Geospatial analysis of cesarean section in Iran (2016-2020): exploring clustered patterns and measuring spatial interactions of available health services

Abstract

Background: The lives of babies and mothers are at risk due to the uneven distribution of healthcare facilities required for emergency cesarean sections (CS). However, CS without medical indications might cause complications for mothers and babies, which is a global health problem. Identifying spatiotemporal variations of CS rates in each geographical area could provide helpful information to understand the status of using CS services.

Methods: This cross-sectional study explored spatiotemporal patterns of CS in northeast Iran from 2016 to 2020. Space-time scan statistics and spatial interaction analysis were conducted using geographical information systems to visualize and explore patterns of CS services.

Results: The temporal analysis identified 2017 and 2018 as the statistically significant high clustered times in terms of CS rate. Five purely spatial clusters were identified that were distributed heterogeneously in the study region and included 14 counties. The spatiotemporal analysis identified four clusters that included 13 counties as high-rate areas in different periods. According to spatial interaction analysis, there was a solid spatial concentration of hospital facilities in the political center of the study area. Moreover, a high degree of inequity was observed in spatial accessibility to CS hospitals in the study area.

Conclusions: CS Spatiotemporal clusters in the study area reveal that CS use in different counties among women of childbearing age is significantly different in terms of location and time. This difference might be studied in future research to identify any overutilization of CS or lack of appropriate CS in clustered counties, as both put women at risk. Hospital capacity and distance from population centers to hospitals might play an essential role in CS rate variations and spatial interactions among people and CS facilities. As a result, some healthcare strategies, e.g., building new hospitals and empowering the existing local hospitals to perform CS in areas out of service, might be developed to decline spatial inequity.

Keywords: Cesarean section; Geographical information systems; Spatial epidemiology; Spatial interaction; Spatiotemporal analysis.

Fig. 1

Map of the study area including hospital capacity and population density of different counties

Fig. 2

The percentage of CS (%) and mothers' age groups by county

Fig. 3

Rate maps of CS deliveries at the county level in KRP, Iran. Dark-red regions represent the highest CS rates, and light-yellow regions represent the lowest CS rates

Fig. 4

A Purely temporal clusters identified by 50% maximum window cluster size in the study region between 2016 and 2020, note: the light blue banded plot area shows significant temporal clusters; B Global Moran’s I statistics for EBS rates of CS deliveries per 100,000 women of childbearing age within the study region from 2016 to 2020

Fig. 5

A Purely spatial clusters of CS rates identified by the SaTScan approach in the study area; B Bivariate choropleth map comparing hospital capacity and LLR score within the study region from 2016 to 2020; shades of purple show significant proportions of both variables

Fig. 6

Spatiotemporal cluster identified by SaTScan approach between 2016 and 2020

Fig. 7

Clusters of spatial variations in temporal trends map

Fig. 8

A Potential spatial interaction among the counties in the study area. B CS weighted rate flows values (red color lines) between origins (counties centroids) and destinations (hospitals)

Fig. 9

Average distance between each public hospital (destinations) and county centroids (origins) in km in the study area. Note: symbols H1 to H18 indicate the identification number of the hospitals