A cross sectional comparison of postnatal care quality in facilities participating in a maternal health voucher program versus non-voucher facilities in Kenya

In OBA programs it is often assumed that their minimum quality standards for accreditation and the expected competition between health facilities encourage providers to improve their quality of care. Voucher-accredited public facilities do appear to score higher for structural attributes than non-voucher public facilities and all private facilities. In another component of the study, the evaluation team interviewed facility managers to ascertain how they utilised their voucher remittances [41]. Most used their funds on structural improvements such as renovating maternity units, laboratory or operating theatres as well as purchasing drugs, equipment or supplies. Moreover, public health facilities were required to follow strict MoH procedures for disbursed funds with formal requests to their county managers prior to any expenditures. Public facilities were also not able to use the funds for recruiting additional staff or on skills training [41, 42]. This MoH policy may explain why public voucher facilities scored higher on structural attributes but lower on process [13, 41]. Slightly lower scores by private voucher facilities than private non-voucher facilities may provide positive evidence that vouchers were correctly targeted at low income women who visit less sophisticated facilities.

Providers’ knowledge and practice were also mixed. More voucher providers reported clients counselled on the importance of breastfeeding, but more non-voucher providers said they counselled women on family planning, which is unexpected due to the fact another voucher for family planning is available for poor women. From our review of facilities’ remittances we know that little is spent on provider training [41], and we recommend that providers (both public and private) have more access to training sessions and contraceptive technology updates organised by the Ministry of Health.

Although these high structural attribute scores could be taken as synonymous with availability or readiness to provide services and commodities, they are incomplete, because clinical or process attributes for the provision (and receipt) of comprehensive postnatal care are not included [32]. A 2010 review and design of an extension to the OBA programme also concluded that postnatal care had yet to change significantly by virtue of reimbursement funds alone [42].

All facilities’ composite scores totalled less than one third of the maximum score, well below the standards of postnatal care provision in the national guidelines [34, 43]. Although voucher facilities scored significantly worse than non-voucher facilities, no significant differences overall were observed for public and private facilities or hospitals and lower level facilities. Non-voucher facilities performed better in history taking and counselling on fertility and infant care, but these scores were not significant when controlled for clustering. Public facilities scored higher on counselling for HIV and STI risk assessment and family planning than private facilities, but private facilities had higher scores for rapport.

The overall low performance in clinical care (or process attributes) at both voucher and non-voucher sites could be explained by several factors. Although providers reported some clinical updates for a range of maternal and newborn care, it is possible providers were unable to translate this knowledge into practice [44]. Secondly, although women are entitled to postnatal care with the Safe Motherhood voucher, providers often only immunise an infant rather than providing a comprehensive package of care for both mother and child. The difficulty of conducting a check up for a mother in a busy immunisation clinic, without privacy or examination couches, can contribute to infant-focused care. Thirdly, women may not be aware of the comprehensive package of postnatal care, as over 60 % of women in this study came to postnatal care for infant immunisation. Elsewhere in the Kenya voucher evaluation we found communities with misunderstanding about what services vouchers actually cover [26]. Finally, although there are now separate vouchers for each of the four antenatal care visits as well as childbirth, there is no specific paper voucher for postnatal care, so facilities can only claim reimbursement for this service as an extension of the delivery voucher (i.e., if a woman give birth in a facility), but women are only encouraged to return to a facility after six weeks for infant immunisation. Nevertheless, it was expected that the OBA programme would improve the overall quality of maternal care including postnatal care.

If voucher facilities are not reimbursed for providing comprehensive postnatal care, this may hinder providers’ motivation for offering any postnatal services for mothers and infants, especially in private facilities that do not receive updates from MoH. There are clearly defined guidelines and standards about the content and timing of mother and infant care immediately after birth, at one or two weeks postpartum, and up to six weeks after birth. Many providers, specifically those from private facilities, appear to be unaware of these national postnatal guidelines, and this contributes to service discontinuity after pregnancy and delivery [3, 45].

More than four fifths of all women were seen within 48 h, and significantly more newborns at voucher facilities received care in their first 48 h in addition to BCG vaccinations than those at non-voucher facilities. Evidence suggests that newborns seen within the first two days of birth by a provider have a greater chance of survival [46]. Overall, postnatal women seemed satisfied with their care, waiting times, and time spent with a provider, regardless of facility type. Clinical competence is less easily judged by clients, and often they evaluate providers on the amount of time spent with them, and their caring attitude, than on technical skills [32]. Although the output attributes based on clients’ experience may not necessarily translate into adequate clinical care, such information presents an opportunity for the health system to understand women’s perceptions of quality postnatal care and what motivates them to seek services.

A comprehensive postnatal care package should include routine visits in the days following childbirth, when risks are high for both mother and baby, complemented by promotion of healthy behaviours (e.g., exclusive breast feeding), for identifying complications and facilitating referrals [3]. Kenya’s MoH recommends one visit within 48 h of delivery, another within one week, a third between four and six weeks postpartum, and a fourth at six months [39]. Postnatal registers have been in existence in Kenya since 2005, but few facilities routinely record or collect this information, or report it through the government health information system. An attempt to deduce potential increased workloads associated with increased in attendance at voucher facilities and its effect on quality of care was not possible due to these recording gaps. The OBA programme’s voucher management agency is in the process of converting the paper vouchers to a ‘smart card’ that may capture all services offered to and received by voucher clients, including postnatal care.

A review of the literature on the quality of private and public health care in low and middle income countries indicates that improving quality of care in a health system is a long term effort and requires attention to various aspects including the incentive structure and providers’ training [47]. Supervision and clinical audits with resultant recommendations, especially if combined with training, have been found to be effective in improving quality [48]. The OBA project itself did not provide any technical training, and expected health facilities to improve their own staff skills through technical updates. Apparently this did not happen sufficiently, although providers received some related maternal and infant care training within the past 24 months.

For the adequate clinical care of postnatal women, and the diagnosis and early treatment of any complications, an evidence-based schedule for visits is critical. While Kenya has detailed the timing and content for postnatal visits in policy guidelines, it appears there is a policy and care provision gap: Many health care providers do not perceive the importance of the postnatal period, although four fifths of women interviewed sought care within 48 h postpartum. A community-based approach may be required to ensure that women discharged after delivery receive adequate information, support, and follow up. Providers should receive more training updates specifically for the postnatal period, and any training should be followed by supportive supervision. Policy guidelines should be more easily available for private and faith-based facilities.

In other results-based financing initiatives, supply side incentives reward improvements in structural and process indicators for quality. These programs develop verification mechanisms that routinely confirm service deliveries and measure quality indicators at contracted facilities. Most OBA voucher programs do not have an explicit financial incentive for quality improvements, and it might be worthwhile for the Kenya program to consider one [18].

There are limitations to this analysis. Its cross sectional design, coupled with a non-random facility selection, limits any tests for causality. Control facilities were, however, matched by administrative type and similar infrastructure, staffing and service characteristics, along with a qualitative assessment of each local healthcare market and access. Ideally, the voucher and non-voucher facilities would be alike in every way except for the OBA voucher contract. Because facilities differed according to ownership status it is possible other differing factors, such as financing and catchment size, could affect quality of care outcomes. Secondly, although comparison sites were carefully selected, support by other partners for maternal and newborn health may also have confounded results. We did control for clustering at facility level, provider type, and level of care. Thirdly, the study population of postnatal women attending a health facility is not representative of Kenya’s general postnatal population, where only 44 % of newly delivered women attend postnatal care [20]. These findings do, however, represent the services received by recently delivered women in most Kenyan health facilities.