Introduction

Pregnancy and childbirth are normal physiological processes. For most women in high-income countries, pregnancy is associated with a feeling of pride and immense joyous expectation [1]. However, for millions of women and their families in low- and middle-income countries (LMICs), where emergency obstetric care is limited, pregnancy and childbirth are a major cause of fear and anxiety [2]. While most women have normal pregnancies and safe deliveries, unanticipated obstetric complications and emergencies sometimes occur. Many causes of maternal mortality such as severe bleeding during and after childbirth, post-delivery infections, obstructed labor, and blood pressure disorders are preventable or treatable conditions [3]. In resource-poor settings, where many women deliver at home or in inadequately equipped health facilities, ensuring that those who develop obstetric emergencies during childbirth are quickly transported to facilities where they can receive quality emergency obstetric care can be the difference between life and death for the pregnant woman and her fetus. Unfortunately, referral to needed emergency obstetric care may not be possible for a plethora of reasons, including geography, cost, and lack of transportation [4].

Delays in reaching healthcare facilities for emergency obstetric care in LMICs can be reduced through implementation of transportation programs [5]. Transportation interventions for emergency obstetric care may include financing schemes that enable pregnant women to overcome barriers of transportation to health facilities for emergency obstetric care during labor and delivery. This may take the form of direct provision of transportation to healthcare facilities for pregnant women in need of emergency obstetric care. Examples include motorbike ambulances specially engineered for use in rough terrains in resource-limited communities, bicycle ambulances, cycle rickshaws, wheeled stretchers, canoes, and ox carts. [6, 7, 8, 9]. Transportation interventions seek to decrease delay in reaching a health facility for emergency obstetric care, and they may contribute to reductions in adverse pregnancy and birth outcomes, including maternal deaths, stillbirths, and neonatal mortality in LMICs [10].

In their seminal paper, “Too far to walk: Maternal mortality in context”, Thaddeus and Maine [11] presented a three-delay framework for analyzing maternal mortality in LMICs (Figure 1). Phase 1 delay refers to delay in the recognition of potentially life-threatening complications/emergencies and decision to seek care at a healthcare facility; Phase II refers to delay in time to reach a healthcare facility; Phase III delay refers to delays in receiving care once a woman reaches a healthcare facility [11, 12, 13, 14]. Numerous social factors influence the decision to seek care, including lack of knowledge about the seriousness of complications, not knowing where to receive care, and/or waiting to receive permission from the husband or other family decision-makers [15]. Furthermore, an analysis of Demography and Health Survey (DHS) data from 41 countries showed that the most common obstacles to seeking obstetric care were financial barriers (>50%), challenges with transportation (37%), and distance (37%) [16]. Lack and high costs of transportation, poor road conditions, and time to arrange transport may also increase the time to reach a health facility [17, 18, 19]. Emergency obstetric transportation interventions are designed to address Phase II delays, i.e., delays that occur after the decision to seek care is made and before obtaining obstetric care. Thus, this review focused on assessment of the effects of emergency transportation interventions that were implemented to address Phase II delays aimed at reducing adverse pregnancy and birth outcomes in LMICs.

Faced with the challenge of unacceptably high maternal mortality rates, community organizations in many LMICs mobilize to provide free emergency obstetric transportation for pregnant women in need. For example, in several communities in northern Nigeria, where the maternal mortality ratio is more than twice the national average [20, 21], the National Union of Road Transport Workers (NURTW) in conjunction with the Amalgamated Commercial Motorcycle Riders Association of Nigeria (ACOMORAN) operate jointly to provide emergency transportation for pregnant women. Other types of obstetric emergency transportation schemes include community health insurance and pre-payments, conditional cash transfers, vouchers, loans, and revolving funds aimed at alleviating the cost of transportation to needed emergency obstetric care [21, 22, 23, 24, 25, 26]. In Kenya, the Maternal and Newborn Improvement (MANI) project uses a transport voucher to assist poor pregnant women to access health services [27]. Por et al [28]. established a voucher scheme and other financial incentives aimed at increasing access to skilled birth attendants for poor women in three rural health districts in Cambodia. Evaluation of this program showed that the scheme increased facility-based deliveries from 16.3% to 44.9% over a two-year period. This marked increase in skilled birth attendance was attributed to lessened financial burden on families [28, 29]. Similar maternal voucher schemes aimed at reducing transport barriers for poor pregnant women have been described in Bangladesh [5] and Pakistan [30].

Although many obstetric emergency transportation interventions are being implemented in LMICs, there is limited empirical evidence to show their effect on reducing adverse outcomes associated with labor and delivery. The only available study on the subject is a 2015 systematic review that focused on only community-based loan funds for transportation during obstetric emergencies in developing countries [31]. This review demonstrated that compared to women in the control communities, those in sites where community-based loan funds were implemented experienced less maternal mortality, a higher rate of facility-based deliveries, and increased utilization of emergency obstetric care. This review is limited in scope, however, given that loan funds represent only one type of intervention that can be implemented to reduce the financial barriers that women in LMICs face in accessing transportation for obstetric care. To increase the scope of available evidence, this review summarized and critically appraised available data on the effect of all forms of interventions and financing mechanisms to promote transportation for emergency obstetric care in LMICs. Findings from this review may help to inform the global debate on access to routine and emergency obstetric services in LMICs.

Methods

The protocol for this review was registered in PROSPERO (ID: CRD42017080092), and the review followed standard systematic review methods [16] and the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) [32] [Supplement 1]. This study did not require approval from the Internal Review Board because it used data from published studies.

Search Strategy

We searched the following databases through December 30, 2019: Medline/PubMed, EMBASE, Web of Science, EBSCO (PsycINFO and CINAHL), SCIELO, LILACS, JSTOR, POPLINE, Google scholar, the Cochrane Pregnancy and Childbirth Group’s Specialized Register, and the Cochrane Central Register of Controlled Trials. No date or language restrictions were applied. The search strategy is described in detail in a previously published review protocol [33].

Inclusion Criteria

We included quasi-experimental studies, randomized controlled trials, controlled before-and-after studies, and cohort studies with control that assessed the effect of transportation interventions on pregnancy outcomes in LMICs. The target population were women who had prenatal, intrapartum, or post-natal care for an obstetric complication and were referred from the community or a primary health care center to a higher-level facility that could provide emergency obstetric care. The interventions of interest in this review included direct provision of transportation services as well as financing schemes or in-kind initiatives that enabled financial challenged pregnant women to overcome barriers of transportation to health facilities for emergency obstetric care in the prenatal period, during labor, delivery, or up to 42 days after delivery (postpartum period). Detailed description of the interventions of interest and conceptual framework is described in this review’s published protocol [33] and summarized in Figure 2. Comparison groups were women who had no transportation interventions for prenatal, intrapartum, or post-natal obstetric complications. Primary outcomes included mortality (stillbirth, maternal mortality, and neonatal mortality). Secondary outcomes included reduced delay in access to care, facility referral rates, time taken to reach appropriate healthcare facility, facility delivery rates, and met need for emergency obstetric care (i.e., receipt of emergency obstetric care by those that needed it). This review was restricted to studies conducted in countries designated as LMIC according to the World Bank’s classification [34].

Study Selection

A study eligibility form was used to screen studies for inclusion [33]. Two reviewers independently screened the titles and abstracts of citations to assess their eligibility for inclusion. Thereafter, the full texts of eligible studies were independently reviewed by two review authors. Disagreements were resolved by discussions within the review team.

Data Extraction

Using a modified Cochrane Collaboration’s data extraction form, two reviewers independently extracted data from eligible studies [35]. Data were extracted on study setting, design, participants’ characteristics, interventions, controls, and duration of follow-up. Study sample size, age, and data collection methods were recorded (Tables 1, 2, 3). Where necessary, authors of included studies were contacted for additional information or missing data.

Quality Assessment

Two review authors assessed the quality of included studies using the Risk Of Bias In Non-randomized Studies of Interventions (ROBINS-I) tool [36]. The domains of the tool include confounding, participant selection, intervention classification, deviation from intended intervention, missing data, measurement of outcomes, and selection of results reported. Disagreements between the two assessors were resolved by discussion and consensus, with arbitration by a third reviewer as required. In line with ROBINS-I guidelines [36], each criterion was scored as “low risk,” “moderate risk,” “serious risk,” “critical risk,” or “no information” (Table 4). An overall risk of bias judgment was made in accordance with the guidelines of ROBINS-I [36].

Data Analysis

There was significant clinical heterogeneity of included studies, which made it impossible to perform a meta-analysis. Therefore, a systematic review of the nine eligible studies was conducted by summarizing, comparing, and contrasting the extracted data. We did not do a GRADE evidence summary of the systematic review because some of the included studies have multiple intervention sub-components, which might make a summary of evidence misleading. For instance, regarding maternal mortality, a primary outcome, an included study (De Costa et al. [37]) provided financial support, while another study (Patel et al. [38]) used a combination of three-wheeled motorcycles and dual-SIM phones to achieve reduced maternal mortality.

Results

The search identified 742 titles, whose screening, along with their abstracts, resulted in 49 potentially eligible studies. Following full text review, 40 were excluded for not meeting the inclusion criteria, resulting in the nine that were included in this review (Figure 2) [37, 38, 39, 40, 41, 42, 43, 44, 45].

Sample Characteristics

As shown in Table 1, the included studies, all published in English between 2000 and 2019, were conducted in Asia, (India [37, 39, 40]) and Africa (Malawi [41, 42], Mali [43], Uganda [44, 45], Zambia [45], and Ghana [38]).

Three controlled before-and-after studies [37, 38, 44], four uncontrolled before and after studies [40, 42, 43, 45], one cluster randomized controlled trial [39], and one case-control study [41] were included. Eight of the nine studies were conducted in rural settings. Participants were identified through obstetric cases and hospital records. There was significant clinical heterogeneity in sample characteristics, including size (157 to 6926 pregnant women), participants’ age (15–45 years), exclusion criteria, and a marked variation in length of follow-up (6–36 months).

Intervention Characteristics

Interventions used various strategies: (a) transportation financing strategies, (b) communication and transport systems, and (c) community mobilization (Table 2). Six studies (66.7%) provided a new form of transportation to communities including four-wheel drive, motorcycle, or bicycle ambulances [38, 40, 42, 44], while one improved an already existing ambulance service [43]. The remaining two studies mobilized communities to arrange local means of transportation [37, 39]. Communication was enhanced by providing radios and mobile cell phones to health workers and drivers in six studies [38, 39, 40, 43]. All included studies attempted to address cost of transportation as a barrier by offering free transportation services (n = 5) [30, 38, 42, 44], transportation vouchers (n = 1) [45], or using a cost-sharing strategy such as community emergency funds or savings groups (n = 5) [37, 39, 41, 43, 45].

Effects of Intervention

Discussion

This systematic review found limited but promising evidence that emergency transportation interventions in LMICs may be effective in: i) reducing maternal and child mortality, ii) increasing health facility delivery significantly, iii) increasing caesarian sections for women in need of such service, and iv) reducing referral delay. Evidence suggests that integrating emergency obstetric transportation with complementary maternal health interventions (e.g., improved communication) reduces adverse pregnancy and childbirth outcomes. Two of the nine included studies reported maternal mortality reduction, and two showed reductions in stillbirths as well as neonatal and perinatal mortality. Six studies reported increases in facility deliveries, ranging from 12–50%; one study demonstrated 19% reduction in home delivery. There was a significant increase in caesarian sections in two studies, while 23.8% of women referred eventually used the service. There was also a reduction in referral delay by 2 to 4.5 hours with the use of motorcycle ambulances compared to car ambulance. However, the type, quality, and scope of transportation interventions varied significantly, and a critical mass of complementary activities is needed to achieve maximum impact.

Different emergency transport services were used, including bicycle ambulances, motorcycle or motorized tricycle ambulances, 4-wheel drive vehicles, and formal ambulances, each with its own advantages and disadvantages. Formal ambulances and 4 × 4 vehicles, while able to accommodate multiple passengers, were costly, both in terms of vehicle procurement and maintenance. For example, in the Ugandan RESCUER program, the proportion of supervised births from 1995 to 1998 increased from 15% to 27%, and hospital-based maternal case fatalities were reduced by 50% [46]. However, by 2005, when the program was scaled up to 56 districts, the high cost of vehicle maintenance made sustaining the program difficult because of insufficient funds. The one included study that reported on cost found that using motorcycles was a budget-friendly option. In Ghana, motorcycle ambulances were effective, culturally acceptable, and able to navigate small roads in rough terrain, allowing drivers to reach women in locations otherwise inaccessible by larger vehicles [38]. However, in Malawi, cultural beliefs that publicizing labor could summon evil spirits resulted in infrequent use of bicycle ambulances [41]. For both motorcycle and bicycle ambulances, customized vehicle features like an enclosed carriage/privacy screen, mattress, seat belts, and/or extended mirrors offered improved safety and comfort. These findings suggest that transportation interventions must be adapted according to local physical, social, cultural and economic environments.

Evidence from the included studies suggest that emergency obstetric transportation interventions were more effective when integrated within an enhanced referral system or when additional strategic interventions aimed at improving the quality of care at service delivery points are present. This underscores the need for a pragmatic approach to strengthening the healthcare systems at facility and community points while addressing context specific emergency transportation barriers [47, 48]. For example, the intervention described in De Costa et al. [37] included financial support for referrals, incentives for early registration of pregnancy, and training of paramedical staff and traditional birth attendants (TBAs). Similarly, in Uganda, transportation interventions included free-of-charge ambulance service and the provision of 2-way radios, landlines, or cell phones between health facilities [44]. Patel et al. [38] attributed reduced maternal mortality and increased referrals to the use of a 3-wheeled motorcycle and dual-SIM mobile phones. However, no significant effect was observed on the number of facility deliveries and the caesarian delivery rate due to low rates of adherence to some care protocols. While transportation and communication were addressed, the remaining gap in delivery of quality care may have hampered the overall outcomes [38]. The conceptual framework we adopted for this review highlights how underpinning factors of socioeconomic status, accessibility of facilities, and quality of care correlate with each phase of the three-delay model [11].

Conclusion

Considering the scarcity of evaluative studies on emergency transport interventions, this review highlights the need for future studies to employ a variety of study designs (experimental or quasi-experimental designs of large, multi-sited programs) to assess the impact of transport interventions on mortality rates and other related secondary outcomes. Additional operational research is needed to determine the sustainability and cost-effectiveness of emergency transport interventions before recommending wide-scale implementation. Poor quality of healthcare services and financial constraints hinder optimal use of maternal and newborn health services in many LMICs [57, 58]. It is important to examine the relationship between transportation, health care utilization after transportation, and mortality rates. Several challenges have been identified, including high vehicle and maintenance costs, establishing effective communication systems in remote settings, maintaining driver coverage, ensuring equitable access to transportation, and sustainability within a resource-constrained health system. New technologies, such as alternative transportation vehicles or mobile phones, are becoming available in low-income settings and should be evaluated in the context of maternal-newborn health transport systems. Finally, studies need to consistently record process indicators to track program effectiveness such as time to reach referral hospital, which has been found to be critical in reducing maternal and neonatal mortality [59, 60].

Additional File

The additional file for this article can be found as follows: