This exploratory qualitative study was guided by a constructivist and interpretivist research paradigm [36]. Through one-on-one interviews, we sought to explore the perspectives of our study participants within the unique contexts of their roles and experiences. We recognize the reflexivity of our research team in our data collection and the interpretation of our study findings, where research team members included bachelor-trained and doctoral-trained health services researchers. The reporting of our findings is guided by the Standards for Reporting Qualitative Research checklist [37].

We conducted one-on-one interviews with patients and providers from 2 obstetrics and gynecology ambulatory care clinics at The Ohio State University Wexner Medical Center, an academic medical center (AMC), between September 2020 and January 2021. Patients who were 18 years or older, pregnant or up to 90 days postpartum, and spoke English were eligible to participate. Eligible providers included physicians and nurses who worked in either or both clinics. Purposeful sampling was used to recruit patients to the study by telephone call inviting their participation, while providers were recruited by email. Emails sent to providers included a Qualtrics survey link to indicate their interest in participating and their preference for a telephone or videoconference interview.

Separate semistructured interview guides were used to support data collection from patients and providers (Multimedia Appendices 1 and 2). These interview guides were developed by 3 members of the research team (ASM, SRM, and NF) who are doctoral-trained health services researchers. Questions for patients asked about their experiences and preferences for sharing information with their providers, including self-reported symptoms and physiologic measures. The patient interview guide was tested with 2 individuals—1 pregnant and 1 post partum—before finalizing the interview questions. Provider interviews asked about current strategies for collecting patient-generated information, opportunities to implement new strategies to collect this information in the future, and potential challenges faced when collecting this information.

Patient interviews were conducted by 2 female members of the research team who were either doctoral-trained or bachelor-trained (SRM and Abigail Petrecca). Patients did not know their interviewer prior to their interview. Patients were contacted by telephone and were given the option to complete the interview at the time of the call or schedule their interview at another time that was convenient for them.

Providers who were interested in participating in the study from the initial recruitment email provided a preferred time to schedule either a phone or videoconference (ie, Zoom) interview. Provider interviews were completed by 2 members of the research team, 1 female and 1 male, who were both doctoral trained (SRM and NF). Most providers did not know their interviewer prior to their interview; however, some providers had met their interviewer in a professional capacity prior to their interview. All interviews were audio-recorded and transcribed verbatim to permit rigorous qualitative analysis.

We used thematic analysis to code the transcripts from all interviews [38]. First, we created separate preliminary coding dictionaries for the patient and provider interview transcripts based on questions from the semistructured interview guides (Multimedia Appendix 3). Three doctoral-trained or bachelor-trained members of the research team (SRM, NF, and Holly Heffer) then coded 2 patient transcripts and 2 provider transcripts using these preliminary coding dictionaries. The coders met to discuss discrepancies in initial coding and refined the coding dictionary. Two of the initial coders (SRM and Holly Heffer) individually coded the remaining transcripts using the refined coding dictionary, meeting frequently to ensure consistency between the coders. After this deductive coding, the 2 coders met to discuss emergent subthemes they identified during the coding process and inductively developed subcodes that were added to the coding dictionary. One coder (Holly Heffer) then applied these subcodes to all the transcripts. The coding team reached thematic saturation based on consensus among coders that themes fully covered emergent topics across transcripts [39]. ATLAS.ti software (ATLAS.ti Scientific Software Development GmbH) was used to support the coding and analysis process.

The Ohio State University’s institutional review board approved this study (2020B0038). Participation was voluntary and all participants provided verbal informed consent. Interview transcripts were deidentified to protect participant privacy and confidentiality. Patients and nonphysician participants received a US $25 gift card in appreciation for their participation.

Interviews were conducted with 30 patients and 14 providers. Patients had a mean age of 30 years and included 22 pregnant individuals (mean gestational age: 28.6 weeks) and 8 postpartum individuals (mean weeks postpartum: 6.3). Providers included 9 physicians and 5 nurses. Patient interviews lasted an average of 16 minutes, while provider interviews lasted an average of 26 minutes.

Our qualitative analysis provides insight into patients’ and providers’ perspectives on the value and challenges of collecting and using PGHD to inform obstetric patient care. While the use of PGHD during pregnancy has most commonly been reported for objective physiologic measurements (eg, blood glucose and blood pressure) [40,41], our patient and provider participants also expressed interest in the collection of subjective measurements including anxiety, depression, and symptoms indicative of pregnancy complications. Recognizing the interest in PGHD beyond physiologic measurements is important, as studies have demonstrated that the collection of a range of different types of PGHD can improve understanding of the unique context of each patient’s pregnancy. For example, pregnant women with heart disease identified general well-being, mental health, fatigue, and quality of life as important topics for which PGHD, in addition to clinical outcomes specific to cardiac function, may be critical information to share with their providers [42]. Similarly, in a study of pregnant women with gestational diabetes, the collection of PGHD including patient perspectives of anxiety, self-knowledge about diabetes, and social support has been proposed to provide a more comprehensive picture of the patient that helps explain why some patients experience challenges with blood glucose management [43]. Increasing holistic awareness of the patient through the collection of PGHD may also help providers target interventions (eg, education) as a means to improve maternal health outcomes [44].

The growing emphasis on patient-centered care during pregnancy has brought to light the potential of PGHD as an important factor that can enhance the delivery of value-based care [23,28,45-47]. Evidence is growing surrounding the use and effectiveness of strategies for using PGHD to inform obstetric care. For example, uploading self-reported measures of blood glucose to a patient portal has been associated with improvements in glycemic control in a sample of pregnant and nonpregnant patients [13], and continuous monitoring of blood glucose has been associated with improved neonatal outcomes [46,48]. Evidence will continue to grow from several ongoing clinical trials on this topic, with the greatest focus of these studies on patient-generated physiologic measurements [49,50].

Despite interest in expanding the use of PGHD during pregnancy, little work has investigated the perspectives of stakeholders in the use of PGHD, including those of pregnant individuals providing this information and the providers receiving and using this information to improve obstetric care. While patients and providers in our study commented about the clinical benefits of using PGHD, for which there is a mixture of evidence in the literature [15,51,52], they also suggested the additional benefit of improving patients’ involvement in their care through the collection of PGHD. In nonpregnant populations, the collection of PGHD has been recognized to create an opportunity for self-reflection among patients that can encourage patient-provider communication, which, in turn, increases provider awareness of patient concerns [53]. Similarly, others have suggested additional benefits of using PGHD beyond their impact on health outcomes, such as improved patient satisfaction with care as a result of improved patient-provider communication [29,45,54].

In imagining the expanded use of PGHD to inform obstetric care, providers in our study explained several considerations in the practical implementation of strategies to collect and use this information. These considerations included questions surrounding the patient’s role in reporting PGHD, the provider’s role in using PGHD, and the potential unintended consequences of collecting PGHD. These considerations bring to light how challenging it can be to introduce PGHD into clinical practice. Furthermore, it will be important to evaluate these considerations within the unique context of the multiple ways in which PGHD can be collected, synthesized, and displayed. Examples of methods in operation or development for these purposes include the use of electronic health records and patient portals [12], automated SMS text messages [55-57], mobile health apps [29], wearable devices [58,59], and remote patient monitoring systems [60,61].

While interest in expanding the use of PGHD during pregnancy was expressed by providers in our study, there is little guidance on the collection of PGHD and a paucity of research on using PGHD in this patient population [33,34]. Furthermore, the infrastructure to measure and synthesize PGHD is not extensive in the United States [45]. For example, challenges noted in the literature include how PGHD can be collected, how to best interpret PGHD, and how meaningful thresholds of PGHD are indicative of an action required by the patient or provider [62-64]. Notably, many of the approaches to address these challenges will be specific to providers, their health care systems, and the patient populations they serve.

One mechanism to increase our collective knowledge surrounding the collection and use of PGHD in pregnancy is through the existing call to action for randomized controlled trials in maternal health to standardize the selection, collection, and reporting of outcomes based on PGHD that reflect the perspectives of study participants [65]. Such an endeavor could facilitate the translation of research into clinical practice for PGHD collection and use while increasing contextual knowledge about patient health and health care experiences that can be used to improve the patient-centeredness of obstetric care. Future research should focus on the evaluation of methods for using PGHD in obstetric care to begin to build evidence and inform implementation strategies for this practice.

One limitation of our study is that the participant population was drawn from a single AMC. Our findings, therefore, represent perspectives specific to the patient and provider populations at our clinical sites. In addition, while our study population included both patients and providers, our analytical approach was not designed to identify convergence or divergence of perspectives across groups. Future research that explicitly assesses similarities and differences in the perspectives of patients and providers on these topics is important to inform interventions for PGHD collection and reporting that meet the needs and preferences of both stakeholder groups. We conducted this study during the COVID-19 pandemic, a period during which pregnant individuals may have experienced changes in their clinical care due to the many impacts of the pandemic. Due to restrictions on in-person research at this time, study participants were recruited by phone or email, which may have influenced their willingness to participate. The remote nature of the interviews, conducted by phone or videoconference, may have also impacted the sharing of information, as compared with an in-person interview. Furthermore, changes in care experienced due to social distancing and other restrictions may have increased interest in remote communications, including those that involved the collection of PGHD. Finally, we did not systematically collect participant clinical characteristics, such as pregnancy complications, that could impact participants’ perspectives about PGHD. Including the collection of these data in future studies may be important to understand the perspectives of individuals based on their specific clinical needs.

Our qualitative analysis presents patient and provider perspectives on the collection of PGHD and its use during pregnancy. Participant responses collectively supported the use of PGHD to improve obstetric care. While several opportunities to expand the use of PGHD during pregnancy were mentioned, many considerations for implementing strategies to use PGHD were also noted, shedding light on the challenging nature of collecting and using this information in clinical practice.