Influenza and other respiratory viruses circulate in Kenya throughout the year rather than following a single Northern or Southern Hemisphere season. The Kenya Ministry of Health monitors severe acute respiratory infections (SARI), including influenza-associated hospitalizations, through a sentinel surveillance network. To support assessment of seasonal impact on hospital services and inform public health responses, investigators developed an approach to quantify seasonal severity by deriving intensity thresholds for respiratory illness– and influenza-associated hospitalizations using multi-year surveillance data.
Data for the assessment were contributed by nine sentinel surveillance sites in Kenya. Sites provided weekly counts of hospital activity for pediatric (0–13 years) and adult (≥14 years) medical wards. The abstract notes that sites had rolling start dates and that not all sites had sufficient historical data for every assessment year; specifically, four sites had enough historical records to generate thresholds for the 2024 assessment, and eight sites contributed to the 2025 assessment.
The dataset covered January 2019 through December 2025, with data collected by retrospective medical-record review for the period prior to June 2023 and by SARI surveillance platform reporting from June 2023 onward. Weekly indicators included all-cause admissions, respiratory illness admissions, and laboratory influenza testing among hospitalized patients. Age groups were categorized as pediatric (0–13 years) and adult (≥14 years). The abstract does not report the complete case definitions, laboratory testing methods, or the exact start dates for each sentinel site, nor does it provide site-level denominators or counts.
The moving epidemic method (MEM) was used to calculate intensity thresholds (ITs). Thresholds were defined for three severity levels: moderate (50th percentile, IT50), high (90th percentile, IT90), and very high (98th percentile, IT98). The MEM produced both annual and biannual thresholds for the percentage of hospitalizations attributable to respiratory illness and to laboratory-confirmed influenza, calculated overall and stratified by the two age groups. The abstract does not supply the numeric threshold values, nor the precise parameters used in the MEM implementation.
For 2024 and 2025, respiratory illness–associated hospitalizations remained below the IT50 for most weeks across age groups. Across the two years, there were typically only one to two weeks when the percentage of hospitalizations for respiratory illness exceeded the IT50. The abstract summarizes these findings as indicating low to moderate severity in both years but does not present weekly counts, site-specific patterns, or absolute hospitalization numbers.
Influenza-associated hospitalizations also remained below the annual IT50 in both 2024 and 2025. However, influenza activity exceeded the biannual IT50 for several weeks coinciding with peaks of each wave. The abstract highlights that influenza waves produced short periods where activity rose above biannual moderate thresholds, but detailed timing, influenza subtype composition, or laboratory positivity rates at peak weeks were not reported in the abstract.
The authors interpret the 2024–2025 experience as reflecting low to moderate severity for both respiratory illness– and influenza-associated hospitalizations when assessed against the derived ITs. They propose that this threshold-based approach can guide public health activities in Kenya, contribute to global influenza surveillance efforts, and provide a model for assessing influenza severity in settings without clear Northern or Southern Hemisphere seasonality. The framework is presented as useful for sentinel-based monitoring of seasonal intensity and for supporting public health decision-making.
The abstract does not report several details that would be needed to fully evaluate and reproduce the assessment: the specific numeric thresholds (IT50/IT90/IT98), site-level weekly counts and denominators, the full case definition(s) used for respiratory illness and SARI, laboratory testing algorithms and positivity rates, and the exact timing of rolling site starts. The abstract indicates that historical data sufficiency limited inclusion for the 2024 assessment (four sites) and that eight sites were available for 2025, but it does not list which sites were included or the reasons some sites lacked adequate historical records. These details may be available in the full text and underlying surveillance reports.