This study attempts to evaluate the functions of a newly constructed maternity ward in KC-PRH, which was supported by Japan grant aid to strengthen the Cambodian health system. Our findings are categorised into two viewpoints, one from the tertiary hospital and contrasted with that of the province health system. Key findings from the former viewpoint are 1) a drastic increase in the number of deliveries after renovation of the maternity ward, 2) overcrowding of the maternity at 187% of bed occupancy rate in 2017, 3) half of the occupancy was for normal vaginal delivery, and 4) the presence of a broad catchment area for normal vaginal deliveries. From the viewpoint of the province health system, this study finds that 1) half of the deliveries in the two provinces were conducted in first-line or secondary public facilities, 2) some district hospitals dealt with more than 20% of the expected births in each district, and 3) at a district in the provincial capital one in four births was conducted in KC-PRH. Since KC-PRH is the only tertiary level facility in the two provinces [11], its principle role is management of complicated cases. However, our findings imply that the burden of handling a substantial proportion of normal vaginal births could hamper its function, and that inequality in access to health facility was widened.
Congestion of referral hospitals is an old but unsolved problem in many low- and middle-income countries [12]. It causes the so-called ‘third-delay’ in the three-delays model: delay in reception, diagnosis, and management of a patient within a hospital [13]. This delay can lead to the death of a parturient woman if she has a life-threatening complication [14]. Overcrowding in a hospital also affects women and babies in the normal process of delivery, because they could receive less attention from health care providers. All situations mentioned above deteriorate the six aspects of quality of care: safety, effectiveness, timeliness, patient-centredness, efficiency, and equity [15]. Therefore, in the following sections we discuss the congestion with respect to possible underlying causes, consequences, and solutions.
Background factors affecting the overcrowding include self-referral and free choice in deciding the place of delivery. Studies have suggested that self-referral is a product of a lack of confidence in the quality of care or unavailability of services in first-line facilities [12, 16–19]. However, this may not be the case in our study site, because health centres are the most used facility for childbirth. Our findings show that first-line services were functioning in the area, implying that those who gave normal birth in KC-PRH actively bypassed health facilities in their districts. Since half of the bed occupancy (182%) in KC-PRH was normal vaginal births, the elimination of uncomplicated cases from the referral hospital could normalise its function. However, there is a freedom in the choice of health services in Cambodia, although the government has set hierarchal tiers of healthcare from the community to district and province levels. Every public facility does not have criteria to restrict patients. Moreover, an increase in the number of patients benefits hospital revenue via ‘user-fees’ which can be used for hospital management as well as supplement of staff salary [20, 21]. A study in Denmark revealed that liberal choice of facilities widens inequity; and that women who belong to higher socio-economic groups or received higher education tend to use higher-level health facilities [22]. These findings may be observed also in our study site despite the differences of study settings.
Considering the indirect medical costs such as transportation fees to and from the provincial capital and opportunity costs of birth companions with a parturient woman, users of KC-PRH would belong to relatively higher socio-economic subgroups within their community. This disrupts vertical and horizontal aspects of equity in service access and utilisation. Horizontal equity means the provision of equal service to those who have equal need [23]. For normal vaginal deliveries, one person may have to stay in their community and use a health centre there, while another can give birth in the newly furbished hospital in the provincial capital. Vertical equity means that the provision of health service is not uniform but served according to the need of individuals [23]. Referral hospitals should be established for and used by those who require higher level services; however, in this study the beds were occupied by a range of users. The maternity ward in KC-PRH is owned by the Cambodian government and was realised by Japan’s official development assistance. Therefore, the maternity is public property in a dual sense, and proposed policies should be cautious of the equity issue. An extensive study on facility-based deliveries has shown that an intervention without the intention of universal distribution often tends to increase service utilisation in the rich rather than the poor [24]. Another study in Senegal shows that a mere intervention on the supply side increased inequity between the rich and poor [25]. The grant aid in this study is a typical example of supply side intervention. Since an important function of a health system is to provide dialogue to balance supply from the provision side and demand from the user side, it is therefore necessary to integrate how to control demand when grant aid is implemented.
A main role of a district health system is to provide equitable, comprehensive, and integrated health services to a defined population [26]. A functioning district health system contributes to ensure the proximity of health services to the target population, and thus through prompt management reduces morbidity and mortality in mothers and babies [27]. However, this study indicated that service utilisation for normal vaginal births skewed toward some district and provincial hospitals. This partial dysfunction of a health system may have been aggravated after completion of the new maternity ward in KC-PRH. Therefore, it is recommended to create a network of first-line facilities with a district hospital to ensure equity in access to health facility as well as to strengthen the health system rather than the mere construction of higher level facilities, especially in a rural setting.
This study has two limitations. Firstly, the number of deliveries in public health facilities did not contain information on the residence of the parturient. In calculating the proportion of deliveries shown in Table 3, we assumed that every facility served only those who live in the same district. Therefore, there might be an over- or underestimation of proportions, and the direction of this bias cannot be determined. However, this potential error would be minimal, because there are similarities in structure and staff among health centres in the target area. Secondly, we could not investigate the indications of caesarean sections in KC-PRH. It is known that a considerable proportion of caesarean sections are performed for non-medical but social reasons or even induced by demands. Since unnecessary medical interventions hamper the timely management of severe complications and increase overall costs [28], an evaluation of appropriateness of interventions is required to assess the functions of hospital and health systems as well as the effect of official development assistance.