Health system redesign for maternal and newborn survival: rethinking care models to close the global equity gap

Assumption 1: risk stratification in pregnancy can be used to select women for hospital care

Though risk stratification is a core function of high-quality antenatal care, many life-threatening complications first manifest at the time of delivery and cannot be predicted.12 13 Studies from HICs suggest that approximately 30% of women categorised as low-risk still develop complications and that available population risk models perform poorly in predicting outcomes or guiding treatment at the individual level.14 15 In the Netherlands more than half of nulliparous women and one in five multiparas starting labour in midwifery practices are referred to obstetric units despite being categorised as low-risk.16 Risk stratification is even more problematic in LICs where detectable antepartum risk factors such as multiple pregnancy, breech presentation, or pre-eclampsia are frequently missed due to poor-quality antenatal care.10 Furthermore, evidence shows that poor women are more likely to receive low-quality antenatal care that fails to deliver essential screening and treatment and that even with excellent antenatal care, it is impossible to predict all intrapartum complications.12 13 17–20 Risk stratification should be used to identify complicated pregnancies that require more specialised levels of care (eg, regional hospitals), but if 30% of those deemed low-risk still develop complications then it may not be adequate for selecting women to safely deliver remote to advanced care.

Assumption 2: primary care clinics can provide good quality birth care

Basic emergency obstetric and newborn care facilities—primary care clinics with no surgical, blood transfusion or advanced neonatal services—have been deemed appropriate for low-risk birth in LICs. However, quality of maternal and neonatal care in primary care clinics has repeatedly been shown to be weak and access to delivery services in such settings is not associated with better outcomes when compared with giving birth at home.21 Furthermore, extensive efforts to improve quality of care in primary care clinics in LICs through coaching, feedback, checklists, decision support and technologies have been disappointing, with modest or null results in large-scale studies.22 23 Even if these strategies were effective, their scale-up across thousands of small facilities would be exceedingly difficult.

Assumption 3: emergency referral is an effective response to complications

To save lives when obstetric and newborn complications arise, women should receive definitive treatment within 30 min or less.24 25 The standard approach if complications arise during childbirth in primary care settings is to refer women to hospitals. However, transporting acutely ill patients is challenging even in high-income settings with fully equipped advanced life support ambulances, reliable communication and good roads. Transport is only attempted if absolutely necessary and ideally, should occur before and not during delivery.26 27 In LICs, which lack these conditions, intrapartum transport is often lethal for mother and/or baby as is immediate postnatal transport of sick neonates with compromised cardiorespiratory function.28 29 Studies show that for women who develop complications, starting labour at a primary care clinic results in loss of precious time and increases the likelihood of maternal and perinatal mortality.30–33

Assumption 4: pregnant women cannot get to hospital for delivery

While historically access to hospitals was limited in rural parts of LICs, rapid expansion in roads, transportation and health facility infrastructure has dramatically reduced travel time to hospital. A recent study in Haiti, Kenya, Malawi, Namibia, Nepal and Tanzania found that the vast majority of women now live within 2 hours of a hospital and that shifting all deliveries to hospital would reduce 2-hour access between 0.6% and 10%, depending on country.34 For many, though not all, women, hospitals are closer than 2 hours: the average travel time to hospitals in the study varied from 22 min in Haiti to 82 min in Tanzania. The finding is consistent with previous studies of access to advanced care.9 35 Moreover, LICs are rapidly urbanising, bringing advanced facilities within easier reach of the majority of women.36 This new demographic landscape is an opportunity for redesigning systems to equitably offer advanced care to all.

Assumption 5: women prefer to give birth in nearby primary care clinics

While some women value local delivery, quality of care is more important than the proximity of the clinic for many women. Women routinely bypass local facilities for childbirth in search of higher quality care even if they incur additional costs.37 Several studies have found that wealthier women were more likely to bypass, exacerbating inequities in quality.9 38 39

Strengthen lifesaving, respectful care in hospitals

For the majority of women, midwife-led care, supported by physicians with obstetric training, leads to good health outcomes, avoids over-intervention and creates clinical environments that respect women and promote their agency.57–59 In the context of redesign, both onsite and near-site midwife-led birth units can make woman-centred care possible while ensuring immediate access to emergency obstetric intervention in case of complications (see table 1). However, in order for midwifery to meet its potential as the cornerstone of a redesigned health system, current gaps in the preservice and in-service education of this cadre need urgent attention.60 61 This is especially important as facility childbirth rates rise and as evidence of disrespectful treatment of labouring women is growing.52 62–64 International guidelines suggest that midwifery education should take a human-rights approach that produces clinicians who are competent to support physiological childbirth for women of all backgrounds, to empower them through the continuum of care, to learn and use evidence and to respond quickly to complications and refer to obstetric colleagues.65

High hospital maternal and neonatal case fatality rates and perioperative mortality rates in LICs also point to needed improvement in surgical services and neonatal care for sick and premature newborns.66–69 Redesigned hospitals will require neonatal care units with sufficient supply of equipment, supplies, medication and human resources.70 These units should be physically separated from other patient care areas, be staffed by providers with training on the care of sick and premature neonates and be able to accommodate and engage families.71 In countries where specialists are in short supply or concentrated in urban areas, telemedicine and telementoring options will need to be explored.

Curriculum reform and twinning relationships between well-performing facilities and local facilities can raise quality standards across these disciplines and interactive, in-service training, including simulation, can be used to refresh infrequently used skills alongside coaching and mentoring.72

Boost primary care

As part of service delivery redesign, health systems will need to clearly articulate the core services best offered in primary care facilities and elevate primary care such that it is recognised as an area of expertise, not a minimalist version of hospital care. Primary care clinics will need to improve the provision of evidence-based antenatal and postpartum services, as well as maintain registries of pregnant women, work with women and couples on birth planning and coordinate care with higher-level facilities. One area for urgent improvement is detection and mitigation of maternal and fetal risk, including anaemia, malaria, HIV, multiple pregnancy and the like. Primary care, through community health worker programmes, is also best placed to follow-up with the mother–baby dyad in the community after discharge, especially in settings where short postpartum hospital stays are the norm.73 From the patient’s perspective, primary care should be the nucleus of their maternal and newborn health services.

In a redesigned health system, primary care facilities providing antenatal care should be linked to an advanced care facility to allow for sharing of care across the continuum and efficient communication. Further, advanced care facilities and their linked primary care centres could create learning collaboratives that meet regularly to review complex cases and solve problems.74

Promote equitable access to care

Women who live far from a hospital will need to travel before the onset of or during early labour. Some remote areas may require upgrading health centres to provide advanced care, new roads and bridges, more reliable transportation options and/or patient-centred maternity waiting homes.75–78 Financial and other socioeconomic barriers must also be addressed, for example, through vouchers, cash transfers and comprehensive health insurance benefits. Novel solutions, including ride-sharing and Airbnb-like or other maternity home waiting options, need to be tested (see table 2). Without fully integrating geographic access interventions, redesign could exacerbate childbirth disparities and leave the most remote and vulnerable families behind.79 Although intrapartum referral will be needed less under redesign, emergency referral systems will need to be strengthened for transport of women and newborns with severe complications to specialised facilities.

Ignite demand and construct accountability channels

In addition to lowering barriers to access, health system leaders will need to raise demand for higher-level maternity care by informing women and families of the rationale and specifics of redesign and by raising their expectations of delivery and newborn care.10 80 Women, families and communities will need to understand what high-quality care for childbirth means and where this care can be received. Community outreach and social marketing through radio, television, and community drama can be effective means of educating the community and increasing demand for quality. Participatory methods, including women’s groups practicing learning and action cycles, are also a good way to meaningfully engage people in improving systems.81 These efforts should include the whole family, not just women, as well as key social and religious leaders who may be influential. In many countries, providing health services close to communities is politically important, and communicating the rationale for removing delivery services from primary care clinics will be essential. Redesign should be accompanied by means for feedback and redress at all levels of the system, for example, through dedicated phone lines or text systems. More research is needed to understand how best to improve accountability and share information about quality of care with communities.82

Political leadership and policy change

Service delivery redesign is fundamentally a political choice and must be led by political leaders who believe that a double standard for women and newborns in LICs is no longer acceptable. Political commitment is central to determining whether redesign is realistic and how quickly it can be implemented. Though a daunting task, we are already seeing such commitment from some countries. In Kakamega County, in Kenya, the county government has committed to implementing redesign; this is after a feasibility assessment was conducted to determine readiness for redesign. The government is now moving into a planning phase where strategies will be developed to close the identified gaps and ensure that no woman is left behind. The government also plans to implement redesign in a phased manner to allow for adaptation as necessary. Policy review and change will be necessary in many countries where current guidelines and policies reflect the assumptions underlying the two-tiered model for LICs.