A scoping review methodology was used to systematically map the evidence from the included studies. The review process involved formulating the review question, identifying relevant studies, selecting eligible studies, charting the data, collating the results, and summarizing the findings.

Eligibility criteria were established based on the JBI population-concept-context (PCC) framework [11]. Studies were included if all PCC elements were met, along with additional criteria for outcomes, study characteristics, language, and publication period. The review focused on English-language studies published between January 2005 and May 2025 to encompass the period of widespread adoption of mobile- and internet-based technologies in SEA. Table 1 summarizes the full criteria.

An academic librarian collaborated with the review team to develop and iteratively refine the search strategy. The search combined controlled vocabulary and free-text keywords related to maternal health, DHIs, and SEA. PubMed, CINAHL, and Scopus were searched as the primary databases. A supplementary search of Google Scholar and the WHO Global Index Medicus, using the Index Medicus for the South-East Asia Region (IMSEAR), was conducted to expand the coverage of regional and gray literature. The reference lists of relevant review articles were also screened for additional studies. Full search strings, filters, dates, and source-specific procedures are provided in Multimedia Appendix 1.

All records retrieved through database and supplementary searches were imported into Covidence (Veritas Health Innovation) for management and duplicate removal. The searches retrieved 1133 records from PubMed (n=455), Scopus (n=300), CINAHL (n=110), Google Scholar (n=200), and IMSEAR (n=68). Of these, 234 duplicate records were removed, leaving 899 records for title and abstract screening.

The primary reviewer (EKY) screened all titles and abstracts for eligibility using Covidence, and each record was independently screened by either ML or HP-C as the second reviewer. During the full-text screening, each article was independently assessed by at least 2 of the 3 reviewers (EKY, ML, and HP-C). Prior to screening, the reviewers completed a calibration exercise using 20 abstracts to practice applying the eligibility criteria. Interrater agreement was assessed using Cohen κ, as generated by Covidence. At the title and abstract screening stage, pairwise agreement was κ=0.35 (n=116) for EKY and ML and κ=0.69 (n=783) for EKY and HP-C. At the full-text screening stage, pairwise agreement was κ=0.66 (n=70) between EKY and ML, κ=0.70 (n=72) between EKY and HP-C, and κ=0.75 (n=17) between ML and HP-C [14]. Disagreements between reviewers were initially resolved through discussion among the 3 reviewers (EKY, ML, and HP-C). When consensus could not be reached, the senior author (SSK) was consulted to make the final inclusion decision.

Full-text screening identified 25 studies that met the inclusion criteria. The primary reasons for exclusion were non–SEA LMIC settings (n=37), wrong populations (n=36), the absence of DHI components (n=28), wrong outcomes (n=11), and wrong study designs (n=6). One study was excluded because its full text could not be retrieved despite several attempts.

A narrative synthesis was performed using descriptive and categorical methods. First, the study and DHI characteristics were described. Second, maternal health outcomes were mapped using the adapted PID framework. Lastly, the target populations and intervention settings in which the DHIs were implemented were summarized. The synthesis revealed patterns and identified evidence gaps. No meta-analysis was conducted.

A quality appraisal or risk-of-bias assessment was not conducted, which is consistent with the purpose and methodology of a scoping review [10,15]. This review used publicly available literature and did not involve human participants or primary data collection; therefore, ethics approval was not required.

Figure 1 presents the full study-selection process. Table 2 summarizes the characteristics of the included studies, DHI modality, contact frequency, intervention duration, follow-up timing, and reported maternal health outcomes. The 25 included studies were published between 2008 and 2025, with 18 (72%) published from 2020 onward [16-33]. Figure 2 shows the distribution of included studies by publication year.

Studies were conducted in 7 SEA LMICs. Indonesia contributed the largest number of studies (n=15, 60%) [16-28,34,35], followed by Thailand (n=4, 16%) [29-31,36], the Philippines (n=2, 8%) [37,38], and Cambodia [39], Malaysia [32], Timor-Leste [40], and Vietnam [33] (n=1, 4% each). No eligible studies were identified from Lao PDR or Myanmar. Fourteen (56%) studies used quasi-experimental designs [19,21-29,34,35,37,40], and 11 (44%) used randomized controlled trial designs [16-18,20,30-33,36,38,39]. Sample sizes were most often 100 or fewer (n=16, 64%) [16,17,19,21,22,25-29,31,32,34-37], followed by 101 to 500 (n=7, 28%) [18,20,23,24,30,38,39] and more than 500 participants (n=2, 8%) [33,40]. Twenty-two (88%) studies enrolled only pregnant participants [16-19,21-38]. The remaining studies included pregnant and postpartum participants [20], postabortion participants [39], and postpartum participants [40] (n=1, 4% each).

Mobile apps (n=14, 56%) were the most common DHI modality [18-21,23-30,32,35], followed by SMS (n=8, 32%) [16,17,31,33,34,36-38]. The remaining interventions used augmented reality [22], automated interactive voice messaging with phone support [39], or a combination of SMS and voice communication (n=1, 4% each) [40]. Contact frequency varied across studies. Daily contact was reported in 7 (28%) studies [17,20,26-28,32,35], multiple contacts per week were reported in 5 (20%) studies [33,36-38,40], and weekly contact was reported in 3 (12%) studies [16,22,34]. Three (12%) studies used other scheduled frequencies: 3 sessions over 4 weeks [29], 56 messages between 13 and 40 gestational weeks [31], and contact every 2 weeks for 3 months [39]. Five (20%) studies provided 24-hour access without a prescribed frequency [18,19,21,23,30], and 2 (8%) did not report contact frequency [24,25]. Intervention duration also varied among studies. Ten (40%) interventions lasted 1 month or less [16,18,19,26-30,34,35], while 6 (24%) studies lasted more than 1 month and up to 5 months [17,21,22,32,37,39]. Seven (28%) studies defined intervention duration by gestational or postpartum timing rather than a fixed period [20,23,31,33,36,38,40], and 2 (8%) studies did not report duration [24,25].

All but one (96%) study implemented a DHI targeting individuals already connected to their respective health care services. The exception [33] recruited participants from remote communities. Seventeen (68%) studies implemented DHIs in primary care settings [16,17,19-25,27,28,32-35,37,40]. Six (24%) studies were hospital-based [26,29-31,36,38]. One study was conducted in a community-based/nonclinical setting through a safe-abortion hotline [18], and another was conducted through outpatient reproductive health clinics [39].

This scoping review aimed to map maternal health outcomes reported in DHI studies conducted in SEA LMICs using the adapted PID framework. Outcomes were represented in all 3 domains, with intermediate outcomes reported most frequently (23/25, 92%), followed by proximal maternal knowledge outcomes (10/25, 40%) and distal biomedical and clinical outcomes (8/25, 32%). Most proximal knowledge outcomes showed significant improvement, while intermediate and distal outcomes showed mixed results. This pattern is consistent with other LMIC-focused reviews of DHIs in maternal health, which reported greater emphasis on knowledge, self-efficacy, and behavioral outcomes than on clinical outcomes [3,8].

Mobile apps and SMS were the predominant DHI modalities, accounting for 22 of 25 (88%) studies. Interventions varied in both contact frequency and duration. Contact frequency ranged from daily or multiple times per week to less frequent scheduled contact, while some interventions provided continuous access without a prescribed frequency. Intervention duration ranged from 1 month or less to several months, or periods defined by gestational or postpartum timing. These findings reflect diverse DHI design and delivery for maternal health in SEA LMICs. The prominence of mobile apps and SMS is consistent with a scoping review of 63 maternal and child health DHI studies in rural and underserved LMIC settings [3].

Intervention settings were identifiable for all studies. Most interventions were implemented in primary care settings, with fewer in hospitals, community-based nonclinical settings, or specialized outpatient reproductive health care settings. Only one study [33] recruited participants from remote communities without requiring an existing health care connection. Thus, this scoping review provides limited evidence on whether DHIs can expand access among populations not already connected to health care.

The studies were concentrated in a small number of SEA LMICs, with 60% (15/25) conducted in Indonesia and none were from Lao PDR or Myanmar. Indonesia, with a population of more than 280 million, is the world’s fourth-most-populous country and has a substantial potential user base for DHIs [41]. The country has established a national digital health transformation strategy that focuses on health data integration, digital health platforms, and a broader digital health ecosystem [42]. Its large population and expanding digital health environment may partly explain the concentration of DHI studies. Evidence from other SEA LMICs was limited, with only one study each from Cambodia, Malaysia, Timor-Leste, and Vietnam. This limited representation highlights gaps in the evidence base and constrains the generalizability of findings across the region.

For maternal health nurses, midwives, and community health workers, this scoping review shows that DHIs have been integrated into primary care to support education, reminders, self-care, monitoring, follow-up, and continuity of care. While mobile app interventions showed improvements in some distal outcomes [26,29], DHIs should be tailored to the target population, care setting, and intended purpose [3,8]. To broaden the reach of DHIs beyond women who are already connected to health care services, maternal health professionals may consider community-based outreach and enrollment strategies.

DHI implementation should account for device access and connectivity, digital literacy, and privacy and communication needs [3,4]. Clear clinical pathways should specify how information is reviewed and when in-person assessment or referral is needed [4]. Digital approaches should complement, rather than replace, existing maternal health services, and appropriate nondigital options should remain available for individuals unable to participate in DHIs [4].

This scoping review highlights research priorities for maternal health DHIs in SEA LMICs. Attention to distal biomedical and clinical outcomes would broaden the evidence base beyond proximal and intermediate outcomes and provide a more complete understanding of outcomes within the PID framework. Research involving women with limited or no connection to existing health care services is also needed to determine whether DHIs can extend access to maternal health care beyond existing service users. Conducting studies in underrepresented SEA LMICs would strengthen the regional evidence base.

This review provides a region-specific synthesis of maternal health outcomes reported in DHI studies conducted in SEA LMICs and systematically maps the outcomes using the adapted PID framework. Several limitations should be considered. The restriction to English-language literature may have excluded relevant studies. The review may also have missed unpublished or nonindexed literature. Several SEA LMICs were underrepresented in the evidence base, limiting regional representativeness. Variation in study designs, intervention characteristics, follow-up periods, and outcome measures limited comparison and synthesis across studies. Consistent with the scoping review design and objectives, no formal risk-of-bias or certainty-of-evidence assessment was conducted [10,11]. Accordingly, the findings describe the available evidence but do not establish its strength or certainty regarding intervention effectiveness.

All 3 PID domains were represented among the maternal health outcomes reported in DHI studies conducted in SEA LMICs. Intermediate outcomes were reported most frequently, whereas distal biomedical and clinical outcomes were reported least frequently. Studies were concentrated in a limited number of countries. The regional evidence base could be strengthened by giving more attention to distal outcomes and underrepresented countries.