This study aimed to explore the preferences and WTP for smart bracelets among pregnant and postpartum women using a DCE. The participants consisted of women who were pregnant or had given birth within the past 2 years, with a sample randomly selected from the Inner Mongolia Maternal and Child Health Hospital. Demographic diversity was intentionally incorporated into the sample to ensure that the research findings are broadly applicable to various subgroups of pregnant and postpartum women in the region. Using the DCE method, the study systematically assessed the preferences of pregnant and postpartum women for smart bracelets, focusing on the device’s functionality, price, and other design features, while examining their WTP for different combinations of product features. This approach provided valuable insights into their preference patterns and market demand.

This was a cross-sectional study designed to collect data in a single survey, assessing the preferences and WTP for smart bracelets among pregnant and postpartum women. To ensure the accuracy and feasibility of the study, a comprehensive set of inclusion and exclusion criteria was established.

The inclusion criteria were as follows: (1) age ≥18 years, (2) female, (3) Chinese nationality (People’s Republic of China), (4) pregnant or had given birth within the past 2 years, (5) able to complete the online survey independently or with the assistance of the investigator, (6) having basic literacy skills and being able to communicate and interact normally, and (7) willing to voluntarily participate in the study and provide informed consent.

The exclusion criteria were as follows: (1) individuals who were unconscious or had mental abnormalities, (2) individuals with cognitive impairments, (3) individuals currently participating in other similar research studies, (4) individuals unwilling to cooperate, and (5) individuals with other severe health conditions that may affect the study results.

Through this approach, the study provided valuable market and preference data for the use of smart bracelets among pregnant and postpartum women, aiding in the development and promotion of related products.

The study received approval from the Ethics Committee of Inner Mongolia Maternal and Child Health Hospital (review [176], 2024). All participants voluntarily engaged in the study and provided informed consent, confirming their understanding of the study’s objectives, methods, and potential risks. This consent was obtained prior to their participation, adhering to ethical standards for research involving human subjects. Given the involvement of human participants, appropriate ethical oversight was ensured, with all research activities sanctioned by the Ethics Committee. To safeguard privacy and confidentiality, all study data were anonymized and deidentified, thereby protecting participants’ identities. The data were securely stored and accessible only to authorized personnel involved in the research. No compensation was offered to participants for their involvement in this study. To further uphold the confidentiality and anonymity of participants, no identifiable images or data of individual participants were included in the manuscript or supplementary materials.

In this study, we conducted a systematic review to comprehensively identify and assess the key attributes and their levels that pregnant and postpartum women prioritize when selecting a smart wristband. We searched the literature across several key databases, including PubMed, Cochrane Library, Web of Science, CNKI (China National Knowledge Infrastructure), and Scopus, to ensure thorough coverage of relevant research on smart wristbands and wearable devices in maternal health management. A combination of Medical Subject Headings (MeSH) and free-text keywords was used to filter the relevant literature. The search terms included “wearable devices,” “smart wristband,” “fitness tracker,” “wearable sensors,” “maternal health,” “pregnancy,” “pregnant women,” “postpartum,” and “remote monitoring,” along with various combinations (eg, “wearable devices AND pregnancy,” “smart wristband AND maternal health”). These terms were tailored to each database’s search syntax to ensure both sensitivity and specificity in retrieval.

The literature was screened based on preestablished inclusion criteria: the research subjects had to be either pregnant or postpartum women or related health research; the studies needed to involve smart wristbands or other wearable devices; and relevant attributes such as price, privacy protection, or functionality had to be mentioned in the literature. Initially, all retrieved records were screened by title and abstract, followed by a full-text review of potentially eligible papers to confirm their eligibility.

This systematic review encompassed 7 highly relevant studies [13,29-34]. Through analysis of these studies, the research team identified 6 key attributes that pregnant and postpartum women prioritize when selecting a smart wristband: price, hospital background monitoring management, main function, privacy protection, ease of use, and frequency of monitoring report delivery. The levels associated with each of these attributes were also extracted from the literature, establishing a foundation for subsequent expert consultation.

Following the systematic review, expert consultations were conducted to further validate and refine the identified attributes and levels. These consultations involved obstetricians with clinical experience in prenatal and postpartum care (3 experts), health management professionals knowledgeable about maternal health service delivery (2 experts), wearable technology specialists with expertise in product design and human factors (2 experts), and market research experts experienced in consumer preference studies (2 experts). In total, 9 experts participated in 2 rounds of consultation.

During the first round, experts reviewed the preliminary list of attributes and levels. They were asked to comment on the relevance of each attribute, suggest modifications to levels (eg, cost ranges and descriptions of functionality), and propose any missing attributes. Specific expert feedback included the following:

After incorporating these suggestions, a second consultation round confirmed that the refined attributes and levels were appropriate, meaningful, and aligned with the decision context and target population’s concerns. Experts agreed that the defined levels reflected realistic choices that pregnant and postpartum women would encounter when choosing a smart wristband.

Through the systematic review and expert consultation, the 6 key attributes and their levels, along with the definitions of each attribute, that pregnant and postpartum women focus on when selecting a smart wristband were ultimately determined (Table 1).

On the basis of the attributes and levels determined earlier, our team used a fractional factorial design method to determine the optimal number of attributes that pregnant women prefer when selecting a smart bracelet. Given that the 6 attributes in this study each contain 2 to 4 levels, a full factorial design would result in 1152 combinations (4×2×4×3×4×3=1152), which is clearly impractical to generate. Therefore, we used a fractional factorial design to determine the optimal number of choice sets. This method is based on 2 principles: the orthogonality principle and the balance principle [35,36]. The choice sets are generated using orthogonal experiments in SPSS AU software (Beijing Qingsi Technology Co). An orthogonal strategy is selected based on the orthogonal table manual, and ineffective designs are eliminated through internal validity checks, leading to the final design. Suitable options were selected as controls to form multiple-choice sets. Each participant faced 12 choice sets, with each choice set containing 3 options: smart bracelet A, smart bracelet B, and “none.” To assess the validity of the DCE questionnaire, a choice set with a clear dominant option (a lie detector question) was included in the questionnaire. Only when participants select the correct option from this set will their data be included in the subsequent analysis. Meanwhile, according to the empirical rule proposed by Johnson and Orme [37,38], the minimum sample size for a DCE study can be calculated using the formula:

where N is the minimum sample size, representing the number of participants required in the study; c is the largest number of levels for any attribute (c=4), t is the number of choice tasks (t=12), and a is the number of alternatives (a=2).

The final calculation resulted in a minimum sample size of 84.

Table 2 is an example of a DCE choice set.

Data were collected using an online survey platform, such as WenjuanXing. The study commenced in September 2024 and concluded in May 2025, lasting approximately 9 months. Recruitment took place from February 2025 to April 2025. The data collection method involved researchers identifying eligible participants offline and requesting them to scan a WenjuanXing link to complete the online survey. During the recruitment period, researchers provided eligible pregnant and postpartum women with comprehensive information about the study and invited them to participate by completing the online questionnaire.

The survey was divided into 2 distinct parts. The first part gathered essential demographic information from participants, including age, education level, and place of residence, to gain a comprehensive understanding of their backgrounds. The second part focused on the preferences of pregnant and postpartum women concerning smart bracelets, using the DCE method for data collection. The DCE questionnaire featured multiple-choice scenarios, each presenting 2 different combinations of smart bracelet attributes, from which participants were asked to select the combination that best aligned with their preferences. The attributes of the smart bracelets encompassed device functions (such as heart rate monitoring, activity tracking, and sleep tracking), price, ease of use, and privacy protection, all chosen based on prevalent features available in the current market. To mitigate sequence effects, all choice scenarios were randomly arranged.

To ensure data quality and protect participants’ privacy, the survey was conducted anonymously, using encryption technology to safeguard the data. All participants voluntarily completed the questionnaire after receiving an informed consent form, which ensured their understanding of the study’s purpose and their involvement [39]. Researchers provided necessary assistance to facilitate the completion of the survey. During data collection, logical checks and additional measures were implemented to ensure the validity and consistency of the data.

Descriptive statistics were used to summarize the sociodemographic characteristics of the participants, including age, income, education, occupation, pregnancy stage, parity, and the presence of pregnancy complications. Means and SDs were computed for continuous variables, while frequencies and percentages were reported for categorical variables. To analyze participants’ preferences regarding smart bracelet attributes, a mixed logit model was used. This model accommodated preference heterogeneity among individuals by allowing the attribute coefficients to vary randomly. Each participant’s utility for a specific smart bracelet profile was modeled as a linear combination of attribute levels. The estimated parameters included the fixed effects (β) of each attribute level, their SEs, and corresponding P values. The relative importance scores quantify the significance of each attribute, calculated by dividing the difference in utility between the lowest and highest levels of that attribute by the sum of the differences for all attributes. Cost was treated as a continuous variable to facilitate the estimation of participants’ WTP for different attributes. WTP values were derived by dividing the attribute coefficient by the negative of the cost coefficient, reflecting the monetary value participants assigned to specific smart bracelet features. Subgroup analyses were conducted to investigate differences in preferences and WTP across various demographic and clinical subgroups, including income levels, employment status, pregnancy stages, parity, pregnancy complications, and age. All statistical analyses were executed using Stata 15.0 (StataCorp LLC). A 2-sided P value of <.05 was deemed statistically significant.

A total of 480 pregnant women were initially recruited for the study; after excluding those with incomplete data or failing logical consistency checks, 464 participants were included in the final analysis, yielding a high data validity rate of 96.67%. Among the 464 participants, the mean age was 31.06 (SD 4.05) years (Table 3). The majority were of Han ethnicity (385, 82.97%) compared to ethnic minorities (n=79, 17.03%). For education level, 90.95% (n=422) had a high level of education. Most participants resided in urban areas (n=446, 96.12%). Income distribution was as follows: ≤4000 (n=102, 21.98%), 4001 to 6000 (n=170, 36.64%), and ≥6001 (n=192, 41.38%). Occupational status included employed (n=353, 76.08%), unemployed or out of work (n=60, 12.93%), and without fixed occupation (freelance, n=51, 10.99%). Pregnancy status was distributed as early pregnancy (1‐13 wk; n=122, 26.29%), midpregnancy (14‐27 wk; n=107, 23.06%), late pregnancy (≥28 wk; n=110, 23.71%), and postpartum (n=125, 26.94%). First-time pregnancy was reported by 71.98% (n=334). Regarding pregnancy complications, 84.05% (n=390) had none.

This study analyzed the preferences and WTP of pregnant and postpartum women for smart wristbands through a DCE (Table 4). The results showed that cost had a negative impact on choice (β=−0.000257; P=.01). For the hospital background monitoring management attribute, the most preferred level was having hospital background monitoring management (β=.226; P=.03); in terms of primary functions, the most preferred function was fetal heart monitoring (β=1.275; P<.001); for the privacy protection attribute, the most preferred level was high-level protection (β=.541; P<.001); in terms of ease of use, the most preferred level was relatively easy to use (β=.973; P<.001); and for the monitoring report frequency, the most preferred level was once a day (β=.882; P<.001).

As presented in Table 5, the main function was the most important attribute.

Table 6 presents the estimated WTP for changes in specific attributes of smart bracelets, with WTP values quantifying participants’ implicit monetary trade-offs between improvements in these attributes. In DCE studies, WTP reflects the marginal additional amount a respondent is hypothetically willing to pay for an enhancement in a given attribute level relative to a reference level. This is derived from the ratio of attribute coefficients to the cost coefficient in the mixed logit model, effectively capturing the relative valuation of attribute changes. Compared to the reference level of activity tracking, participants were willing to pay an additional ¥2975.168 (95% CI 743.556-5206.781) for sleep quality, ¥4806.745 (95% CI 990.014, 8623.475) for vital sign monitoring, and ¥4967.454 (95% CI 1212.358, 8722.550) for fetal heart monitoring. For the privacy protection attribute, participants were willing to pay ¥847.659 (95% CI 162.234-1533.085) for standard protection and ¥2109.294 (95% CI 551.576-3667.013) for high-level protection, both compared to no special protection. Regarding ease of use, relative to the category of “very difficult to use,” participants were willing to pay ¥2581.683 (95% CI 598.881-4564.485) for relatively difficult-to-use devices, ¥2802.522 (95% CI 539.744-5065.300) for devices that are very easy to use, and ¥3792.292 (95% CI 860.861-6723.723) for relatively easy-to-use devices. Finally, concerning monitoring frequency, compared to the reference level of “once a week or less often,” participants were willing to pay ¥2225.933 (95% CI 148.084-4303.781) for using it once every 2 days and ¥3437.094 (95% CI 713.160-6161.028) for using the device once a day.

We further conducted subgroup analyses based on demographic and clinical characteristics, including income, employment status, pregnancy stage, parity, presence of pregnancy complications, and age (Tables S1-S7 in Multimedia Appendix 1). The results showed that these characteristics significantly influenced pregnant and postpartum women’s preferences for the attributes of smart bracelets. In terms of the price attribute, pregnant and postpartum women with low income (income ≤ ¥4000), those who were unemployed, those in early or late pregnancy, those with previous births, those without pregnancy complications, and those of younger age (age ≤ 30 y) exhibited a significantly stronger preference for lower-priced smart bracelets (P<.05); whereas women with moderate income (¥4001‐¥6000), those without fixed employment (self-employed), primiparous women, those without pregnancy complications, and younger women showed a significantly stronger preference for smart bracelets with hospital background monitoring functions (P<.05). In terms of specific feature preferences, pregnant and postpartum women with low and high income, those who were employed, those in mid to late pregnancy, both primiparous and multiparous women, women with or without pregnancy complications, and older women (age >30 y) exhibited a significant preference for smart bracelets equipped with fetal heart monitoring functions (P<.05), whereas women with moderate income, those without fixed employment, and those who were unemployed, women in early pregnancy or postpartum, and younger women exhibited a significant preference for smart bracelets with vital sign monitoring functions (P<.05). Except for unemployed pregnant and postpartum women, all other subgroups showed a significant preference for smart bracelets with high levels of privacy protection (P<.05). With respect to ease of use, pregnant and postpartum women without fixed employment and those in late pregnancy showed a stronger preference for very easy-to-use devices, while the remaining subgroups all preferred devices that were relatively easy to use (P<.05). In addition, all subgroups of pregnant and postpartum women exhibited a significant preference for a once-daily notification frequency. Overall, factors such as income, employment status, pregnancy stage, parity, presence of pregnancy complications, and age influenced Chinese pregnant and postpartum women’s preferences for different smart bracelet attributes to varying degrees.

Our results indicate that the functionality of smart bracelets is the most influential attribute affecting participant preferences, followed by ease of use, frequency of monitoring reports, and price. Participants favored smart bracelets that were affordably priced, offered hospital backend monitoring, included fetal heart rate monitoring, ensured a high level of privacy protection, were relatively user-friendly, and provided frequent monitoring reports. Furthermore, our findings reveal that participant preferences for smart bracelets varied based on sociodemographic characteristics, gestational stage, parity (first-time pregnancy), and the presence of pregnancy complications. Understanding the preferences of participants within these subgroups may facilitate the development of tailored smart bracelets.

The preference for lower-cost smart bracelets aligns with previous research [40]. Our subgroup analysis further demonstrated that participants with lower incomes and those who were unemployed exhibited a significant preference for lower-priced smart bracelets. Jacob et al [41] found that low-income individuals display greater price sensitivity compared to their high-income counterparts, a sensitivity that is exacerbated by income disruption among the unemployed. Additionally, we observed that participants in both early and late pregnancy showed a stronger inclination toward lower-priced smart bracelets. The increased price sensitivity among women in early pregnancy may be attributed to their anticipation of forthcoming childbirth-related expenses, such as delivery fees and newborn care [42]. Notably, heightened price sensitivity during late pregnancy could be linked to the expectation of additional costs related to the imminent delivery [43]. Furthermore, our study revealed that participants who were not experiencing their first pregnancy also exhibited heightened price sensitivity, potentially due to the stress of reallocating household resources [44]. Younger pregnant individuals demonstrated a higher level of price sensitivity, likely reflecting their position in the early stages of their careers, where income levels are generally lower compared to those in mid-to-late career stages. This underscores the necessity of providing cost-effective smart bracelets, particularly for lower-income, unemployed, and first-time pregnant individuals.

Participants expressed a clear preference for smart bracelets equipped with hospital backend monitoring. This finding aligns with Bachiri’s research on the integration of mobile health systems, which demonstrated an increase in maternal adherence when hospital data connectivity was enabled [45]. Subgroup analysis revealed that participants with moderate incomes and those experiencing their first pregnancy exhibited a significant preference for hospital backend monitoring, likely due to their reliance on medical authority [46,47]. Additionally, self-employed participants showed a heightened preference for hospital backend monitoring, indicating a perceived need for professional medical support in nontraditional work settings [48]. Younger participants also demonstrated a notable preference for hospital backend monitoring, reflecting a reliance on institutionalized medical care within this demographic [49]. Therefore, incorporating hospital backend monitoring into smart bracelets would address the preferences of first-time mothers, self-employed individuals, and younger users, thereby enhancing trust in the device’s medical relevance. For the design of smart bracelets, it is crucial to integrate fetal heart rate monitoring for late-stage pregnancy while also recognizing the importance of vital sign monitoring for early pregnancy and postpartum users.

Regarding the functionality of smart bracelets, participants favored those with fetal heart rate monitoring. This aligns with prior research, such as Ahmed et al [50], which indicated that fetal heart rate monitoring was the significantly preferred feature among pregnant women, directly related to their sensitivity to pregnancy health risks. Other studies have shown that abnormal fetal heart rates are important early warning signs of pregnancy complications, and the continuous monitoring capabilities of smart bracelets can increase the detection rate of anomalies [13,51]. However, subgroup analysis revealed that both low-income and high-income groups preferred fetal heart rate monitoring, while the middle-income group tended to favor vital sign monitoring. This differential preference among the middle-income group may stem from their heightened health management awareness, leading them to favor multidimensional health data integration [52]. This nonlinear relationship may also arise because high-income individuals prioritize fetal safety, while low-income individuals rely on basic monitoring as a substitute for prenatal care, and middle-income individuals seek comprehensive health management [53-55]. Our research also found that freelancers and unemployed participants preferred vital sign monitoring, while employed participants preferred fetal heart rate monitoring. This may be because self-employment or unemployment allows more time to focus on one’s own physiological changes, such as adjusting lifestyles by monitoring sleep quality or stress levels [56]. Unemployed individuals may experience higher health anxiety due to financial pressures, leading them to prefer vital sign data for preventing potential health risks. For employed individuals under work pressure, fetal heart rate monitoring is viewed as a direct indicator of fetal health, enabling quick identification of anomalies and reducing the frequency of medical visits [7,57]. Additionally, employed individuals may have access to professional prenatal care through workplace medical benefits, thereby reducing their need for vital sign monitoring [58,59]. Participants in early pregnancy and postpartum preferred vital sign monitoring, while those in the mid-to-late stages of pregnancy preferred fetal heart rate monitoring. This may be because, during the initial bodily adaptation phase of early pregnancy, women are more concerned with their own physiological changes (eg, morning sickness and fatigue) [60,61] and early warning signs of pregnancy complications (eg, anemia and hypertension) [62,63], and there is an increased need for health monitoring during the recovery period, such as postpartum depression risk, blood pressure fluctuations, and wound healing tracking [64,65]. Literature indicates that continuous vital sign monitoring is feasible within 7 months postpartum [66]. The mid-to-late stages of pregnancy are critical periods for fetal development. Fetal movement becomes regular after 20 weeks, and fetal heart rate monitoring becomes a core indicator for assessing fetal development, directly related to pregnancy complications such as preeclampsia and fetal growth restriction [67-69]. Pregnant women in late pregnancy experience a significant increase in anxiety about fetal safety, and fetal heart rate data can provide immediate psychological reassurance [70].

Regarding privacy protection, participants generally preferred smart bracelets with a high level of privacy protection, indicating that data security is a core concern for users. This finding aligns with existing research highlighting users’ privacy concerns related to wearable health devices [71,72]. Additionally, participants placed a high value on ease of use. Notably, while very easy-to-use bracelets were favored to some extent, the intensity of preference was lower compared to “relatively easy-to-use” bracelets. This may be due to participants feeling that very simple bracelets lack sufficient functionality or intelligent experience, reflecting a synergistic need for functional depth and operational convenience within the pregnant and postpartum population [73]. This observation aligns with human-computer interaction research on wearable devices, where overly simplified interfaces can undermine user trust in data reliability [74,75]. Younger participants preferred vital sign monitoring, while older participants favored fetal heart rate monitoring. Younger pregnant women perceive their pregnancy risk as lower, focusing more on their own health rather than fetal-specific indicators. In contrast, older mothers face a significantly higher incidence of pregnancy complications (eg, gestational hypertension and premature delivery), leading them to consider fetal heart rate monitoring as a necessary risk management tool [16,76]. Therefore, smart bracelet design should prioritize user privacy and ease of use, balancing simplicity with functionality to build trust among all user demographics.

Regarding the frequency of monitoring report delivery, participants generally preferred smart bracelets capable of providing more frequent updates. This preference indicates that regular health data updates are a significant requirement for pregnant and postpartum individuals seeking to manage their health through real-time feedback. However, it is crucial to acknowledge that high-frequency monitoring may increase health anxiety [77]. Therefore, it is advisable to mitigate the false alarm rate through AI-assisted interpretation [78].

The design of smart bracelets must comprehensively consider key factors such as functionality, price, ease of use, privacy protection, and report frequency. This approach is essential for accurately aligning with the diverse preferences of pregnant and postpartum individuals across various gestational stages, income levels, employment statuses, and parity. Priority should be given to integrating functions for monitoring fetal heart rate and vital signs, with differentiated configurations tailored to user characteristics. While ensuring data privacy and operational convenience, the human-computer interaction interface should be optimized to avoid overly simplified functions that could undermine user trust. Furthermore, establishing appropriate monitoring frequencies and incorporating AI-assisted interpretation can enhance the practicality and reassurance of data feedback. Only by addressing these multifaceted needs can smart wearable devices be effectively tailored to meet the demands of the maternal population.

This study has several limitations. First, the sample was primarily drawn from a specific regional population of pregnant and postpartum individuals, which may not fully represent the preferences of users across diverse regions, cultural backgrounds, and health care environments. Second, the data were based on self-reported responses, which are subject to potential biases, such as subjective perception and social desirability, and may not accurately reflect actual user behavior during device usage. Third, although detailed subgroup analyses were conducted, other potentially influential variables—such as digital health literacy and family support levels—were not explored. Additionally, the selection of smart bracelet features was based on current mainstream product configurations, which may change as technology evolves. Furthermore, while a mixed logit model was appropriately used to estimate preferences and accommodate heterogeneity across the overall sample, some individual subgroups (eg, “unemployed or out of work” and “no fixed occupation (freelancer)”) had relatively small sample sizes. Small subgroup counts can increase uncertainty when interpreting subgroup-specific effects or interactions, potentially yielding unstable estimates or larger SEs for those categories. For this reason, findings related to these small subgroups should be interpreted with caution and not overgeneralized. Future research should consider expanding the sample scope and incorporating longitudinal studies and real-world usage data to further validate and enrich the findings, as well as examine additional factors that may influence preferences for wearable health technology.

In conclusion, preferences for smart bracelets among pregnant individuals are influenced by various factors, including functionality, price sensitivity, ease of use, and privacy concerns. The most significant features identified are the primary function and ease of use. Design considerations should account for the diverse needs across different sociodemographic groups, stages of pregnancy, and potential complications. Specifically, cost-effective options, robust privacy protections, and user-friendly interfaces are essential to ensure widespread adoption. Furthermore, integrating real-time monitoring with AI-assisted feedback can enhance the usability of smart bracelets, addressing concerns related to health anxiety and improving the overall user experience. Ultimately, a personalized and comprehensive approach to smart bracelet design can better support the health and well-being of pregnant women during both their pregnancy and postpartum periods.