We performed a systematic literature review (27) on mHealth interventions aimed at reducing preventable maternal, neonatal and under-five mortality in Sub-Saharan Africa and Southern Asia. The review was conducted in accordance with the Preferred Reporting Items for Systematic review and Meta-Analysis Protocols (PRISMA-P) (28, 29) to ensure a transparent and complete reporting of the study. The study protocol was developed a priori and published in the International Prospective Register of Systematic Reviews (PROSPERO) [registration number CRD42019109434] (30).

The scope of the review and eligibility criteria were formulated using the PICOS framework (participants, interventions, comparisons, outcomes, study type) (31) (Table 1).

Relevant articles were searched in Medline, Embase and Web of Science. The search strategy was based on the PICOS framework and developed using selected keywords as well as thesaurus terms of each database. The search strategy (Supplementary Table 1) included terms relating to or describing mHealth interventions according to the eligibility criteria. The search strategy was tested before starting the formal screening. The results of each database search were stored into a single reference database (Endnote). Duplicate references were removed. The electronic search was performed by one review team member (EB).

Titles and abstracts were screened, and studies which did not meet the inclusion criteria were excluded. The full texts of the selected studies were then retrieved and independently assessed for eligibility by two review team members (EB, PZ). Any disagreement over the eligibility of particular studies was resolved through discussion and the involvement of another reviewer (MJ). A standardized data extraction form was developed, piloted and used to extract information from the full texts for evidence synthesis (Supplementary Table 2). For each article, results were extracted for the primary and secondary outcomes (Table 2).

The overall quality related to the conduction of quantitative studies and reporting of the results is not always satisfactory (32). Studies of low methodological quality typically tend to report better treatment effects than studies of high quality (33–35). The Effective Public Health Practice Project (EPHPP) Quality Assessment Tool for Quantitative Studies was used to assess the quality of the eligible studies (36), since the EPHPP tool has proven to have a higher inter-rate agreement compared to the Cochrane Collaboration Risk of Bias tool when evaluating quantitative studies (37). The quality of each study was assessed in terms of selection bias, study design, confounders, blinding, data collection methods, and withdrawal and dropouts. A rating is assigned to each of the six components, and a final global rating (strong, moderate, weak) was then assigned. Studies with weak ratings were defined by the quality assessment tool as those with more than three weak component ratings. According to the recommendations by the Cochrane handbook (31) and the EPHPP Quality Assessment Tool for Quantitative Studies (36), studies of weak methodological quality were excluded from the final analysis to improve the quality of the results of this systematic literature review.

Data extracted on the characteristics and findings of the included studies were structured in form of a tabular summary. The findings were analyzed and organized according to study types, mHealth interventions and outcomes.

A total of 1,299 articles were identified from the search strategy (Figure 1). After removing duplicates and initial screening, 154 full-text articles were assessed for eligibility. Following further inspection and quality assessment (Supplementary Table 3), 23 articles were included in this review (19, 23, 38–58).

The characteristics of the included studies, including study type, population, interventions, level of care, and setting, were structured in a tabular summary (Table 3). There were 20 randomized controlled trials (RCTs), one non-randomized controlled trial, one quasi-experimental study and one retrospective cross-sectional survey. Sixteen studies were undertaken in Sub-Saharan Africa, while only seven were undertaken in Southern Asia. All studies except one were conducted in primary care settings.

The vast majority of the studies reported mHealth interventions used for educational purposes. The most common form of delivery was SMS, used either alone as reminders or in combination with a mobile application (Table 4).

This study provides a comprehensive overview of mHealth interventions implemented in Sub-Saharan Africa and Southern Asia with outcomes addressing, directly or indirectly, the Target 3.1 and 3.2 of the SDG 3. We identified 23 relevant studies published from 2010 to 2020, which demonstrate a growing effort in evaluating the effects of mHealth interventions on maternal and child health outcomes. Low access to maternal health services, including skilled birth attendance at delivery, ANC attendance, PNC attendance, and vaccination and immunization coverage, represents a major cause of maternal and neonatal mortality. Interventions aimed at addressing those outcomes can therefore contribute to alleviate the burden of maternal and neonatal deaths targeted by the SDG 3. Despite few studies addressed mortality outcomes directly, the results from this systematic review showed that mHealth interventions are effective at changing the behavior of pregnant women, caretakers and health workers, and at increasing ANC attendance, PNC attendance, childhood vaccination/immunization rates, and skilled delivery attendance. Most of the mHealth interventions covered by this review were delivered with minimal resources and consisted of simple SMS or voice message reminders which are particularly suitable for resource-limited settings.

Only two studies (50, 51) reported results on neonatal mortality, while none of the included studies reported results on maternal and under-5 year mortality. The two studies with results on neonatal mortality found that SMS reminders for pregnant women and smartphone-based applications for education of health workers can be effective in reducing neonatal deaths in resource-limited countries. The evidence of the impact of mHealth interventions on neonatal mortality is, however, limited since it is based on two studies only, both conducted in the Eastern part of Sub-Saharan Africa. More studies conducted in the two SDG regions covered by this review, and especially in Southern Asia, are still needed to support the evidence of the effects on neonatal mortality. This is also necessary to avoid any divergence in the progress toward the SDGs between the two regions. Overall, only seven of the 23 studies included in this review were conducted in Southern Asia. A number of studies from Southern Asia were excluded from this review since they obtained a weak rating in the quality assessment (Supplementary Table 3).

Skilled birth attendance is recognized as a crucial intervention to avert preventable obstetric complications, which are one of the main causes of maternal and perinatal mortality (59). The proportion of births or deliveries attended by skilled birth personnel is a vital indicator for harnessing the SDG Target 3.1. Access to skilled birth attendance is also addressed by the Global Strategy for Women's, Children's and Adolescents' Health 2016–2030 and by the framework for ending preventable maternal mortality 2015–2030 (60). The coverage of skilled birth attendants in Sub-Saharan Africa is lower compared to all other regions in the world (61). Moreover, the MDG tracking report (62) stated that the proportion of births in the presence of a skilled attendant in Sub-Saharan Africa increased only by 7% between 1990 (41%) and 2008 (48%).

Evidence from this systematic review demonstrated that mHealth interventions, such as the Wired Mothers, have the potential to support skilled birth attendance which, in turn, can contribute to reduce the global maternal mortality (48). Other studies concluded that mHealth interventions can support behavioral change and influence women's choice of delivery attendance (63, 64). The Safe Delivery Application study conducted in Ethiopia demonstrated that mHealth interventions can improve knowledge and skill acquisition of healthcare workers as well as quality care during emergencies in child birth (51). Such findings are instrumental in addressing the challenges in the provision of quality care due to the absence of continuing education programs. Findings from another study conducted in Ethiopia on the use of a mHealth intervention to improve delivery and PNC (58) supported an earlier study conducted in Zanzibar testing the ability of SMS reminders to improve skilled birth delivery at births (48).

Having access to ANC and PNC has a significant impact on neonatal mortality as well as an indirect role in reducing maternal mortality by encouraging women to deliver with assistance of a skilled birth attendant or in a health facility (65). Interventions encouraging frequent ANC contacts allow pregnant women to prepare for delivery and be acquainted with educational information on the warning signs of poor maternal or infant health during pregnancy and childbirth. Likewise, interventions aimed at promoting PNC allow mothers and babies to be assessed during visits and can mitigate the risks of maternal and infant morbidity and mortality, which are highest in the days and weeks following childbirth (66–68). mHealth interventions have the potential to support the uptake of these maternal and child health services, thus contributing to reduce regional differences in infant and maternal mortality.

The Wired Mothers (48) and the SUSTAIN-MNCH (45) studies indicated that mHealth interventions delivered through smartphones can contribute to attain the World Health Organization's (WHO) recommendation of four ANC visits in the course of pregnancy (at weeks 14, 24, 30, and 36) and delivery at a health facility (69, 70). Such interventions are especially useful in disadvantaged settings burdened with poverty and low rates of facility delivery. Additionally, the SUSTAIN-MNCH intervention and the Safe Delivery Application confirmed the importance of training of health workers through smartphone applications to enhance uptake of health facility delivery and skills improvement on newborn resuscitation, thus supporting the global effort in reducing neonatal mortality. The results from the Safe Delivery Application support the findings from a review showing that neonatal resuscitation reduces neonatal and perinatal mortality (71). Moreover, ensuring that every child delivery is done in the presence of skilled health personnel is a key to reduce the global maternal deaths (72). The current systematic review showed that access to skilled birth delivery and health facility delivery, as part of mHealth interventional strategies in the continuum of care, can contribute to reduce mortality rates as addressed by the SDG Target 3.1 and 3.2.

A UNICEF report in 2015 acknowledges childhood diseases as the cause of most neonatal and under-five deaths in Sub-Saharan Africa (73). A simple and effective way of protecting children and minimizing the spread of diseases is immunization. The main target of the immunization agenda 2030 is to make immunization achievable for everyone, everywhere, at every age, by 2030. A number of studies have confirmed that immunization is a cost-effective public health intervention capable of contributing significantly to the SDG 3.2 (i.e., end avoidable deaths of newborns and children under five) (74–76).

The results from this review are indicative of how simple mHealth interventions, including SMS reminders (40, 44, 46), voice message reminders (42, 52, 54) and voice calls (41), are effective at improving vaccination and immunization coverage in resource-limited settings where coverage is low, such as Sub-Saharan Africa and Southern Asia. The available evidence also shows that mHealth interventions can contribute to reduce delays in immunization and improve immunization compliance rate (40, 41), thus increasing adherence to timely receipts of immunization and immunization completeness (43, 54). This will ultimately contribute to the Global Vaccine Action Plan (GVAP), which aims to achieve at least 90% national vaccination coverage and herd immunity level of immunization coverage where the chain of disease transmission is broken. Improved vaccination and immunization coverage will, in turn, support the SDG Target 3.2 of ending avoiding deaths of newborns and children under-5 years of age.

According to the scope of this systematic literature review and its eligibility criteria, only papers published in peer-reviewed journals were included. As a consequence, other sources which might include useful results on maternal and child mortality, such as reports from non-profit organizations or white or gray literature, were excluded. Moreover, the study included only papers published in English. It is possible that studies reporting results of mHealth interventions conducted in Sub-Saharan Africa and Southern Asia were published in another language than English. It is also worth noting that there was an overlap among some of the studies included in this review. In particular, three papers reporting on use of SMS reminders were linked to the same program in Zanzibar, Tanzania (48–50). The generalizability of the findings from this review is affected by the variety of the mHealth interventions described in the included studies.

This systematic literature review provides a useful contribution toward the role of mHealth interventions in achieving the UN's SDG 3. The results showed an increasing number of evidence-based interventions aimed to make pregnancy and childbirth safer for both mother and child by alleviating the burden of maternal and neonatal deaths. The findings from this study can inform policy makers and serve as a basis to provide recommendations on the implementation of mHealth interventions in resource-limited settings, including SMS messages to improve maternal health-seeking behavior and reminders to caretakers in the health delivery structure.

With most maternal, neonatal and child deaths taking place in Sub-Saharan Africa and Southern Asia, a need for action is required to achieve the SDG Target 3.1 and 3.2. If innovative interventions supporting maternal, neonatal and child health are implemented in the healthcare delivery system, more than half of the global 56 million deaths under-5 year could potentially be prevented between 2018 and 2030 in Sub-Saharan Africa and Southern Asia (77). The potential presented by mHealth interventions is promising and offers new hope for the future. The progress demonstrated in the two SDG regions included in this systematic review shows that the maternal and child health can be improved through mHealth interventions aimed to increase access to skilled delivery attendance, vaccination coverage, immunization uptake, ANC and PNC. Increasing women's access to quality care, from the antenatal stage to the post-natal period, also contribute to improving equity and reducing health disparities between high-income and low-income countries. This ultimately contributes to a growing evidence that mHealth can support the SDG 3 (78) by reducing maternal and neonatal mortality in the post 2015 agenda of the UN.