Geospatial analysis of COVID-19 vaccine access in Kenya: the interplay of travel time, perceived availability, and vaccine uptake.

Abstract

Background: COVID-19 vaccines have reduced severe illness and mortality, yet inequitable distribution delayed rollout in many low- and middle-income countries (LMICs), including Kenya, where only 36% of adults were fully vaccinated by October 2022. While sociodemographic and attitudinal determinants of uptake are well documented, role of geographic access remain underexplored at individual level. This study assessed population coverage by travel time to COVID-19 vaccination sites in Kenya and examined how travel time was associated with perceived vaccine availability and vaccine uptake.

Methods: We conducted a secondary analysis of cross-sectional data from Round 3 (October-December 2021) of the COVID-19 Gendered Risks, Impact, and Response Survey. Vaccination site locations were obtained from the Kenyan Ministry of Health. Travel times from constituency population-weighted centroids to the nearest site were estimated using a cost-path distance algorithm in AccessMod. Two binary outcomes were analysed: perceived vaccine availability and vaccine uptake. Logistic regression with restricted cubic splines modelled associations with travel time in the full sample and a restricted sample excluding respondents who reported vaccines as 'unavailable' despite modelled travel time <15 minutes.

Results: Nationally, 55% of the population lived within 30 minutes of a vaccination site, 69% within 60 minutes, and 21.6% more than 120 minutes away, with shorter travel times concentrated around major cities and longer times in eastern and northern Kenya. In the full analytic sample, travel time was not significantly associated with perceived availability or uptake. In the restricted sample, longer travel time showed strong, nonlinear associations with lower perceived availability and uptake.

Conclusions: Geographic barriers, captured by modelled travel time, are important determinants of perceived vaccine availability and uptake in Kenya. Combining high-resolution geospatial modelling with individual-level data can guide expansion of vaccination services to improve vaccine equity and preparedness for future health emergencies in Kenya and similar LMICs.