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 Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication on https://www.jmir.org/, as well as this copyright and license information must be included.
Alcohol consumption in pregnancy has been associated with serious fetal health risks and maternal complications. While previous systematic reviews of digital interventions during pregnancy have targeted smoking cessation and flu vaccine uptake, few studies have sought to evaluate their effectiveness in preventing alcohol consumption during pregnancy.
This systematic review aims to assess (1) whether digital interventions are effective in preventing alcohol consumption during the pregnancy/pregnancy-planning period, and (2) the differential effectiveness of alternative digital intervention platforms (ie, computers, mobiles, and text messaging services).
PubMed, Embase, CINAHL, and Web of Science were searched for studies with digital interventions aiming to prevent alcohol consumption among pregnant women or women planning to become pregnant. A random effects primary meta-analysis was conducted to estimate the combined effect size and extent to which different digital platforms were successful in preventing alcohol consumption in pregnancy.
Six studies were identified and included in the final review. The primary meta-analysis produced a sample-weighted odds ratio (OR) of 0.62 (95% CI 0.42-0.91;
Overall, our review highlights the potential for digital interventions to prevent alcohol consumption among pregnant women and women planning to become pregnant. Considering the advantages of digital interventions in promoting healthy behavioral changes, future research is necessary to understand how certain platforms may increase user engagement and intervention effectiveness to prevent women from consuming alcohol during their pregnancies.
Alcohol consumption during pregnancy is a major public health concern, and it has explicit links to fetal alcohol spectrum disorders (FASDs) and adverse birth-related outcomes like miscarriage and stillbirth [
Barriers to alcohol abstinence during pregnancy range from lack of awareness about health consequences to low socioeconomic status and/or ability to access necessary health care services [
Social inequalities are also a fundamental risk factor for alcohol consumption during pregnancy, with women of low socioeconomic status and racial/ethnic minority backgrounds at greater risk of bearing children with severe forms of FASDs like fetal alcohol syndrome [
With digital technologies having considerable potential to deliver health care interventions at a low cost and with easy accessibility [
To our knowledge, no systematic review to date has evaluated the effectiveness of digital interventions for preventing alcohol consumption among pregnant women. By contrast, multiple systematic reviews and meta-analyses have examined the effectiveness of digital interventions for smoking cessation [
This systematic review sought to (1) identify the current studies describing the above-mentioned digital interventions, (2) assess whether these digital interventions are effective in preventing alcohol consumption among the target population, and (3) examine the extent to which digital interventions on various platforms, such as computers (web-based, internet, eHealth, etc), mobiles, and text messaging services, may vary in their degree of effectiveness in preventing alcohol consumption.
Studies that discussed digital interventions to prevent alcohol consumption among pregnant women or women planning to become pregnant were identified by searching MEDLINE/PubMed (National Library of Medicine, NCBI), Embase (Elsevier), Cumulative Index of Nursing and Allied Health Literature (CINAHL Plus, EBSCO), and Web of Science Core Collection (Clarivate). Controlled vocabulary terms (ie, MeSH, Emtree, and CINAHL subject headings) were used when available and appropriate. The search strategies were designed and executed by a librarian (CM). Searches were not limited to a specific region, language, study design, or time period. The exact search terms used in each of the databases, and corresponding result numbers, are provided in
We included studies that (1) targeted pregnant women or women planning to become pregnant, (2) measured the use of a digital intervention aiming to prevent alcohol consumption during pregnancy, (3) involved a digital interaction between the patient and a health care provider or professionally developed service (social media where subjects communicated with one another were excluded), and (4) reported rates of alcohol abstinence.
Using these eligibility criteria, 2 independent investigators (SSO and DC) examined all studies reporting the use of a digital intervention to prevent alcohol consumption among pregnant women. All studies were screened at the title and abstract levels and excluded if the main target population did not consist of pregnant women or women planning to become pregnant, or if they did not include a digital intervention or a control group/preintervention comparison group. Subsequently, full-text reviews were performed to ensure that all articles measured and reported alcohol abstinence, and involved a digital interaction with a health care provider or professionally developed service. Any discrepancies were resolved by discussion. For the extraction of data regarding intervention characteristics and outcome measures (effect size), an online data extraction sheet was employed so that 2 independent investigators (SSO and JYM) could extract the necessary information. Regarding interrater reliability, kappa values (
Rates of alcohol abstinence during pregnancy were extracted and presented as crude odds ratios (ORs) to maximize similarity between different studies. To examine the extent to which a digital intervention was effective, a random effects primary meta-analysis was conducted to determine the combined effect size and extent to which each digital intervention affected overall alcohol abstinence. An exploratory subgroup analysis was carried out to determine whether different platforms of digital interventions differed in the extent to which they affected the effect size. A random effects model was adopted for all meta-analyses to estimate intervention effects with 95% CIs that fall on a distribution of effect sizes. The Cohen
We assessed study quality in terms of potential bias using the Cochrane Collaboration tool for randomized controlled trials to assess the validity of the included studies [
The PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flowchart in
PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flowchart of the literature search.
Of the 48 articles assessed for eligibility, 42 were excluded for the following reasons: (1) weak study design in terms of the absence of a control group (pertaining to usual care or a preintervention baseline) or no targeting of alcohol consumption prevention during pregnancy (n=18); (2) no targeting of currently pregnant women or women with plans to become pregnant (n=10); (3) no outcome measure for alcohol abstinence (n=9); (4) no use of a digital intervention (n=4), and (5) no report of the outcome of interest (n=1). Ultimately, 6 studies were included in our final review.
Characteristics of the included studies.
Author | Country | Sample size, n | Mean age (years) | Population sample | Control | Digital intervention | Follow-up assessment |
Evans et al, 2014 [ |
United States | 459 | 26.5 | Pregnant military health care beneficiaries aged 18-45 years presenting for care (>14 weeks’ gestation) | Usual care only | Text4Baby: Text messaging service on nutrition, smoking, taking vitamins, alcohol use, flu shots, health care appointments, health information seeking, and related risk prevention behaviors | 4 weeks (pilot study) |
Evans et al, 2012 [ |
United States | 86 | 27.6 | Pregnant women first presenting for care at the Fairfax County, Virginia Health Department | Usual care only | Text4Baby Pilot: Text messaging service with immediate “just-in-time” tips about prenatal and postpartum health outcomes | 28 weeks of the baby’s gestational age |
Ingersoll et al, 2018 [ |
United States | 71 | 27.8 | Pregnant women |
Patient education | CHOICES intervention: Automated internet intervention providing 6 web-based cores of information, videos, and interactive activities (eg, diaries) regarding alcohol-exposed pregnancies | 24 weeks posttreatment |
Ondersma et al, 2015 [ |
United States | 48 | —a | Pregnant women seeking services at a prenatal care clinic affiliated with the Henry Ford Health System in Detroit, Michigan | Time-matched (20 minutes) and moderately interactive intervention focused on infant nutrition, with no mention of alcohol use during pregnancy | e-SBI intervention facilitating self-change and/or treatment-seeking through a 20-minute interactive session, using techniques such as education about alcohol-related pregnancies and feedback regarding proactive problem-solving | Postpartum, for the past 90 days (22-23 weeks) |
van der Wulp et al, 2014 [ |
The Netherlands | 258 | 32.6 | Pregnant women seeking services at midwifery practices in the Netherlands | Usual care only | Both computer tailoring internet-based feedback and offline health counseling based on the I-Change model (promote awareness, motivation, and action for behavioral change) | 24 weeks posttreatment |
Wernette et al, 2018 [ |
United States | 50 | 24.4 | Pregnant women visiting a prenatal clinic in a large inner-city hospital | Time- and attention-matched control group (watched segments of popular television shows and received brochures about health risks during pregnancy postintervention) | Computer-delivered single-session brief motivational intervention plus booster session addressing both substance use and sexually transmitted infection risk | 4 months posttreatment |
aNot reported.
Two studies delivered digital content via a text messaging service called “Text4Baby,” which provides weekly tips about prenatal care, emotional support, alcohol and drugs, infectious diseases, and exercise to pregnant women and new mothers [
Four studies included computer/internet-based interventions consisting of interactive counseling sessions, educational videos, and interactive activities (ie, diary writing, meditation, etc) [
Three studies used usual care in the form of a standard physician, obstetrician, or nurse-midwife/midwife providing advice [
Alcohol consumption during pregnancy was employed as the primary outcome. Studies administered self-reported questionnaires via telephone/email asking participants whether or not they had consumed any alcohol during the pregnancy period (eg, “Since you found out about your pregnancy, have you consumed alcoholic beverages?” [yes/no]). Participants were mostly questioned at 16 [
A primary meta-analysis including 6 trial arms from 6 studies was performed. The sample-weighted OR indicated that digital interventions decreased the odds of alcohol consumption during pregnancy compared with control groups (OR 0.62, 95% CI 0.42-0.91;
A stratified analysis examining the influence of different intervention platforms revealed that computer-based interventions (OR 0.59, 95% CI 0.38-0.93) were effective for preventing alcohol consumption; however, too few studies of text messaging (OR 0.85, 95% CI 0.29-2.52) were available to draw a conclusion regarding the effect of this platform (
When studies were stratified according to each publication’s quality risk of bias, point estimates (OR 0.62) were identical across study quality (
Effectiveness of digital interventions for preventing alcohol consumption in pregnancy [
Effectiveness of digital interventions by platform [
Effectiveness of digital interventions by quality risk of bias [
A summary of the quality assessment can be found in
In a high-risk study, the authors reported that the blinding of both participants and researchers was not possible because they had to keep track of whether participants received additional counseling from their midwives or tailored feedback via the computer [
Results from the Egger test for funnel plot asymmetry were not statistically significant (
Risk of bias summary [
Funnel plot assessing publication bias.
In this systematic review, we found that digital interventions for preventing alcohol consumption during pregnancy may be effective in preventing alcohol consumption, especially on computer/internet-based platforms. Excluding a pilot Text4Baby study [
Regarding text messaging platforms, there were too few studies in our review to draw a conclusion regarding their effectiveness as digital platforms for alcohol abstinence. Previous studies on the use of text messaging services to raise awareness about smoking cessation and flu vaccinations among pregnant women have shown mixed results, with some studies reporting promise compared to nontailored or internet platforms [
As expected, the most effective interventions in our review were those that incorporated both offline house counseling and internet or mobile-based feedback (ie, “blended” care) for individuals [
As the first systematic review to question the effectiveness of digital interventions for preventing alcohol consumption during pregnancy, the findings of this review are novel. However, there are some limitations of our review. While our assessment of funnel plot symmetry did not formally detect publication bias (by significance testing), the sample of studies was small. It is possible that underpowered studies with null results are missing (a “file drawer” problem). Cultural differences between the United States and the Netherlands may also have affected study outcomes; while both the United States and the Netherlands officially recommend that pregnant women completely abstain from alcohol, the prevalence of alcohol consumption during pregnancy is higher in the Netherlands (19%-21%) than in the United States (15%) [
Most concerningly, the primary outcome for alcohol abstinence during pregnancy was self-reported, and follow-up methods/timing differed among all studies. The absence of validation by biomarkers to assess abstinence was a fundamental limitation of the included trials, which is concerning as self-reports of alcohol consumption may be affected by memory loss from alcohol abuse and underreporting due to a fear of negative consequences like being reported to Child Protective Services [
Intervention duration, quality, and intensity could not be controlled for, with some studies, such as the e-SBI trial, specifically targeting high-risk individuals via professional counseling methods (eg, motivational interviewing) [
Future studies would benefit from controlling for discrepancies among varying trials regarding the quality of usual care provided in the control group, assessment of alcohol abstinence, and intervention duration/quality. However, in our study, this was not possible due to the limited descriptions provided by the included studies regarding these factors. In future studies when more trials targeting alcohol abstinence during pregnancy are available for review, a more consistent and thorough subgroup analysis of intervention techniques, involving video, counseling, blended care, etc, is warranted.
More studies are required to assess the extent to which digital interventions targeting pregnant drinkers may be effective for women from disadvantaged backgrounds and/or a low socioeconomic status. While few programs and trials are currently available to review, digital technologies are being embraced rapidly for personalized health care. Future studies would benefit from assessing how better allocation of both online and offline resources may help pregnant women and women planning to become pregnant avoid consuming alcohol and other teratogenic substances during their pregnancies.
Summary of database search terms.
electronic screening and brief intervention
fetal alcohol spectrum disorder
odds ratio
This research was supported by a grant of the Korea Health Promotion R&D Project, funded by the Ministry of Health & Welfare, Republic of Korea (grant number: HS21C0037).
IK and SSO conceptualized and supervised the study. IK, SSO, JAL, and JYM developed the methodology. CM provided the literature search. SSO, JAL, and DC screened the studies and performed the formal analysis/quality assessments. SSO and IK performed the validation. SSO, JAL, CM, and IK wrote, reviewed, and edited the manuscript. All authors read and agreed to the published version of the manuscript.
None declared.