Effective non-drug interventions for improving outcomes and quality of maternal health care in sub-Saharan Africa: a systematic review

The range of interventions to improve quality of maternal care and maternal outcomes included in this paper is summarized in Table 1. The interventions were broadly classified into mobile and electronic health interventions (n = 5; 7 %); financial incentives in the form of user fee exemptions and payments to health facilities based on performance (n = 13; 18 %); clinical audits (n = 13; 18 %); health systems and infrastructure development (n = 23; 32 %); community mobilization and peer-based programs (n = 8; 11 %); home visits and counseling and educational and health promotion programs (n = 5; 7 %); emergency transportation, communication, and referrals (n = 4; 6 %); PMTCT (n = 4; 6 %); and task shifting interventions (n = 2; 3 %). Some of the studies covered in the review were included in more than one category. The summary of results of the review is presented in Table 1. Below, we highlight the key findings of the studies within each of the broad classifications.

Mobile health (mHealth) refers to the application of wireless, portable information and communication technologies to support health and health care [18, 19]. These interventions often involve the use of devices such as phones, computers, personal digital assistants, and digital point-of-care testing devices [20]. Although evidence of the potential value of mHealth for maternal health is just emerging, donors and development agencies continue to give considerable attention to it as a means of improving maternal health [21, 22].

Three studies on mHealth were included in this review. These studies were conducted in Nigeria [23], Rwanda [24], and Zanzibar [21]. The mHealth interventions sought to provide two- or three-way communication between women and the health care system, through provision of mobile phone messages reminding pregnant women to attend clinic days. Overall, evaluations of the mHealth interventions show that they improved utilization of, and access to maternal health services, as well as women’s attendance at primary health care facilities and services such as antenatal care (ANC). In Zanzibar, the provision of mobile phones, together with redeemable vouchers for pregnant women, reportedly improved the timing of ANC services as well as the number of women seeking and receiving preventive health services, attending ANC late in pregnancy (i.e., more ANC visits and those treated for antepartum complications). In Rwanda, the short message service (SMS)-based alert program called the Rapid SMS-MCH was used to monitor pregnancy by allowing a three-way interactive communication among a community health worker (CHW) following pregnant women in the community, a health care worker at a health facility, and an ambulance driver who would be called in to facilitate emergency transport for obstetric care. The project led to an increase in facility-based deliveries from 72 to 92 % at the end of a 12-month pilot phase.

An electronic clinical health decision system (eCDSS) implemented in Ghana [25] and the Basic Antenatal Care Information System (Bacis) implemented in South Africa [26] were shown to improve health workers’ compliance with established maternity care protocols. In Ghana, health workers’ compliance with maternal care protocols reportedly increased the detection of pregnancy complications during ANC in the health centers, consequently decreasing the number of labor-related complications. The Bacis system in South Africa improved health workers’ responsiveness and quality of services offered to their patients. It also improved the promptness and quality of services offered to mothers younger than 18 years and their retention in ANC beyond week 20. On final assessment, the Bacis program showed minimal health data input error (13.2 %) compared to the District Health Information System (DHIS) that was in use (25 %).

Mobile health and electronic health interventions may improve health care processes through lower failed appointments for services like ANC and quicker diagnosis and treatment of obstetric complications. Unreliable electricity infrastructure and supply and mobile network failure were the major challenges to the mHealth and e-health initiatives identified in this review. The interventions considered were also small in scale and at high risk of dilution of possible impact between intervention and control communities.

Financial incentives to improve access and utilization of maternal health care and service delivery have also been implemented in a number of countries in SSA [27‐29]. Interventions involving financial incentives take the form of supply-side and demand-side incentives and seek to enhance equitable access to health care while improving the market supply and quality of services available to underserved populations [30, 31]. Supply-side incentives include pay for performance (P4P), performance-based financing (PBF), and various community-based insurance schemes (CBHI) and while demand-side incentives include vouchers that can be redeemed for health services, user fee exemptions, conditional cash transfers, and subsidies or cost-sharing arrangements.

We reviewed 12 studies that reported on the effectiveness of demand-side financial incentives including user fee exemptions (majorly for caesarean section deliveries), cost-sharing programs between the public and the health care facilities, and output-based approach (OBA) vouchers covering costs for ANC visits, and facility-based deliveries. Other demand-side financial incentives implemented were vouchers redeemable for emergency transport to facilitate timely movement of mothers requiring emergency obstetric care and vouchers to health service providers to redeem costs incurred for providing delivery services in their facilities. The OBA vouchers supported targeted subsidies to underserved populations in the urban slums of Nairobi increased facility-based deliveries [32, 33]. Results showed that beneficiaries of the OBA voucher were more likely to deliver a subsequent child in a health facility compared to non-beneficiaries. Fee exemption for caesarean sections increased caesarean birth rates in cities in Mali, which reportedly reduced the risk of maternal deaths from obstructed labor [34]. The Safe Motherhood Initiative (SMI) provided vouchers to facilitate access to health facility deliveries for women in rural Kenya. This program significantly decreased the number of home births by about 10 % and added to the number of facility-based deliveries by a similar margin [35‐37]. In Ghana, exemptions from community-based health insurance program premiums increased the registration of female community members, enhancing uptake of health services and leading to improved ANC attendance, and facility-based deliveries [38]. The premium exemptions also increased access to a spectrum of other services offered at health facilities, including clinical and diagnostic services during ANC visits, counseling on safe motherhood, education on recognition of pregnancy-related danger signs, and vaccination against tetanus. A voucher system that catered for costs for ANC services and hospital delivery in Uganda raised demand for facility-based deliveries by 52.3 %, of which about 9 % were new health facility users [39]. In Nigeria, a co-financing program for maternal health between the government and the community resulted in a 60 percentage point increase in the utilization of maternal health services (from 26.7 to 85.6 %) [40], while a user fee exemption program in the same country resulted in increased uptake of institutional deliveries by up to 88 %, resulting in a decline in institutional MMR from 532 to 371 per 100,000 live births [41].

Burundi [41, 42], Ghana, Mali and Senegal [38, 43], Rwanda [44], and Uganda [39] have implemented community-based health insurance schemes, in some cases, combined with the exemption of pregnant women from the premiums payable to these schemes. The P4P scheme in Rwanda resulted in a 23 % increase in the number of institutional deliveries while in Burundi, the PBF initiative led to significant improvements in the quality of ANC services received by women at health facilities that participated in the intervention.

Financial incentives have the potential to increase access to and utilization of maternal health services especially among the poor, thereby facilitating facility-based deliveries, use of skilled care, and reduction of maternal mortality [45]. However, evidence on the impact of financial incentives on the quality of and equitable access to maternal health care in SSA is inconclusive. Some of the studies reviewed relied on small sample sizes or did not have comparison group and therefore did not offer high-quality evidence.

Several studies examined the use of audit tools including criterion-based clinical audits (CBCA) and maternal death reviews (MDR) to improve the quality of maternal care in health facilities. CBCA is a tool for measuring and improving the quality of obstetric care at health facilities with a focus on five life-threatening obstetric complications: hemorrhage, eclampsia, genital tract infection, obstructed labor, and uterine rupture [46, 47]. MDR, on the other hand, employs a qualitative approach to investigate the causes of and circumstances surrounding maternal deaths at health facilities [48]. Both approaches are combined with feedback and target setting sessions with health care providers, to improve the management of obstetric cases, improve quality of care, and reduce maternal mortality. The main element of CBCA is that performance is audited to ascertain whether guidelines were followed. The tool facilitates improvements on areas of weakness in situations where the right processes in the management of obstetric complications are not followed.

CBCA implemented in a health facility in Nigeria [49] resulted in an improvement of the overall care for obstetric complications by 20 % (from 61 to 81 %) for obstetric hemorrhage; 54 to 90 % for eclampsia; 82 to 94 % for obstructed labor and from 66 to 85 % for genital tract sepsis. Zongo et al. [50] reported that women treated at facilities where CBCA was implemented in Mali and Senegal had better clinical examination and postpartum monitoring practices. Similarly, in a study in Malawi [51, 52], the use of CBCA decreased the incidence of uterine rupture by 68 % (from 19.2 to 6.1/1000 deliveries), improved the diagnosis of maternal morbidity, and increased the number of women receiving blood transfusions and caesarean deliveries. Other studies reported that CBCA decreased the aggregate obstetric case fatality rate (CFR) by 33 % in Senegal [53] and contributed to an increase in first ANC visits, facility deliveries (from 31.8 to 34.7 %), and the quality of women-friendly services and satisfaction of women (by 9 % in Uganda) [54]. In addition, CBCA was shown to reduce institutional MMR from 250 to 182 per 100,000 live births, decrease the CFR from 3.7 to 1.5 % [55‐57], improve documentation by health care workers of case notes and maternity registers, and increase caesarean delivery rate from 1.9 to 3.3 % in Burkina Faso [58].

The multifaceted QUARITE (quality of care, risk management, and technology in obstetrics) study conducted in Mali and Senegal [50] and another study on facility-based maternal death reviews conducted in Senegal [59] reported that MDR and on-site training of health care providers on EmOC reduced maternal mortality among high-risk women who delivered through caesarean section. In addition, MDR reportedly decreased obstetric complications and the overall hospital-based MMR by a massive 410 in 100,000 and led to a drop in the proportion of women who had not received any ANC services from around 11 to 4 % at the end of the intervention [59].

Although CBCA and MDR have been shown to be effective in improving the quality of EmOC and the management of obstetric cases, leading to a reduction in obstetric case fatality rates, their strength of evidence is limited. The techniques are often health facility-based interventions and do not provide a comprehensive overview of all cases of maternal deaths in a community. In addition, variability in standard checklists used in clinical audits potentially limits the replicability and comparability of clinical audits.

Strengthening health systems was considered a key pillar to achieving the Millennium Development Goals (MDGs) and will still be key in achieving the Sustainable Development Goals (SDGs) [60]. Inequitable access to quality sexual and reproductive health care in SSA is exacerbated by acute shortage and unequal distribution of skilled and specialized health personnel such as physicians, midwives, and nurses [61, 62]. Efforts to address shortages in specialized care, and inequities in the distribution of skilled health personnel, have sought to build the capacity of available maternal health services providers in the provision of basic and comprehensive emergency obstetric care and in conducting vital quality assurance exercises, such as maternal death reviews [63, 64].

The CARE/AMDD (Averting Maternal Death and Disability) program collaboration is a comprehensive intervention implemented in three high maternal mortality countries in Africa (Rwanda, Tanzania, and Ethiopia) aimed at increasing the availability and quality of EmOC services [65, 66]. The program supported the three countries mainly by training health care workers to provide EmOC services, strengthening existing health information systems (HIS) to monitor change, and identifying gaps in quality of care. Technical leaders and policy makers in these countries were involved in the development and formative stages of internal quality review systems, to establish nationally acceptable standards and guidelines for maternal health care. Following implementation, there was a reported increase in the availability and utilization of EmOC in Tanzania, where the met need for EmoC—defined as a percent of all women with major direct obstetric complications who are treated in a health facility providing emergency obstetric care in a given reference period—was shown to have increased slightly from 14 to 19 % over 4 years, while in Rwanda it reportedly increased met need for EmOC from 16 to 25 %. Obstetric case fatality rates were also shown to have declined by between 30 and 50 % in all the three countries. There was a reported general increase in the level of preparedness for emergencies and the ability of health care workers in these countries to manage common obstetric complications according to evidence-based practices.

Training of health care workers on emergency obstetric care was shown to have impacted positively on the availability and quality of EmOC and improved delivery and care skills among midwives in Somalia [67]. It was also associated with improved knowledge and confidence in carrying out clinical audits among health care professionals. The program reportedly reduced hospital-based mortality in first-level referral hospitals in the country [68]. In Kenya, training of health care workers on EmOC was associated with decreased rates of postpartum hemorrhage [69]. Similarly, health providers trained on providing care for postpartum hemorrhage (PPH) under the Advanced Life Support in Obstetrics (ALSO) program in Tanzania led to a reduction in the incidence of PPH from 32.9 to 18.2 % and that of severe PPH from 9.2 to 4.3 % [70].

The Integrated Maternal Newborn and Child Health Program in Nigeria (an intervention that was based on a network of CHWs, who connect between households and the health facilities) [71]; the FIGO Save the Mothers Initiative in Ethiopia that aimed to establish basic and comprehensive EmOC in order to increase the availability and utilization of quality obstetric care based on UN indicators; and the Maternal Health in Ethiopia Partnership (MaNHEP) in Ethiopia that was based on community-oriented model for improving maternal and newborn health [72‐74] were all reported to have improved access and the delivery of quality caesarean section to women and decreased obstetric case fatality rates. A skilled care initiative was also shown to have led to a 34 % decline in pregnancy-related mortality and the reduction in adverse pregnancy outcomes (obstetric complications and deaths) in Ethiopia and in Burkina Faso, respectively [75, 76].

The quality improvement packages, incorporating, among other things, quality assurance and provision of equipment were shown to have improved performance of health care providers in counseling women on recognizing and managing maternal complications in Kenya [77]. In Nigeria, an obstetric service quality assurance [78], together with improvements in infrastructure, provision of equipment, and continuous maternal and fetal data collection, analysis, and feedback, was shown to have lowered mean institutional MMR from 1790 per 100,000 births in 2008 to 940 per 100,000 births in 2009.

In Burkina Faso, health care providers were trained on routines and the provision of EmOC services, available referral systems were strengthened, health infrastructure was upgraded, and equipment as well as medical supplies including essential drugs were provided. This strengthening of community and health facility systems increased facility-based deliveries from 24 to 56 % and attendance by skilled care providers from 33 to 65 % among the women who reported an obstetric complication. The authors reported an increase in first ANC visits by women and their partners [58, 79]. A study on a participatory quality improvement intervention conducted in 21 health facilities in South Africa enhanced the uptake of CD4 testing from 40 to 97 %. In addition, it showed an improved uptake of maternal and infant nevirapine from 57 to 96 % and 15 to 68 % respectively, as well as an increase in the 6-week polymerase chain reaction (PCR) testing from 24 to 68 % [80], adding to efforts to prevent mother-to-child transmission of HIV.

Improvements in infrastructure, human resource development, transportation and communication systems, and management of health facilities were associated with the growth of fully functional EmOC facilities from 4 to 18 in Mozambique, a steady increase in number of complications treated, and a 50 % decline in deaths from obstetric complications. The met need for EmOC also reportedly increased 3-fold (from 11.3 to 32.8 %) in all facilities while births from caesarean section deliveries increased steadily, doubling at the end of the intervention. There was also a reduction of aggregate obstetric CFR by almost half (2.9 to 1.6 %); with concomitant decline in the deaths from obstetric hemorrhage, obstructed labor, and postpartum sepsis [81]. A community-based basic ultrasound service significantly reduced referrals to a regional center for fetal surveillance and delivery in South Africa [82]. In Uganda, health facility capacity strengthening was shown to have significantly improved the number of pregnant women delivering in the health facility from 55.2 to 99.3 % following the provision of mama-kits (delivery kits) [54]. A quality improvement package incorporating quality assurance for health facilities was linked to enhanced counseling and technical aspects of service provision in family planning and antenatal care in Uganda [83], and in Ghana [84], a 34 % decrease in maternal mortality, a decrease from 3.1 to 1.1 % of case fatality rates for preeclampsia and hemorrhage, and a reduction in stillbirths by 36 %. An overall decrease in maternal mortality ratio from 496 per 100,000 live births in 2007 to 328 per 100 000 in 2009 was also reported.

Community mobilization (CM) interventions seek to empower communities to modify power dynamics and form sustainable environments for better health [85]. CM may improve equitable access to health care services through individual capacity building skills, advocacy, and community actions [86‐88]. Similarly, peer-based interventions or peer support are becoming popular as a strategy to improve access to and utilization of maternal health care especially for pregnant women living with HIV (WLH) [89, 90].

In rural Malawi, coupling facility quality improvements and community activities involving the sensitization and mobilization of the women using participatory approaches was shown to enhance the quality of maternal care provided [91]. A community-based safe motherhood intervention by skilled birth attendants also reportedly promoted the utilization of obstetric care in rural Tanzania [92], while family meetings in addition to labor and birth notification in Ethiopia were associated with an increase in the uptake of institutional-based deliveries [73]. In Kenya, the community health strategy implemented by the government used community health workers to deliver maternal and newborn care through household visits to screen pregnant mothers for danger signs and to provide support for birth planning and referrals. The initiative was reported as having significantly increased ANC attendance among pregnant women from 39 to 62 % as well as deliveries by skilled birth attendants from 31 to 57 %. The program was also associated with increased HIV testing during pregnancy from 73 to 90 % [93].

Community mobilization efforts to sensitize women to deliver in health facilities and encourage participation of men in the ANC attendance and HIV testing were conducted in rural Tanzania [92, 94]. Women were taught to make birth plans including places of delivery, savings to offset costs, and preparations for complications during pregnancy, labor, delivery, and the postnatal period. Findings indicate that the intervention improved both the utilization of health units for delivery and postnatal care within one month of delivery, without any significant negative effect on providers’ and women’s satisfaction with the ANC they provided or received respectively. Women in the intervention were shown to have discussed most elements of the birth plan with their providers, including how to recognize danger signs, identifying a delivery place, and making financial and transport arrangements for delivery. Women in the intervention arm were shown to be 16.8 % more likely to deliver in a health unit than women in the control arm and were 22.7 versus 2.2 % more likely to attend postnatal care within the first 48 h.

Other studies conducted in Burkina Faso, Uganda, and Zambia also found that community mobilization increased the level of health information about danger signs and risk factors in pregnancy and first ANC visits and health facility use and deliveries [54, 95, 96]. Community mobilization also reduced perinatal mortality rates. Similarly, peer mentor interventions by WLH on pregnant WLH at the antenatal and primary health care clinics were linked to significantly higher likelihood of completion of both maternal and infant ARV therapy regimens, increased adherence to all PMTCT tasks through 1.5 months after delivery, increased likelihood of asking partners to test for HIV, and lower likelihood of reporting depressed mood [90].

Home visits by health care workers, and usually community health workers (CHWs), is another strategy used to promote access to maternal health care in SSA for underserved communities [97‐99]. In South Africa, home visits by CHWs to mothers in their neighborhood [89] reportedly improved adherence to condom use among pregnant WLH [100]. Similarly, volunteer peer counselor health education by women’s groups was shown to have decreased MMR by up to 52 % in poor rural populations in Malawi [101]. Birth plan counseling and health education during ANC were associated with increased uptake of skilled delivery and post-delivery care in Tanzania without negatively affecting women’s and providers’ satisfaction with available ANC services [94]. Job aids which are pictorial or illustrated cards used in counseling by nurses and midwives in Benin were said to be positively perceived by providers and pregnant women and associated with significant improvements in birth preparedness, danger sign recognition, safe and hygienic delivery, and even newborn care [102].

An intervention that provided free-of-charge 24-h ambulance and communication services between patients and health care providers reportedly increased access to and utilization of maternal health services, particularly caesarean delivery services from 0.57 to 1.21 % in Uganda. Also reported was a 50 % increase in the number of hospital deliveries, with a slight reduction in the average hospital stillbirths [103]. In South Africa, an inter-facility transportation scheme was associated with decreased institutional-based maternal mortality from 279 per 100,000 live births before implementation of the intervention in 2011 to 152 per 100,000 live births during implementation in 2012 [104]. CURGO (Centre d’UrgenceGyneco-Obstetric), an emergency obstetric care patient transfer system in Burundi, involving the relocation of patients requiring emergency obstetric care from public hospitals to a referral facility set up by the Médecins Sans Frontières was said to have averted 74 % of maternal deaths in Kabezi district [105]. A maternity referral system that included basic and comprehensive emergency obstetric care, transportation to obstetric health services and community cost-sharing schemes in Mali [106] was associated with an increase in the number of women receiving emergency obstetric care and caesarean sections performed for absolute maternal indications from 0.13 to 0.46 %. Institutional deliveries also increased from 19 % at baseline to 39.4 % at the completion of the intervention. Overall, the intervention was said to have reduced the risk of death by half in women treated for an obstetric emergency, with nearly half [47.5 %] of the reduction in deaths attributable to fewer deaths from hemorrhage.

Prevention of mother-to-child transmission (PMTCT) of HIV programs often come as a set of interventions and are reported to be highly effective in the eradication of vertical transmission of HIV [107, 108]. While PMTCT coverage and utilization has remarkably increased over the past decade in SSA, it is still far from the recommended targets [108]. Community mobilization, as previously discussed, was employed in Uganda to encourage male partners’ acceptance of HIV testing through couple attendance at ANC. Findings showed a 16 % improvement in the number of women who attended ANC together with their partners, while 95 % of male partners who attended the ANC together with their spouses were tested for HIV [54, 109, 110]. Training of health care workers to offer PMTCT services and equipping health facilities with supplies in Mali and Senegal was shown to have produced significantly better quality obstetric care results [110, 111]. The number of pregnant women attending first ANC visit was said to have increased 11-fold in Northern Uganda, while the number of male partners counseled, tested, and given results for HIV together with their wives at the first ANC visit also rose significantly. In Cote d'Ivoire, a significant rise in the number of pregnant women delivering in the health facilities was reported following the intervention [112].

Task shifting has been shown to be effective in improving maternal health care provision for patients in LMICs, most notably in the administration of ARV therapy [113, 114], non-communicable disease management [115‐117], and mental health care [118]. Task shifting interventions involve equipping a cadre of staff in the health care system with the appropriate skills to provide services that would otherwise be provided by higher cadre providers, who are often scarce. In some countries in SSA, non-physician clinicians (NPCs) have been used to fill various roles that are conventionally handled by physicians [119‐121]. In Benin, antenatal counseling by lay nurse aides was shown to have resulted in improvements in maternal knowledge among women in prenatal care compared to those counseled by nursing midwives. It also reportedly improved birth preparedness and recognition of danger signs in pregnancy [122]. In Ethiopia, NPCs were said to have performed 63 % of emergency caesarean section deliveries, reducing the number of maternal and fetal deaths experienced and reducing the length of hospital stay for patients in the country [123].