This paper is a systematic review of systematic reviews. The review addressed the following research questions: (1) What mHealth apps are used for maternal health care delivery globally? (2) What is the effectiveness of mHealth apps for maternal health care? and (3) What are the barriers and facilitators in the use of mHealth apps for maternal health care delivery?

We conducted this study in line with the updated guidelines for PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses; Multimedia Appendix 1) statement and the methodological considerations when including existing systematic reviews [35].

We developed a protocol to guide the conduct of the study. The protocol was registered with the International Prospective Register of Systematic Reviews (PROSPERO).

Our search focused on published systematic reviews that focused on specific mHealth apps used in delivering any maternity health care (ie, prenatal, birth, or postnatal) and newborn care, subsumed under maternal care. A newborn refers to a neonate, thus, a child within 28 days of birth, as conceptualized by the WHO [36]. A systematic search for systematic reviews was executed in 5 databases: Scopus, Web of Science (Core Collection), MEDLINE or PubMed, CINAHL, and Cochrane Database of Systematic Reviews, using a predeveloped search strategy (Multimedia Appendix 2). The aforementioned databases were used due to their dominance in biomedical reviews. Searches were done on October 20, 2022. Besides, relevant references of retrieved papers were searched manually to retrieve other relevant papers.

This paper included systematic reviews involving primary studies of all study designs (eg, randomized controlled trials [RCTs], pre- and posttest designs, nonrandomized trials, and observational studies). We included reviews focusing on the effectiveness of all available mHealth apps used for maternal and newborn health care delivery globally, for example, service delivery apps. A paper was considered a systematic review if it had these five characteristics: (1) clearly defined aim or research question, (2) eligibility criteria for included studies, (3) appropriate search strategy, (4) appraisal or quality assessment, and (5) analysis or synthesis [37]. Besides, the Population, Intervention, Comparison, and Outcomes framework was applied in determining the eligibility. Thus, the population of interest was pregnant women, women in labor or at birth, postnatal women, and neonates. The intervention of interest was mHealth as defined in the Introduction section. Where reported, the comparison was the women or neonates who were not treated with the assistance of any mHealth app, and the outcomes were the reported results that emerged after using mHealth app for any maternity or neonatal condition. Only studies in the English language were considered without year limits. Due to the language competency of authors, only papers published in the English language were considered as English is the working language of the authors. We also excluded papers focusing on nonmaternal health issues or reviews on interventions other than mHealth.

Reviews were imported to EndNote (Clarivate), and duplicates were removed, after which title and abstract screening were performed by 2 authors (EKA and PAA) using the inclusion and exclusion criteria. This was done while blinding each other’s decision, and all discrepancies were discussed for resolution. Initial screening was on the titles and abstracts retrieved from the search. All suitable papers were retrieved for full-text evaluation. Afterward, we conducted a full-text assessment to determine the eligibility for inclusion. The journal, authorship, or years were not blinded in the process. Reference lists of included papers were searched for additional suitable papers, and we had no reason to contact authors as our results were exhaustive.

We extracted data following a standardized extraction form (EKA), after which the data were reviewed by a senior researcher (PAA). Specifically, we extracted the following data: author information and year of publication, setting or context, objective, time range of included papers, databases searched, number of included papers, design of included papers, and whether the meta-analysis was conducted or not (Table 1). In addition, we extracted data on the description of included mHealth apps, the target population, the effectiveness of the mHealth apps, and other relevant outcomes of interest as shown in Table 2. All discrepancies were discussed for resolution. By way of definition, (1) mHealth apps refer to the specific mHealth technique and tools that were used; (2) the target population could either be pregnant women, women in labor or childbirth, postnatal women, and neonates; and (3) effectiveness refers to the outcomes reported after using the reported mHealth technique or intervention. We presented a narrative synthesis of the findings by highlighting the specific mHealth apps, where they are implemented, and their effectiveness or outcomes toward various maternal conditions, such as the impact on maternal health care use. Each of the authors (EKA, PAA, and OE) independently assessed the quality of the reviews using the critical appraisal instrument for systematic reviews and research synthesis by the Joanna Briggs Institute (Multimedia Appendix 3 [38]). This is a 12-point critical appraisal tool for assessing the quality of a systematic review to determine if specific papers could be included or otherwise. Reporting was guided by the Preferred Reporting Items for Overviews of Reviews (PRIOR).

Our rigorous systematic search retrieved 2516 papers, with 168 duplicates thereby leaving 2348 papers (Figure 1). An additional 11 papers were identified through cross-referencing and free-hand search. As a result, we screened the titles and abstracts of 2359 papers, and 58 papers were eligible for full-text screening. Of these, 19 papers were deemed eligible for inclusion. However, 3 papers were excluded due to limited data, hence 16 (8.1%) reviews out of 198 studies were included in the final review (Figure 1), as no study was excluded based on quality assessment outcome. Of the 198 studies, 29 (14.6%) studies were included in more than 1 review and they were counted once. These reviews were published between 1996 and 2020 (Table 1), and 11 were meta-analyzed [27,34,39, 43,45-50,52]. In all, 8 included papers on RCT designs only [27,39,40,45,46,49,50,52]; 7 had studies of RCTs and other designs such as controlled clinical trials, cohort studies, non-RCTs, quasi-RCTs, and pilot RCTs [34,41-44,47,48]; and 1 review had no RCT design [51].

Most of the reviews were conducted in HICs (13/16, 81%), predominantly from the United States (n=11) [27,34,40-47, 49,50,52], United Kingdom (n=7) [39,41-43,46, 47,49], Australia (n=6) [27,39,40,42,47,49], the Netherlands (n=4) [27,46,47,50], and Norway (n=2) [41,49]. Only 3 of the reviews were conducted in LMICs (such as Tanzania, Thailand, Kenya, South Africa, Ethiopia, Nigeria, and India) [48,51,52]. In total, 11 (69%) of the 16 reviews focused on pregnant women alone [27,34,40-45,48-50], 2 focused on both pregnant and postpartum women [39,51], 1 was on both pregnant women and women intending to be pregnant [46], and another was focused on pregnant women and their partners [47] (Table 2).

Different forms of mHealth interventions are implemented to augment the delivery of maternity health care globally. As illustrated in Table 2, the dominant channels or platforms by which mHealth interventions operate are mobile phones [42,43,51] usually in the form of SMS text messages [45-47,49-51]. Quite a significant proportion (3/16, 19%) of the mHealth interventions are delivered through websites [27,39,45]. With respect to the specific maternal conditions, a substantial proportion of the mHealth interventions were used to assist pregnant or postnatal women in overcoming anxiety and depression [27,39,44,50], assist with diabetes in pregnancy [41,45], and gestational weight [42].

Some were also used to help mitigate smoking cessation [43,47] and alcohol prevention or substance use in pregnancy [44,46,49] or to support ANC attendance and facility delivery [48,51,52]. One review focused on asthma in pregnancy [40]. For instance, in the review by Ming et al [45], they assessed modem transmission of blood glucose readings to a centralized health facility’s telephone system that interpreted blood glucose readings into audio tones by transmitting them to a database. In the case of Farzandipour et al [42], a greater proportion of their studies used the telephone for prenatal weight monitoring, and provision of educational information. The review by Griffiths et al [43] involved interventions delivered by video or DVD, computer (either PC or laptop), mobile phone, or any portable handheld device [43].

Considering that different mHealth interventions targeted different conditions in pregnancy and postnatal phases, varied effectiveness outcomes emerged. Meanwhile, all the studies revealed that indeed mHealth interventions are effective for maternal health care, although at varying magnitudes or levels of effectiveness. On depression, there was a uniform agreement among all studies on this condition, demonstrating the extent to which mHealth interventions help mitigate the condition (Table 2). All the reviews on depression highlighted that mHealth interventions were helpful in subsiding depression in pregnancy and childbirth. In the review by Bayrampour et al [39], 4 out of 5 studies revealed positive impacts, with 1 reporting that only 18% of the pregnant and postpartum women met the clinical criteria for depression relative to 79% in the control group, after a 12-week assessment. Similar findings emerged from the review by Lee and Cho [44]. Meanwhile, 2 reviews concluded that mHealth intervention had minimal effect on depression among pregnant women [27,50].

mHealth interventions were reported to be effective in enhancing maternal health care use, as concluded by all the 3 studies focusing on maternal health use. These manifested in ANC enhancement [52], skilled birth attendance at birth leading to a decline in perinatal mortality [51], or both ANC and the use of health facilities for childbirth [48]. Relatedly, the mHealth interventions focusing on alcohol or drinking and smoking showed effectiveness in diverse ways as indicated by all 4 reviews focusing on these [43,44,46,49]. Another area where significant impacts were recorded was obesity or overweight management, especially during pregnancy [42,47] and diabetes management in pregnancy and newborn outcomes [41,45]. For instance, 66% of studies included in the review by Farzandipour et al [42] revealed positive impacts on gestational weight gain. Some studies in the review by Eberle et al [41] also reported marginally higher neonatal birth weight in the intervention group relative to the nonintervention group.

In total, 3 reviews reported that women were satisfied with the mHealth interventions, partly due to the positive user experience and the benefits [39,45,47]. Among the 4 reviews, 1 highlighted dissatisfaction arising from text messaging–based mHealth interventions due to fatigue caused by multiple messages [51]. The same review indicated that the use of mHealth interventions is compromised by privacy issues in instances where the mobile device (eg, mobile phone) is coshared among family members, when women do not own mobile phones, or when there is illiteracy. Two reviews indicated that acceptability and use of mHealth is enhanced by providing information in varied local languages, using locally based software, and when the target is broadened to include women’s significant others [44,51]. Meanwhile, low literacy rate, low accessibility to mobile phones among women in rural locations in LMICs, and uncertainty regarding whether a text is received by the women were some of the identified barriers [51].

There is sufficient evidence that mHealth apps effectively mitigate diverse conditions that pregnant and postpartum women encounter. All 16 reviews converged that mHealth apps are useful for maternity health conditions. The specific areas where positive impacts manifested include increased use of maternal health care services (ANC, skilled birth, and postnatal care), reduction in perinatal deaths, weight management in pregnancy, alcohol or drinking and smoking mitigation, as well as diabetes management. These conditions represent some of the leading causes of maternal and newborn morbidity and mortality [53].

The positive impact of mHealth interventions on maternity health resonates with prevailing evidence on the implications of mHealth on overall health care such as the review by Marcolino et al [30], which also revealed positive impacts after synthesizing evidence from available systematic reviews. Additionally, consistent evidence has emerged from different parts of the world [54-56]. Continuous and effective use of mHealth interventions in maternal health care can, therefore, constitute a cornerstone strategy for achieving the first and second targets of the third sustainable development goal, which enjoins all countries to attain fewer than 70 maternal deaths per 100,000 live births and reduce neonatal mortality to no more than 12 deaths per 1000 live births by 2030.

Most of the reviews reported evidence from HICs, which is suggestive of limited use of mHealth interventions for maternal health care services in LMICs. As we did not restrict our search to HICs, it is unlikely that evidence from LMICs is missed due to eligibility issues. Hence, we can infer that there is limited use and evidence about mHealth and maternal health services across LMICs. The United States, United Kingdom, and Australia were the dominant locations where mHealth interventions were implemented and assessed. This situation is quite worrying as little evidence exists on the use and effectiveness of mHealth interventions in LMICs, where worse maternal and neonatal morbidities and mortalities occur. Estimates from the World Bank indicate that the United States has a maternal mortality ratio (MMR) of 19, while the United Kingdom and Australia have an MMR of 6 each, as of 2017 [57]. In the same year, African-based LMICs such as South Sudan, Chad, and Sierra Leone recorded MMRs of 1150, 1140, and 1120, respectively [57]. Hence, LMICs with alarming MMRs need to create a more conducive environment for mHealth to thrive. These may include conscious and continuous strengthening of internet connection and electrification, especially in deprived settings.

Several factors could account for the divide in mHealth use for maternal health care between HICs and LMICs. Mobile devices and the internet are essential prerequisites for mHealth activities to thrive. Meanwhile, mobile phone ownership and internet access are quite difficult in most LMICs. For instance, 1 of the 3 reviews that focused on LMICs reported that some women share mobile phones with other family members and this was a setback toward the maximization of the benefits of mHealth interventions [51]. A recent report on mobile phone surveys in 9 LMICs (Colombia, Ghana, India, Indonesia, Kenya, Mozambique, Nigeria, Rwanda, and South Africa) showed that 90% of people in LMICs do not have a decent internet connection [58]. The report also revealed that less than half of the people in LMICs have access to basic internet. The extent to which limited internet access compromises the operation of mHealth in LMICs has been echoed [59-61]. Undeniably, mobile phone access and use in HICs far exceed LMICs [62] and seem to have no relevance without the internet. All these factors might have culminated in the preponderance of mHealth interventions in HICs relative to LMICs.

mHealth apps are yielding enormous positive outcomes for maternal well-being. However, the use of mHealth is skewed toward HICs, relative to LMICs where worse maternal conditions emerge from. This is, however, not surprising given the disparity in internet access between LMICs and HICs [63]. Another contributory factor may be the relatively low government expenditure on health care in LMICs, manifesting in limited resource allocation to health care. As of 2019, LMICs averagely spent 5.32% of the gross domestic product on health care as compared with 12.49% average health care expenditure in HICs [64-66]. Hence, an increment in the overall allocation to health care could make it feasible to invest in the most cost-effective mHealth facilities that can offer women the opportunity to benefit from some mHealth interventions that are implementable given the context and acceptability. For instance, simple text message–based mHealth interventions are implementable with relatively minimal resources relative to more advanced options like audio and video conference interventions [67,68].

The use of mHealth interventions typically requires some basic literacy skills [33].

The overall literacy rate of female individuals may, therefore, have to be enhanced for them to benefit from the mHealth interventions. Countries intending to enhance the benefits of mHealth, especially those in LMICs that do not currently have universal compulsory education, ought to give fair consideration to compulsory education or literacy skills training to ensure that all persons attain at least second-cycle education. This could better place female individuals and their partners to use current and future technological innovations not only in health care, but also in all spheres of life. We are not oblivious to the reality that some LMICs already have universal compulsory education systems [69,70]; for those countries, the fortification of such educational arrangements can enhance gains in the long run. More importantly, backing this with a formidable policy framework can guarantee sustainability and effective implementation of mHealth.

For instance, considering internet connectivity challenges in LMICs compared with HICs, text messaging mHealth interventions may be suitable and relatively easily implementable with political commitment backed by a rigorous policy framework. Second, as mHealth interventions require some level of literacy, there is a need for a conscious effort to reduce illiteracy rates. By this, all persons especially women will be well positioned to use mHealth interventions and ascertain the associated benefits. Globally, governments might have to build partnerships or intensify partnerships with the private sector, particularly the telecommunication companies to widen access, as this can offer some leverage for maximizing the use and impact of mHealth interventions for the well-being of women and newborns.

All countries intending to implement mHealth or maximize the benefits of mHealth could implement national mHealth policies that prioritize maternal issues and also invest in internet connectivity and electrification. These are the critical factors that can bridge the gap between HICs and LMICs as well as the rural and urban residents [71,72]. When women and their partners have uninterrupted internet and electricity and a generally suitable environment for mHealth use, minimal advocacy from ANC providers, nongovernmental agencies, and central government can translate into high use and acceptability. By so doing, the alarming MMR battling LMICs could be reduced, with the bridged disparity between HICs and LMICs, leading to a global decline in MMR.

Future research on this subject could focus on how mHealth interventions work out for special populations such as the people with physical disabilities; persons living with HIV in stereotyped settings; teenagers or adolescents; and persons identified with lesbian, gay, bisexual, transgender, and intersex sexual orientations. These populations are sometimes either not comfortable or have truncated access to services meant for the general population [73-76]. Identification of workable mHealth interventions for special populations will augment efforts to attain the global commitment encapsulated in the third sustainable development goal “ensure healthy lives and promote wellbeing for all at all ages” [77].

To our knowledge, this is the first systematic review of systematic reviews that has exclusively investigated the effectiveness of mHealth apps or interventions on maternal health globally. Besides, we included all available reviews irrespective of study design and year of publications. This aided in identifying a wide array of mHealth interventions and the myriad of maternity conditions they are used for. Notwithstanding the strengths, the study is not devoid of limitations. First, despite the comprehensive search, we only included reviews written in the English language, which happens to be the working language of the authors. Second, the search in PubMed did not include Medical Subject Headings terms. However, in addition to the database searches, we performed citation searches, specifically forward and backward citation searches, to identify additional reviews that might have been missed in the database search. This was done to identify additional reviews that might have been missed in the database search.

The effectiveness of mHealth apps in maternal health care is well established in the literature. This manifests in the positive impact toward management and reduction in anxiety and depression, diabetes in pregnancy, and gestational weight. Other aspects include smoking reduction or cessation during pregnancy and enhancement in maternal health care use. Meanwhile, mHealth apps are skewed toward HICs, likely due to the advancement in technology and other resources required to optimize the inherent utility of mHealth apps. Considering that mHealth apps dominate in HICs, LMICs must consider pragmatic approaches to improve their availability and use. Such approaches may include the adoption of workable mHealth policies and political commitment through increment in budget allocation for health care to offer cost-effective mHealth interventions for maternal health care delivery (eg, nonandroid text messages).