Role of Chatbots on Supporting Women in Perinatal Period: A Scoping Review

Abstract

Introduction:

Perinatal women need continuous, individualized, accessible care, which traditional models often fail to provide. Chatbots can offer education, symptom tracking, and psychological support, but evidence is fragmented and requires systematic synthesis to assess effectiveness.

Aim:

To understand the current studies on the application of chatbots in the perinatal period and to explore their contributions, limitations, and future directions in this field.

Methods:

Relevant studies were systematically identified through a comprehensive search of major databases, including PubMed, Scopus, Web of Science, Embase, CINAHL, Joanna Briggs Institute, Cochrane Library, CNKI, SinoMed, VIP, and WanFang, covering all records up to June 2025. Two independent researchers screened and evaluated the retrieved citations, and data were extracted in a structured format to maintain methodological rigor and inter-reviewer consistency.

Results:

A total of 17 studies were included in this review, which indicated that chatbots offer a range of functionalities: optimizing follow-up and revisit processes, monitoring health status, providing psychological and counseling support, offering educational resources, and serving as interactive tools. Most of the included studies demonstrated that chatbots offer beneficial effects, including disease monitoring, enhanced parenting self-efficacy, mental health support, facilitation of role adaptation, improved health literacy and health behaviors, and increased user engagement and satisfaction.

Conclusion:

Most of the included studies indicate that chatbots have a positive impact on pregnancy and postpartum care outcomes. However, the majority of these studies remain in the preliminary stages, and further validation of chatbot effectiveness in this context is still needed through large-scale, multicenter, randomized controlled trials.

Impact:

Deepening knowledge of perinatal chatbots’ uses and limits enables targeted digital health strategies to improve care quality and maternal, neonatal, and family experiences.

Patient and Public Contribution:

Chatbot could be a potentially valuable tool in perinatal care to enhance health education, psychological support, and communication between care providers and women.

Introduction

Pregnancy and postpartum care encompasses medical and nursing services during pregnancy, childbirth, and the postpartum period, with the aim of ensuring maternal and infant health.^1,2 During this critical period, women undergo not only profound physiological changes but also face considerable psychological stress and emotional challenges. ³ World Health Organization reports about 300,000 annual maternal deaths from pregnancy- or childbirth-related complications, about 80% of which are preventable with timely, adequate care. ⁴ About 10%–13% of women during pregnancy and postpartum are affected globally, with higher rates in developing regions.^5–7

Previous studies have shown that personalized care has a positive impact on pregnancy and postpartum outcomes, ⁸ including improved maternal satisfaction and psychological well-being, enhanced adherence to prenatal and postpartum care, and better clinical outcomes such as early identification and management of postpartum complications, reduced hospital readmissions, and improved breastfeeding.^9–12 At present, pregnancy and postpartum care mainly consists of face-to-face, telecommunication counseling, home care by nurses, group seminars, and so on.^13–15 However, these traditional care models face several limitations, including shortages of health care personnel, uneven regional distribution of services, and inadequate timeliness in doctor–patient communication. ¹⁶ These challenges hinder continuous, personalized health support in pregnant and postpartum care.

As an emerging digital health technology, ¹⁷ chatbots have demonstrated considerable potential in pregnancy and postpartum care. They are mostly powered by rule-based artificial intelligence (AI) and utilize natural language processing (NLP) to interact with users conversationally. ¹⁸ Characterized by their high feasibility, scalability, and relatively low cost, chatbots offer a practical solution for expanding access to health care services, particularly in resource-constrained settings. ¹⁹ Previous studies have highlighted their effectiveness in various domains, including enhanced chronic disease management, better health promotion, increased access to patient education, and improved patient self-management capabilities.^20,21

Chatbots have been used in perinatal care, offering continuous, tailored, timely information, and psychosocial support to improve pregnancy and postpartum well-being. However, some research focuses on individual functional modules,^22–27 which do not systematically evaluate the overall system architecture or the interactive effects of multi-module integration. ²⁸ Additionally, while 17 included studies demonstrated the potential of chatbots to improve medical accessibility and enhance user engagement, their long-term impact, sustainability, and user acceptance have not been fully validated.^25,29–33

Given that pregnancy and postpartum care span multiple stages, existing evidence tends to be fragmented and lacks a comprehensive perspective on the full continuum of care. ²⁹ Therefore, this review summarizes evidence on chatbots in perinatal care, examining their effectiveness and evaluation.

Methods

Study design

The review process was guided by the Joanna Briggs Institute (JBI) methodology, and the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) framework was followed to ensure transparent reporting.^10,34 As the review relied on previously published studies, ethical approval was not required.

Research question

The research questions were identified as follows:

• 1. What chatbots currently exist for supporting pregnant and postpartum women?
• 2. What features do these chatbots provide during the perinatal period?
• 3. How about the effectiveness of the chatbots in supporting pregnancy and postpartum care?

Search methods

Eligibility criteria

Following the methodological framework of the JBI, the study established its eligibility criteria by referencing the three core dimensions of population, concept, and context. ³⁴ (1) Population: Pregnant individuals at any gestational stage or postpartum individuals. Participants were required to be ≥18 years; (2) Concept: The chatbot is defined as a software application designed to interact with humans through text-or speech-based interfaces. Its primary objective is to simulate human conversation and provide real-time responses to user inquiries. ²⁸ (3) Context: Settings providing pregnancy and postpartum health care, including hospitals, centers, and communities. (4) Types of evidence sources: Peer-reviewed articles published in English or Chinese language, including randomized controlled trials, quasi-experimental studies, observational studies, development studies, and mixed-method studies.

Studies were excluded if they (1) focused solely on technical development or optimization of chatbot systems without clinical application; (2) were non-primary sources such as conference or workshop papers, reviews, commentaries, letters, or protocols; (3) were duplicates or had incomplete, irrelevant, or inaccessible data.

Search strategy

A comprehensive search was performed across 11 electronic databases, including PubMed, Scopus, Web of Science, Embase, CINAHL, JBI, Cochrane Library, CNKI, SinoMed, WanFang, and VIP, from their inception to June 2025. To refine the search strategy, preliminary searches were first conducted in PubMed for English publications and in CNKI for Chinese studies. The final search terms were developed through this pilot phase and further validated in consultation with university librarians to ensure both precision and completeness. In PubMed, the strategy incorporated a combination of free-text keywords and Medical Subject Headings, with equivalent adaptations made for other databases according to their specific indexing structures. To maximize coverage, reference lists of the included studies were also screened using a snowballing approach to capture additional relevant records. Details of the database-specific search terms, strategies, and retrieval results are provided in Supplementary Appendices A1 and A2.

Data management

Data selection

All search results were imported into Zotero for reference management, and duplicate records were automatically identified and removed. Two reviewers (R.H. and Y.J.), both trained in evidence-based research methods, independently screened the titles and abstracts against the predefined inclusion and exclusion criteria. Full-text articles that satisfied these criteria were subsequently examined to confirm eligibility in relation to the research questions. Any discrepancies between reviewers were resolved through discussion or, when necessary, consultation with a third reviewer (J.L.).

Data extraction

Data extraction was performed using Microsoft Excel (version 7). R.H. and Y.J. developed the extraction template and pilot tested it: R.H., Y.J., and J.L. independently extracted data from two studies (representing ∼10% of the total) to refine fields (title, year, authors, country, chatbot name, platform, interaction type, study design, objectives, population, sample size, intervention/control details, data collection time points, intervention duration, outcomes, and statistics). The remaining 15 studies were independently extracted by R.H. and Y.J. Disagreements were resolved through discussion or by adjudication from a third reviewer (J.L.) when consensus could not be reached.

Data synthesis and analysis

Data were analyzed descriptively in accordance with the aims of this scoping review. Given the heterogeneity of study designs, findings were integrated using a narrative synthesis approach to provide an overall summary of the evidence. Quantitative data were extracted and tabulated to provide a descriptive overview of study characteristics and key findings. Final findings were reviewed by the full research team to ensure coherence, transparency, and consistency with the study objectives.

Results

Search strategy results

From 11 databases, 5,170 records were identified. After duplicates were removed, 3,885 records underwent title and abstract screening, leaving 33 for full-text review. Of these, 17 did not meet the eligibility criteria, and one additional study was located through reference searching, resulting in a total of 17 included studies. The PRISMA flow of study selection is presented in Figure 1.

FIG. 1.

Preferred Reporting Items for Systematic reviews and Meta-Analyses flow diagram of study selection. JBI, Joanna Briggs Institute.

Characteristics of included studies

Publication years of included studies ranged from 2020 to 2025. Among the 17 studies included in the analysis, 15 studies^{22–25,27,29–33,35–37} were published in English journals, while the remaining 2 studies were from Chinese journals.^38,39 Most were conducted in United States (n = 6)^{26,30–33,40} and Singapore (n = 3),^27,36,37 while the remaining studies were conducted in China (n = 2),^38,39 Brazil (n = 2),^23,24 Norway (n = 1), ²⁵ Sri Lanka (n = 1), ²⁹ Italy (n = 1), ²² and Korea (n = 1). ³⁵

These studies were classified into five categories based on their design, including mixed methods (n = 8),^{22–24,30,31,35,36,40} randomized controlled trial (n = 4),^26,32,33,37 quasi-experimental study (n = 3),^27,38,39 development study (n = 1), ²⁹ and observational study design (n = 1). ²⁵ Among them, the largest proportion of participants were pregnant women during the antenatal period (n = 7),^{22–26,29,39} parents/couples (n = 4),^27,35–37 perinatal women (n = 5),^{30,31,33,38,40} and postnatal women (n = 1). ³² Sample sizes ranged from 5 to 258; one study ²⁵ did not report sample size. More details are presented in Table 1.

Table 1.

Characteristics of the Included Studies (N = 17)

First author	Year	Country	Study design	Sample size	Participant characteristics	Chatbot name	Interactive type	Rule-based versus AI chatbot	Platforms
Chua	2024	Singapore	RCT	236	Heterosexual couples, gestational age >24 weeks	Parentbot	Text	Rule-based	Mobile app
Chua	2024	Singapore	Mixed methods	118	Heterosexual parents	Parentbot	Text	Rule-based	Mobile app
Wong	2021	Singapore	Quasi-experimental study	26	Parents	ClaimIt	Dialogue	Rule-based	ClaimIt app
Mancinelli	2024	Italy	Mixed methods	5	General pregnant women	Juno	Buttons, links, images	Rule-based	Telegram
Li	2023	China	Quasi-experimental study	92	Pregnant women with gestational diabetes	NR	Voice	Rule-based	NA
Kang	2020	China	Quasi-experimental study	199	Pregnant women	NR	NA	Rule-based	NA
Chung	2021	Korea	Mixed methods	15	Perinatal women/partners	Dr. Joy	Text, buttons	Rule-based	Kakao Talk
Montenegro	2023	Brazil	Mixed methods	25	Pregnant women and health care professionals	BotMaria	Predefined buttons, open-ended questions	Rule-based	Facebook Messenger
Montenegro	2022	Brazil	Mixed methods	13	Pregnant women and health care professionals	NA	Buttons	Rule-based	App
Suharwardy	2023	United States	RCT	152	Women (≥18 years) with a live birth and access to a smartphone	Woebot	Text	Rule-based	App
Nguyen	2024	United States	RCT	29	Minority women (≥14 years, pregnant/postpartum ≤6 months)	Rosie	Text	Rule-based	App
Mane	2023	United States	Mixed methods	109	Women of color (≥14 years, pregnant/having child ≤3 years)	Rosie	Text	Rule-based	App
Inkster	2023	United States	Mixed methods	61	Postpartum women (≤1 year)	Wysa	Text	Rule-based	App
Sagstad	2022	Norway	Observational study	NA	GDM patients	Dina	Text, buttons	Rule-based	Digital health platform
Maduwantha	2021	Sri Lanka	Development study	6	Pregnant women	Baby Chatbot	Text	Rule-based	App
McAlister	2025	United States	Mixed methods	181	Pregnant or Postpartum women (<1 year); perinatal women	Moment for Parents	Dialogue	Rule-based	Web app
B.Su	2025	United States	RCT	258	One live fetus was present, weeks of gestation (<20 weeks); 18–35 years (up to delivery)	iPrenatal	Text, decision buttons, images and videos	Rule-based	App

GDM, gestational diabetes mellitus; RCT, randomized controlled trial; NA, not available; NR, not reported.

Characteristics of chatbots

Among the interactive types, chatbots in 16 studies had two or more mixed forms “Text and buttons” (n = 8)^{22–27,30,32} was the most prevalent, followed by “text, images, audio, video, and buttons” (n = 2),^36,37 “text, buttons, and text message” (n = 2),^31,33 “text, images, buttons, and artificial response” (n = 1), ³⁹ “voice, image, video, and buttons” (n = 1), ³⁸ “text, images, and buttons” (n = 1), ³⁵ and “text and voice” (n = 1). ²⁹ However, one study of chatbots had only one form of interaction: text (n = 1). ⁴⁰ Regarding the form of support, app-based platforms were the most commonly utilized (n = 11),^{24,26,27,29–33,36,37,40} comprising app (n = 7),^{24,29–33,40} mobile apps (n = 2),^36,37 Claimlt application (n = 1), ²⁷ and web app (n = 1). ²⁶ Messaging platforms represented the second most frequent category (n = 3),^22,23,35 including Telegram (n = 1), ²² Kakao Talk (n = 1), ³⁵ and Facebook Messenger (n = 1). ²³ The remaining studies employed a digital health platform (n = 1), ²⁵ while two studies^38,39 did not specify the platform used. More details are presented in Table 1.

Features of chatbots in supporting pregnancy and postpartum care

Chatbots were categorized into five domains: follow-up and revisit (n = 3),^33,38,39 health status monitoring (n = 8),^{25,27,30,33,35,36,38,39} psychological support (n = 13),^{22,24,27,29–33,35–37,39,40} consulting support (n = 14),^{22–24,26,29–33,35–39} and educational support (n = 16).^{22–26,29–33,35–40} More details are presented in Figure 2 and Table 2.

FIG. 2.

Features of chatbots in supporting pregnancy and postpartum care.

Table 2.

Outcome Measures of Chatbots Supporting Pregnancy and Postpartum Care (N = 17)

First author	Year	Country	Study design	Outcome measurements	The main outcome results in the intervention group			Statistical value
					Function	User experience	Post-use evaluation indicators
Chua	2024	Singapore	RCT	Self-efficacy (primary outcome), stress, depression, anxiety, social support, parent–child bonding, and parenting Satisfaction (secondary outcomes)	Education enter: Multimedia resources (text, audio, video). Smile enter: Gratitude/reflection journals, mindfulness videos. Rule-based chatbot: Answers common questions; transfers unresolved queries to “Ask an Expert” platform. Accessible 24 weeks of gestation until 1-month postpartum	The intervention had a generally positive effect on parents’ well-being and could significantly improve parenting self-efficacy at one month postpartum and mothers’ state of anxiety at 3 months postpartum.	Improving parenting self-efficacy (primary outcome), stress, depression, anxiety, social support, parent–child bonding, and parenting satisfaction (secondary outcomes) among parents across the perinatal period.	p: Parenting self-efficacy (p = 0.04), mother anxiety state (p = 0.03)
Chua	2024	Singapore	Mixed methods	User experience	Answers common questions; transfers unresolved queries to “Ask an Expert” platform. Accessible from 24 weeks of gestation until 1-month postpartum.	The PDA provided parents with informational, socioemotional, and psychological support and could be used to supplement the perinatal care provided for future parents.	New babies, new troubles, and new wonders; support system for the parents; reshaping perinatal support for future parents	p: NA
Wong	2021	Singapore	Quasi-experimental study	Anxiety/depression changes; user experience and engagement; user satisfaction and feedback to Juno	Chatbot interactions in stress, sleep, and feeding domains	Parents who completed the study gave positive feedback on their user experience with the chatbot as a tool to express their thoughts and concerns	Real-time conversation data on stress, sleep, and infant feeding from a group of 45 parents.	p: NA
Mancinelli	2024	Italy	Mixed methods	Anxiety/depression changes; user experience and engagement; user satisfaction and feedback to Juno	6-week behavioral activation intervention via Juno, with post-interview 10–14 days after	Focusing on the app itself, women all agreed on the convenience (convenient app) of Telegram as an interface, allowing them to avoid downloading another app and describing it as an optimal channel that they already knew and that is easy to use.	The tool’s capacities, risks, and limitations need to be clearly reported (point 5). In addition, safety measures were limited to self-reported depression, anxiety, and stress levels.	p: NA
Li	2023	China	Quasi-experimental study	Blood glucose control; Self-management; Health literacy and compliance	Disease monitoring (blood glucose, blood pressure, weight). Provide diet/exercise guidance and health education. Capable of providing intelligent consultation and continuous self-management support services. The back-end can collect usage data	Users gave a positive overall evaluation of the use experience of AI robots in home management of gestational diabetes mellitus, especially in terms of ease of use, practicality and relevance of content.	Ease of use; satisfaction	p: Pregnant women’s self-management ability (p < 0.01); FBG pass rate (p = 0.043); 2 h PBG pass rate (p = 0.012); weight control rate (p = 0.023); the incidence of adverse perinatal outcomes (p < 0.05); pregnant women are anxious (p < 0.01).
Kang	2020	China	Quasi-experimental study	Knowledge of health knowledge; satisfaction; follow-up times of pregnant women	Including voice-based vital sign input, follow-up management, and intelligent Q&A for health management.	Users have a good experience with the robot, and their satisfaction is significantly higher than that of the traditional telephone follow-up group	Satisfaction; users’ subjective experience	p: Maternal health care awareness was significantly better than traditional telephone follow-up (p < 0.01)
Chung	2021	Korea	Mixed methods	USE Questionnaire	7-day usability test: ≥3 questions/day to Dr. Joy	To satisfy user intent by monitoring both male and female user utterances.	Produce both utilitarian and hedonistic value during the 7-day contextual UT period.	p: The correlation between usefulness (USE), ease of use (EOU), satisfaction (SAT), and total availability (Total) (p < 0.01), association between perceived benefits and SEE (p < 0.001), association between perceived benefits and SHA (p = 0.004
Montenegro	2023	Brazil	Mixed methods	Satisfaction, usability, content reliability, perceived educational impact	BotMaria provides nutrition education	Ensure positive support for pregnant women, clarify certain issues for the public, and remove some knowledge barriers.	To satisfy user intent by monitoring both male and female user utterances.	p: No significant statistical differences were found
Montenegro	2022	Brazil	Mixed methods	Unified Theory of Acceptance and Use of Technology	Health education information was provided, covering topics including prenatal and postnatal care, dietary guidelines, minor complications, and physical activity.	The intended design could ensure positive support for pregnant women, clarify certain issues for the public, and remove some knowledge barriers.	The highest level of agreement was Expectation of Performance, in which pregnant women agreed that the chatbot had taught them something new.	p: NR
Suharwardy	2023	United States	Randomized controlled trial	Assessed by anxiety and depression scale, Degree of satisfaction PHQ-9 and EPDS	To optout of daily check-ins, in addition to usual care by their obstetric care team	80% of study participants reported being comfortable with the use of a mobile smartphone application for mood management.	A change in depression symptoms at 6 weeks: PHQ-9 and EPDS. Anxiety symptoms: Generalized Anxiety Disorder-7, and satisfaction and acceptability using validated scales.	p: PHQ-9 scores (p = 0.025)
Nguyen	2024	United States	Randomized controlled trial	Changes in depressive symptoms (PHQ-9 score); frequency of infant emergency department visits. User satisfaction with Rosie and usage frequency.	Personalized health information, covering topics, infant development, daily health tips, and stores the most recent 7 days’ tips. Provides resource information and triggers.	Provide a personalized, flexible tool to increase the timeliness and accessibility of high-quality health information to individuals during a period of elevated health risks for the mother and child.	Pregnancy outcomes (birth weight and gestational age), emergency room visits for infants, PHQ-9 depression scale, and group-specific questions. How often they used Rosie and the following issues.	p: Depression symptoms improved (p = 0.008)
Mane	2023	United States	Mixed methods	User willingness, satisfaction, and acceptance of Rosie	To provide Q&A services on pregnancy, postnatal health, infant health, and care. Enhance user engagement and experience through in-app features.	About 75% of mothers said they searched online for maternal and infant health information at least weekly. More than 96% of participants expressed likely or very likely to use Rosie.	NA	p: NA
Inkster	2023	United States	Mixed methods	Changes in PHQ-9 depressive symptom scores; Themes of user-chatbot interaction behaviors.	Users engage with the chatbot to articulate emotions, explore concerns, and obtain personalized recommendations. Users may choose to complete the PHQ-9 depressive symptom assessment, with biweekly reminders provided.	These users also provided feedback sessions within the app with overall sessions seeing 97% of users with high or mid-satisfaction.	As a post hoc analysis, we explored depressive symptom reductions between the groups across increasing PHQ-9 severity.	p: Depression symptoms improved (p < 0.05)
Sagstad	2022	Norway	Observational study	NA	Dina provides information related to GDM, including blood glucose, diet, diagnosis, and exercise.	The chatbot was able to answer about 9 out of 10 of all questions from the users.	NA	p: The difference between the two phases (p = 0.013)
Maduwantha	2021	Sri Lanka	Development study	Satisfaction	Simulate conversations; pregnancy calendar; reminder system; Kick counter; emergency dial system; relaxation mini-games; pregnancy health tips.	NA	Complete a survey about their opinions about the MumCare, what they liked and disliked about it, the drawbacks and then the features they willing to have in the future.	p: NA
McAlister	2025	United States	Mixed methods	Chatbot usability, engagement and user experience	Emotional personalization intervention; Mental health support	The chatbot effectively alleviates the psychological isolation of women in the perinatal period through lightweight interaction, emotion-matching exercises, and staged support.	Overall participation data, participation rate, relevance and applicability, user demand satisfaction, content and scene adaptability	p: NA
B.Su	2025	United States	Randomized controlled trial	Knowledge score after education, patient satisfaction, patient consultation time, and visit time	Provide education on the availability of prenatal genetic screening and diagnostic testing options.	The positive experience of users on chatbots is demonstrated in terms of user satisfaction, ease of use and interface design, information quality, willingness to use, and information acceptance.	Knowledge level assessment, satisfaction assessment, education experience feedback, consultation time, and visit time.	p: Knowledge level score of the two groups after education (p < 0.001)

AI, artificial intelligence; EOU, Perceived Ease of Use Questionnaire; EPDS, Edinburgh Postnatal Depression Scale; FBG, fasting blood glucose; GDM, gestational diabetes mellitus; NA, not available; PBG, postprandial glucose; PDA, Parentbot-A Digital health care Assistant; PHQ-9, Patient Health Questionnaire-9; Q&A, question and answer; RCT, randomized controlled trial; SAT, Satisfaction Assessment Test; SEE, seek health information; SHA, share health information; USE, Usefulness, Satisfaction, and Ease of Use Questionnaire.

Health status monitoring

Chatbot implementation was summarized, covering monitoring dimensions, technology application, intervention mechanisms, collection of user information, and feedback. Among the 17 studies, 8 studies^{25,27,30,33,35,36,38,39} reported the health monitoring features, which included a total of 14 health monitoring indicators, including sleep quality monitoring,^27,35,38 blood glucose monitoring,^25,38,39 sports monitoring,^25,35,38 diet monitoring,^25,35,38 weight monitoring,^38,39 blood pressure monitoring,^38,39 temperature monitoring,^38,39 monitoring of breastfeeding, ²⁷ heart rate monitoring, ³⁸ height monitoring, ³⁸ vision monitoring, ³⁸ adverse outcome monitoring, ³⁹ abdominal condition monitoring, ³⁵ and drug safety monitoring. ³⁵

Psychological support

Thirteen studies^{22,24,29–33,35–37,39,40} reported the feature of psychological support, covering four indicators: anxiety reduction support (n = 12),^{22,24,29–33,35–37,39,40} depression reduction support (n = 11),^{22,29–33,35–37,39,40} psychological stress reduction support (n = 4),^27,36,37,39 and suicide prevention support. ³¹

Educational support

A total of 16 studies^{22–26,29–33,35–40} reported the feature of education. It involved 14 educational functions, including perinatal care (n = 11),^{22,24,25,30,31,33,35–39} health reminders (n = 11),^{22–25,30,31,35–39} diet and nutrition (n = 11),^{23–25,29–31,35–39} physical activity (n = 10),^{24,25,29–31,35–39} medication safety (n = 8),^{24,25,29,31,35–37,39} mild complications and contraindications (n = 7),^{24,29–31,37–39} pregnancy calendar (n = 6),^{24,29,31,36,37,39} pregnancy health guidance (n = 5),^{22,25,26,30,38} digital health support (n = 2),^36,37 fetal kick counter (n = 2),^24,29 parenting care (n = 2),^24,33 cultivation of emotional competence (n = 2),^32,37 and fetal development (n = 1). ²⁹

Follow-up and revisit

Three studies^33,38,39 featured chatbots with the function of follow-up and revisit processes. The schemes implemented in different regions exhibit distinct strategic emphases in optimizing follow-up and reexamination processes. Two studies conducted in China^38,39 primarily focus on the overall workflow through intelligent and automated mechanisms, including intelligent task allocation, personalized scheduling, feedback systems, convenient appointment booking, report delivery, and health assessment functionalities. In contrast, one study ³³ mainly focuses on responding to health emergencies.

Consultation support

Among the 17 chatbots, the majority (n = 14)^{22–24,26,29–33,35–39} provided consultation services. It uses rule-based AI and NLP systems to analyze maternal input, identify key parameters, retrieve personalized data from the database, and generate humanized responses. The remaining three chatbots (n = 3)^25,27,40 did not offer this function: one as data collection tool ²⁷ and two were information-retrieval only.^25,40

Effect of chatbots in pregnancy and postpartum

Of the included studies, effect of chatbots can be classified into the following six categories, including disease monitoring (n = 6),^{30–33,38,39} parenting self-efficacy (n = 8),^{22,24,27,33,36–39} mental health (n = 13),^{22,24,27,29–33,35–37,39,40} role adaptation (n = 13),^{22–24,27,29–33,35–37,39} health literacy and health behavior (n = 15),^{22–25,27,29–33,35–39} and engagement and satisfaction (n = 17).^{22–26,29–33,35–40} More details are presented in Figure 3.

FIG. 3.

Effect of chatbots in pregnancy and postpartum.

Mental health

Among all included studies, a total of 13 studies reported the effect of improving the mental health of pregnant and postpartum women.^{22–24,29–33,35–37,39,40} Among them, seven studies analyzed the intervention effects,^{22,24,29–32,40} including three studies^30,32,33 had control groups and experimental groups for comparative research, while the other four studies^24,29,31,40 did not have control groups but conducted pre- and post-comparisons on their own. Chatbots showed significant effects in improving depressive symptoms, especially for highly engaged users. The average reduction in Patient Health Questionnaire-9 scores ranged from 1.32 to 3.25 points, with effect sizes ranging from moderate to large. Additionally, four studies^24,29,31,40 did not have control groups and mainly evaluated effectiveness through qualitative feedback and usage rates, with users having better acceptance and information acquisition capabilities. Improving the following five major mental health issues of pregnant and postpartum women: anxiety (n = 10),^{24,29–32,35–37,39,40} psychological stress (n = 8),^{22,27,29,31–33,35,36} depression (n = 8),^{22,25,29,32,33,35,36,40} emotion management (n = 3),^29,30,32 and preventing suicide (n = 2).^22,32

Health literacy and health behavior

Among included studies, a total of 15 studies^{22–25,27,29–33,35–39} reported that chatbots could enhance the health literacy of pregnant and postpartum women and support changes in their health behaviors. However, in most studies (n = 10),^{25,27,29–32,35–37,40} chatbots have indirectly enhanced the health literacy and supported changes in healthy behaviors. Only a few (n = 6)^{22–24,33,38,39} showed direct improvements.

Disease monitoring

Some of the chatbots in studies (n = 6)^{30–33,38,39} could enhance the disease monitoring ability of pregnant and postpartum women. Through voice interaction, they can efficiently collect individual vital sign data (such as blood sugar levels and weight),^38,39 thereby improving the efficiency of follow-up and personal health management capabilities. At the same time, this technology can effectively help reduce the incidence of infant emergencies. ³³ Studies show that user groups with higher engagement have more obvious effects, but there is still room for further improvement in the overall user acceptance. ³⁰

User engagement and satisfaction

All included studies (n = 17)^{22–27,29–33,35,36,38–40} reported high chatbot engagement and high participant satisfaction. In terms of satisfaction, 14 studies used survey instruments: researcher-developed questionnaire (n = 9)^{23,24,26,29–31,33,39,40}; Mobile App Rating Scale (n = 1) ²² ; electronic usability questionnaires (n = 1) ²⁷ ; Client Satisfaction Questionnaire-8 (n = 1) ³² ; Usefulness, Satisfaction, and Ease of Use (n = 1) ³⁵ ; and What Being the Parent of a Baby is Like–Revised (n = 1) ³⁷ ; three studies did not mention the survey tool.^25,36,38 After analyzing scale scores and user suggestions,^{22–27,29–33,35,36,38–40} all included studies indicate that users are satisfied with functional utility and interactive experience, while 12 studies also show user satisfaction with emotional support (n = 12).^{22–24,29,30,32,33,35–37,39,40}

Parenting self-efficacy

Among all included studies, most studies (n = 9)^{22,24,25,27,33,36–39} on chatbots suggested that they can enhance the parenting self-efficacy. Studies indicate chatbots directly or indirectly enhance parenting self-efficacy: five studies^{25,33,36,37,39} reported direct feedback, and four studies^22,24,27,38 reported indirect feedback.

Role adaptation

Most studies (n = 13)^{22–24,27,29–33,35–37,39} on chatbots suggested that they can enhance the role adaptation ability of perinatal women. Except for the following four studies,^25,26,38,40 none of them mentioned enhancing the role adaptation ability of perinatal women. These studies (n = 13)^{22–24,27,29–33,35–37,39} show that the chatbot directly or indirectly provides feedback to enhance the role adaptation of perinatal women. Five studies^{30,33,36,37,39} reported direct feedback, and eight studies^{22–24,27,29,31,32,35} reported indirect feedback.

Discussion

This scoping review finally included a total of 17 studies, systematically reviewing recent research on chatbot support for pregnancy and postpartum care, revealing their design characteristics, functionalities, and application effects. In general, chatbots facilitate interaction and offer a variety of functions and intelligent services.

Chatbots demonstrate significant potential in the field of pregnancy and postpartum care. Their functions encompass not only pregnancy and postpartum health monitoring,^{22,23,25,26,29,33,35,37} psychological support,^{22–27,30–33,35–38,40} and educational roles,^{23–27,29–33,35–40} but also optimization of follow-up processes between hospitals and pregnant women.^24,29,33 A review of previous studies indicates that the technological advancement of chatbot applications in health care continues to progress, ⁴¹ with their clinical value in pregnancy and postpartum care increasing substantially. ³¹ However, chatbots continue to face several practical challenges at the functional level despite relative technological advancements. For instance, effective health monitoring depends on high-quality sensors and advanced data analysis algorithms, yet the associated hardware costs and technical complexity may limit their widespread implementation. ⁴² Additionally, pregnant women have diverse and individualized psychological needs, which chatbots may struggle to address adequately due to limited capacity for personalized emotional support. ²² Educational functions also present difficulties, as chatbots must continuously update their knowledge base to ensure that the information they deliver is accurate, evidence-based, and aligned with the latest scientific guidelines. ⁴³ Moreover, follow-up functions require deep integration with existing health care systems, which raises concerns about patient privacy, data security, and regulatory compliance. ⁴⁴ Future research could explore ways to enhance chatbot design by improving multilingual and cultural adaptability, ⁴⁵ incorporating multimedia content such as videos and animations to make information more accessible, and applying machine learning techniques to dynamically optimize knowledge bases. Additionally, establishing robust auditing mechanisms is essential to ensure accuracy, timeliness, and credibility of information provided. ⁴⁶

Improvement in mental health is one of the most prominent and frequently reported outcomes associated with chatbot interventions. ⁴⁷ More than half of the reviewed studies demonstrated that chatbots can effectively reduce anxiety, depression, and psychological stress among pregnant and postpartum women.^{22–27,30–33,35–38,40} These positive outcomes are primarily supported by participants’ self-reports and validated psychological scales, suggesting the potential value of chatbots in delivering psychological support. However, some studies reported no significant improvements, indicating that effectiveness of such interventions may depend on design quality, usage frequency, and user adherence.

In terms of health literacy and behavior change, current chatbot-based health education programs primarily focus on pregnancy weight management and blood glucose control.^44,48–50 Most studies have shown that chatbots can enhance pregnancy and postpartum health literacy and promote positive health behavior.^{22–27,29–33,35–38,40} However, these interventions are often short term and lack long-term follow-up data to assess the sustainability of behavioral change, suggesting that existing interventions are predominantly supportive and motivational in nature. ³⁵ Future research should incorporate direct measures and longitudinal tracking to better evaluate the long-term effectiveness of chatbot-driven behavioral interventions.^51–53

Evaluation of chatbots in pregnancy and postpartum care primarily involves user engagement, satisfaction, and technical performance. Most satisfaction assessments rely on qualitative interviews and rating scales, yet they often lack in-depth exploration of emotional support and cultural adaptability. ⁵⁴ Technical performance evaluations serve as a critical component of chatbot assessment, with current advancements demonstrating strong capabilities in language comprehension and voice interaction technologies. ⁵⁵ However, chatbots still face limitations in handling offline scenarios and open-ended queries, which restricts their ability to address more complex user needs. Future evaluations could establish multidimensional and systematic assessment frameworks that integrate both quantitative and qualitative methods, enabling a comprehensive understanding of the practical functions and real-world impact of chatbots in pregnancy and postpartum health care.

Limitations

This review has several limitations. First, the included studies were primarily published in English and Chinese, which may have introduced language and selection bias. Second, most studies relied on quantitative designs,^{22–24,26,27,30–33,35–40} with few qualitative studies, limiting deeper understanding of users’ psychological needs, lived experiences, and contextual factors influencing engagement with chatbot-based interventions. Third, the applicability of chatbot interventions may depend on users having sufficient technical literacy and access to an appropriate digital device and a stable internet connection, which could create barriers for some populations, particularly older adults or socioeconomically disadvantaged groups.^56–58 Fourth, some studies also provided limited reporting on negative user experiences, such as frustration or dissatisfaction, and on instances where chatbots may have delivered inaccurate, incomplete, or potentially misleading information.^59,60 These issues should be more explicitly acknowledged when interpreting the findings. Fifth, the confidentiality and privacy of users’ health information require clearer consideration, as chatbot-based interventions often involve collection, transmission, and storage of sensitive personal data.^56,57,61 Future research should prioritize large-scale, multicenter, randomized controlled trials and longer follow-up periods to generate more robust evidence, while also incorporating qualitative methods to better capture user experience, accessibility challenges, safety concerns, and privacy implications.

Conclusion

This study examines the role of chatbots in supporting pregnant and postpartum women, highlighting their potential to improve self-management, offer emotional support, and enhance access to educational resources. By establishing a systematic evaluation framework and strengthening privacy protections, chatbots can increase user trust and engagement, thereby narrowing health service gaps in pregnancy and postpartum care.

Authors’ Contributions

R.H.: Study design, search strategy formulation, data collection and analysis, writing original draft, and writing review and editing. Y.J.: Study design and search strategy formulation. J.L.: Data collection and analysis. J.Z.: Search strategy formulation. M.M. and W.H.: Writing—review and editing. J.Y.-H.W.: Project administration, supervision, and methodological guidance.

Abbreviations

AI

artificial intelligence

CINAHL

Cumulative Index to Nursing and Allied Health Literature

CNKI

China National Knowledge Infrastructure

EPDS

Edinburgh Postnatal Depression Scale

EOU

Perceived Ease of Use Questionnaire

FBG

fasting blood glucose

GDM

gestational diabetes mellitus

JBI

Joanna Briggs Institute

NA

not available

NLP

natural language processing

NR

not reported

PBG

postprandial blood glucose

PDA

Parentbot-A Digital health care Assistant

PHQ-9

Patient Health Questionnaire-9

PRISMA

Preferred Reporting Items for Systematic reviews and Meta-Analyses

Q&A

question and answer

RCT

randomized controlled trial

SAT

Satisfaction Assessment Test

SEE

seek health information

SHA

share health information

SinoMed

Chinese BioMedical Literature Database

UNICEF

United Nations Children's Fund

UNFPA

United Nations Population Fund

USE

Usefulness, Satisfaction, and Ease of Use Questionnaire

VIP

VIP Chinese Science and Technology Periodicals Database

WHO

World Health Organization

Cite this article as: Huo R, Jiang Y, Lei J, Zhou J, Miu M, Hu W, Wong JYH (2026) Role of chatbots on supporting women in perinatal period: A scoping review, Women’s Health Reports 7, 1–15, DOI: 10.1177/26884844261453447.