Remember me
Women have been entering pregnancy less healthy than previous generations []. Maternal morbidities such as preeclampsia and gestational diabetes, and severe maternal morbidities such as postpartum hemorrhage all have implications for the long-term health of mothers well beyond the postpartum period. Preeclampsia is associated with cardiovascular disease for decades following pregnancy, including chronic hypertension, stroke, and age-adjusted overall mortality []. Gestational diabetes mellitus (GDM), a condition that in many cases resolves itself after delivery, can still affect mothers beyond the puerperium []. Severe maternal morbidity (SMM) encompasses myriad conditions, including acute myocardial infarction, eclampsia, and hemorrhage. Those who have experienced SMM are more likely to die at any point after delivery and into the decades beyond the postpartum period []. Preterm birth (PTB; delivering before 37 weeks gestation) is associated with long-term cardiovascular complications in the mother, including ischemic heart disease, stroke, and atherosclerosis [].
Informatics interventions offer tools that can help to prevent perinatal health conditions that have long-term health consequences for mothers, monitor these conditions in the perinatal period so that mother and baby remain healthy, and follow mothers post partum to ensure that they continue to receive the health monitoring they need. Technology-based tools that can be used for these purposes include mobile apps, wearable technology, physician decision support, and telehealth, among others. As artificial intelligence (AI) technology becomes more common in health care, a range of interventions can be used to predict, diagnose, and treat maternal morbidities that lead to maternal complications, with the hope of reducing maternal morbidity and mortality []. While some patients are hesitant for their physicians to use AI to diagnose or treat medical problems, physicians have demonstrated growing acceptance and belief that AI will improve interactions with their patients []. As the landscape of health informatics shifts substantially with the advent of new technologies, it is important to assess what technologies have already been tested to prevent, diagnose, and treat common maternal morbidities so that gaps can be identified and addressed in future research [].
In this scoping review, we seek to identify what informatics interventions have been designed and tested that address maternal morbidity. In addition, we aim to assess the extent, range, and nature of informatics interventions that have been tested to prevent, diagnose, and treat maternal morbidities known to have long-term health consequences for mothers. The review summarizes discrete populations addressed in the literature, types of informatics interventions tested in the literature, conditions addressed by these interventions, and outcomes measured.
Our scoping review methodology was guided by Arksey and O’Malley’s [] framework, stages 1-5, and Joanna Briggs’s Manual for Scoping Reviews. We asked what informatics interventions have been designed and tested for SMM. Using the PCC (Population, Concept, and Context) framework outlined in the Manual for Scoping Reviews, the population was adult women with SMM as defined by the Centers for Disease Control and Prevention (CDC), the concept was informatics interventions, and the context was prospective studies anywhere in the world. We developed a review protocol to guide the process. The literature searches were led by an expert information specialist (JCS) in consultation with the research team. Four electronic databases were searched from inception until June 6, 2022: MEDLINE and Embase (searched simultaneously on Ovid), Cochrane Library (Wiley), and Engineering Village (Elsevier). One database was searched from inception until June 7, 2022: IEEE Xplore (IEEE). The MEDLINE (Ovid) search was peer-reviewed by another expert librarian using the PRESS (Peer Review of Electronic Search Strategies) checklist and modified as required []. The search strategy was limited to the English language and encompassed all years of publication. Embase and MEDLINE citations were deduplicated in Ovid before exporting to Covidence online review software. The full search strategies for each database are publicly available in searchRxiv []. Lastly, review results are reported using PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews) []. This checklist can be found in . Our protocol was registered retrospectively in Open Science Framework Registries [].
Study SelectionThe inclusion and exclusion criteria are described in . Studies were included that prospectively tested a medical informatics intervention on either a group of women or physicians who treat women. Study methodologies could include both quantitative and qualitative approaches. Only studies analyzing primary data to describe a health-related outcome were included. Studies were included that tested medical or clinical informatics interventions on human subjects to prevent, monitor, or treat maternal conditions that have long-term health consequences for mothers within and beyond the perinatal period. Conditions included preeclampsia or hypertensive disorders of pregnancy, GDM, PTB, new mental health diagnoses, and SMM as defined by the US CDC [].
Textbox 1. Criteria for inclusion and exclusion.Inclusion criteria
Test a medical informatics intervention prospectivelyMedical or clinical informatics interventionsAdult women or physicians who treat womenPreeclampsia, gestational diabetes mellitus, preterm birth, perinatal mental health diagnoses, severe maternal morbidity as defined by the Centers for Disease Control and Prevention (21 indicators)Physical health, patient-centered outcomes, mental health, health behavior, health knowledge or attitudes, health care use, quality of care outcomesExclusion criteria
Systematic reviews, scoping reviews, literature reviews, opinion pieces, commentaries, proposals, reports, conference papersPilot/feasibility studiesBioinformatics studies, including scans, ultrasounds, biomarkers, and predictive algorithmsAdolescentsPreexisting conditions: type 1 or 2 diabetes, chronic hypertension, preexisting mental health conditionsFeasibility, acceptability, user experienceSystematic reviews, scoping reviews, and literature reviews were excluded together with opinion pieces, commentaries, proposals, reports, or gray literature. Studies were also excluded if they focused primarily on bioinformatics, effectively eliminating imaging (eg, ultrasounds), biomarkers, and predictive algorithm development.
Article references were entered into review software for screening and data extraction. The first and senior authors conducted an initial title, abstract, and keyword screening. If any author recommended inclusion, then the article underwent full-text screening. All authors except for the information specialist (JCS) then conducted a full-text screening, with two team members reviewing all articles. Studies with two votes to include were moved forward for extraction; those with two votes to exclude were excluded from the study. During the full-text screening, Cohen κ was used in the initial training to gauge agreement until reviewers reached a κ of 0.80. After training, disagreements were resolved by the first author.
Data ExtractionA subset of four authors extracted data using Covidence review software. Reviewers were given an initial set of 10 articles and met with the first author for consensus before reviewing the remaining articles. All studies were reviewed for extraction by the first author and one additional author. Most data extraction categories were developed a priori, and selection options were developed by the first and senior authors after reviewing all articles for trends. The following data categories were added after the protocol had been developed: aim of the study, years of data collection, population inclusion and exclusion criteria, and total number of participants. These categories were added to the data extraction form iteratively to provide more context to the reviewed studies.
Demographic data collected for each article by reviewers included title of article, year published, lead author surname, and country in which the study was conducted. The following characteristics of included studies were extracted: informatics technology tested, health problem, aim of study, study design, years of data collection, population description, study participation inclusion criteria, study participation exclusion criteria, total number of participants, and health outcome. Reviewers were given the option of selecting multiple answers for each study characteristic. An open-ended “other” option was available for the following study characteristics: informatics technology, health problem, study design, and outcome. Following extraction, the first and senior authors also extracted data on whether a significantly positive result was found in the studies, as well as racial diversity in participant samples for those studies conducted in the United States.
Data were synthesized by tallying the options from each category (study design, region, technology, health outcome, study population, study outcome, and results) and calculating the corresponding frequencies for each option.
The PRISMA flow diagram for this scoping review can be found in . A total of 3799 records were identified for title and abstract screening; 652 records underwent full-text screening for eligibility. Of these, 572 were excluded. Common reasons for exclusion were study design (eg, retrospective, systematic or other review, narrative), type of study outcome (user experience), or the wrong intervention (eg, machine learning algorithms or predictive models not tested with patients or providers). A total of 80 studies underwent full data extraction. A list of the studies and their characteristics can be found in .
Frequencies for synthesized data can be found in .
Figure 1. PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) diagram. Table 1. Studies selected for extraction.aGDM: gestational diabetes mellitus.
bRCT: randomized controlled trial.
cDSS: decision support system.
dHDP: hypertensive disorders of pregnancy.
ePTB: preterm birth.
Table 2. Outcomes measured.aEHR: electronic health record.
bDSS: decision support system.
cGDM: gestational diabetes mellitus.
dHDP: hypertensive disorders of pregnancy.
Included studies were mostly RCTs (n=60). Other study designs included nonrandomized experimental (n=9), cohort (n=6), case control (n=2), retrospective (n=2), and qualitative (n=1). Included studies represented a range of countries and regions. Most of the studies were conducted in North America (n=24), East Asia and the Pacific (n=22), and Western Europe (n=20). Fewer studies were authored from the Middle East and North Africa (n=9), sub-Saharan Africa (n=3), and South Asia (n=2). No studies focused on Latin America or the Caribbean.
Informatics TechnologiesThe specific informatics technologies tested by the study were described. Descriptive categories included telehealth or telemedicine (remote consultation via telephone or video chat that may replace medical consultation), remote monitoring (patient-performed measurements of vital signs and blood glucose), electronic health record decision support system (tools within the electronic health record that assist providers with clinical decisions), smartphone or tablet applications, SMS text messages (between providers and patients or between patients), web-based programs (patient portals and other programs administered through a web browser, not through a smartphone or tablet application), and other (with the technology charted manually). Studies tested a range of informatics technologies. The most frequently tested technologies included smartphone/tablet applications (n=28) and telehealth/telemedicine (n=23). Technologies tested less frequently included remote monitoring (n=15), web-based programs (n=10), and SMS text messaging (n=8). Technologies assessed in very few studies included electronic health record decision support (n=3), wearable technologies (n=2), and other (cognitive aid on an anesthesia machine; n=1).
Health ProblemsArticles were categorized by the health problems treated by the informatics technology; some tested more than one. These included preeclampsia or hypertensive disorders of pregnancy, hypertension, GDM, PTB, hemorrhage, and new (perinatal) mental health diagnoses. Although our protocol searched for “severe maternal morbidity as defined by the CDC” and included a list of 21 conditions [], after reviewing the articles, it was clear that the only condition from the CDC list that was frequently tested by informatics technology was postpartum hemorrhage. Thus, any additional conditions were added manually in the “other” option. Studies primarily used informatics technologies to address new mental health conditions (n=30) and GDM (n=30). Other health problems addressed by the studies include preeclampsia or hypertensive disorders of pregnancy (n=12), hypertension (n=5), and hemorrhage (n=3). One study each addressed preterm birth and myocardial infarction.
Population TypePopulation type was categorized as either: women or medical providers who treat women. Included studies overwhelmingly tested informatics technologies on women, with only 3 studies testing informatics technologies on providers. US studies were also coded for sample demographics.
OutcomesFinally, outcomes were categorized to gain an understanding of the measures upon which the studies focused; many tested more than one measure. These included medical outcomes (eg, weight loss, blood pressure, blood glucose), number of remote measures (eg, patient-collected and -reported blood pressure or blood glucose measurements), mental health outcomes (eg, Edinburgh Postnatal Depression Scale score, subjective well-being measures), health behaviors (eg, dietary habits, exercise frequency), health knowledge or attitudes (eg, patient’s understanding of health condition, patient-reported attitude toward care), health care use (eg, number of emergency room visits, number of prenatal care visits), quality of care (eg, Hospital Consumer Assessment of Healthcare Providers and Systems [HCAHPS] scores, safety of care, effectiveness of care, efficiency of care), or other. Studies measured myriad outcomes, and many studies measured more than one outcome. Many assessed medical outcomes (n=39), mental health outcomes (n=30), and patient health behaviors (n=21). Fewer measured quality of care (n=19), number of remote readings (n=13), and health care use (n=12). Very few measured health knowledge and attitudes (n=5) or other outcomes (n=1).
EffectStudies were coded as to whether a significant positive effect was exhibited for at least one health outcome measure (yes/no). Most (n=63) studies found that the informatics technology yielded at least one positive effect for a health-related (ie, not user experience) outcome. Some (n=17) studies showed no significant positive effect of the informatics technology tested.
This scoping review was performed to characterize what informatics interventions have been tested to address the maternal morbidities that result in long-term negative health outcomes for mothers. Our results show that most of the tested informatics technologies were tested in North America, East Asia, and Western Europe. They overwhelmingly focused on patient populations diagnosed with GDM, mental health conditions, and preeclampsia. Most were patient focused and most aimed to prevent, diagnose, or treat physical or mental health outcomes.
Women have been entering pregnancy less healthy than previous generations. As the world population grows increasingly obese, high BMI presents critical risks for birthing people, with obesity linked to such perinatal complications as gestational diabetes [], preeclampsia [], and postpartum hemorrhage []. Women also currently enter pregnancy in poorer cardiovascular health than previous generations, with prepregnancy hypertension especially pronounced in rural populations of childbearing age []. This scoping review surveys the landscape of informatics interventions tested to treat these increasingly common conditions.
Most of the technologies were tested in North America, East Asia or the Pacific, and Western Europe. Despite some evidence of resistance to new technologies [], studies generally show perceptions of mobile health to be positive in sub-Saharan Africa, especially toward maternal health applications []. There may be opportunities to test technologies developed in North America, Asia, and Europe in underserved parts of the world. In addition to improving maternal health in these underserved areas, testing in diverse types of populations may yield compelling findings that could be translated into US, Asian, and European contexts through a process of reciprocal innovation [].
Our results show that most of the tested informatics technologies focused on patient populations diagnosed with GDM; mental health conditions; and, to a lesser extent, preeclampsia. Far fewer focused on preventing or treating postpartum hemorrhage. Although prevalences of GDM and mental health conditions are high, preeclampsia and hemorrhage are leading causes of maternal mortality and pregnancy-associated deaths in both the United States and abroad []. Postpartum hemorrhage, in particular, is the leading cause of maternal death worldwide, causing 94% of maternal deaths [,].
There are at least two possible reasons for this imbalance in testing of interventions. First, diabetes mellitus and mental health conditions are also common in the general population, unlike preeclampsia and hemorrhage. Thus, preexisting interventions may be more easily transferrable to pregnant populations. Second, GDM and postpartum depression can be diagnosed through routine screening, with GDM diagnosed via prenatal glucose testing and postpartum depression diagnosed using the Edinburgh Postnatal Depression Scale. Indeed, 9.3% of interventions measured the extent/accuracy with which patients review their own biodata, including blood glucose levels. Screening for GDM could be seen as a parsimonious method to improve women’s health. On the other hand, postpartum hemorrhage is less predictable, typically being diagnosed through an urgent/emergent presentation. Thus, the ease of routine screening practices may lend certain diagnoses to being more amenable to the implementation of informatics interventions.
Most of the interventions in this review aimed to prevent, diagnose, or treat physical or mental health outcomes, and almost all were patient focused. Very few interventions (n=3, 4%) were tested on providers who treat women. Of those studies, two of the interventions focused on midwives and one used a decision support system to assist physicians in screening for maternal depression. Most interventions in this review were tested on women and involved active participation from patients (eg, taking blood glucose readings, logging symptoms in an app, participating in an online support group). In other words, most of the interventions entrusted patients with aspects of their own health.
This focus on outcomes is logical given the maternal mortality crisis both within and outside of the United States. Mental health conditions are the leading cause of pregnancy-related deaths in the United States according to state Maternal Mortality Review Committees [] and of non-Hispanic White and Hispanic women when stratified by ethnicity. Cardiac conditions are the leading cause of pregnancy-related death for Black/African American women in the United States, but the focus on interventions aimed at patients themselves minimizes the role provider knowledge and practice within a health system plays in the crisis. Research shows that lack of standardized emergency obstetric care, maternity care coordination [], and systemic racism [] can play a role in the safe and effective delivery of care. The results of this scoping review show a large gap in the literature related to using informatics to improve the quality of care for maternal morbidities. Interventions such as telehealth have been shown to improve obstetric quality, including perinatal smoking cessation and breastfeeding []. It is possible that expanding such technologies and examining the way they are used by physicians can help improve the quality of care and ultimately save the lives of women.
The need for informatics interventions does not end with birth. Prior research suggests that many physicians, including obstetricians, are unaware of how maternal morbidity can influence maternal health in the decades following pregnancy []. Because of this lack of awareness, patients often do not know to monitor their own health in the long term, and they may not receive referrals for proper follow-up services. The paucity of provider-focused technologies identified in this review may also present opportunities for the development of informatics-based interventions targeting physicians and affiliated providers—especially in the treatment and prevention of postpartum preeclampsia and hemorrhage—that could use technologies like decision support to improve maternal health.
LimitationsOur study has several limitations. Because this is a scoping review, we did not systematically assess the quality of individual studies. The studies reviewed may vary in quality or contain potential biases or methodological limitations. Therefore, this study is not intended to assess the quality of the evidence supporting informatics-based interventions on the given perinatal health conditions. Additionally, the nature of a scoping review is not to analytically aggregate studies so as to make claims about effect sizes. When additional, more comparable studies have accumulated, further research is needed to examine which informatics interventions are most effective and the size of the impact on maternal and perinatal health.
In addition, we only considered published studies that tested an informatics intervention and presented health-related outcomes. We excluded many studies because they tested feasibility or acceptability. It is possible that such excluded feasibility studies will eventually test health-related outcomes. Overall, this scoping review summarizes trends in the populations, geographic areas, technologies, and conditions targeted by informatics interventions tested and disseminated in the available literature.
ConclusionsThis scoping review paints a picture of the landscape of informatics interventions aimed at preventing and treating maternal morbidity. Most interventions identified in this study were tested in North America, East Asia and the Pacific, or Western Europe. Most tested either smartphone/tablet applications or telehealth/telemedicine, and most technologies tested for new mental health conditions and GDM. Almost all the studies tested technologies on populations of women and reported medical, mental health, and patient behavior outcomes. Results suggest that there may be opportunities to use informatics technologies to target providers who treat women as well as conditions such as postpartum hemorrhage that are more likely to lead to mortality. As the landscape of informatics applications in health care continues to expand, maternal health is poised to be an important target of these interventions.
The authors would like to acknowledge the VA Medical Informatics Fellowship. This work was funded by the Department of Veterans Affairs Office of Academic Affiliations. The views expressed in this article are those of the authors and do not reflect the position or policy of the US Department of Veterans Affairs, the US government, or the authors’ affiliated institutions.
The search strategies used in this study are openly available at searchRxiv [,-].
Investigation: RCS (lead), SS (equal), SAEA (supporting), OO (supporting), BB (supporting)
Generative artificial intelligence was not used in the writing of this manuscript.
None declared.
Edited by T de Azevedo Cardoso; submitted 27.07.24; peer-reviewed by A Morgan, R Appiah-Kubi; comments to author 01.10.24; revised version received 20.12.24; accepted 21.01.25; published 25.03.25.
©Jill Inderstrodt, Julia C Stumpff, Rebecca C Smollen, Shreya Sridhar, Sarah A El-Azab, Opeyemi Ojo, Brendan Bowns, David A Haggstrom. Originally published in the Interactive Journal of Medical Research (https://www.i-jmr.org/), 25.03.2025.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Interactive Journal of Medical Research, is properly cited. The complete bibliographic information, a link to the original publication on https://www.i-jmr.org/, as well as this copyright and license information must be included.
Comments (0)