Background: Hypertensive disorders of pregnancy (HDP) are associated with high maternal mortality in Ghana and globally. Evidence shows that there is poor availability of pregnancy-related point-of-care (POC) tests in Ghana’s primary healthcare (PHC) clinics (health centre or community-based health planning services facilities). Therefore, we employed geographic information systems to estimate the geographical distribution of and physical accessibility to HDP POC testing services in the Upper East Region (UER), Ghana. Methods: We collected data on 100 out of 365 PHC clinics, public hospitals providing HDP testing, PHC clinic type, ownership, and availability of urine dipsticks and blood pressure (BP) devices. We also obtained the geo-located data of the PHC clinics and hospitals using the global positioning system. We employed ArcGIS 10.4 to measure the distance and travel time from the location of each PHC clinic without HDP POC testing services as well as from all locations of each district to the nearest hospital/clinic where the service is available. The travel time was estimated using an assumed motorised tricycle speed of 20 km/hour. We further calculated the spatial distribution of the hospitals/clinics providing HDP POC testing services using the spatial autocorrelation tool in ArcMap, and Stata version 14 for descriptive statistical analysis. Results: Of the 100 participating PHC clinics, POC testing for HDP was available in 19% (14% health centres and 5% community-based health planning services compounds) in addition to the 10 hospitals use as referral points for the service. The findings indicated that the spatial pattern of the distribution of the health facilities providing HDP POC testing was random (z-score = -0.61; p = 0.54). About 17% of the PHC clinics without HDP POC testing service were located > 10 km to the nearest facility offering the service. The mean distance and travel time from PHC clinics without HDP POC testing to a health facility providing the service were 11.4 ± 9.9 km and 31.1 ± 29.2 min respectively. The results suggest that if every 19% of the 365 PHC clinics are offering HDP POC testing in addition to these 10 hospitals identified, then the estimated coverage (health facility-to-women in fertility age ratio) in the UER is 1: 3,869. Conclusions: There is poor physical accessibility to HDP POC testing services from PHC clinics without HDP POC testing in the UER. Mothers who obtain maternal healthcare in about 17% of the PHC clinics travel long distances (> 10 km) to access the service when needed. Hence, there is a need to improve the availability of HDP POC diagnostic tests in Ghana’s rural clinics.
We employed geographic information systems to estimate the geographical distribution of and physical accessibility to HDP POC testing services as an integral part of ANC in the UER of Ghana based on the findings of a previous cross-sectional survey [18]. The cross-sectional study before the current study found poor availability of pregnancy-related POC diagnostic tests in the UER [18]. A detailed description of the study setting has been described and published elsewhere [17–19]. The study was conducted from February to March 2018 in all 13 districts using a random sample of 100 out of a total 365 public PHC clinic providing maternal healthcare services from all the districts. A multistage sampling strategy involving stratified, probability proportionate to size, and simple random sampling techniques. A detailed description of this study’s sampling strategy is published elsewhere [18]. We obtained the geographic information (geo-located data) of the PHC clinics and their referral health facilities from the Upper East Regional Health Directorate, and the use of a global positioning system. We then applied the world geodetic system Zone 30 north coordinate system to all spatial data to allow for the results of spatial processes in a chosen unit of meters. Topographic data included roads, rivers, and the digital elevation model. The collectors interviewed the midwives/nurses in-charge of the antenatal clinics or the heads of the clinics for information about the clinic type, ownership, availability of urine dipsticks, BP monitoring devices, and name of the nearest hospital (referral facility) and geographical location (Town/village) using a questionnaire (published elsewhere) [18]. We also cross-checked the information about the referral health facilities to be sure of the availability of urine dipsticks for proteinuria testing and BP monitoring device. To ensure data quality, the principal investigator closely supervised the data collection and data entry activities for consistency and completeness of information throughout the study period. The availability of urine dipsticks for proteinuria testing and BP monitoring devices, or laboratory services. Proximity to the nearest health facility where a POC testing service for HDP is available. Estimated time likely to be spent by an expectant mother traveling from a PHC clinic or her settlement location to the entrance of a health facility providing HDP diagnostic services. We used PHC clinics that lacked urine dipsticks and BP monitoring devices from all 13 districts, as inputs to measure proximity to the nearest health facilities providing HDP testing services. We computed the Euclidean distance from the PHC clinic to the nearest HDP diagnostic services using the near function of analysis tools in ArcGIS 10.4 software. Data on the health facilities, area, and the geographic coordinates of the PHC clinics and their referral health facilities were linked to ArcGIS 10.5 software and a base map. Travel time was estimated using a motorised tricycle transport system (“motor king”) since it is the most used public transport in the UER. The model and procedure used to estimate the travel time for this study have been published elsewhere [19]. We employed ArcMap 10.5. to calculate the spatial autocorrelation or Moran’s Index (MI) of the health facilities providing HDP diagnostic services and, the z-score and p-value reported. MI value of 0, or very close to 0, was considered as random distribution, MI value less than zero was interpreted as dispersed distribution, and MI value greater than zero was considered as a clustered distribution. Data on the distance and travel time to the nearest health facility providing HDP diagnostic services from the clinics were exported to Stata version 14.0 and the mean distance and travel time calculated for each district. We considered PHC clinics located > 10 km from the nearest health facility providing HDP POC testing services as areas with poor physical accessibility. Research has shown that physical access to healthcare beyond 10 km is significantly associated with higher risks of poor health outcomes [22].
N/A