A call for action to change age structure in Kenya´s health data management system to enhance malaria surveillance and interventions

¹Department of Biostatistics and Health Data Science, Indiana University School of Medicine, Indianapolis, Indiana, United States of America, ²Centre for Community Health and Wellbeing, Jaramogi Oginga University of Science and Technology, Bondo, Kenya

^&Corresponding author
Felix Manuel Pabon-Rodriguez, Department of Biostatistics and Health Data Science, Indiana University School of Medicine, Indianapolis, Indiana, United States of America

Perspectives    

Malaria is a communicable disease transmitted to humans by infected female Anopheles mosquitoes and caused by the Plasmodium parasites. Despite all efforts made to reduce malaria infections and associated deaths, this mosquito-borne disease remains a public health concern in sub-Saharan Africa (SSA). At least 50% of the world´s population is at risk of malaria [1]. Malaria is preventable, but if untreated, it can lead to several health complications and eventually death. Several successful interventions exist that have shown reductions in malaria prevalence. These intervention methods include case management, early diagnosis and treatment, health education, use of malaria vaccine (region- and age-specific), intermittent preventive treatment in pregnancy, and vector control strategies such as mass distribution of long-lasting insecticide-treated nets, indoor residual spraying, and larval source management to reduce mosquito breeding sites [1]. Even with this comprehensive list of intervention and control measures, SSA bears the highest burden of malaria worldwide, with 94% of cases and 95% of deaths globally [1].

For decades, the primary targeted groups for many of these preventative and intervention programs have been pregnant women and children under five years old. However, research studies carried out in SSA [2,3], including a study done by some authors of this manuscript, focusing on Siaya County in Western Kenya [4], report an age-specific epidemiological shift in the risk of malaria towards school-age children. Other studies in Uganda have also observed similar patterns [5]. School-age children, defined as children in the range of 5 to 16 years old, have recently been suffering more from malaria infections compared to their counterpart group (under 5 years old), and in some regions, this also applies to adults (>16 years old) [5]. Across SSA, malaria parasite prevalence among school-age children usually ranges from about 20% to over 50% in moderate-to-high transmission areas, often matching or surpassing that in children under five [6]. As malaria control reduces early-childhood mortality and incidence, the median age of infection increases, shifting residual clinical burden toward older children and adolescents [7]. In mainland Tanzania, multilevel survey analysis of 2015-2023 data demonstrated persistent infection among school-age children, with adjusted odds ratios for infection exceeding unity relative to younger children in multiple survey years and regional prevalence remaining above 15-20% in higher transmission zones [8]. These quantitative patterns highlight the limitations of binary <5 vs ≥5 surveillance categories. New research and tools should be adapted to aid researchers and users in continuing their work towards malaria elimination.

In this paper, we make a call to action directed to the Kenya Health Information System (KHIS) development team, to make modifications to their platform to easily allow extraction of data in user-specified strata, to reflect developments in research for enhanced surveillance, and focused or targeted interventions, specifically to allow malaria researchers to define and extract data using user-defined ranges of ages unrelated to current standard guidelines. Beyond a technical enhancement, this change represents a step toward more adaptive, evidence-driven, and equitable surveillance infrastructures in SSA.

Age-specific malaria risk shift

In recent studies, increased evidence from many SSA countries shows that school-age children (5-16 years old) have become a group at high-risk for malaria infection [6,8,9,10,11]. New or improvement of existing interventions may be needed to address this shift in malaria burden towards this group and to reduce malaria prevalence. Also, the reporting guidelines might also need to change to reflect the shift of the malaria burden. For instance, children in SSA are the most vulnerable to malaria infections and mortality, but school-age children carry a higher yet overlooked burden of malaria [11]. In 2010, it was estimated that over 500 million school-age children were at risk of malaria worldwide, with 200 million in SSA [6,10,11,12]. The burden of malaria in children aged between 5 and 16 years serves as an important source of transmission of malaria parasites from humans to mosquitoes [11]. This not only fosters the onward transmission between humans and mosquitoes, but also negatively impacts health and education. Malaria parasites invade an individual´s red blood cells, leading to fever and anemia. If a child does not get a timely diagnosis and treatment, their health status can deteriorate, leading to severe complications such as convulsions, coma, and eventually death. Due to serious health conditions following malaria infection, other aspects of children´s lives also get impacted, such as impaired attention, decreased cognitive function, and increased school absenteeism [13,14].

Data in the Kenya Health Information System (KHIS)

The Kenya Health Information System (KHIS), a national implementation of the District Health Information Software 2 (DHIS2), is a free, open-source, web-based platform designed for the collection, validation, analysis, and presentation of aggregate and patient-based statistical data across multiple health programs [15]. With its flexible metadata model and configurable user interface, KHIS allows tailoring of health information systems without custom programming. The scientific community can use this system (KHIS) to explore a diverse range of important research questions and even merge extracted data files with other relevant information according to a particular context [16], for example, extracting malaria cases for a certain region, and linking data to climate and economics, to explore how climate has impacted malaria cases, and how this, by transitivity, has impacted the economy in the area.

However, KHIS currently structures malaria case data into standard age categories such as <5 and ≥5 years. These predefined groupings do not align precisely with the age brackets now recognized as epidemiologically important for malaria control: <5, 5-16, and >16 years [6]. Although it is technically possible to approximate custom age groups by aggregating existing bands through category option groups, this approach is imprecise because certain bands (e.g., 15-19 years) overlap with both the 5-16 and >16 categories. For researchers and public health practitioners, this lack of alignment hinders the ability to conduct accurate comparisons, evaluate interventions targeted at school-age children, and track evolving transmission patterns. The problem is further compounded in settings where individual-level data collection uses DHIS2 Tracker. Disaggregation in Tracker often requires creating separate data elements for each desired age band, which increases reporting complexity and discourages field implementation of nonstandard but scientifically justified age groupings. If health information systems do not evolve to reflect changing epidemiology, they risk obscuring important findings and slowing the development of age-specific surveillance and interventions.

Theoretical framing - data infrastructures and adaptive governance

Literature on Information Systems (IS), Information and Communication Technologies for Development (ICT4D), and digital health (DH) highlight that digital infrastructures are social as well as technical systems that influence governance, visibility, and resource allocation [17-20]. Different studies from researchers in these spaces of IS, ICT4D, and DH have shown that health information systems embody implicit assumptions about what counts as relevant data, whose needs are prioritized, and how accountability is enacted [21,22]. Platforms like KHIS are not merely repositories of information but “infrastructures of knowing” that shape the contours of public health decision-making [21]. Within the ICT4D literature, digital systems are also seen as sites where global standards and local realities meet. DHIS2, for instance, operates through standardized metadata models designed for global comparability, but these can inadvertently constrain local responsiveness [21]. When surveillance infrastructures fail to evolve alongside epidemiological transitions, they risk reproducing blind spots in both knowledge and policy [19,20]. Viewing KHIS through this socio-technical lens highlights that reconfiguring age bands is not only a matter of data convenience, but also an act of adaptive governance, enabling the system to represent new forms of health vulnerability and guide equitable intervention.

A call to action: modify age strata to reflect new changes

We therefore call on the KHIS development community, including the Kenya Health Management Information Systems (KeHMIS) project, now in its third phase (KeHMIS III) [23], to enable more flexible age stratification in malaria data collection and reporting. Specifically, the platform should allow users to define custom age groupings at the point of analysis and extraction, rather than forcing researchers to rely solely on fixed global defaults. This could be accomplished by: a) Implementing configurable age-band definitions in both aggregate and tracker modules that do not require redefining data elements from scratch; b) allowing dynamic re-aggregation of existing age bands into user-defined groupings without data loss or approximation errors; c) supporting consistent metadata updates so that age group changes in one module automatically propagate to analytics tools and dashboards. Such modifications would enable malaria programs, researchers, and policymakers to better monitor the shifting burden of malaria among school-age children, evaluate the impact of interventions, and align surveillance data with emerging scientific evidence.

Implications for research and policy

The proposed modification to the KHIS has broader implications beyond malaria surveillance. For research, customizable age stratification would facilitate more accurate modeling of malaria transmission dynamics and enable researchers to integrate epidemiological, behavioral, and environmental data at finer levels of granularity. It would also support comparative studies across regions, helping to identify patterns of age-specific vulnerability and intervention effectiveness. For policy, more flexible data structures would empower national malaria programs to design evidence-based interventions targeted to populations most at risk. Recognizing school-age children as a distinct epidemiological category will influence the allocation of preventive tools, such as insecticide-treated nets, vector control, intermittent preventive treatment, and vaccination, toward this neglected group [24-27]. For digital health governance, this case illustrates how system design decisions can either reinforce or challenge outdated paradigms. By embedding adaptability into KHIS, developers and policymakers can create infrastructures that learn from emerging evidence, fostering data systems that are not only functional but also reflexive and responsive to public health realities.

There are precedents where recognition of substantial malaria burden among school-age children led to differentiated intervention strategies. In Malawi, school-based screening and treatment programs were implemented after epidemiologic data demonstrated high parasite prevalence and anemia among school-age children. In two studies, repeated school-based screening and treatment significantly reduced Plasmodium falciparum infection prevalence and anemia in both high- and moderate-transmission settings, suggesting potential transmission benefits [28]. These interventions targeted a population that would otherwise be subsumed within the ≥5 category in routine reporting systems. This illustrates how finer age-specific data can inform alternative delivery platforms, such as school-based interventions, distinct from traditional household or strategies focused on the under-five.

Although many malaria interventions are delivered at household or community level, improved surveillance granularity would influence programmatic decision-making in three ways: (i) enabling justification and planning of school-based interventions such as intermittent preventive treatment for school-based screening initiatives; (ii) informing resource allocation where malaria prevalence among school-age children is disproportionately high; and (iii) enabling monitoring of intervention coverage and impact specifically within the 5-16 age group. Surveillance refinement does not itself constitute an intervention but strengthens evidence-based targeting and accountability.

We acknowledge that revising national age-disaggregation structures entails operational and governance considerations, such as national-level metadata updates, indicator recalculation, dashboard modification, and staff training. These represent transitional operational investments rather than new software or infrastructure costs. Implementation would also require alignment with National Malaria Control Program reporting standards and careful management of historical trend comparability. A phased pilot implementation would allow assessment of feasibility, data quality, operational burden, and programmatic value before consideration of national scale-up.

Concluding remarks

Malaria surveillance systems must keep pace with evolving epidemiological patterns. Evidence from SSA countries shows that school-age children now carry a disproportionate burden of malaria, a finding that challenges long-held assumptions about disease risk being concentrated primarily among children under five. KHIS, as the leading health information platform used by malaria programs worldwide, is well-positioned to adapt quickly to these findings. By incorporating configurable age stratification, KHIS can empower both national programs and the global research community to generate more precise, actionable insights and accelerate progress toward malaria elimination.

Competing interests     

The authors declare no competing interests.

Authors' contributions     

FMPR and GA were primary contributors to this work; they contributed equally. MOO and VK assisted with editing. All the authors read and approved the final version of this manuscript.

References