This study was assessed and approved by the Medical Ethical Committee of the Erasmus University Medical Center, Rotterdam, The Netherlands (MEC-2020-0974). Informed consent was obtained from all participants via email. Considering the low risk of this study, composing a Data Safety Monitoring Board was deemed unnecessary.

Our aim was to include 40 participants in this study. Between September 2021 and June 2022, 337 women registered for this study, of which 102 (30.3%) were eligible for inclusion. Due to a higher-than-expected confidentiality, integrity, and availability (CIA) Triad classification (a risk score regarding user information safety [29]), additional security demands were necessary. Fulfilling these demands delayed the launch of Pregnant Faster from September 2021 to November 2021, leading to a loss of recruited eligible women. From the launch onward, 47 (46.1%) women were included, of which 39 (83%) completed the intervention (Figure 2).

Recruitment took place through the Sneller Zwanger website [30], which was distributed through posters and flyers, advertisements on the social media platforms Facebook and Instagram, and midwifery practices that provide primary care to all pregnant women in the Netherlands. Additionally, a collaboration took place with the Dutch influencer Midwife Mother (Dutch Verlosmoeder) on Instagram [31-33]. Participants filled in a survey with their first name, age, telephone number, email address, zip code, educational level, and when they planned on trying to become pregnant (currently pregnant, currently trying, or trying in ≤3, >3-12, or >12 months). Pregnant women were asked to fill in their estimated or calculated due date.

Participants were selected based on the following inclusion criteria: assigned female at birth, 18-45 years old, actively trying to become pregnant within now and 12 months or pregnant with a gestational age of <8 weeks at the start of the intervention, a low-to-intermediate educational level (prevocational or vocational education), and able and willing to download and evaluate the app.

Exclusion criteria were as follows: insufficient proficiency in the Dutch language, not in possession of a smartphone or tablet suitable for the app, and refusal to download or evaluate the app. All excluded women received a free coupon for the Dutch or English version of Smarter Pregnancy [34].

Pregnant Faster was developed by the Erasmus University Medical Center’s research group Periconception Epidemiology at the Department of Obstetrics and Gynecology, in collaboration with TJIP The Platform Engineer and the event bureau Improve. A detailed description of the cocreation and design process of Pregnant Faster and the study protocol has been published in JMIR Protocols [28].

Eligible women were sent the patient information folder, in which the intervention was explained. Inclusion was finalized after a telephone conversation in which further clarification could be provided. An email was sent with instructions on how to download and install Pregnant Faster from Apple App Store (iOS) or via a link (Android). If more than 3 days passed between inclusion and downloading, participants were approached twice by email and telephone and once by a text message to provide further support with installation.

The first log-in marked the start of the 4-week intervention. During this period, participants logged in with their email address and a password, which yielded 1 coin per day per log-in. The first log-in yielded 50 coins as a reward for installation and to immediately stimulate participants to further engage with the app. After log-in, a dashboard appeared, containing 5 buttons: (1) “Earn coins,” (2) “Overview coins,” (3) “See a midwife!,” (4) “This study,” and (5) “Rewards” (Figure 3).

Button 1, “Earn coins,” led to a timeline where new blogs and tips appeared daily (Multimedia Appendix 1). Reading this information yielded 4-8 coins. In the same timeline, a daily questionnaire appeared in which participants could tick a box if they ate sufficient fruit and vegetables, exercised, and took folic acid supplements that day. Each ticked box yielded 2 coins per day. Button 2, “Overview coins,” displayed when and how coins were earned and how many coins were spent on which products. Button 3, “See a midwife!,” contained information regarding what PCC is and who it is for, stimulating participants to register through the app for a PCC consultation. Registering consisted of filling in their phone number, which immediately yielded 25 coins. An additional 75 coins were allocated after the visit was confirmed by the midwife. Button 4, “This study,” contained information about the study itself and contact details for support. Button 5, “Rewards,” contained an in-app shop where participants could order rewards, including (but not limited to) folic acid supplements, fruit, nail polish, mascara, ovulation or pregnancy tests, and newborn clothing. Rewards were sent to their home address to arrive within 5 business days. If a participant had not logged in for 7 days, they received a manually sent text message and email, encouraging them to read up on the newly offered blogs and tips, earn more coins, and order rewards.

At the end of the intervention, participants were offered a coupon for Smarter Pregnancy via the timeline, which would yield 25 coins upon use and provide them with an additional 26 weeks of coaching. Furthermore, they received an email regarding finalization of the study and available means of support. Earned coins could be spent up to 2 weeks after the intervention ended. The blogs and tips remained accessible for as long as the app remained installed. Figure 4 provides an overview of the study flow.

During the study, registration and inclusion rates were tracked and a log book was kept to note any encountered barriers and changes to the protocol.

Participants’ baseline characteristics were collected prior to inclusion through the selection survey, and home addresses were collected via email after obtaining informed consent. If selected participants did not respond to attempts to include them, they were contacted twice by email, twice by telephone, and once by a text message.

After using Pregnant Faster for 4 weeks, participants filled in a modified version of the 18-item mHealth App Usability Questionnaire (MAUQ), which uses a 7-point Likert scale (Multimedia Appendix 2) [35]. In addition, the first 10 (25.6%) participants went through a semistructured interview (Multimedia Appendix 3) in which they elaborated on their experiences with the app and, if applicable, the PCC consultation. The audiotapes of these interviews were used to compose the Experience Questionnaire (ExQ; Multimedia Appendix 4) that consisted partly of questions using a 5-point Likert scale. The ExQ was offered to the remaining participants, and the first author filled in the ExQ for the first 10 (25.6%) participants, using the audiotaped data provided during the interviews. If a question in the ExQ was not clearly answered in the interview, the participant was approached by telephone to provide an answer. The answers to the open questions in the ExQ were evaluated for notable and recurring comments.

At the end of the study period, data were collected on the number of coins earned, the types of rewards that were chosen, and the number of booked and visited PCC consultations.

To evaluate the inclusion strategy, the following percentages were calculated: (1) eligibility percentage, (2) inclusion percentages, and (3) intervention completion percentages. The eligibility percentage was determined by dividing the number of women eligible for inclusion by the total number of women who registered for the study. The inclusion percentages were calculated by dividing the number of included participants (who provided informed consent) by the number of total registrations and the number of eligible women who were approached for inclusion. Completion of the intervention entailed completing the evaluation of the app. The intervention completion percentages were obtained by dividing the number of participants who completed the intervention by the number of eligible women and the number of participants who provided informed consent.

The baseline of the study population is presented in tabular form using the median (IQR) for continuous data and n (%) for categorical data. Data obtained through the MAUQ and the ExQ are presented in bar charts. Notable answers to open questions are presented in narrative form. Data on feasibility from the researchers’ point of view are presented as bullet points.

All calculations were carried out using SPSS Statistics 25 (IBM Corporation), charts were created using Excel 2016 (Microsoft), and figures were created in PowerPoint 2016 (Microsoft).

A total of 337 women registered for the intervention, of whom 102 (30.3%) were eligible for inclusion. Informed consent was signed by 47 (46.1%) women, and 39 (83%) of the 47 women participated and completed the intervention. Table 1 displays participants’ baseline characteristics.

A 2-month gap arose between the start of recruitment and the intervention, due to the app’s CIA Triad classification [29]. A low classification was expected, but the combination of 40 intended participants and their registering their first name and email address in the app warranted a slightly higher classification for confidentiality and therefore additional security demands. Despite frequent updates to keep eligible women engaged, 55 (53.9%) of 102 women did not respond when inclusion commenced. Next, we describe these events and the inclusion process in detail.

Between September 2021 and January 2022, 212 (62.9%) women from the total 337 registrations reached in June 2022 registered for the study. Of these 212 women, only 9 (4.2%) were included. To boost the registration and inclusion rates, more flyers and posters were distributed, and the choice was made to include women who were trying to become pregnant within 12 months as opposed to within 3 months, as originally intended. Furthermore, the intervention was expanded from the municipality of Rotterdam to nationwide, delivering rewards through the postal service instead of by car. Midwives throughout the Netherlands were actively approached to ask whether they were interested in participating in the study and were offered support in setting up PCC consultations in their practices. Subsequently, the collaboration with Midwife Mother was renewed, who uploaded another post and multiple stories regarding PCC and Pregnant Faster to her Instagram. All women who were previously excluded based on not living near Rotterdam were contacted and asked to participate.

These efforts showed a limited effect. By May 2022, an additional 54 (16%) registrations and a total of 16 (34%) inclusions were obtained, which led to the decision to drop the inclusion criterion of living in a deprived neighborhood, thereby lowering the chance of including women likely to have a VHS. This choice was made to allow for further development and testing of Pregnant Faster while searching for a more effective way to recruit the intended target group for the planned larger cohort. Another social media campaign was conducted, and all women who were previously excluded based on their neighborhood’s median income were invited to participate.

Between May 2022 and July 2022, 71 (21%) more women registered for the study, adding up to the total of 337 registrations. Registration was closed after no new women registered for 2 weeks. From May onward, 31 participants were included, adding up to a total of 47 inclusions, of which 39 (83%) completed the intervention and 8 (17%) dropped out. Of these 8, 3 (37.5%) women who provided informed consent did not respond to instructions on how to install the app or attempts to reach them; 2 (25%) women were unable to install the app: in one case, an iPhone with Belgian settings preventing download from the Dutch Apple Store, and in the other case, the woman who had an Android device was scared by the warning prompted by download of an app outside of Google Play Store. Of the 3 (37.5%) remaining dropouts, 2 (25%) stopped trying to become pregnant and 1 (12.5%) found the questionnaires too burdensome. Table S1 in Multimedia Appendix 5 provides an overview of the eligibility, inclusion, and intervention completion percentages.

A total of 17 (43.6%) of 39 participants registered for a PCC consultation, and 16 (41%) consultations were conducted by 9 midwifery practices. One consultation was performed via telehealth by the first author because the midwife chosen by the participant had no experience in providing PCC and did not wish to implement PCC in her practice. The participant who registered for PCC but did not attend a consultation was worried about her health care insurance not covering the costs. Stating to feel overwhelmed, she declined additional support as well as a free telehealth consultation.

The most often reported reason to register for a PCC consultation was to obtain more personalized information (14/17, 82.4%), followed by being curious about what a consultation entails in practice (6/17, 35.3%). The most frequent reason not to register for a consultation was simply not being interested in doing so (6/22, 27.3%). A visual overview of participants’ motivation regarding registration for PCC can be found in Figures S7 and S8 in Multimedia Appendix 6.

All participants who visited a PCC consultation agreed that registering through Pregnant Faster is easy (2/16, 12.5%, agree; 14/16, 87.5%, strongly agree) and were glad they had done so (3/16, 18.7%, agree; 13/16, 81.3%, strongly agree).

During the study, participants could earn a maximum amount of 468 coins. Together, they earned a total of 11.791 coins (mean 284, SD 109 per participant; median 276, IQR 221-358; range 79-443). In total, 344 rewards were ordered during the study period (mean 8, SD 6 per participant; median 7, IQR 3-12; range 0-22). One participant did not wish to order rewards because she was happy with “just the app.” She stated that she did not feel it was morally objectionable to be rewarded but just that she was not interested in receiving rewards.

The most popular reward was a €10 (US $10) book voucher, with 19 (48.7%) of 39 participants ordering the voucher at least once. The second-most popular reward was fruit, with 17 (43.6%) participants ordering fruit at least once and a total of 87 orders (87/344, 25.3%). Bananas were the most popular fruit, amounting to 32 (36.8%) of 87 fruit orders. The third-most popular reward was a set of 2 home pregnancy tests, with 16 (41%) participants ordering this reward at least once. Most participants ordered a reward more than once, displaying their personal preferences and satisfaction regarding their previous order. Table S2 in Multimedia Appendix 5 displays all rewards and their order frequency and percentage.

During the study, 11 issues were noted that should be considered before attempting to establish Pregnant Faster’s (cost-)effectiveness in a larger cohort study:

In this pilot study, we aimed to determine the feasibility of the app-based nudge Pregnant Faster, which is designed to fit the needs of women with a low SES and a high likelihood of having a VHS, who have a higher risk of adverse pregnancy outcomes. The aim of Pregnant Faster is to encourage these women by nudging them to adequately prepare for pregnancy through education by making healthy lifestyle choices and engaging in PCC, which will help improve their pregnancy outcomes.

Pregnant Faster has shown to be feasible from the users’ point of view, showing high user satisfaction with a rating of 7.7/10 and PCC uptake by 16 (41%) of 39 participants. Notably, 27 (69.2%) participants stated to have learned a lot about pregnancy preparation and 28 (71.7%) felt motivated by the app to make healthy lifestyle choices. After the intervention ended, 12 (30.8%) participants stated that they more often make healthy choices than prior to using Pregnant Faster.

With regard to the 55 (53.9%) of 102 eligible participants who did not respond when inclusion commenced after the 2-month delay, we suspect that a loss of interest and perhaps of trust in the intervention played a role. The amount of lost eligible women suggests that time is a limiting factor, impacting women’s willingness to participate in the intervention. This emphasizes the necessity of quickly responding to their willingness to participate and acceptance of offered care.

Feasibility from the researchers’ point of view was satisfactory as well but only with regard to practical conduction, as adjustments to the inclusion criteria were made to up the number of inclusions. Dropping the criterion of living in a deprived neighborhood likely impacted the chance of including women who actually have a VHS. The feasibility of Pregnant Faster from the researchers’ point of view can be improved by developing a new method of finding the target group, making the app available via both Apple App Store and Google Play Store, and automating (parts of) the inclusion process and coin allocation, which will limit the number of administrative tasks.

To the best of our knowledge, Pregnant Faster is the first mHealth intervention that aims to encourage adequate pregnancy preparation and increase the uptake of PCC, promoting blended lifestyle care, by nudging participants with a loyalty program consisting of earning and spending coins. During study conduction and after evaluation, important knowledge was gained concerning the strengths and limitations of this intervention and how best to proceed with a larger cohort study.

Despite our earlier experiences regarding the recruitment of women who likely have a VHS, we did not manage to conduct this study adhering to the original inclusion criteria [28]. It is possible, therefore, that the user feasibility would have been different had the full study population met the intended criteria.

Pregnant Faster has been designed through iterative cocreation, actively involving the target population in its development [28]. Even though adjustments were made, we consider this pilot study to be another step in the iterative cocreation process, as the results will be used for further development of the app and nearly half (19/39, 48.7%) of the study population met the criteria of living in a deprived neighborhood.

The insights gained through this study have prompted us to re-evaluate which characteristics to use as a proxy for a low SES and the associated VHS. To improve recruitment of the target group, we have hosted meetings with health care professionals specializing in health-related vulnerability and adverse pregnancy outcomes to gain more insight and develop new inclusion criteria for a larger cohort study. At this moment, we are researching (combinations of) different inclusion criteria based on self-reported vulnerability markers, such as high stress; financial insecurity; addiction to alcohol, drugs, and tobacco; and lack of social support, which are also associated with unfavorable health outcomes [36]. Through developing these new criteria, we aim to be more inclusive and provide support to all women with a certain degree of health-related vulnerability, instead of limiting support to those with a high likelihood of having a VHS based on the educational level and neighborhood deprivation.

We aim to continue promoting Pregnant Faster on social media platforms, such as Instagram and YouTube. These platforms have been known to use algorithms that successfully reach target audiences and prove to be effective tools with regard to providing people with support and education [37,38]. Using these platforms, therefore, will not only support recruitment and benefit the target population but also allow Pregnant Faster to contribute to pregnancy-related health in the general population.

The knowledge gained through this pilot study has inspired us to research different methods of information transfer to ensure the app fits the needs of the target group and improve Pregnant Faster’s accessibility for those who experience limited literacy [39]. For the planned cohort study, for example, we have created an audio version and infographic of the patient information folder and informed consent form. Furthermore, we are currently creating additional content for the app, again focusing on multiple methods of information transfer, such as podcasts, videos, and infographics.

On a technical level, the inclusion process and content management system will be adjusted to reduce manual tasks and promote feasibility. Furthermore, we plan to change the log-in procedure to a pin code or fingerprint and enhance the app with daily notifications.

Regarding focusing more on participants’ partners, we have chosen to not adhere to this suggestion at the current time, as the tips and blogs already contain information for partners and we are still in the process of establishing (cost-)effectiveness and further developing Pregnant Faster. In research concerning reproductive health, it is known that partners may sometimes take on a more passive role [40], which places the burden of preparing for pregnancy largely on the person who will carry the baby. For future development, therefore, we will consider the possibility of adding personalized settings to allow users to fill in characteristics that will adjust the app’s content accordingly, such as relationship status, gender and sexual orientation, and, if applicable, gestational age and the use of donor semen.

In the future, we wish to investigate the possibility of offering Pregnant Faster to all who wish to become pregnant, possibly with rewards if cost-effectiveness is established.

With this pilot study, we have demonstrated that the app-based nudge Pregnant Faster provides a feasible way to stimulate the uptake of PCC and boost participants’ motivation to adequately prepare for pregnancy. We will use the knowledge we have gained through this pilot study to create an updated version of the app, which will be named Pregnant Faster 2. Our next step consists of determining the (cost-)effectiveness of Pregnant Faster 2, for which we will conduct a cohort study of 400 women with a VHS based on newly devised inclusion criteria.