Poor and rural women in Pakistan are disadvantaged in terms of their access to institutional delivery services, according to a geographic analysis of data from nine districts.1 On average, married women who reported giving birth within the past three years lived seven kilometers by road from the nearest health facility offering delivery care; the average distance ranged from five kilometers among the wealthiest women to 10 kilometers among the poorest. Women’s odds of delivering in a facility decreased by 3% per additional kilometer of distance from the nearest facility, although this association held only in rural areas. Living within 10 kilometers of a facility offering basic or comprehensive emergency obstetric care was positively associated with institutional delivery (odds ratios, 1.8 and 1.7, respectively).
To examine relationships between geographic and economic access to a health facility and institutional delivery, researchers linked survey data for individual women and health facilities in study districts with data on distance between the primary sampling unit of women’s residence and health facilities in those districts. The individual data were drawn from a 2005 survey that asked married women in selected households in primary sampling units about their health care before, during and after delivery, as well as about their social and demographic characteristics. Household wealth was measured in quartiles (lowest, lower middle, upper middle and highest) on the basis of such criteria as ownership of certain household amenities and types of construction materials used for the household’s dwelling. The sample of women was restricted to the 4,435 women who reported giving birth in the past three years.
Facility data were drawn from a 2008 survey of all health facilities, ranging from large public teaching and district hospitals to private registered clinics and community pharmacies. A structured questionnaire was used to ask individuals in charge of facilities about the type and level of available reproductive health care; it included questions about staff, equipment and client amenities. The sample included all 763 facilities that offered obstetric care ser- vices before 2005. Of those, 547 were considered able to provide only "normal delivery" care (i.e., assisted vaginal delivery in a health facility that lacked an operating room), 91 could also provide "basic emergency obstetric care" (assisted deliveries, manual extraction of the placenta and removal of retained products) and 125 could further provide "comprehensive emergency obstetric care" (cesarean sections and blood transfusions).
In addition, as part of the individual and health facility surveys, the exact locations of primary sampling units and facilities were determined using global positioning system (GPS) devices. The researchers used the GPS data to create georeferenced digital maps that included health facilities, district and subdistrict boundaries, rivers and roads, and then used global information system software to calculate exact road distances between primary sampling units and health facilities within each district. Two measures of geographic access to services were created: the distance from each primary sampling unit to the nearest health facility, and the highest level of delivery care within 10 kilometers of each primary sampling unit. Multilevel mixed-effects logistic regression analyses were conducted to examine associations between geographic access, wealth status and institutional delivery.
Most women (84%) lived in a rural area; 20% lived in households in the lowest wealth quartile, 24% in the lower middle, 27% in the upper middle and 29% in the highest. On average, the primary sampling unit in which women lived was seven kilometers from the nearest health facility offering delivery care. Twenty-eight percent of women lived within 10 kilometers of a facility offering only normal delivery care, 14% within 10 kilometers of a facility offering basic emergency obstetric care, and 37% within 10 kilometers of a facility offering comprehensive emergency obstetric care; the remaining 21% of women lived more than 10 kilometers from a facility.
Overall, the distance between a woman’s place of residence and the nearest district health facility was smaller among wealthier women than among poor women. The average distance to a facility ranged from five kilometers among the wealthiest women to 10 among the poorest; 88% percent of the wealthiest women lived within 10 kilometers of a health facility, compared with 62% of the poorest. Geographic access to a health facility also varied by urban or rural residence. The average distance to a facility was eight kilometers in rural areas and only one kilometer in urban areas. In addition, although 75% of rural women lived within 10 kilometers of a facility, only 27% lived that close to comprehensive emergency obstetric care; in comparison, all urban women lived within 10 kilometers of a facility, and 85% lived that close to the highest level of care. The difference in geographic access to health facilities by wealth held in rural areas, but not urban ones: For rural women, the average distance to the nearest facility ranged from six kilometers in the highest wealth quartile to 12 in the lowest, whereas for urban women, the average distance to the nearest facility was similar across wealth quartiles (1.1–1.6 kilometers).
Thirty-two percent of women reported that their most recent delivery had occurred in a health facility, while 68% had delivered at home. In multilevel regression analyses, the odds of institutional delivery decreased by 3% per kilometer from the nearest health facility (odds ratio, 0.97); in separate analyses by rural and urban residence, this association was significant only among rural women (0.98). Women who lived within 10 kilometers of a facility offering basic or comprehensive emergency obstetric care had greater odds than women who lived further than 10 kilometers from any health facility of delivering in an institution (1.8 and 1.7, respectively); no association was found between living within 10 kilometers of a facility offering only normal delivery services and institutional delivery. Increased wealth and education were independently and positively associated with institutional delivery in all multilevel regression models (odds ratios, 1.3–4.5).
The authors acknowledge several limitations of their study, including that their measures of geographic access were based on distances to health facilities within a district, and that they lacked information on travel cost, travel time and quality of care. Even so, the findings suggest that geographic distance, poverty and low education are all barriers to Pakistani women’s use of institutional delivery services, especially in rural areas. The authors conclude that the "disadvantages can be minimized by upgrading existing facilities at district and subdistrict levels to provide comprehensive emergency care and by expediting the transport of poor rural women to these facilities when life-threatening childbirth complications occur."—J. Rosenberg
REFERENCE
1. Jain AK, Sathar ZA and ul Haque M, The constraints of distance and poverty on institutional deliveries in Pakistan: evidence from georeference-linked data, Studies in Family Planning, 2015, 46(1):21–39.