevention methods at households. Integrated malaria prevention requires regular education and reminding community members of how to integrate the various methods [ 48 ]. An earlier pilot project on malaria prevention showed that households observed various practices with support from CHWs who carried out regular community sensitization on the different methods [ 49 ]. CHWs, who spend vast time with members of the community, could spearhead campaigns of carrying out routine health education on using multiple methods to prevent malaria. Indeed, CHWs are known to play an instrumental role in health promotion particularly in rural settings.

In our study, many of the participants and community members did not know the importance of using several malaria prevention methods. This finding is similar to previous quantitative studies where few methods were being integrated due to lack of knowledge [ 45 – 47 , 50 ]. Similar to our study, knowledge and use of LLINs was prominent in previous qualitative and quantitative studies including in Kenya and Tanzania [ 40 , 44 , 47 , 49 ]. Adequate knowledge on LLINs among populations may be attributed to the mass campaigns on use of mosquito nets including by the Ugandan Ministry of Health [ 8 ]. Future malaria initiatives should also emphasize the importance of using several other methods alongside LLINs to increase knowledge of other malaria prevention practices which could contribute to increased utilisation. Furthermore, holistic malaria prevention activities can also support the shortfalls in use of LLINs such as low uptake and sustainability in communities [ 20 ].

Participants mentioned that some malaria prevention methods such as body repellents were inaccessible in their communities. A quantitative study done in Ethiopia found that 29.9% of the participants did not use malaria prevention methods due to accessibility challenges [ 50 ]. Inaccessibility of malaria services undermines collective efforts to eradicate the disease by 2050 [ 51 ]. Furthermore, some of the malaria prevention methods used such as LLINs (in cases where free ones are not provided by the government), electric mosquito traps, and body repellents were considered as expensive. Views of some malaria prevention methods being unaffordable have been expressed in other qualitative and quantitative studies [ 41 , 45 , 47 , 50 ]. This could be because rural communities have many competing financial priorities yet they are generally poor [ 41 ]. The use of electric mosquito traps is best suited for communities with an electric power grid, which may not be the case with many households in rural areas. Furthermore, the few houses that are connected to the electricity in rural settings cannot afford the high-power tariffs hence rendering electric devices unsustainable. Given that the average number of persons in households in Uganda is 5 [ 52 ], frequent purchase of body repellents in rural areas is unsustainable. Addressing the social determinants of health in rural communities such as poverty is likely to improve malaria prevention practices in the short and long term.

Poor housing quality was another barrier to integrated malaria prevention. Concerns of mosquito entry in houses due to open eaves, as well as gaps in doors and windows have also been documented in several countries including Gambia [ 53 ] and Ethiopia [ 54 ]. In our study, participants stated that people were sleeping in incomplete and poorly built houses because of the costs involved. It is characteristic of rural houses to have open eaves and incomplete walls which allows entry of mosquitoes [ 41 ]. An earlier study on reducing malaria by mosquito proofing suggests that the importance of screening houses has been ignored in many areas [ 55 ]. In our study, household members improvised with the use of iron sheets and cloth materials to provide temporary screening in windows and doors. However, these materials are not effective in preventing mosquito entry into houses. Indeed, recent case studies in Uganda and Tanzania have recommended improving the design and structures of houses to reduce mosquito entry [ 48 ].

Despite participants agreeing that the use of LLINs, IRS, and spraying of insecticides were core methods to integrated malaria prevention, concerns were expressed of these chemical-based methods being linked to allergic reactions and other negative effects on health. Quantitative studies elsewhere have documented allergic reactions such as skin irritation and discomfort due to the use of these methods [ 56 , 57 ] even among pregnant women [ 58 ]. However, it should be noted that such reactions are usually short-term [ 59 , 60 ] and are mostly caused by failure to observe recommended guidelines of hanging nets outside for 24 hours before usage [ 61 ]. In addition, staying outside the house for at least 2 hours after IRS is complete to allow the insecticide to dry could minimise side effects of the intervention [ 62 ]. Continuous health education on the recommended usage of chemical-based malaria prevention methods in homes by health workers and CHWs is necessary especially during community LLIN and IRS campaigns.

The use of photovoice in our study not only allowed participants to identify facilitators and barriers to holistic malaria integration but also enabled them to take leadership in identifying malaria prevention practices within their communities. Engaging 20 participants who were diverse enabled the collection of a wide range of perspectives and experiences. However, our study was carried out in only one predominantly rural area in Wakiso district hence the findings may be different within other geographical contexts hence not generalisable to other settings. In addition, evaluation of the community’s knowledge and practices on malaria prevention following the participatory approach was not assessed. Future studies on integrated malaria prevention including non-conventional methods using both qualitative and quantitative methods are recommended in peri-urban and urban settings in Uganda.

Several facilitators to using integrated malaria prevention were found such as low cost and accessibility of several methods, and support provided for certain methods. However, inadequate knowledge, inaccessibility, and unaffordability of some methods, as well as allergic reactions to chemical methods impeded the use of several malaria prevention methods holistically. The findings from the study could support future research, as well as inform policy and practice related to integrated vector management.

Supporting information

S1 data. study dataset..

https://doi.org/10.1371/journal.pgph.0002469.s001

Acknowledgments

The authors acknowledge the contribution to the study from: the community mobilisers (John Bosco Matovu, Henry Bugembe, and Henry Kajubi); the administration of Kasanje Health Center III and Kasanje Town Council; Wakiso District Local Government; and the Ministry of Health. We also thank the participants for their involvement in the study.

• View Article
• PubMed/NCBI
• Google Scholar
• 6. Gallup JL, Sachs JD. “The Economic Burden of Malaria.” CID Working Paper Series 2000.52, Harvard University, Cambridge, MA, July 2000.
• 11. World Health Organization (WHO). Global technical strategy for malaria 2016–2030: World Health Organization 1211 Geneva 27 Switzerland; 2015.
• 20. Shretta R, Liu J, Cotter C, Cohen J, Dolenz C, Makomva K, et al. Malaria Elimination and Eradication. In: Holmes KK, Bertozzi S, Bloom BR, Jha P, editors. Major Infectious Diseases. 3rd ed. Washington (DC): The International Bank for Reconstruction and Development / The World Bank; 2017 Nov 3. Chapter 12. https://doi.org/10.1596/978-1-4648-0524-0_ch12
• 23. Espino F, Koops V, Manderson L. Community participation and tropical disease control in resource-poor settings. World Health Organization, 2004.
• 32. Uganda Bureau of Statistics (UBOS). Uganda Bureau of Statistics: National Population and Housing Census 2014, Kampala, Uganda. 2014.
• 37. World Health Organization (WHO). Handbook for integrated vector management. World Health Organization; 2012.
• 48. Okello-Onen J, Mboera LE, Mugisha S. Malaria research and management need rethinking: Uganda and Tanzania case studies. Ecohealth Research in Practice. Insight and Innovation in International Development, vol 1. Springer, New York, NY. https://doi.org/10.1007/978-1-4614-0517-7_13 .
• 52. UBOS. Uganda National Household Survey 2019–2020. In: Statistics UBo, editor.: Kampala Uganda; 2021.
• 61. Ministry of Health, Uganda. Press statement on the mosquito net distribution campaign in Kampala and Wakiso district. Kampala: 2017. Available from: health.go.ug/wp-content/uploads/2019/12/press-statement-Wave-6-Nov-21-1-Final.pdf.
• 62. World Health Organization (WHO). Indoor residual spraying: an operational manual for indoor residual spraying (IRS) for malaria transmission control and elimination: World Health Organization; 2015.

An official website of the United States government

Here's how you know

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. A lock ( ) or https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Keyboard Navigation

• Agriculture and Food Security
• Anti-Corruption
• Conflict Prevention and Stabilization
• Democracy, Human Rights, and Governance
• Economic Growth and Trade
• Environment, Energy, and Infrastructure
• Gender Equality and Women's Empowerment
• Global Health
• Humanitarian Assistance
• Innovation, Technology, and Research
• Water and Sanitation
• Burkina Faso
• Central Africa Regional
• Central African Republic
• Côte d’Ivoire
• Democratic Republic of the Congo
• East Africa Regional
• Power Africa
• Republic of the Congo
• Sahel Regional
• Sierra Leone
• South Africa
• South Sudan
• Southern Africa Regional
• West Africa Regional
• Afghanistan
• Central Asia Regional
• Indo-Pacific
• Kyrgyz Republic
• Pacific Islands
• Philippines
• Regional Development Mission for Asia
• Timor-Leste
• Turkmenistan
• Bosnia and Herzegovina
• North Macedonia
• Central America and Mexico Regional Program
• Dominican Republic
• Eastern and Southern Caribbean
• El Salvador
• Middle East Regional Platform
• West Bank and Gaza
• Dollars to Results
• Data Resources
• Strategy & Planning
• Budget & Spending
• Performance and Financial Reporting
• FY 2023 Agency Financial Report
• Records and Reports
• Budget Justification
• Our Commitment to Transparency
• Policy and Strategy
• How to Work with USAID
• Find a Funding Opportunity
• Organizations That Work With USAID
• Resources for Partners
• Get involved
• Business Forecast
• Safeguarding and Compliance
• Diversity, Equity, Inclusion, and Accessibility
• Mission, Vision and Values
• News & Information
• Operational Policy (ADS)
• Organization
• Stay Connected
• USAID History
• Video Library
• Coordinators
• Nondiscrimination Notice
• Collective Bargaining Agreements
• Disabilities Employment Program
• Federal Employee Viewpoint Survey
• Reasonable Accommodations
• Urgent Hiring Needs
• Vacancy Announcements
• Search Search Search

The U.S. President’s Malaria Initiative (PMI)

KENYA OVERVIEW

Malaria remains a major public health challenge in Kenya. Due to altitude, rainfall patterns and temperature, about 75 percent of the Kenyan population is at risk for malaria. In 2023, Malaria accounted for at least 15 percent of outpatient consultations nationally and 58 percent of outpatient consultations in the eight counties supported by the U.S. President’s Malaria Initiative (PMI). Over the past decade, Kenya has made significant gains in reducing the malaria burden among its 52.4 million population. From 2010-2020, the prevalence of malaria reduced by 50% from 38.1% to 18.9% in the high-burden lake-endemic area, where PMI focuses 70 percent of its investments. Nationally, the malaria burden decreased 49%, from 11.4% to 5.8%, during this same period.

Kenya is a focus country for PMI, which supports efforts by USAID and other U. S. g overnment partners, in collaboration with the Government of Kenya (GOK), to expand malaria prevention, diagnosis and treatment measures. Since 2006, PMI has invested over $ 530 million in Kenya. USAID works closely with the National Malaria Control Program and counties to implement the Kenya Malaria Strategy, promote the use of insecticide-treated nets, and create awareness on the importance of prompt diagnosis and treatment for suspected malaria – particularly among pregnant women and children.

This article is free to read if you register or sign in.

Simply register at no cost and sign in below to proceed.

Questions or problems? Email [email protected] or call 0711 046 000 .

REASSURANCE

Latest results confirm malaria vaccine in kenya safe, latest results show a reduction in hospital admissions with severe malaria by 32%.

Researchers have confirmed ...

Most Popular

State to test children cough syrups for toxins, governors threaten to pull out of talks on doctors strike, put hiv prevention pills in condom dispensers – researchers, kenya latest victim as moderna halts sh26bn vaccine plant.

Kenya Police:

Kenya Medical Board:

Ministry of Health:

Nursing Council of Kenya:

Other Links

Medical Practitioners and Dentists Board

Kenya red Cross

St. John's Ambulance

Nairobi Women's Hospital

The data driven journalism + tools in StarHealth section were kickstarted by Code4Kenya

The code & data for this page are open source. You can re-use it by visiting the above repositories.

Latest Videos

Police in turkana arrest kdf soldiers who slapped, disarmed colleague, the news brief: job losses loom as state targets academic fakes to ..., sign up for the free star email newsletter and receive the latest kenya news daily..

Volume 29, Number 11—November 2023

Research Letter

Plasmodium vivax prevalence in semiarid region of northern kenya, 2019.

Cite This Article

In urban and rural areas of Turkana County, Kenya, we found that 2% of household members of patients with Plasmodium falciparum infections were infected with P . vivax . Enhanced surveillance of P . vivax and increased clinical resources are needed to inform control measures and identify and manage P . vivax infections.

Until recently, little or no endemic transmission of Plasmodium vivax has been reported in sub-Saharan Africa outside of the Horn of Africa ( 1 ). P. vivax was presumed to be largely absent because the Duffy blood group antigen was rare in persons living in the region. However, accumulating evidence of endemic P. vivax has indicated that this parasite might be present in many areas of sub-Saharan Africa, albeit at low levels, and Duffy antigen–negative persons can be infected and contribute to transmission ( 2 ).

Turkana County is in northwestern Kenya and shares a border with Uganda, South Sudan, and Ethiopia. Turkana county’s harsh climate is characterized by an average rainfall of <215 mm/year and daytime temperatures of 40°C. Malaria transmission in this region was predicted to occur in isolated pockets with epidemic potential only after unusual rainfall. However, reactive case detection conducted across central Turkana County documented year-round symptomatic and asymptomatic P. falciparum infections and confirmed perennial endemic transmission of malaria ( 3 ).

We hypothesized that P. vivax might also be circulating in Turkana County because of stable malaria transmission and proximity to Ethiopia, where P. vivax infections are endemic. To test this hypothesis, we extracted genomic DNA from 3,305 dried blood spots collected from household members of patients with P. falciparum infections; household members were enrolled in the study at their homes in catchment areas surrounding 3 rural and 3 urban health facilities in central Turkana County ( 3 ). The study was approved by the Moi University Institutional Research and Ethics Committee and Duke University Institutional Review Board.

We tested each DNA sample for P. vivax by using an established nested qualitative PCR protocol ( 4 ). Gel electrophoresis bands were identified independently by 2 observers. We randomly selected 15 extracts for retesting by probe-based real-time PCR with the same primer sequences to detect the same target; all PCR products were confirmed. For our analysis, we used nested qualitative PCR results.

Figure . Prevalence of Plasmodium vivax infection in communities along the Turkwel River in study of P . vivax prevalence in semiarid region of northern Kenya,...

The percentage of household members infected with P. vivax was 2.1% (69/3,305); of those, 45% (31/69) were co-infected with P. falciparum ( Table ). We detected P. vivax infections across our study transect throughout most of the year ( Figure ; Appendix Figures 1, 2); the highest (5.8%, 28/485) prevalence was recorded near an urban facility in the town of Lodwar. Infections were present across all age groups, but we observed a slightly higher (1.6%, 8/490) percentage of P . vivax monoinfections in children <5 years of age ( Table ). Ten P . vivax –infected participants reported malaria-like symptoms when they were screened; 7 of those were co-infected with P . falciparum . Only 3 P . vivax –infected participants had a malaria-like illness within 1 month before enrollment; none reported taking antimalarial drugs. None of the P . vivax –infected participants reported traveling outside of their subcounty within 2 months before enrollment; 16% (11/69) reported having a net for their sleeping space, which was slightly less than uninfected participants (19.7%, 468/2,376) who had a net.

The burden of P . vivax infections in sub-Saharan Africa remains unclear; infections are rarely diagnosed in a clinical setting and might often be asymptomatic. The recommended rapid diagnostic test in most countries of sub-Saharan Africa is P . falciparum –specific. Consequently, P . vivax infections might be underestimated or undocumented.

Strategies designed to eliminate P . falciparum are undermined by P . vivax because dormant P . vivax hypnozoites that can cause relapse and sustain transmission are difficult to detect and treat ( 5 ). Furthermore, P . vivax infections generate gametocytes before symptom onset, making detection and treatment challenging before onward transmission occurs. P . vivax infections could present a growing challenge in Kenya, even as P . falciparum is brought under control, a process that has been observed in co-endemic malaria settings in Southeast Asia ( 6 ).

We did not test participants for Duffy antigen expression, which could have affected their susceptibility to P . vivax . Estimated Duffy antigen positivity in Kenya is 5%–10% ( 7 ). P . vivax infections in Duffy-negative subjects have been documented in Africa ( 2 ). Characterization of Duffy antigen expression will be needed to understand the threat of P . vivax infections in Kenya.

Anopheles stephensi mosquitoes have been identified in Kenya (E.O. Ochomo et al., unpub. data, https://doi.org/10.21203/rs.3.rs-2498485/v1 ), and the potential expansion of this highly competent vector, which survives in urban and manmade habitats, could dramatically change malaria transmission patterns. Continued spread of this invasive vector into sub-Saharan Africa would place ≈126 million persons at risk for malaria ( 8 ). Identification of An. stephensi mosquitoes in Djibouti was linked with a >100-fold rise in malaria cases, including the first autochthonous cases of P. vivax reported in 2016 ( 9 ).

In conclusion, if emerging An. stephensi mosquitoes become established across Kenya in the presence of confirmed P . vivax cases, malaria elimination in Kenya will be substantially more difficult to achieve. Enhanced surveillance for both An . stephensi mosquitoes and P . vivax will be needed to inform control measures, and increased clinical resource allocation will enable detection and effective treatment of patients with P . vivax malaria.

Dr. Prudhomme O’Meara is a scientist with joint appointments at Duke University and Moi University. Her research interests focus on malaria transmission dynamics and prevention, control, and elimination strategies in remote communities facing new malaria threats.

• Battle  KE , Lucas  TCD , Nguyen  M , Howes  RE , Nandi  AK , Twohig  KA , et al. Mapping the global endemicity and clinical burden of Plasmodium vivax , 2000-17: a spatial and temporal modelling study. Lancet . 2019 ; 394 : 332 – 43 . DOI PubMed Google Scholar
• Wilairatana  P , Masangkay  FR , Kotepui  KU , De Jesus Milanez  G , Kotepui  M . Prevalence and risk of Plasmodium vivax infection among Duffy-negative individuals: a systematic review and meta-analysis. Sci Rep . 2022 ; 12 : 3998 . DOI PubMed Google Scholar
• Meredith  HR , Wesolowski  A , Menya  D , Esimit  D , Lokoel  G , Kipkoech  J , et al. Epidemiology of Plasmodium falciparum infections in a semi-arid rural African setting: evidence from reactive case detection in northwestern Kenya. Am J Trop Med Hyg . 2021 ; 105 : 1076 – 84 . DOI PubMed Google Scholar
• Singh  B , Bobogare  A , Cox-Singh  J , Snounou  G , Abdullah  MS , Rahman  HA . A genus- and species-specific nested polymerase chain reaction malaria detection assay for epidemiologic studies. Am J Trop Med Hyg . 1999 ; 60 : 687 – 92 . DOI PubMed Google Scholar
• Olliaro  PL , Barnwell  JW , Barry  A , Mendis  K , Mueller  I , Reeder  JC , et al. Implications of Plasmodium vivax biology for control, elimination, and research. Am J Trop Med Hyg . 2016 ; 95 ( Suppl ): 4 – 14 . DOI PubMed Google Scholar
• Price  RN , Commons  RJ , Battle  KE , Thriemer  K , Mendis  K . Plasmodium vivax in the era of the shrinking P. falciparum map. Trends Parasitol . 2020 ; 36 : 560 – 70 . DOI PubMed Google Scholar
• Howes  RE , Patil  AP , Piel  FB , Nyangiri  OA , Kabaria  CW , Gething  PW , et al. The global distribution of the Duffy blood group. Nat Commun . 2011 ; 2 : 266 . DOI PubMed Google Scholar
• Sinka  ME , Pironon  S , Massey  NC , Longbottom  J , Hemingway  J , Moyes  CL , et al. A new malaria vector in Africa: Predicting the expansion range of Anopheles stephensi and identifying the urban populations at risk. Proc Natl Acad Sci U S A . 2020 ; 117 : 24900 – 8 . DOI PubMed Google Scholar
• Seyfarth  M , Khaireh  BA , Abdi  AA , Bouh  SM , Faulde  MK . Five years following first detection of Anopheles stephensi (Diptera: Culicidae) in Djibouti, Horn of Africa: populations established-malaria emerging. Parasitol Res . 2019 ; 118 : 725 – 32 . DOI PubMed Google Scholar
• Figure . Prevalence of Plasmodium vivax infection in communities along the Turkwel River in study of P. vivax prevalence in semiarid region of northern Kenya, 2019. Household members of patients with...
• Table . Plasmodium falciparum and P. vivax infections according to age groups of household members in study of P. vivax prevalence in semiarid region of northern Kenya, 2019

DOI: 10.3201/eid2911.230299

Original Publication Date: October 01, 2023

Table of Contents – Volume 29, Number 11—November 2023

Please use the form below to submit correspondence to the authors or contact them at the following address:

Wendy Prudhomme O’Meara, Duke Global Health Institute, Duke University, 310 Trent Dr, Durham, NC 27708, USA

Comment submitted successfully, thank you for your feedback.

There was an unexpected error. Message not sent.

Exit Notification / Disclaimer Policy

• The Centers for Disease Control and Prevention (CDC) cannot attest to the accuracy of a non-federal website.
• Linking to a non-federal website does not constitute an endorsement by CDC or any of its employees of the sponsors or the information and products presented on the website.
• You will be subject to the destination website's privacy policy when you follow the link.
• CDC is not responsible for Section 508 compliance (accessibility) on other federal or private website.

Article Citations

Highlight and copy the desired format.

Metric Details

Article views: 1325.

Data is collected weekly and does not include downloads and attachments. View data is from .

What is the Altmetric Attention Score?

The Altmetric Attention Score for a research output provides an indicator of the amount of attention that it has received. The score is derived from an automated algorithm, and represents a weighted count of the amount of attention Altmetric picked up for a research output.

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Trop Med Health

Prevalence and associated determinants of malaria parasites among Kenyan children

Marufa sultana.

1 Health Economics & Financing Research Group, International Centre for Diarrhoeal Disease Research, Bangladesh (icddr,b), 68 Shahid Tajuddin Ahmed Sharani, Mohakhali, Dhaka, 1212 Bangladesh

Nurnabi Sheikh

Rashidul alam mahumud, tania jahir, ziaul islam, abdur razzaque sarker.

2 University of Strathclyde, Glasgow, UK

Associated Data

The datasets of 2015 Kenya Malaria Indicator Survey (KMIS) are available in Demographic and Health Survey website.

Approximately 80% of deaths attributed to malaria worldwide occurred mainly in Africa in 2015. Kenya is one of the major malaria endemic countries, making malaria the leading public health concern in this country. This study intended to document the prevalence of malaria and determine associated factors including socioeconomic status among children aged 6 months to 14 years in Kenya.

This study analyzed the secondary data extracted from the 2015 Kenya Malaria Indicator Survey (KMIS), a cross-sectional country representative survey. Associations of demographic, socioeconomic, community-based, and behavioral factors with the prevalence of malaria in children were analyzed using multivariable logistic regression analysis.

Data from 7040 children aged 6 months to 14 years were analyzed. The prevalence of malaria showed an upward trend in terms of age, with the highest prevalence among children aged 11–14 years. Prevalence was also higher among rural children (10.16%) compared to urban children (2.93%), as well as poor children (11.05%) compared to rich children (3.23%). The likelihood of having malaria was higher among children aged 10–14 years (AOR = 4.47, 95% CI = 3.33, 6.02; P <  0.001) compared with children aged under 5 years. The presence of anemia (AOR = 3.52, 95% CI = 2.78, 4.45; P  < 0.001), rural residence (AOR = 1.71, 95% CI = 1.31, 2.22; P <  0.001), lack of a hanging mosquito net (AOR = 2.38, 95% CI = 1.78, 3.19; P <  0.001), primary education level of the household head (AOR = 1.15, 95% CI = 1.08, 2.25; P <  0.05), and other factors, such as the household having electricity and access to media such as television or radio, were also associated with the likelihood of infection.

This study demonstrated the need to focus on awareness programs to prevent malaria and to use existing knowledge in practice to control the malaria burden in Kenya. Furthermore, this study suggests that improving the information available through the mass media and introducing behavior change communication and intervention program specifically for those of poor socioeconomic status will help to reduce malaria cases.

Malaria is an entrenched global health challenge and is a major public health concern in many countries including Kenya [ 1 ]. It is endemic in over 100 countries, and almost half of the worldwide population is at risk of malaria, where approximately one million people die from malaria each year [ 2 ]. This life-threatening disease is transmitted in humans from one person to another indirectly via the bite of female mosquitoes of the genus Anopheles, which harbors one of five species of parasites belonging to the genus Plasmodium [ 2 , 3 ].

In 2015, malaria transmission had been noted in over 95 countries and territories, constituting approximately 214 million cases. However, approximately 80% of all deaths due to malaria were concentrated in just 15 countries, mainly in the African region [ 2 ]. Children and pregnant women are most vulnerable to morbidity and mortality associated with malaria. Globally, approximately 306,000 children under the age of 5 died that year due to malaria, and approximately two thirds of these deaths occurred in the African region [ 4 ]. According to the World Health Organization (WHO), it is estimated that 9 out of 10 deaths in children were caused by malaria in Africa [ 2 ]. Transmission of malaria highly depends on the temperature, humidity, and rainfall [ 5 ]. High temperature and heavy rainfall in summer season leads the highest malaria transmission, especially in Africa [ 6 ]. Despite those climatic factors, malaria transmission is also determined by the socioeconomic conditions and knowledge of and access to malaria prevention tools as well as the healthcare services [ 6 ]. In Africa, malaria transformation is comparatively higher among the rural setting than urban areas which may be because of the higher vector density, lower housing quality, and the poor drainage systems in rural settings [ 7 ]. Malaria is a major threat to public health and is the leading cause of morbidity and mortality in Kenya [ 8 ]. Out of 34 million Kenyans, approximately 25 million are estimated to be at risk of malaria, which is more than 70% of the population at risk [ 6 ]. An estimated 6.7 million new clinical cases each year, with 4000 deaths occurring particularly among children, make malaria a major health burden for Kenya [ 9 ].

Pregnant women and children are high-risk groups for malaria as they are typically affected most severely by this disease, and prevention efforts typically target these vulnerable groups in Kenya [ 2 ]. School-aged children (age 5–15) bear the most significant burden of malaria in terms of having the highest prevalence rate [ 10 ]. Although a number of studies have been conducted on malaria among children and young adults, most of them are clinical-, treatment-, and prevention-based studies [ 1 , 9 , 11 ].

It is evident that sociocultural context and community attitudes and perceptions have a significant role in prevention and control of malaria cases. However, such studies are rarely examined in the context of Kenya specifically [ 12 , 13 ]. Furthermore, in the African region, it was demonstrated that community awareness is generally very poor at preventing malaria cases, although it is the cardinal tool currently used for malaria prevention strategies [ 14 ]. Greater knowledge, attitudes, and active practices regarding malaria disease are critical in establishing effective control measures. The introductions of mass media and behavior change communication (BCC) to malaria control are well documented and proven interventions which increases the possibility of a better return on malaria programmes [ 15 ]. BCC campaigns can create demand among the families to use and hang their nets regularly and can improve malaria prevention as well as treatment behaviors, especially among the vulnerable groups [ 16 ]. A study in Kenya regarding the practice of malaria control in a specific division showed above average practices, but approximately 30% of respondents had household members who failed to use the control method properly. That study was based in a specific area and recommended further studies on health care promotion, intervention, and better communication regarding sustainable behavior changes [ 17 ]. Another study conducted in Kenya observing the efficacy of text-message reminders found that text-message reminders can increase a child’s compliance with respect to follow-up after anti-malarial treatment [ 18 ]. However, the effects of existing knowledge, attitude, and practices (KAP) of Kenyan populations with regard to malaria prevention strategies are rarely examined. Identifying key risk factors by socioeconomic context and incorporating the effect of existing knowledge and practice of malaria prevention are crucial for the effective implementation of prevention and health intervention programs. It is also essential for policy formulation and for the assessment of resource requirements, specifically for low-resource settings, and for intervention prioritization by regions. Furthermore, a better understanding of the association between malaria and sociodemographic factors related to poverty is needed because the financial protection necessary to take remedial measures is a major challenge for some households [ 9 , 11 , 12 ]. Therefore, the intention of this study was to determine the prevalence of malaria among children under 15 years of age, to examine associated determinants considering socioeconomic status, and to examine the effect of knowledge and attitudes concerning malarial disease and its prevention strategies among Kenyan households. The knowledge generated by this study can contribute to the formulation of malaria control programme among the young children of Kenya. Understanding causal association of malaria in sociodemographic context along with the knowledge, attitudes are vital issues which can provide essential insight into malaria burden and helps policy level decision on malaria control strategies.

This study used cross-sectional survey data from a secondary source extracted from the Kenya Malaria Indicator Survey (KMIS), 2015. The survey was based on a nationally representative sample drawn from the four epidemiological zones in Kenya (highland epidemic-prone areas, endemic areas (lake and coast), semi-arid seasonal malaria transmission areas, and low-risk malaria areas) [ 8 ]. Data were collected using a two-stage cluster sampling design, based on the sampling frame of the Fifth National Sample Survey and Evaluation Program (NASSEP V), which itself is based on the 2009 Population and Housing Census (PHC) Enumeration Areas (EAs) created by the Kenya National Bureau of Statistics (KNBS) [ 8 ]. The sampling frame was divided into four equal sub-samples, from one of which the 2015 KMIS sample data were drawn. In the first stage, a total of 246 clusters (EAs were selected as sample clusters numbering 131 and 115 for rural and urban, respectively) with equal probability of selection were chosen from the NASSEP V master sample. In the second stage, using a systematic sampling technique, a uniform sample of 30 households from each of the selected clusters were selected for the study. The data were collected from 6 July to 15 August 2015, using three types of questionnaires (a Household Questionnaire, a Woman’s Questionnaire, and a Biomarker Questionnaire) that covered a sample of 7313 households based on household surveys executed by the National Malaria Control Programme (NMCP) of the Ministry of Health (MOH) and the Kenya National Bureau of Statistics [ 8 ]. From the children aged 6 months to 14 years in the selected households, blood samples were collected for testing anemia and malaria. Hemoglobin analysis was carried out to detect the presence of anemia in the children. Severe anemia was considered to be a hemoglobin level < 8.0 g/dl, and moderate anemia was between 8.0 and 9.9 g/dl. Other anemia classifications varied by age group as follows: children 6–59 months: mild anemia 10.0–10.9 g/dl, no anemia > 11.0 g/dl; children 5–11 years: mild anemia 10.0–11.4 g/dl, no anemia > 11.5 g/dl; children 12–14 years: mild anemia 10.0–11.9 g/dl, no anemia > 12.0 g/dl [ 8 ]. Since microscopic examination is the gold standard for the diagnosis of malaria, for this study children were considered as malaria positive or negative based on the result of this test only [ 19 – 21 ]. Approval was obtained from the Demographic Health Survey (DHS) website to use the 2015 KMIS data. A total of 7040 children aged 6 months to 14 years were analyzed for the current study.

Statistical analysis

All statistical analyses were performed using the statistical package Stata/SE 13.0 and significant associations have been measured at 5% alpha level ( p  < 0.05). Based on the MIS instruction, sampling weight was used for cluster adjustment. Both bivariate and multivariable statistical analyses were conducted during data analysis. Bivariate analysis was carried out to explore the prevalence of malaria compared to different selected variables and to the knowledge and the attitude of the respondent. The chi-squared test of independence was used to determine any significant associations between positive blood smear test results and attitudes, knowledge, and measures in terms of the P value. Based on significant associations with the results, variables were chosen for the multivariate analysis [ 6 ]. Binary logistic regression model was used to trace the significant determinants for malaria, and the results were presented in terms of odds ratio (OR) that controlled for multiple confounders (with 95% confidence interval). A binary logistic regression model was used in this analysis because the outcome variable has a binary response of malaria positive or negative. This outcome variable was re-coded as “0” for children who did not have malaria and “1” for those who had malaria. Both adjusted and unadjusted ORs were considered for finding the single and multi-factorial (covariates) effects in the model [ 22 ]. Age, sex, and anemia levels of children, sex, and education of household head, household electricity status, media exposure, number of living room, net hanging status for sleeping, residence, and wealth index were used as independent variables in multivariate regression model. Some independent variables were used as per the original dataset and some were re-coded depending on research interests. Socioeconomic status was measured by a wealth index, which is a composite measure of a household’s cumulative living standard calculated using data on the household’s selected assets by generating a weight or factor score through principal components analysis [ 8 ].

Prevalence of malaria

Malaria prevalence increases with increasing age of the children in this population. Comparatively aged children (10–14 years) suffer more due to malaria in terms of the prevalence (10.22%), whereas malaria prevalence was 4.83% among children under the age of 5 (Table ​ (Table1). 1 ). Malaria prevalence was considerably higher among male children (8.23%) than female children (8.04%).

Distribution of the prevalence malaria among 6-month to 14-year children on sociodemographic characteristics ( N  = 9040)

Malaria prevalence varies between urban and rural areas, with children from rural areas having a higher prevalence (10.16%) compared to children from urban areas (2.93%). Among all children, approximately 19% were found to be anemic, and malaria prevalence was more than two times higher among these children (16.92%, 95% CI = 15.23, 18.77) compared to non-anemic children (6.06%, CI = 5.54, 6.63). Education of the household head appeared to be an important factor to control malaria among children since malaria cases were higher among children with less-educated (10.98%) and illiterate (7.82%) household heads. In this analysis, approximately 92.11% of the respondents reported using a mosquito net for sleeping, and malaria prevalence was observed to be more than two times higher among households that did not use mosquito nets (23.11%) compared to net user households (8.05%). Our study reveals a higher malaria prevalence (11.05%) among children from poor communities, followed by children from middle class (9.09%) and rich (3.23%) communities. Furthermore, malaria prevalence was higher in children whose household did not have access to media, such as radio and television, with prevalence rates of 9.99 and 10.40% for not having radio or television compared to 7.22 and 2.36% in those having access to radio or television, respectively (Fig. ​ (Fig.1 1 ).

Prevalence of malaria among media exposure and non-exposure households

Knowledge and attitude towards malaria

The knowledge and attitudes of the study respondents regarding malaria are presented in Table ​ Table2. 2 . According to this, approximately 69.76% of the study respondents remarked that hanging a mosquito net is extremely important for preventing malaria while only 1.24% reported that it is not important at all. However, the malaria prevalence was higher for children in those households (9.70%) who mention that the hanging net is extremely important, showing variation between knowledge and practice. In terms of the ‘importance of having children sleep under a treated net’, 75.02% of the respondents regarded it as extremely important, where 24.86% regarded it as just important. However, the prevalence was higher among the households’ children whose respondents answered that a treated net was extremely important (9.61%) compared to those who answered just important (4.74%). Approximately 67% of the respondents agreed that people in the community should sleep under an insecticide-treated net (ITN) all the time while 32% disagreed with this statement. The children of respondents that disagreed had the lowest malaria prevalence (4.10%), with a prevalence of 10.24% for those who agreed. Higher malaria prevalence was found (8.84%) among the children whose household respondents disagreed with the risk of malaria in the rainy season than among those who agreed (7.19%).

Chi-square test of association on respondent’s knowledge, attitude on malaria and mosquito net ( N  = 9040)

Determinants of malaria

Logistic regression models were constructed to examine sociodemographic determinants of malaria prevalence and are reported in Table ​ Table3. 3 . The age of the children, presence of anemia, education of the household head, household having electricity, access to television, residence type (rural or urban), and mosquito net use behavior for sleeping were found as significant determinants for malaria in children by both models, and the effects of these variables were directly predicted using the odds ratio [ 23 ]. From model II, the child’s age was one of the most significant factors for malaria among children aged 6 months to 14 years. The odds ratio of children aged 5–9 years (OR = 2.29, 95% CI = 2.23, 3.82) and 10–14 years (OR = 4.47, CI = 3.33, 6.02) demonstrated that these two age groups were more vulnerable to malaria than children under 5 years (reference group). Anemic children were 3.52 times more likely to have malaria compared to their non-anemic counterparts (OR = 3.52, 95% CI = 2.78, 4.45). The analysis of odds ratios also verified that children from households with uneducated heads (OR = 1.15, CI = 1.08, 2.25) and primary educated heads (OR = 1.82, CI = 1.35, 2.44) were more prone to have malaria than children of secondary and higher educated household heads. The children of households having no electricity were also more likely to have malaria (OR = 3.08, CI = 1.77, 5.34) compared to those in households with access to electricity. It was also found that the probability of having malaria was more likely among rural children than urban children (OR = 1.17, CI = 1.31, 2.22) and for the children of households with no access to television (OR = 1.63, CI = 1.01, 2.63) than their counterparts. Children of the households not using a bed net for sleeping were 2.38 times more susceptible to malaria (OR = 2.38, CI = 1.78, 3.19) compared to net users.

Multivariable logistic regression model on sociodemographic determinants of malaria

**p <  0.05, ***p <  0.01

Variable included in the multivariate model (model II): malaria test result either test positive (coded 1) or test negative (coded 0) for both model I and model II. Age, sex, and anemia levels of children, sex and education of household head, household electricity status, media exposure, number of living room, net hanging status for sleeping, residence, and wealth index were used as independent variables in multivariate regression model (model II)

This study identified the prevalence and examined sociodemographic and knowledge-based factors that determine the likelihood of malaria infection among children aged 6 months to 14 years in Kenya based on country representative secondary data from the 2015 Kenya Malaria Indicator Survey (KMIS). The government of Kenya tries to ensure improved health service delivery with a high priority for malaria prevention and control [ 11 ]. The government has developed several effective strategies for monitoring and evaluating malaria control on a regular basis, mainly focused on the reduction of malaria morbidity and mortality by 2018 [ 11 ]. This study enables understanding of the association between malaria and sociodemographic factors, revealing that factors such as age, education, economic status of the household, media access, knowledge, and attitude have potential impacts for affecting the prevalence of malaria among the Kenyan children.

This research exposed that malaria prevalence was lower among children less than 5 years old, and susceptibility tends to increase along with increasing age. This is consistent with the finding of another study in Africa, where malaria prevalence was found to be higher among children aged 5–18 years [ 24 ]. Male children were found to have a higher prevalence than female children. Several studies also reported similar findings, which may be due to female children being less biologically vulnerable to infectious diseases than male children [ 1 , 10 ]. The significantly higher prevalence of malaria among the anemic children was also in line with other studies concerning malaria [ 20 , 25 ].

Higher prevalence was also found among the children whose household heads were less educated. This might be because higher educated heads of household could take more protective measures to reduce exposure that would prevent malaria infection [ 1 ]. Similar to other studies, this study also revealed that rural children experienced more malaria cases than urban children, which might be due to less availability of health care facilities and lack of proper social mobilization concerning malaria prevention [ 26 ]. Malaria prevalence was also higher among poor children than their rich counterparts. A previous study shows that people in poor households in Kenya generally sleep on the floor and are more vulnerable to be infected with malaria. Additionally, because of their low economic means, they are not able to bear the expenses associated with taking preventive action against malaria [ 10 ]. Malaria is also well acknowledged as a disease of poor communities because of their vulnerability and decreased financial means to buy malaria control tools [ 12 ]. Similar to an earlier study, using a net could be an effective preventive measure against malaria among children in Kenya [ 27 ]. The interesting findings noted here concerning the knowledge and attitude about malaria prevention among Kenyans show a contradictory relationship between knowledge and prevalence of malaria. There was a higher percentage of malaria occurrences among the members of households with higher knowledge regarding malaria prevention, revealing a gap between knowledge and practices. A lack of practice of both indoor and outdoor vector control measures is strongly related to higher malaria prevalence [ 13 ]. Another study revealed that although participants had knowledge about prevention strategies against malaria, it was rarely seen in their practices, which supported the significant differences between knowledge and practices of malaria prevention [ 28 ]. One study addressed the reasons behind not using mosquito nets, with discomfort (primarily due to heat) being the most reported reason by participants [ 27 ]. The contradictory results which may be because of the weak association of malaria knowledge with the use of bed nets and ITNs [ 29 ]. Additionally, other socioeconomic and household factors may be responsible for this contradictory relationship between knowledge and inconsistent behavior which leads higher prevalence of malaria [ 29 , 30 ].

Preventive education campaigns are recommended focusing on the translation of knowledge into practices [ 13 ]. Reinforcement of good protective vector control behavior is needed in these circumstances [ 2 ]. This study found that media played an important role in the prevention of malaria as its prevalence is lower among those who have watched television or listened to radio programs addressing malaria intervention programs, which supports the positive findings of the influence of mass media for eliminating malaria in African settings [ 31 ].

There are several limitations of the study. The study used data from a secondary source based on a cross-sectional design and thus had limited opportunities to measure any causal association between malaria and other factors. Information collected from respondents was self-reported and might be affected by recall bias when highlighting knowledge, perception, and practices. Microscopy test may affect the prevalence of malaria, especially in endemic populations with the low transmission of infection [ 20 , 21 , 32 ]. Despite these limitations, this study generates distinctive information regarding determinants of malaria from country representative data, which could be helpful for formulating further steps to implement interventions.

Findings of this study revealed that malaria still remains a public health problem, especially for children under 15 in Kenya. It also demonstrated some significant risk factors with independent effects on the prevalence of malaria among Kenyans. This study also found a gap in translating knowledge into practice to prevent the potential infections. However, improvements in these factors with proper practice of preventive measures might have a positive effect in reducing malarial infection. Based on the findings in the present study, multi-modal programs are needed to control malaria in Kenya. Furthermore, need-based innovative interventions and introducing Behavior Change Communication program (BCC) to prevent and treat malaria are recommended to reduce the health burden caused by malaria. Education and awareness programs are suggested to use existing knowledge in practice to control malaria. Communications should be employed by a combination of radio and television programs, posters at local health facilities or identified public places, the formulation of groups of local stakeholders, and interventions such as the distribution and use of insecticide-treated mosquito nets, especially for households with poor socioeconomic status. These interventions are strongly suggested for prevention of malaria cases in Kenya. This study also suggests taking actions towards income-generating interventions (e.g., poultry raising, farming) among the rural poor community, which can improve their financial means to buy safety measures to control malaria.

Acknowledgements

icddr,b acknowledges with gratitude the commitment of Bill & Melinda Gates Foundation to its research efforts. Current donors providing unrestricted support include Government of the People’s Republic of Bangladesh; Global Affairs Canada (GAC); Swedish International Development Cooperation Agency (Sida), and the Department for International Development (UKAid). We gratefully acknowledge these donors for their support and commitment to icddr,b’s research efforts. We are also grateful to the authority of the Kenya Demographic and Health Survey (KDHS) for permitting and providing nationally representative Malaria Indicator Survey (MIS) data.

Authors did not receive any financial help to prepare and publish this manuscript.

Availability of data and materials

Abbreviations, authors’ contributions.

MS and ARS contributed to conception and designed the study. MS, NS, RAM, TZ, ZI, and ARS contributed to acquisition and drafting the manuscript. MS, ARS, NS, and RAM analyze the data. All authors critically revised the manuscript and gave final approval.

Ethics approval and consent to participate

Data generated by DHS are publicly available. For this study, data were made available to us upon request to Measure DHS. Ethical clearance for conducting the DHS was obtained from the Measure DHS and the Ethics Committee of ICF Macro (Calverton, MD, USA).

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Contributor Information

Marufa Sultana, Phone: +880-2-9827001-10, Email: gro.brddci@afuram .

Nurnabi Sheikh, Email: [email protected] .

Rashidul Alam Mahumud, Email: [email protected] .

Tania Jahir, Email: [email protected] .

Ziaul Islam, Email: gro.brddci@aiz .

Abdur Razzaque Sarker, Email: gro.brddci@euqazzara .

2024 Global Health Equity Challenge

Digital Community Malaria Control

Dr. Yonatan Fialkoff

Our organization.

Zzapp Malaria

What is the name of your solution?

Provide a one-line summary of your solution..

Empowering Women-led Community Efforts with Digital Tools to Combat Malaria-Carrying Mosquitoes

In what city, town, or region is your solution team headquartered?

In what country is your solution team headquartered, what type of organization is your solution team.

For-profit, including B-Corp or similar models

Film your elevator pitch.

What specific problem are you solving.

According to the World Health Organization (WHO) , in 2022 malaria posed a threat to nearly half of the world's population, causing sickness to 209 million people and claiming 608,000 lives. Africa accounts for 95% of all malaria cases and deaths, with the remainder of cases occurring in India (2% of the global burden) and in various countries in South America and Asia. Malaria is the number one cause of death and disease in many developing countries, where pregnant women and children under the age of five are the groups most affected. In addition, it is responsible for the loss of millions of workdays and school days. It has been estimated that malaria severely impedes economic development, to the extent that countries affected by malaria have a per capita income only 30% as high as non-affected countries.

Currently, the most common methods for fighting malaria are insecticide-treated nets (ITNs) and indoor residual spraying (IRS), both of which have significantly decreased malaria cases in the last few decades, but whose effectiveness has been dwindling due to mosquito adaptation to outdoor biting and their growing resistance to the insecticides used in these methods, as well as limited uptake and compliance and other reasons. 

Decades ago, many countries eliminated it using an environmentally benign method known as larviciding, namely, treating the water bodies in which mosquitos breed. However, in order to achieve malaria eradication, it is necessary to find and treat a very high percentage of the millions of water bodies (in many cases, puddles) around people's houses. Studies have shown that even experienced fieldworkers miss about 40% of the water bodies in the areas they are assigned to search. 

Digital tools have significantly enhanced the cost-effectiveness of larviciding operations in urban and semi-urban areas, streamlining the identification and treatment of mosquito breeding sites. However, in rural areas, the scenario is markedly different. The challenges here are compounded by the vast distances between remote communities and operational centers, leading to increased costs related to transportation and the additional working hours field workers spend in transit. While drones have demonstrated potential in improving coverage and detection of breeding sites, they cannot always be applied effectively in these settings and, by themselves, do not constitute a comprehensive solution. Therefore, there is a pressing need for a tailored solution that addresses the unique challenges faced by rural communities. These areas are in particularly acute need of such interventions, as they carry the main burden of malaria, suffering higher incidence rates and facing more significant obstacles in accessing healthcare and preventive services.

What is your solution?

Our solution is a comprehensive initiative designed to empower women in rural areas for effective malaria control through community-based larviciding, leveraging a designated mobile app. This project will train women to utilize the app for identifying and treating mosquito breeding sites, deploying traps and monitoring results. It will harness the pivotal role women play in their communities, particularly in healthcare, by equipping them with the digital tools and skills necessary for sustainable, eco-friendly mosquito control.

The methodology will involve recruiting capable women from the community, providing them with smartphones, and offering condensed training. This training will cover understanding malaria transmission, mosquito breeding habits and the use of WHO-approved larvicides. The app, central to our solution, will employ GPS technology for tracking and reporting larviciding activities and mosquito population data. This information will be uploaded to an online dashboard for remote monitoring by project managers, facilitating real-time support and assessment of the initiative's effectiveness.

By focusing on water bodies as potential breeding sites, women will apply larvicides, cover or dry these areas as needed, incorporating domestic mosquito control advocacy within their communities. Compensation for their work will be based on hours recorded by the app, with additional bonuses for achieving benchmarks in mosquito and malaria reduction. This approach will not only address public health concerns but also promote socio-economic development and empowerment of women in these communities.

Moreover, the project will emphasize gender equality and human rights, ensuring the safety and well-being of participants throughout. It will include pre-training gender assessments, secure training environments, selection of communities with consideration for women's safety and provision of appropriate safety gear for field activities. This initiative is about empowering women with digital skills and tools for malaria control, fostering their economic growth, personal development, and enhancing their roles in healthcare and community leadership.

Who does your solution serve, and in what ways will the solution impact their lives?

Malaria is almost exclusively a disease of the underprivileged. To begin with, it affects developing countries, almost all of which in Africa, and many of which with a large proportion of the population living in extreme poverty (i.e., on less than $1.90 per day). Within these countries, rural communities bear the brunt of malaria, with vulnerable populations disproportionately affected by the disease: children under the age of 5, pregnant women and– more than any other significant health issue–impoverished people . Not surprisingly, malaria is profoundly underfunded. According to the WHO, the annual resources needed in order to reach its malaria control milestones for 2025 were estimated at $6.8 billion in 2020. However, the actual funding was a mere $3.5 billion, and the funding gap is ever growing. Based on a Lancet article from 2016 , the average expenditure of an African country on controlling malaria was $3 PPP. 

The flipside of this grim situation is that even incremental improvements result in multiple lives saved. Indeed, contributing to malaria prevention efforts results in one of the most lives saved per dollar donated . Our solution, which, as mentioned, has been shown to be twice as cost-effective as today's leading anti-malaria intervention, thus holds the potential of saving more than 140,000 lives annually even with existing funding.  The benefits of fighting malaria extend beyond saving lives, since they also contribute to the improvement of education and economic conditions. When children suffer from malaria, they often miss school days, which hinders their educational progress (this is particularly true for girls, who are often required to stay home to take care of sick family members). By reducing the incidence of malaria, children can attend school more consistently and concentrate better in class, ultimately leading to better educational outcomes. In addition, the economic situation of affected communities improves as healthier individuals can work more efficiently, boosting productivity and overall economic growth. Furthermore, in the case of digitally managed field operations to combat malaria, there are additional benefits to the community. Not only do these operations support local labor, but they also promote the acquisition of digital skills. The use of modern technology in these operations helps empower local communities, fostering self-reliance and resilience, while simultaneously providing valuable tools and resources to improve the overall effectiveness of malaria prevention initiatives.

How are you and your team well-positioned to deliver this solution?

Zzapp was founded in 2016 by Arnon Houri-Yafin after he spent three months in hospitals in India conducting research on a malaria blood test. Witnessing the devastating impact of the disease, Arnon decided to use technology not only to diagnose malaria but to eradicate it. Zzapp team members are highly skilled professionals that are deeply committed to the cause and possess years of field experience. Zzapp’s team members hold advanced degrees, including two PhDs, from prestigious universities such as the London School of Hygiene & Tropical Medicine, Georgetown University and the Hebrew University. With a wealth of experience in diverse fields ranging from operation management and data analysis to software development and community engagement, our team has received numerous academic excellence awards and grants. Several team members have either contracted malaria themselves or lost siblings to the illness.

Despite our extensive professional and personal experience, we acknowledge that effective and sustainable solutions require a community-driven approach. We therefore continuously gather feedback through formal and informal methods. Feedback collection, led by our implementation expert, Alexandra Wharton-Smith, starts during the 3-day training that we offer fieldworkers. Based on our experiences in these sessions, we have enhanced our system, guidance materials, and teaching methods. During operations, we encourage managers and fieldworkers to share comments and suggestions via a designated WhatsApp group. We also actively seek their input during field visits, casual conversations, and structured interviews and focus groups. Overall, we have received overwhelmingly positive feedback from both managers and workers. They have reported finding the app useful and even fun (similar to a scavenger hunt or a game of Pokémon GO…). In some instances, workers voluntarily continued working after hours to complete their assigned "chunks," even after being assured that it was not expected of them. Many workers also uploaded photos of themselves standing next to water bodies they had identified, demonstrating a clear sense of pride in their accomplishments.

These are a few of the reviews that we received:

“Despite the technology’s sophistication, the app and the dashboard are very intuitive and user-friendly. In an operation in the Amhara region in 2019, we located all the water bodies, which is usually a great challenge. Fundamentally, the technology helped us in saving time and energy. It also helped us in prioritizing severely affected villages.” (Dr. Abebe Asale, the International Centre of Insect Physiology and Ecology [ICIPE], Ethiopia, 2019  "The app has come at the right time. The app comes in handy in terms of helping us map all the municipality and also find the breeding sites for treatment." (Kwame Desewu, entomologist, AngloGold Ashanti Malaria Control[AGAMal], Ghana, 2017).

Which dimension of the Challenge does your solution most closely address?

Which of the un sustainable development goals does your solution address.

• 1. No Poverty
• 3. Good Health and Well-Being
• 5. Gender Equality
• 10. Reduced Inequalities
• 11. Sustainable Cities and Communities

What is your solution’s stage of development?

Please share details about why you selected the stage above..

We have a field-tested, fully functioning system – in the development of which more than $1M has been invested – whose core principle is applying local data, micro planning and thorough management on a nation-level scale. We continue to improve our tools, refine our working protocols and deepen our entomological, epidemiological and operational knowledge,  As determined by an external consultant for scalability, we are well prepared to scale up to protect hundreds of districts and towns with a population of hundreds of millions across Africa, thanks to several factors: the atomization our system facilitates; the outstanding results we have already achieved; the highly competitive price we offer; our strong ties with research institutions, implementing partners and governments; and high market demand. 

Our forthcoming solution, a women-led larviciding initiative guided by our existing digital app, represents a critical innovation that will enable the scaling of our efforts to rural areas, unlocking the full potential of larviciding in combating malaria on a national level.

Why are you applying to Solve?

As a Business-to-Government (B2G) company, we work with governments, but also with donors, NGOs, and policymakers, including the WHO. The involvement of multiple and varied stakeholders often results in a complex and extended sales cycle. We believe that Solve can help us surmount these challenges by facilitating vital partnerships to accelerate our efforts, validate our business model's impact and sustainability, and ultimately broaden our solution's reach to more communities in need. In addition to addressing these technical and market barriers, we seek networking opportunities to assist in completing a funding round. We have secured a term sheet from an investor (who is actually not an impact investor), contingent on the involvement of another VC in the round. We are confident that Solve can aid in identifying and connecting us with a suitable investor (VC or angel) to participate in our funding round and further advance our mission.

In which of the following areas do you most need partners or support?

• Financial (e.g. accounting practices, pitching to investors)
• Public Relations (e.g. branding/marketing strategy, social and global media)

Who is the Team Lead for your solution?

Arnon Houri-Yafin

What makes your solution innovative?

Our solution is distinguished in the battle against malaria by its integration of approach, functionality and technology, underpinned by a strong emphasis on community-driven efforts.

Approach:  Moving beyond the prevalent reliance on insecticide-treated nets and indoor residual spraying, our strategy focuses on community-led larviciding at outdoor water sources. This method draws on the successes of historical eradication campaigns. By mobilizing community members, in these initiatives, we leverage local knowledge and commitment, significantly boosting the impact and durability of our efforts.

Functionality : Traditional larviciding operations often face challenges with efficient management, which our solution overcomes through the strategic use of mobile technology. Our app facilitates precise coordination of treatment areas and real-time progress monitoring, ensuring comprehensive coverage and the collection of dependable data. The engagement of community members, guided by our technology, ensures an organized and effective approach, markedly improving operational performance.

Technology:  Our system employs cutting-edge neural networks to analyze satellite imagery for accurate mapping of dwellings, then scrutinizes various satellite data to identify areas suitable for mosquito breeding. Armed with this data and an optimization algorithm, we meticulously plan larviciding activities, delegating manageable tasks to community-based teams. Additionally, our spatial-agent-based malaria simulator fine-tunes intervention strategies, enabling each action to achieve maximum efficacy. This technological sophistication supports community operatives, enabling them to make informed decisions, even without internet access and maintain momentum and data accuracy.

By centering our efforts on community involvement, our solution not only innovates in malaria prevention but also empowers local populations, providing valuable skills and job opportunities while tackling a major health issue. This community-centric approach seeks not just to significantly cut down malaria transmission but also to foster broader positive changes, advocating for community empowerment and collective action in public health endeavors. Our model aims to revolutionize the market landscape, showcasing the power of community-led, technologically empowered health interventions on a broad scale.

Describe in simple terms how and why you expect your solution to have an impact on the problem.

Reduce malaria within a large population of rural communities by at least 50% for a cost of no more than 40 cents per person, thereby improving public health, enhancing economic productivity and increasing the quality of life among these populations.

Assumptions 

1. Larviciding is highly cost-effective compared with other mosquito control methods, especially in urban areas ( Worrall & Fillinger, 2011 ).

2. Effective larviciding can significantly reduce malaria ( Damabach et al., 2020 ;  Fillinger et al., 2009 ). 

3. Zzapp's digital tools increase the effectiveness of larviciding operations (Vigodny et al., 2023). 

4. Women play an important role in fighting malaria. 

5. Community involvement increases the effectiveness of larviciding operations.  

1. Financial Resources: Funding for purchasing larvicides, developing and maintaining the digital monitoring platform and covering operational costs. 

2. Human Resources: Recruitment of women as primary agents for identifying and treating mosquito breeding sites. In addition, a dedicated team comprising program coordinators, health educators, data analysts and trainers.  3. Technological Resources: Adaptation of our existing mobile app so as to support remote management of fieldwork in rural communities.  4. Training Materials: Educational and training materials designed to effectively convey information about malaria prevention, the use of the app and the application of larvicides.   5. Community Access: Cooperation with local authorities and community leaders to ensure program access to all areas and households within the target communities.  6. Logistical Support: Transportation and logistical arrangements to facilitate the distribution of larvicides, deployment of training teams and organization of community engagement sessions. 

Program Activities 1. Training Community Members: Women from rural communities will be recruited and trained in using a mobile app and in identifying and treating mosquito breeding sites using eco-friendly larvicides, with a focus on regular and widespread coverage.

2. Digital Monitoring and Data Collection: The program team will utilize digital tools to map breeding sites, track larvicide application and monitor mosquito population dynamics, ensuring real-time data availability and accuracy. 

3. Community Engagement and Education: Health educators and community leaders will conduct awareness sessions on malaria prevention, symptoms and the importance of environmental management.

Outputs 1. Increased Capacity for Malaria Control: A network of trained women capable of identifying and treating mosquito breeding sites.

2. Enhanced Monitoring System: A digital platform that provides real-time data on mosquito populations and larvicide application effectiveness.

3. Improved Community Awareness: Increased knowledge and proactive behavior among community members regarding malaria prevention and environmental management.

Outcomes 1. Reduced Mosquito Population: A decrease in the mosquito population of at least 50% due to the consistent application of larvicides, as evidenced by data collected through the digital monitoring system.

2. Increased Community Engagement: Community members actively participate in malaria prevention efforts.

3. Improved Health Outcomes: A reduction in malaria incidence as community members adopt preventive behaviors and benefit from a decreased mosquito population.

Final Outcome

A sustainable reduction in malaria cases by 50%+ in the targeted rural areas, demonstrating that the technology and community-led approach can be scaled nationally with cost-effectiveness significantly better than existing solutions. 

What are your impact goals for your solution and how are you measuring your progress towards them?

Our primary impact goal is to significantly reduce malaria incidence in rural areas by at least 50% on average, at a cost of less than $0.45 per person protected, demonstrating our solution's scalability and cost-effectiveness. This target reflects our commitment to transforming lives through innovative malaria control strategies, aligning with the UN Sustainable Development Goal (SDG) 3.3 to end the malaria epidemic by 2030. We aim to showcase our technology's potential for national-level application at a cost-effectiveness that doubles that of today's leading solutions. Moreover, our approach is poised to unlock the potential for nationwide malaria elimination at the average expenditure an African country currently allocates to merely control the disease.

Measuring progress towards impact goals :

1. Malaria Incidence Reduction: We rely on official data regarding malaria incidence per 1,000 population, aligning with the UN's indicator for SDG 3.3. This allows us to directly measure our impact on reducing malaria cases in the communities we serve.

2. Mosquito Population Control: Through before, during and after comparisons of mosquito trap catches, we assess the effectiveness of our larviciding operations in reducing the adult mosquito population.

3. Mosquito Larvae Reduction: By comparing water body sampling results from mosquito breeding sites, we gauge the success of our interventions in diminishing mosquito larvae populations.

4. Cost-Effectiveness: The cost per person protected is a critical metric for us, ensuring our solution remains affordable and scalable. Achieving our goal of less than $0.45 per person underscores the financial viability and broader applicability of our approach.

Describe the core technology that powers your solution.

The core of our technology comprises an AI planning tool, a spatial-agent-based mosquito simulator, a mobile app and an online dashboard. First, our AI component extracts from satellite images the location of houses and demarcates the general area for the intervention. Then, integrating data on climate, topography and land use, it runs numerous scenarios in the simulator. Based on predicted effectiveness outcomes, it selects the specific strategy for each village, recommending where to scan for water bodies, which houses to spray, where to place mosquito traps and what implementation order to employ. The system assists with implementing the selected strategy using a GPS-based mobile app that allocates treatment areas to workers, monitors their location in the field to ensure they cover the entire area, and tracks schedules for water body treatment and other tasks. The app was designed to work in Africa’s field conditions: it has a low battery consumption, works well even on simple smartphones and can operate offline. Information from the fieldworkers, as well as that which has been obtained from drones, is automatically uploaded to the dashboard, allowing managers to monitor the operation in real time. The dashboard also produces issue-specific reports – for example, fieldworkers’ productivity or the correlation between water body type and positivity – that improve monitoring of the operation and facilitate post-operation analysis and lesson learning.

Which of the following categories best describes your solution?

A new application of an existing technology

Please select the technologies currently used in your solution:

• Artificial Intelligence / Machine Learning
• GIS and Geospatial Technology
• Software and Mobile Applications

If your solution has a website or an app, provide the links here:

zzappmalaria.com

In which countries do you currently operate?

Which, if any, additional countries will you be operating in within the next year, how many people work on your solution team.

Full time: 7 

Part-time: 4

How long have you been working on your solution?

Zzapp has been engaged in the fight against malaria for seven years. Although we have incorporated community engagement in all of our efforts, we have not yet executed a large-scale operation of the type outlined in the current proposal.

Tell us about how you ensure that your team is diverse, minimizes barriers to opportunity for staff, and provides a welcoming and inclusive environment for all team members.

Diversity serves as a catalyst for innovation. It generates multiple perspectives on issues and fosters the development of creative solutions. Moreover, the ability to understand and empathize with others enables better addressing of the needs of customers from various cultural backgrounds. And, simply, it helps everyone feel comfortable with who they are. As a company operating in diverse cultural settings, with solutions that significantly benefit from community engagement, it is only natural that our team reflects the advantages of diversity. Our team members represent diverse genders, sexual orientations, religions, religious observances and nationalities. In this context, adopting a gender lens when addressing malaria is particularly important. Women are not only disproportionately affected by malaria (e.g., they face higher risks of mortality during pregnancy and often care for sick family members), but they also play a crucial role in malaria control efforts. Women make up 70% of the community health worker force and are instrumental in adopting malaria control practices. These practices range from ensuring that their children receive seasonal preventive medications and use bed nets, to seeking medical care at healthcare facilities when their children fall ill. It has been shown that gender integration improves malaria control. Acknowledging the importance of this perspective, we took part in writing the Gender & Malaria Community of Practice Advocacy Agenda (“Eradication Through Equity: An Advocacy Agenda for a Gender Transformative Approach to the Fight Against Malaria”). We also apply a hiring policy in which if one of two equally competent candidates is a woman, she will receive the position. Indeed, women have been holding leading positions in Zzapp with regards to AI, policy and operations. Gender equality extends beyond hiring and staffing; it also involves creating supportive working conditions. For example, when our employee once had a work-related trip to a field operation in West Africa, Zzapp covered the full costs of flights and accommodations for the baby and her nanny.

What is your business model?

Our business model focuses on providing value to the populations we serve in terms of impact and revenue through comprehensive anti-malaria vector control and surveillance services. We tailor our solutions to each location's specific needs and budgets, offering services at a competitive cost while maintaining effectiveness and profitability. 

We provide our customers, which include governments, municipalities, local authorities and implementing partners, with products and services that support the implementation of operations ranging from the municipal to the nationwide scale. Besides the software, we offer a wide gamut of services, including training, operational management, quality assurance, monitoring and community engagement. Our resources comprise digital technology, materials and equipment, logistical centers, offices, entomological labs, human resources, data and company know-how. 

The gains for citizens in affected countries are immense, as large-scale prevention or elimination of malaria can significantly improve public health, reduce morbidity and mortality and contribute to economic growth. Implementing partners benefit from our AI-generated tailored strategies, automated cost estimation and budgeting, improved management, highly cost-effective outcomes, increased accountability and post-analysis insights, addressing their main pain points in planning, execution and results analysis. 

We establish customer relationships through various channels, such as conferences, embassies, direct engagement and contact with funders like The Global Fund to Fight AIDS, Tuberculosis and Malaria and the President’s Malaria Initiative (PMI). Our pre-sales process involves analyzing customers' budgets and needs and tailoring solutions accordingly. We offer a range of sales options, including district pilots, nationwide elimination, surveillance, and additional services such as modeling and consultation. 

Our services are delivered to citizens in collaboration with governments, local/municipal authorities, and implementing partners. We generate revenue as a B2G company, charging quarterly fees based on the number of licensed workers involved in an operation and the amount and level of services we provide. Our cost-effective solutions enable us to offer competitive pricing to our customers. 

Do you primarily provide products or services directly to individuals, to other organizations, or to the government?

What is your plan for becoming financially sustainable, and what evidence can you provide that this plan has been successful so far.

Our goal is to achieve financial sustainability through a revenue stream from paying customers. Our primary service, accounting for 95% of revenues until EOY 2026 and 81% of gross profit, is nationwide malaria elimination operations—a unique offering in the market. Additional services include district control for low-burden countries, urban larviciding for comprehensive city water body coverage and future platforms to combat mosquito-borne diseases like dengue and Zika.

Our nationwide malaria elimination operations will typically begin with an eight-month, single-district paid pilot to showcase the effectiveness of our method within the client's local context. The pilot helps us tailor our system and methods to the country's specific ecosystem and society, facilitates government decision-making and bidding processes, and aids in obtaining loans from the World Bank. Following a successful pilot, we plan a three-year, two-stage nationwide campaign: a first-year "blanket phase" and a two-year "targeted phase." The final cost and structure depend on factors such as area size, population and climate. 

Based on our experience and mathematical models, we estimate that a typical 3-year nationwide elimination operation will cost the government an average annual $3.33 PPP, from which we will deliver immediate annual governmental cost savings of $1.67 PPP and major additional long-term cost savings. We will form a joint venture together with a selected implementing partner that will manage the whole sum ($300M in the case of a typical 30M population country), with which we will hire labor and transportation (70% of cost), buy aerial imagery and insecticides, run labs, and cover all the other operation costs. We will actively manage all the operational aspects of the operation. For this service, our gross profit will be $0.43 PPP or $39M (13% of the operation's total 3-year cost; according to service providers with whom we spoke, including companies and NGOs, the typical gross profit of service-givers for World Bank-funded projects is 25% of the project’s total expenditure. We aim for gross profit of between only 7% and 17% depending on the type of service and program complexity working in close concert with profit-sharing local implementation partners). 

So far, our efforts to achieve financial sustainability have been met with positive outcomes, as evidenced by various grants, loans, investments and awards that we have received over the years. In 2018, we were awarded a $0.08M non-dilutive grant from the Bill and Melinda Gates Foundation via the Innovative Vector Control Consortium (IVCC). In 2019, we received a $0.18M loan from the Israel Innovation Authority (IIA). Between 2016 and 2019, our parent company, Sight Diagnostics LTD, invested a total of $0.36M in equity. In 2021, we won $3.015M in non-dilutive prize money from the IBM Watson XPRIZE AI for Good competition and an additional $0.25M from the Cisco Global Problem Solver challenge. In 2022, we received a $0.3M grant investment from the Future Fund Regranting Program. We are currently in the process of completing an investment round to further support our financial sustainability efforts. 

Solution Team

The Solve team will review your report and remove any inappropriate content.

Why is this item inappropriate?

We use cookies.

We use cookies and other tracking technologies to improve your browsing experience on our website and to understand where our visitors are coming from. By browsing our website, you consent to our use of cookies and other tracking technologies.

• Open access
• Published: 07 April 2023

Malaria control and elimination in Kenya: economy-wide benefits and regional disparities

• Zuhal Elnour 1 , 2 ,
• Harald Grethe 1 ,
• Khalid Siddig 1 , 3 , 4 &
• Stephen Munga 5  

Malaria Journal volume  22 , Article number:  117 ( 2023 ) Cite this article

8594 Accesses

2 Citations

9 Altmetric

Metrics details

Malaria remains a public health problem in Kenya despite several concerted control efforts. Empirical evidence regarding malaria effects in Kenya suggests that the disease imposes substantial economic costs, jeopardizing the achievement of sustainable development goals. The Kenya Malaria Strategy (2019–2023), which is currently being implemented, is one of several sequential malaria control and elimination strategies. The strategy targets reducing malaria incidences and deaths by 75% of the 2016 levels by 2023 through spending around Kenyan Shillings 61.9 billion over 5 years. This paper assesses the economy-wide implications of implementing this strategy.

An economy-wide simulation model is calibrated to a comprehensive 2019 database for Kenya, considering different epidemiological zones. Two scenarios are simulated with the model. The first scenario ( GOVT ) simulates the annual costs of implementing the Kenya Malaria Strategy by increasing government expenditure on malaria control and elimination programmes. The second scenario ( LABOR ) reduces malaria incidences by 75% in all epidemiological malaria zones without accounting for the changes in government expenditure, which translates into rising the household labour endowment (benefits of the strategy).

Implementing the Kenya Malaria Strategy (2019–2023) enhances gross domestic product at the end of the strategy implementation period due to more available labour. In the short term, government health expenditure (direct malaria costs) increases significantly, which is critical in controlling and eliminating malaria. Expanding the health sector raises the demand for production factors, such as labour and capital. The prices for these factors rise, boosting producer and consumer prices of non-health-related products. Consequently, household welfare decreases during the strategy implementation period. In the long run, household labour endowment increases due to reduced malaria incidences and deaths (indirect malaria costs). However, the size of the effects varies across malaria epidemiological and agroecological zones depending on malaria prevalence and factor ownership.

Conclusions

This paper provides policymakers with an ex-ante assessment of the implications of malaria control and elimination on household welfare across various malaria epidemiological zones. These insights assist in developing and implementing related policy measures that reduce the undesirable effects in the short run. Besides, the paper supports an economically beneficial long-term malaria control and elimination effect.

Malaria remains a public health problem in Kenya despite the scale-up of intervention tools [ 1 ]. Every year, nearly 6.7 million clinical cases of malaria are reported in Kenya, with 70% of the population being at risk of malaria [ 2 ]. It is estimated that approximately 4000 people die from malaria annually, most of them being children. Besides, malaria is responsible for 13–15% of outpatient consultations [ 3 ].

Climate change and farming practices, such as deforestation, are expected to increase malaria incidence [ 4 ]. Malaria transmission and infection risk in Kenya are closely related to altitude, temperature, and rainfall patterns. Consequently, malaria prevalence varies considerably across seasons and regions [ 5 , 6 ]. The country has been classified into five malaria epidemiological zones to address the varied risks, as shown in Fig.  1 . These zones include: (a) coastal endemic, (b) lake endemic, (c) seasonal malaria transmission, (d) malaria epidemic-prone areas of western highlands and epidemic-prone areas, and (e) low-risk malaria areas.

Source: Author's compilation based on [ 6 ]

Malaria epidemiological map for Kenya.

Climatic conditions of the lake endemic and coastal endemic zones are suitable for high survival rates of the malaria vector. As a result, malaria transmission is intense throughout the year in these two regions, with high annual entomological inoculation rates. In 2015, malaria parasite prevalence was 27% and 8% in the lake endemic and coastal endemic zone, respectively [ 6 ].

Malaria transmission in the highland epidemic zone is seasonal, with a considerable annual variation. Transmission intensity increases under favourable climatic conditions for the malaria vector with sustained minimum temperatures around 18° C, which sustain vector breeding. In the regions where malaria occurs regularly, fatality rates during an epidemic can be up to 10 times greater than in the lake endemic and coastal endemic zones. In 2015, malaria parasite prevalence in this zone was 3% [ 6 ].

The seasonal malaria transmission zone includes the arid and semi-arid areas in Kenya's northern and south-eastern parts. It experiences short periods of intense malaria transmission during the rainy season. In 2015, malaria parasite prevalence was approximately 1% [ 6 ]. Under extreme climatic conditions like flooding, the zone is expected to experience malaria epidemics with high morbidity due to the low immunity of the population.

In the low-risk zone, malaria parasite prevalence in 2015 was less than 1% [ 6 ]. Low temperatures in this zone prevent the malaria parasite in the vector from completing the sporogonic cycle. Climate change, e.g., increasing temperatures and changes in the hydrological cycle, is likely to increase the areas suitable for malaria vector breeding, introducing malaria transmission in regions where it did not exist before [ 6 ].

Malaria morbidity and mortality comes along with costs of treatment, control and prevention, thus establishing a substantial economic burden. This results in reduced economic growth mainly by reducing the labour force. Moreover, malaria causes health care spending at private and public levels. Consequently, malaria restrains long-term economic growth and sustainable development.

The interest in economy-wide assessments of the economic costs of malaria and government health intervention policies has recently increased. Such assessments provide insight into the long-run economy-wide effects of malaria on economic growth and development and may support policymakers in adopting measures that would eliminate malaria. Such economy-wide assessment does not yet exist for Kenya.

This paper assesses the economy-wide impacts of the efforts toward malaria control and elimination in Kenya, referring to the objective of the Kenya Malaria Strategy (KMS) (2019–2023). Further, it measures the benefits of declining malaria incidences and deaths, while on the other hand, it captures the costs of increasing government expenditure on malaria treatment, control, prevention and eradication.

Economic costs of malaria

The economic costs of malaria can be classified into direct and indirect costs [ 4 , 7 , 8 ]. First, the direct malaria costs contain a combination of household and government expenditures on treating and preventing malaria, as illustrated in Fig.  2 . Household expenditure on malaria treatment consists of individual or family spending on consultation fees, drugs, transport and the cost of subsistence at a distant health facility, and costs of accompanying family members during hospital stays [ 9 ]. Household expenditure on malaria-related prevention includes costs of buying preventive means, for instance, mosquito coils, aerosol sprays, bed nets and mosquito repellents [ 7 ]. According to the malaria-endemic degree, these means can be used differently across regions and counties.

Source: Authors' compilation based on [ 4 , 7 , 9 ]

Types of economic costs of malaria.

Government expenditure on malaria-related treatment, control and prevention includes spending on maintaining health facilities and health care infrastructure, publicly managed vector control (e.g., insecticide-treated bed nets, indoor residual spraying, larviciding, community-wide campaigning), education and research. The indirect costs of malaria consist of losses in productivity or income due to illness or deaths [ 7 , 9 ], as shown in Fig.  2 . Losses due to sickness can be measured as the cost of lost workdays due to illness or caring for sick family members. In contrast, losses due to deaths can be calculated as discounted future lifetime earnings of those who die.

A study reviewing the empirical evidence on the indirect costs of malaria in Africa found considerable variation across the results of the reviewed studies, depending on the methods used for measuring and valuing time lost [ 7 ]. Indirect costs are often estimated using the wage method. This method defines the costs as the estimated time (workdays) lost multiplied by the working day wage (income). The time cost is determined as the sum of the opportunity cost of time foregone by the sick individual and the opportunity cost of healthy family member's time spent treating or caring for sick persons or accompanying them for treatment.

On average, time lost per episode for a sick adult ranges from 1 to 5 working days. The variation in time loss by episode depends on the prevalence of different malaria species, immunity levels in adults, accessibility to treatment services, type of economic activity, and mode of remuneration [ 7 ]. In Ethiopia, for example, the indirect costs of malaria account for approximately 78% of the total malaria costs incurred by private households [ 10 ].

Several empirical studies have assessed the direct economic costs of malaria in Kenya. One such study estimates that the total economic costs of malaria for children aged 5 years for the year 2009 is about US$ 251 million [ 8 ]. Total direct malaria costs represent 44% of total estimated costs. Malaria treatment expenditure accounts for 27%, of which private households cover about 68%. Indirect costs, including losses due to deaths, account for 57%.

The total cost of malaria hospitalization is approximately US$ 58 per person in Kenya, of which government costs represent 72% [ 4 ]. Total cost of malaria intermittent screening and treatment in school per child in 2010 is estimated at US$ 7 [ 11 ]. About 47% of these costs represent intervention costs, which comprises redeployment of existing resources, including health worker time and hospital vehicle use.

An empirical study that evaluated the effects of malaria on wage income in Kenya, concluded that a 10% increase in malaria prevalence reduced the monthly individual wage income by 3.3% to 3.8% [ 12 ]. Besides, total economic malaria costs represent (on average) 1% of total household income [ 13 ]. Moreover, nearly 170 million working days are lost annually due to malaria in Kenya [ 14 ].

Since 2004, The Kenyan government has implemented sequential Kenya malaria strategies to control and eliminate malaria [ 6 ]. Each strategy is developed based on the recommendations and evaluation of the malaria programme review of the previous strategy. The shared vision of these strategies is to free Kenya from malaria by directing and coordinating efforts through effective national and international partnerships. The current KMS (2019–2023) aims to reduce malaria incidence and deaths by at least 75% of the 2016 levels by 2023. To achieve this target, approximately Kenyan Shillings (Ksh) 61.9 billion will be spent in total over a period of 5 years on malaria control and elimination programmes.

Depicting malaria effects in CGE models

Computable General Equilibrium (CGE) models depict the economy as a whole as a system of equations. These cover production (based on the standard assumption of profit maximization) and consumption (based on the standard assumption of utility maximization) of goods and mechanisms governing the economy as a whole such as the balance of government income and expenditure and a balanced exchange with the rest of the world. Such models allow for assessing the economic implications of complex and simultaneous changes in exogenous variables such as health expenditure and the labour force due to changes in human health. Especially, they can depict indirect effects on sectors and household groups. Such indirect effects, which are mediated through changes in product and factor prices, are often important in case of shocks affecting the economy as a whole. Several empirical studies attempt to estimate the economic costs of malaria, particularly in developing countries [ 7 , 15 , 16 , 17 ]. Nevertheless, few CGE studies assess the economy-wide implications of malaria [ 18 , 19 , 20 , 21 ].

Climate change-induced changes in human health through malaria and other selected diseases are for example assessed using a global CGE model, namely the Global Trade Analysis Project (GTAP) model [ 21 ]. This study captures the effects of malaria on labour productivity and the demand for health care. Malaria affects labour productivity by changing mortality and morbidity. These effects are incorporated in the model as exogenous shocks. The changes in childhood mortality are determined by changes in the prevalence of vector-borne diseases resulting from a one-degree increase in global mean temperature. The relative annual loss of labour productivity equals the number of additional malaria deaths plus the additional years of working time lost, divided by the total population. In addition, malaria's effects on health care demand are captured in the model by changing the productivity of health services for private and public final demand.

Another CGE-based study evaluates the health and economy-wide impacts of malaria transmission in Ghana using an integrated epidemiological-demographic CGE model [ 20 ]. The epidemiological component depicts malaria infections and prevalence and calculates clinical outcomes for infected individuals. The clinical outcomes are used to estimate the effects of malaria on mortality and morbidity rates. These rates are applied in the demographic component for calculating changes in demographic structure due to malaria transmission. The demographic component classifies population by age group and gender type. It also includes international and interregional migration specifications. The CGE component is a recursive-dynamic model, which explicitly covers capital accumulation in different sectors over time. The key link between the three components is the determination of the labour force and ownership by population demographics based on two malaria-related morbidity rates. These are (a) the rate of female adults caring for sick children and (b) the rate of sick adults. Morbidity effects on the labour force are determined as the affected gender-specific working-age population group multiplied by gender-specific labour market participation rates, labour factor skill shares, rates of reduction in annual labour supply per malaria episode, and the average number of malaria episodes per person per year. The skill shares and participation rates are estimated using secondary data, i.e., labour force data from household surveys or the World Development Indicator database.

In contrast, the rate of reduction in annual labour supply per malaria episode is estimated endogenously in the CGE component. It is a function of intervention effective coverage rates and fixed morbidity rates associated with and without effective intervention treatment. Intervention effective coverage rates are defined by private and public malaria-related composite intervention commodities, underlying regional population levels in the case of prevention interventions and by number of uncomplicated episode cases.

This paper uses a static CGE model to assess the economy-wide impacts of the efforts to control and eliminate malaria in Kenya, following the recommendations from the Kenya Ministry of Health (KMoH) [ 6 ]. This policy benefits the Kenyan economy by decreasing the indirect costs of malaria, i.e., a reduction in malaria incidences and deaths. As a result, household labour endowment increases, which is incorporated in the model as an exogenous increase in household labour supply. In addition, implementing this policy increases the direct costs of malaria. Hence, the government expenditure on malaria control and elimination is expanded, which is depicted as an exogenous increase in government expenditure on health services.

A social accounting matrix (SAM), which is an economy-wide database, is developed for Kenya for the year 2019 [ 22 ]. The SAM contains data related to malaria epidemiological and agroecological zones, as shown in Fig.  3 . These zones include: (a) arid and seasonal transmission, (b) coastal endemic, (c) high rainfall and highland epidemic, (d) high rainfall and lake endemic, (e) high rainfall and low epidemic, (f) high rainfall and seasonal transmission, (g) semi-arid and coastal zone, (h) semi-arid and seasonal transmission, (i) semi-arid and highland epidemic, and (j) semi-arid and low-risk epidemic.

Source: Author's compilation based on [ 6 , 21 ]

Kenyan malaria epidemiological and agroecological zones included in the 2019 SAM.

The SAM is an update and extension of a 2017 SAM for Kenya [ 23 ] based on data from a 2019 SAM for Kenya [ 24 ] and domestic sources, including the Kenya National Bureau of Statistics (KNBS) [ 25 ], the KMOH [ 6 ], and the Central Bank of Kenya (CBoK) [ 26 ].

The 2019 micro-SAM for Kenya has 186 accounts representing the Kenyan economy. It identifies 51 production activities, of which 22 are agricultural, 19 are industrial, and the rest and the rest are services.

The SAM has 34 production factors: two capitals (agricultural and not), two lands (irrigated and not) and 30 labour categories. Labour is classified based on skill level into three categories: skilled, semi-skilled and unskilled labour. The three labour categories are regionalized based on the ten malaria epidemiological and agroecological zones (Fig.  3 ).

Households are disaggregated into 40 representative groups using three criteria: (a) ten malaria epidemiological and agroecological zones, (b) two residence places (rural and urban), and (c) two income levels (poor and non-poor). The remaining nine accounts consist of enterprises, trade and transport margins, government, four tax accounts (i.e., sales tax, import tax, production tax and income tax), one capital account (savings and investment), and the rest of the world account.

STAGE, a static CGE model, is used to assess the economy-wide impacts of malaria control and elimination in Kenya [ 27 ]. This type of model evaluates the effect of policy changes by comparing the equilibrium state of the economy before and after the reform. However, it does not show the process of the economy's transition from the initial equilibrium to the new equilibrium after a shock.

STAGE is based on standard microeconomic theory: productive activities maximize profits, and consumers maximize utility. Production is modelled as a three-level system of nested Constant Elasticity of Substitution (CES) and Leontief production functions. On the top level, activities combine aggregate primary production factors and aggregate intermediate inputs using a CES function. The different groups of production factors are aggregated using CES functions at different levels, while the intermediate input component is aggregated using a Leontief production function. Producers decide to sell their products either in the local market or the export market depending on relative prices according to a Constant Elasticity of Transformation (CET) function. Households supply their production factors to productive activities through factor markets (e.g., the labour market) against wages, which constitute a major source of their incomes. They spend their income on purchasing goods and services after paying taxes and making savings. The demand system is derived from a Stone-Geary utility function whereby households choose optimum mixes of commodities and services subject to their purchase prices and the constraints of preferences and income.

Simulation design

Against the model base representing the Kenyan economy in 2019, two counterfactual scenarios ( GOVT and LABOR ) are developed to depict the effects of the KMS (2019–2023).

First, the GOVT scenario assesses the isolated effects of increasing the Kenya government expenditure on malaria control and elimination programs without considering the impacts on the labour force. The Kenyan government plans to spend around Ksh 12.4 billion annually over 5 years. This policy increases the total government expenditure on health services by 7% annually. This additional expenditure represents 0.9% of total government expenditure and 0.1% of current Gross Domestic Product (GDP). This scenario depicts a typical year during the "investment phase" of the strategy.

Second, the LABOR scenario simulates the effects of reducing malaria incidences by 75% in all epidemiological malaria zones without accounting for the changes in government expenditure. This reduction is expected to happen by the end of the policy implementation period. This scenario thus depicts a typical year during the "pay-off phase" of the strategy, which comprises the benefits of reduced malaria after the increased government expenditure in GOVT . The LABOR scenario translates the reduction in malaria into increasing working days by boosting household labour endowment, as explained in Table 1 , based on a set of assumptions. These assumptions are developed using a database and reports for Kenya from Demographic and Health Surveys (DHS) produced by the United States Agency for International Development (USAID).

Malaria prevalence across different household groups in Kenya is presented in Table 2 . Numbers in this table are calculated using the malaria indicator survey database for the year 2020 [ 28 ]. The table shows that rural (poor and non-poor) households lose more due to malaria than urban (poor and non-poor) households. This can be attributed to the availability and easy access to health services in urban areas [ 29 ]. Skilled households are less affected by malaria compared to semi-skilled and unskilled household groups (Table 2 ). This can be attributed to the access of skilled households to knowledge on prevention methods and incomes able to afford the costs of malaria treatment and prevention means [ 29 ].

The changes in household labour endowment due to implementing the KMS (2019–2023) are calculated in Table 1 . The relative malaria prevalence across household groups is assumed to be similar across different malaria epidemiological zones (columns 1–3). Malaria prevalence across different malaria epidemiological zones (column 4) and malaria prevalence at the national level (column 5) is obtained from the Kenya malaria indicator survey report for the year 2015 [ 30 ]. Reduction in malaria prevalence is obtained from the current implemented strategy documentation (column 6) [ 6 ].

Based on the above information, changes in household labour endowment are estimated (columns 7–9), as shown in Table 1 . For instance, rural poor skilled households in the lake endemic zone lose 7.54% of total working days due to sickness or taking care of a sick family member. This figure is calculated by first multiplying malaria prevalence in the corresponding household group (2.98%) by the ratio of malaria prevalence in the lake endemic zone (27%) over the national malaria prevalence average of 8%. The resulting figure is multiplied by 75%, generating the reduction in lost working days due to malaria at the end of the KMS implementation period.

The differences across labour categories in the increase in labour service availability in response to malaria control and elimination (Table 3 ) are associated with regional and household differentiation. This can be explained by regional variation in skill composition of the labour force. Besides, the scenario considers differences in effects on skill level categories (see Table 2 ), which result from household groups being affected differently due to socioeconomic characteristics (e.g., poor/non-poor). As a result, the provision of labour services increases most for unskilled labour, the most affected labour category by malaria in Kenya.

Closure rules

The scenarios are implemented under the following closure rules. These closure rules reflect the main characteristics of the Kenyan economy. First, the macro closure includes a savings-driven neoclassical approach. This closure describes the relationship between total saving and total investment in the economy. Savings rates of households and enterprises are assumed to be constant, allowing total savings to change according to income changes. Consequently, total investment spending changes to accommodate changes in total savings. Second, the government closure holds the value of government consumption expenditure constant at its initial level. Besides, government savings are fixed in absolute terms and household income taxes vary to clear the government account. Third, the factor market closure assumes full employment of factors in all markets. In addition, all factors are assumed to be mobile across sectors, with fixed overall factor supply and wage rates clearing the market. Fourth, the small country assumption is used to fix world market prices. Besides, the external balance (foreign savings) is kept constant by a flexible exchange rate. Last, the CPI is the model numéraire.

This section reports simulation results as changes in values of model variables relative to their values in the reference scenario. After presenting the effects on factor prices and quantities of domestic production, it discusses the effects on household welfare. The section ends with a sensitivity analysis showing how different choices of financing instruments affect the simulation results.

Factor prices

The effects of counterfactual scenarios on factor prices are illustrated in Fig.  4 . The GOVT scenario captures the annual cost of the KMS (2019–2023), expanding government expenditure on malaria control and elimination programs. This scenario increases labour wages, which is strongest for skilled labour, and is caused by the expansion of the health sector being labour-intensive and especially skilled labour-intensive sector. The expansion of the health sector is funded by increasing taxes and results in overall lower domestic consumption and production. As a result, total demand for capital and land decreases, driving down their prices.

Source: Author's calculations based on simulation results

Effects on factor prices (% change compared to the reference scenario).

The LABOR scenario depicts the benefit of the KMS (2019–2023) in terms of increased labour supply. As a result, wages decline relative to the reference scenario. The wage for unskilled labour drops more than for other labour categories because of its higher increase in supply (shock structure). Simultaneously, prices for complementary factors (i.e., capital and land) increase, driven by a relative supply shortage and increase in demand.

Domestic production

Effects of the two scenarios on quantities of domestic production vary across sectors, as shown in Fig.  5 . The variation across sectors is explained by differences in their cost structure. The GOVT scenario boosts government services production by about 0.5%, driven by expanding expenditures on malaria control and elimination program. However, the expansion of health services is financed by increasing household tax payments. This increase reduces the domestic demand and production of crops, livestock, fishing, and private services. Production of manufacturing, water, electricity, and construction increases because they are important intermediate inputs in the health sector. In contrast, the reduction in capital price and the growth of construction increases slightly production of forests and mining, which are capital intensive sectors and significant intermediate inputs in construction sector.

Effects on quantities of sectoral domestic production (% change compared to the reference scenario).

Under the LABOR scenario, reducing malaria prevalence and incidence drives up sectoral domestic production in Kenya driven by increased total labour supply and lower production costs. Production of crops, construction, and services (i.e., trade and administration) increase more than in other sectors, as shown in Fig.  5 , because these sectors have the most prominent labour share from the lake and coastal endemic zones, the two regions that benefit the most from malaria elimination. Production of forests and mining decreases under this scenario due to increased prices of non-labour factors, which are used intensively by these two sectors.

• Household welfare

Figure  6 presents the effects of the two scenarios on household welfare, which are measured using the Equivalent Variation (EV). The welfare effects vary across scenarios and representative household groups. This variation can be explained by differences in income sources, consumption patterns, and tax payments.

Effects on household welfare (EV as a share of household expenditure in the base situation).

The GOVT scenario expands health services, which increases its demand for production factors, as illustrated earlier in Fig.  4 . In contrast, production costs of non-health sectors increase, driving up their producer and consumer prices. Moreover, because the implemented policy is financed by increasing household income tax payments, disposable household income decreases. Consequently, the welfare of all households drops. Additionally, poor households in rural and urban areas lose more than non-poor households in relative terms (Fig.  6 ), which can be attributed to their lower income shares from skilled labour, for which wages increase most (Fig.  4 ).

In the LABOR scenario, controlling and eliminating malaria increases labour supply and lowers wages. This boosts household income because the increase in labour supply is more significant than the reduction in wage rates. It also boosts domestic production because it reduces the cost of production for most sectors. Subsequently, total household welfare increases (Fig.  6 ).

Figure  6 illustrates that rural household welfare increases more than urban household welfare in the LABOR scenario. This difference is driven by the growth in labour supply, which has a higher share in rural income than in urban income. Furthermore, this scenario boosts poor household income more than non-poor household income, due to higher malaria incidences amongst poor households (shock structure). Rural non-poor households lose because of high-income tax payments and a low increase in their factor incomes, particularly labour incomes.

The effects on household welfare across malaria epidemiological and agroecological zones are presented in Fig.  7 . It shows that the GOVT scenario decreases household welfare in all malaria epidemiological and agroecological zones due to higher income tax payments and product prices compared to the base situation. In the LABOR scenario, lake and coastal endemic households benefit more than others due to the increase in their factor income driven by high labor supply growth. In contrast, the welfare of households in the highland epidemic, low epidemic, arid and seasonal transmission zones improve mainly due to the increased income from the complementary factors of land and capital.

Effects on household welfare across malaria epidemiological and agroecological zones.

Sensitivity analysis

A sensitivity analysis is conducted to assess the choice of instrument to finance the malaria control and elimination interventions under the GOVT scenario. It examines the model's robustness and the sensitivity of results to variations in the closure rules and assumptions. In addition to the household income tax, two alternative instruments are chosen to finance government expenditure on health services ( GOVT scenario): a sales tax and foreign transfers to the government.

Figure  8 illustrates the sensitivity analysis results, taking household welfare as an example. It shows that the magnitude of the results varies only slightly among different tax instruments. Under the foreign transfer to government instrument, all households benefit slightly. This is plausible as domestic taxpayers do not need to fund government interventions.

Sensitivity analysis of different instruments for financing the implemented policy on household welfare.

This study found that implementing the Kenya Malaria Strategy 2019–2023 influences the Kenyan economy in two different and often opposite ways. On the one hand, it increases household labour endowment due to reduced malaria prevalence, which positively influences the economy by increasing domestic production. At the aggregate level, agriculture and services, both labour-intensive sectors, would benefit more than industry (non-labour-intensive). At the sectoral level, sectors that use labour intensively in high endemic zones, such as the crop sector, would benefit more than other sectors. The effects on household welfare would be positive, though they vary across malaria epidemiological and agroecological zones due to the differences in malaria relevance and labour ownership. Consequently, the real GDP increase is driven by the growth in the total labour force.

Implementing this strategy requires expanding government expenditure on health services. Expanding health services increases the demand for production factors and thus factor prices. Producer and consumer prices for products from non-health sectors increase. As a result, the consumption patterns of domestic consumers change. However, the increase in consumer prices exceeds the increase in household income. Hence, household welfare declines. In contrast, the real GDP increases slightly, driven by the significant growth in government consumption.

The sensitivity analysis regarding the financing instruments for expanding health services shows that tax-based financing instruments reduce the welfare of all household categories. In contrast, financing the strategy through increasing foreign transfers to government benefits all households slightly.

Finally, some suggestions for future research in the field are highlighted, which are related to five shortcomings of this study. First, the study focuses on incorporating the benefits and costs of the malaria control and elimination strategy in a comparative static model setup, not considering the time path throughout the implementation process. Such an analysis does not adequately depict the time path from short- to long-run effects. For instance, the adverse effects of expanding government consumption at the expense of private consumption and investment demand are expected to fade out after the end of the spending period. The growth of household labour endowment due to reduced malaria prevalence phases in stepwise and is a permanent effect.

Second, the first objective of the KMS (2019–2023) is to protect 100% of people living in endemic malaria zones through access to appropriate preventive interventions by 2023 [ 6 ]. Consequently, the Kenyan government plans to spend 67% of additional malaria control and elimination expenditure on scaling up malaria prevention and control interventions, e.g., distribution of long-lasting insecticidal nets, indoor residual spraying, larval source management, and establishment of documents for malaria vector control. The rest of the additional elimination expenditure (33%) is planned to be spent on activities related to malaria treatment, elimination, and management. Nevertheless, expenditure on malaria control and prevention programs is aggregated with other health services in our model database. This implies that additional malaria control and elimination expenditures cannot be depicted with adequate detail. A detailed database covering the individual health-related sectors would enhance the analysis.

Third, the study does not incorporate the effects on mortality rates, which prevents capturing changes in population demographics. This would require linking the CGE model to a demographic model to analyse the impacts of demographic and health condition changes on the labour force, as suggested by Jensen et al. [ 18 ]. Future research could benefit from such a model combination.

Fourth, the study does not capture the negative effects of malaria on children's education and school outcomes. This kind of assessment could require including different types of educational cycles in the model and its database and linking labour force via educational outcomes to trace the changes in the educational outcomes. It also would require model specifications such as dynamic CGE model, which considers time dimension and assumes that behaviours of firms and households are derived from intra- and intertemporal optimization. Last, neither variations in labour income due to malaria control and elimination according to the type of employment nor gender were integrated. For instance, people employed in non-permanent jobs, e.g., self-employed in agricultural sectors, may lose more income in high malaria endemic zones than those with permanent jobs because of losses in payment for absent working days due to malaria. To depict such effects would require disaggregating labour according to employment type in each zone.

This paper applied an economy-wide (CGE) model for assessing the economy-wide implications of malaria control and elimination in Kenya, considering regional malaria disparities. Two scenarios are developed based on the costs and benefits of implementing the KMS (2019–2023). The first scenario captures the annual costs of decreasing malaria prevalence by increasing government expenditure on malaria control, treatment and prevention programs (the direct malaria costs), which would prevail for 5 years. The second scenario depicts the benefits of implementing the KMS by increasing household labour endowment (reducing the indirect malaria costs). The results show that applying the KMS (2019–2023) enhances overall economic performance as measured by growth in GDP at the end of the strategy implementation compared to the reference scenario without the implantation of the strategy. In terms of private household welfare, more than 10 years would be needed to compensate for the investment period through a higher labour endowment.

Although the paper does not capture the specific time path of costs and benefits, it provides policymakers with an ex-ante assessment of the implications of malaria control and elimination on household welfare across various epidemiological malaria zones. These insights could assist in developing and implementing related policy measures that reduce the negative effects in the short term, e.g., increasing subsidies or social transfers for those who are more negatively affected by the strategy in the short run.

Availability of data and materials

The datasets generated and analysed are publicly available.

Abbreviations

Central Bank of Kenya

Constant Elasticity of Substitution

Constant Elasticity of Transformation

Computable General Equilibrium

Demographic and Health Surveys

Equivalent Variation

Gross domestic product

Global Trade Analysis Project

Kenya Malaria Strategy

Kenya Ministry of Health

Kenya National Bureau of Statistics

Kenyan Shillings

Social accounting matrix

United States Agency for International Development

World Health Organization

Amboko B, Stepniewska K, Macharia PM, Machini B, Bejon P, Snow RW, et al. Trends in health workers’ compliance with outpatient malaria case-management guidelines across malaria epidemiological zones in Kenya, 2010–2016. Malar J. 2020;19:406.

Article   CAS   PubMed   PubMed Central   Google Scholar  

Sultana M, Sheikh N, Mahumud RA, Jahir T, Islam Z, Sarker AR. Prevalence and associated determinants of malaria parasites among Kenyan children. Trop Med Health. 2017;45:25.

Article   PubMed   PubMed Central   Google Scholar  

Division of National Malaria Programme. Kenya Malaria Programme: Mid-Term Review 2022. Nairobi: Kenya Ministry of Health; 2022.

Google Scholar  

Watts C, Atieli H, Alacapa J, Lee M-C, Zhou G, Githeko A, et al. Rethinking the economic costs of hospitalization for malaria: accounting for the comorbidities of malaria patients in western Kenya. Malar J. 2021;20:429.

USAID. President’s Malaria Initiative: Kenya Malaria Operational Plan Fy 2019. U.S. Global Malaria Coordinator and the national malaria control programs and partners. 2018.5.

National Malaria Control Programme. Toward a malaria free Kenya: Kenya malaria strategy 2019–2023. Nairobi: Ministry of Health; 2020.

Chima RI, Goodman CA, Mills A. The economic impact of malaria in Africa: a critical review of the evidence. Health Policy. 2003;63:17–36.

Article   PubMed   Google Scholar  

Sicuri E, Vieta A, Lindner L, Constenla D, Sauboin C. The economic costs of malaria in children in three sub-Saharan countries: Ghana, Tanzania and Kenya. Malar J. 2013;12:307.

WHO. Economic costs of malaria. Geneva, Roll Back Malaria, World Health Organization. 2011;20.

Tefera DR, Sinkie SO, Daka DW. Economic burden of malaria and associated factors among rural households in Chewaka District, Western Ethiopia. Clinicoecon Outcomes Res. 2020;12:141–52.

Drake TL, Okello G, Njagi K, Halliday KE, Jukes MC, Mangham L, et al. Cost analysis of school-based intermittent screening and treatment of malaria in Kenya. Malar J. 2011;10:273.

Kioko UM, Mwabu G, Kimuyu P. The economic impact of malaria on wage earnings in Kenya: a household level investigation. Int J Educ Res. 2013;1:4.

Gatei CT. Economic costs of malaria in Kenya. Master thesis. School of economics, University of Nairobi. 2006.

World Bank. The World Bank Booster Program for Malaria Control in Africa: scaling-up for impact (SUFI) a two-year progress report; 2007.

Singh MP, Saha KB, Chand SK, Sabin LL. The economic cost of malaria at the household level in high and low transmission areas of central India. Acta Trop. 2019;190:344–9.

Willis DW, Hamon N. Evidence for the impact of malaria on agricultural household income in sub-Saharan Africa. Gates Open Res. 2019;3:9.

Article   Google Scholar  

Willis DW, Hamon N. Eliminating malaria by 2040 among agricultural households in Africa: potential impact on health, labor productivity, education and gender equality. Gates Open Res. 2018;2:33.

Smith RD, Keogh-Brown MR, Chico RM, Bretscher MT, Drakeley C, Jensen HT. Will more of the same achieve malaria elimination? Results from an integrated macroeconomic epidemiological demographic model. Am J Trop Med Hyg. 2020;103:1871–82.

Yerushalmi E, Hunt P, Hoorens S, Sauboin C, Smith R. Exploring the use of a general equilibrium method to assess the value of a malaria vaccine: an application to Ghana. MDM Policy Pract. 2019;4:2381468319894345.

PubMed   PubMed Central   Google Scholar  

Jensen HT, Keogh-Brown M, Smith R, Bretscher M, Chico M. Malaria in Ghana: integrated macroeconomic and epidemiological-demographic impact assessment. 16th Annual Conference on Global Economic Analysis, Shanghai, China. 2013.

Bosello F, Roson R, Tol RS. Economy-wide estimates of the implications of climate change: Human health. Ecol Econom. 2006;58:579–91.

Elnour Z, Siddig K, Grethe H. A 2019 Social accounting matrix for Kenya including malaria epidemiological and agroecological zones. Working paper no. 103/2022. Department of Agricultural Economics, Faculty of Life Sciences, Humboldt-Universität zu Berlin. 2022.

Ferrari E, Mainar Causapé A, Jiménez Calvo S. Social Accounting Matrix of Kenya 2017. European Commission, Joint Research Centre (JRC). PID: http://data.europa.eu/89h/bc924665-e888-4642-9b3b-297f4f6b13d3 . 2020.

Thurlow J. 2019 Social Accounting Matrix for Kenya: A Nexus Project SAM. IFPRI-Data; International Food Policy Research Institute (IFPRI) 2021.

Kenya National Bureau of Statistics. Economic survey 2021. Nairobi, Kenya; 2021.

Central Bank of Kenya. Diaspora Remittances Survey. Nairobi, Kenya. 2021.

McDonald S, Thierfielder K. A Static Applied General Equilibrium Model: Technical Documentation: STAGE Version 2. 2015:1–87.

DHS. Malaria indicator survey database for Kenya for the year 2020. http://www.dhsprogram.com . Accessed 8 Mar 2022.

Chuma JM, Thiede M, Molyneux CS. Rethinking the economic costs of malaria at the household level: evidence from applying a new analytical framework in rural Kenya. Malar J. 2006;5:76.

National Malaria Control Programme (NMCP), Kenya National Bureau of Statistics (KNBS), and ICF International. Kenya malaria indicator Survey 2015. Nairobi, Kenya, and Rockville, USA. 2016.

Division of National Malaria Programme, Ministry of Health, Kenya National Bureau of Statistics. Kenya malaria indicator survey 2020. Nairobi, Kenya. The DHS Program, Rockville, USA; 2021.

Causapé AJM, Boulanger P, Dudu H, Ferrari E, McDonald S. Social accounting matrix of Kenya 2014. EUR 29056 EN, JRC Technical Reports. Publications Office of the European Union, Luxembourg, 2018.

Download references

Acknowledgements

The authors thank the editor and reviewers for helpful comments and suggestions and the members of the DFG Research Unit Climate Change and Health in Sub-Saharan Africa for all the discussions on malaria impact pathways.

Open Access funding enabled and organized by Projekt DEAL. This work was funded by Deutsche Forschungsgemeinschaft (DFG) as part of the Research Unit FOR 2936: Climate Change and Health in Sub-Saharan Africa.

Author information

Authors and affiliations.

Humboldt-Universität zu Berlin, Berlin, Germany

Zuhal Elnour, Harald Grethe & Khalid Siddig

Agricultural Research Corporation, Wad Madani, Sudan

Zuhal Elnour

International Food Policy Research Institute, Khartoum, Sudan

Khalid Siddig

University of Khartoum, Khartoum, Sudan

Kenya Medical Research Institute, Kisumu, Kenya

Stephen Munga

You can also search for this author in PubMed   Google Scholar

Contributions

ZE developed the analytical model, constructed the database, analysed the data, and drafted the manuscript. ZE and HG developed the simulations. HG and KS checked the result's robustness and revised manuscript drafts. SM provided substantial scientific input to the study by reviewing the design of the simulations and the full content. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Zuhal Elnour .

Ethics declarations

Ethics approval and consent to participate.

This study was approved by the Kenya Medical Research Institute, Scientific and Ethics Review Unit (KEMRI/SERU/CGHR/327/3962): Climate Change and Health in Sub-Saharan Africa.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

See Fig.  9 .

Changes in labour supplies in the LABOR scenario (% change compared to the reference scenario).

See Fig.  10 .

Effects on labour price across malaria epidemiological and agroecological zones, Source: Author's calculations based on simulation results

See Fig.  11 .

Effects on domestic producer prices (% change compared to the reference scenario).

See Fig.  12 .

Effects on aggregate economy-wide indicators (% change compared to the reference scenario).

See Table 4 .

See Table 5 .

See Table 6 .

See Table 7 .

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

Elnour, Z., Grethe, H., Siddig, K. et al. Malaria control and elimination in Kenya: economy-wide benefits and regional disparities. Malar J 22 , 117 (2023). https://doi.org/10.1186/s12936-023-04505-6

Download citation

Received : 04 September 2022

Accepted : 18 February 2023

Published : 07 April 2023

DOI : https://doi.org/10.1186/s12936-023-04505-6

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Economic performance
• CGE analysis
• Human health
• Public health policy

Malaria Journal

ISSN: 1475-2875

• Submission enquiries: [email protected]