Defining equity in physical access to clinical services using geographical information systems as part of malaria planning and monitoring in Kenya

Abstract

Distance is a crucial feature of health service use and yet its application and utility to health care planning have not been well explored, particularly in the light of large-scale international and national efforts such as Roll Back Malaria. We have developed a high-resolution map of population-to-service access in four districts of Kenya. Theoretical physical access, based upon national targets, developed as part of the Kenyan health sector reform agenda, was compared with actual health service usage data among 1668 paediatric patients attending 81 sampled government health facilities. Actual and theoretical use were highly correlated. Patients in the larger districts of Kwale and Makueni, where access to government health facilities was relatively poor, travelled greater mean distances than those in Greater Kisii and Bondo. More than 60% of the patients in the four districts attended health facilities within a 5-km range. Interpolated physical access surfaces across districts highlighted areas of poor access and large differences between urban and rural settings. Users from rural communities travelled greater distances to health facilities than those in urban communities. The implications of planning and monitoring equitable delivery of clinical services at national and international levels are discussed.

Figure 1

A district map of Kenya () showing the study area () and maps of each of the four districts showing a surface map of populations overall access to government health facilities () with varying shades of brown representing areas with access distances between <1–5 km and the red colour representing areas >5 km ( <1, 1–2, 2–3, 3–4, 4–5 and >5 km). The populations are represented by light blue point , which are the EA centroids. Almost the entire population in Greater Kisii is within 5 km of a government health facility. In Bondo, Kwale and Makueni, 80%, 65% and 65% of the population were within 5 km of a government health facility, respectively.

Figure 2

Cumulative health service use against distance travelled to government health facility where treatment was sought for the four districts. The line - -■- - represents Greater Kisii, – –■– – represents Bondo, – —■— – represents Kwale and —■— represents Makueni. The y-axis shows the cumulative percentage of patients in each district who travelled to a health facility at the distance (km) value shown on the x-axis. The cumulative percentages were computed by summing all patients who went in a district who used a facility at each distance on an 11-point scale (0–1, 1–2, 2–3, 3–4, 4–5, 5–6, 6–7, 7–8, 8–9, 9–10 and >10 km). Eighty-four per cent in Bondo and 80, 64 and 62 of patients in Greater Kisii, Kwale and Makueni used health facilities within 5 km, respectively.

Figure 3

General decay in use of government health facilities as distance from health facilities increased. The line - -■- - represents Greater Kisii, – –■– – represents Bondo, – —■— – represents Kwale and —■— represents Makueni. The y-axis shows the percentage patients in each district who travelled to a health facility at the distance (km) value shown on the x-axis. The percentage use was based on a discrete number of patients in each district using government health facilities at each distance, from 0–25 km. The logarithmic relationship was worked out using a simple linear regression comparing logarithmically transformed percentage proportion of distance predicted users and actual users at each distance interval. The relationship was significant at 1% level for Bondo, Greater Kisii, Kwale and Makueni with respective R² values of 0.706, 0.879, 0.996 and 0.927.

Figure 4

District maps showing EAs, where patients came from, as represented by centroid of EA polygons, the health facilities they used and the underlying overall access to health services. The EAs have been classified into those where patients travelled ≤5 km (blue centroid point ) and those >5 km (green centroid point ). There was a close relationship between distance travelled by patients and the population's underlying access to GoK health facilities with varying shades of brown representing areas with access distances between <1–5 km and the red colour representing areas >5 km ( <1, 1–2, 2–3, 3–4, 4–5 and >5 km). The figure also shows the facilities catchment areas and the position of the facilities [] used in the study for each district.