Evaluating the MyPath web-based reproductive decision support tool in VA primary care: Protocol for a pragmatic cluster randomized trial

Abstract

Background:

Counseling to identify and support individuals’ desires for family formation is a key component of preventive health care that is often absent in primary care visits. This study evaluates a novel, web-based, person-centered intervention to increase the frequency and quality of communication about reproductive goals and healthcare needs in Veterans Health Administration (VA) primary care.

Methods:

We describe a hybrid type 1 effectiveness-implementation cluster randomized controlled trial in seven VA healthcare systems testing a web-based reproductive health decision support tool (MyPath). VA primary care providers are enrolled and randomized to intervention or usual care arms. Veterans scheduled to see intervention-arm providers receive a text message inviting them to use MyPath ahead of their appointment; Veterans scheduled to see control-arm providers receive usual care. Target enrollment is 36 providers and 456 Veterans. Outcomes are assessed by Veteran self-report after the visit and at 3- and 6-months follow-up. The primary outcome is occurrence of reproductive health discussions involving shared decision making; secondary outcomes include measures of communication, knowledge, decision conflict, contraceptive utilization, and receipt of services related to prepregnancy health. Data on implementation barriers, facilitators and cost are collected.

Results:

The trial is ongoing with no results to report. We have enrolled 36 primary care providers across 7 VA healthcare systems and recruitment of Veterans is ongoing.

Conclusions:

Results will inform efforts to increase the quality and person-centeredness of reproductive healthcare delivery in primary care and to operationalize and scale up use of digital decision support tools in clinical settings.

Trial Registration:

http://ClinicalTrials.gov Identifier: NCT04584294

Trial Status:

Recruiting.

1. Background

Counseling to identify and support individuals’ desires for childbearing is a key component of preventive health care. National guidelines recommend that clinicians routinely engage patients in conversations about their reproductive goals and provide services to help patients optimize health prior to desired pregnancies and prevent unwanted ones. [1–3] Despite these recommendations, conversations about reproductive goals occur infrequently in preventive care settings in the U.S., [4] and data indicate a need for reproductive counseling and care to be more person-centered and responsive to diverse preferences and needs. [5–7]

The Veterans Health Administration (VA) provides care to a diverse and growing population of Veterans who are capable of pregnancy. [8] While VA has made substantial investments in reproductive health services delivery, [9,10] gaps in the quantity and quality of these services remain. In a 2017 national survey, only 38% of Veterans capable of pregnancy reported that they discussed contraception or optimizing health prior to pregnancy with their VA primary care provider (PCP) in the past year. [11] This finding is particularly concerning given that Veterans face elevated risks of adverse pregnancy and birth outcomes due to a higher burden of chronic medical [12,13] and mental health conditions, [14] sexual trauma histories, [15] intimate partner violence, [16] and homelessness, [17]. Furthermore, racial disparities in pregnancy outcomes are well-documented, and nearly half of women Veterans of reproductive age are minority race/ethnicity. [8,18] Access to person-centered contraceptive counseling and prepregnancy healthcare is therefore critical to support Veterans in achieving healthy outcomes and the families they desire.

Previous efforts to increase the frequency and quality of reproductive counseling, such as provider-facing electronic health record (EHR) reminders, have had limited impact likely due to alert fatigue and competing priorities during primary care visits. [19–21] Interactive patient-facing decision support tools present an alternate approach through building patient knowledge, self-efficacy, and engagement in shared decision-making (SDM), a two-way communication process between providers and patients incorporating patients’ values and preferences. [22] While several decision-support tools exist for choosing a contraceptive method, [23,24] the MyPath Reproductive Decision Support tool promotes broader discussions about reproductive goals and desires by including content on prepregnancy health as well as contraception. In pilot testing, MyPath was highly acceptable to both Veterans and PCPs and significantly increased the occurrence of reproductive health discussions at primary care visits compared to usual care. [25]

To date, health care systems have struggled to realize the benefits of web-based tools in routine practice. [26] Pragmatic studies are urgently needed to bridge these promising technologies from research settings into the field. This manuscript describes a multicenter hybrid type 1 effectiveness-implementation trial focused on establishing intervention effectiveness while exploring implementation barriers and facilitators. [27] We aim to test the hypothesis that the MyPath tool, sent to Veterans via text message, will increase discussions about Veterans’ reproductive health during primary care appointments. In addition, the study will collect data to understand potential barriers and facilitators to implementing the intervention in VA primary care.

2. Methods/design

2.1. Study design

The study is a pragmatic, cluster randomized controlled trial among 36 VA PCPs and 456 of their Veteran patients. Veterans will be enrolled prior to a scheduled primary care visit and those in the intervention arm will be sent a text message with a weblink to the MyPath tool prior to their visit. Outcomes will be assessed using surveys after the primary care visit and at three and six months (Fig. 1). The research team used the Pragmatic Explanatory Continuum Indicator Summary-2 (PRECIS-2) framework, [28] which guides researchers in evaluating the extent to which their trial is pragmatic versus explanatory, to maximize the study’s applicability to real-life VA clinical practice settings (Appendix 1). The study was approved by the VA Central Institutional Review Board.

Fig. 1.

MyPath study procedures.

The figure shows the timing of study procedures for Veterans, including Veteran recruitment, intervention text message, the scheduled study visit, and followup surveys.

2.2. Setting and population

The MyPath study will take place within a range of primary care settings, including designated Women’s Health Clinics, hospital-based primary care clinics, and community-based outpatient clinics, at VA healthcare systems in seven states (Pennsylvania, North Carolina, Georgia, Texas, Colorado, Utah and Washington). We will enroll approximately 36 designated women’s health providers, or VA providers (physicians or nurse practitioners) with demonstrated proficiency in sex-specific care (e.g., pelvic exams) and with women Veterans comprising at least 10% of their patient panel. [9] The VA EHR has not historically included sex at birth and gender identity separately, and transgender Veterans’ gender in the EHR may or may not accurately reflect their current gender identity. [29] We use the phrase “women Veterans” throughout this manuscript to refer to Veterans who are identified as women in the medical record, while acknowledging the limitations inherent to these data.

Eligible providers must have a minimum of 30 outpatient visits in the past year with unique women Veterans 18–44 years old or, for new providers, have adequate projected volume based on available data to ensure we are able to meet Veteran recruitment goals. Providers who anticipate leaving VA in the next 18 months and resident physicians will be excluded from participation. We will enroll approximately 456 women Veterans 18–44 years old who are scheduled to see enrolled providers via in-person or virtual care modalities. Veterans will be eligible if they currently have a uterus and are interested in discussing or receiving information about contraception or fertility and excluded if they are currently pregnant, have diagnosed dementia or cognitive impairment, or report previously using MyPath. Additionally, Veterans who have opted out of the VA text messaging program and do not wish to opt in will be excluded because the intervention is delivered via text. Veterans will be withdrawn from the study and excluded from analyses if they do not attend the scheduled primary care visit.

2.3. Recruitment and participants

The MyPath study involves provider recruitment followed by Veteran recruitment. We will send eligible providers study information via email, then follow-up via telephone or electronic communication. We will screen and enroll interested providers verbally using telephone or audio/visual conferencing technology. Next, we will identify all potentially eligible Veterans assigned to enrolled providers’ panels and mail a packet of study information including a letter; summary of the study purpose, procedures, risks, and potential benefits; and postage-paid optout postcard. We will review enrolled providers’ panels and mail study packets to new patients every six months. Weekly, we will generate a list of eligible Veterans with an upcoming appointment and remove any who did not receive a study packet; opted-out of contact; or were enrolled or excluded previously. We will call Veterans during the two weeks prior to their scheduled appointment. Interested Veterans will be screened and enrolled via verbal informed consent processes.

Because the occurrence of reproductive health discussions during the appointment is the study’s primary outcome, we will employ a minor alteration to consent procedures in order to minimize the extent to which enrollment procedures themselves might prime Veterans in both study arms to discuss reproductive health. Prior to enrollment and consent, we will inform Veterans about study procedures, risks, and potential benefits of participation, but we will withhold details about the specific health topics addressed by the intervention until after an enrolled Veteran has attended their appointment. We will inform them this information will be revealed later in the study, and we use a small number of unnecessary screening questions to mask the study’s specific purpose. Once Veterans have completed their appointment and before assessing any outcomes, we will disclose that the intervention and study outcomes are focused on reproductive health and answer Veterans’ questions.

2.4. Randomization and blinding

After enrollment, providers will be randomized to intervention or usual care arm with a one-to-one allocation ratio. Randomization will be stratified by site and clinic type [Women’s Health Clinics (WHCs) vs. other setting]. WHCs are physical locations where women Veterans receive primary care, and are more likely to be located within facilities that have gynecology clinics, academic affiliations, and higher caseloads of women Veterans; [30] thus reproductive health counseling practices may differ among providers in WHCs versus other clinics. A pre-programmed randomization list will be used, allocating enrolled providers to the next treatment assignment in sequence. Allocation concealment will prevent study staff from obtaining information on the sequence of treatment assignment. Blocked randomization with blocks of two will ensure balanced treatment groups. Enrolled Veterans will automatically be assigned to the same study arm as their provider. Staff involved in Veteran recruitment and data collection will remain blinded to providers’ study arm assignments and to which provider each Veteran is scheduled to see during enrollment and baseline surveys, although they may become unblinded to Veteran study arm during follow-up surveys when intervention adherence is assessed. A staff member not involved in Veteran recruitment will manage study arm information and initiate text messages to Veterans in the intervention arm.

2.5. MyPath intervention and usual care descriptions

2.5.1. Conceptual framework

The theoretical basis, development, content, and results of pilot testing the MyPath tool are described in detail elsewhere. [25] In brief, MyPath was designed to support reproductive autonomy by allowing users to first consider their thoughts and orientation toward pregnancy outside of the power dynamics of a clinical encounter. Users can navigate to the sections of the tool that are relevant to them and engage as deeply as they desire with topics including fertility, optimizing health before pregnancy, and/or birth control (Fig. 2). [31] Throughout their use of the tool, users’ responses, health topics of interest, individualized birth control recommendations (when appropriate), and questions are populated in the MyPath Summary which can be printed or emailed to the user and used as a communication guide during clinical visits. In pilot testing, Veterans spent an average of 11 min using the tool. [25]

Fig. 2.

MyPath Welcome Page and Main Menu.

The figure shows the welcome page and main menu of the MyPath decision support tool.

2.5.2. MyPath intervention description

The principal investigator will deliver a 30-min overview of the MyPath intervention to each intervention arm provider, which will include an orientation to the tool itself and key points about how the tool can support shared decision making for patients and providers. VA provides required shared decision-making training to all primary care providers [32]. In order to maintain the scalability of the intervention, existing training of VA providers will be leveraged without extensive new required trainings or prompts. Intervention providers will also receive online and hard-copy resources supporting person-centered reproductive health counseling. Utilization of these resources will be optional and not measured formally. Providers will not be notified which of their patients are enrolled in the study. Intervention arm Veterans will be sent a one-way text message containing a weblink to the MyPath website one to two business days prior to a scheduled primary care visit. At baseline, Veterans will be encouraged to visit the weblink if they receive the text message, but will not be provided other guidance about how to use MyPath.

2.5.3. Usual care

Usual care providers will be instructed to continue to provide usual care and will not receive an orientation. Veterans in the usual care arm will not be sent the study text message.

2.6. Outcome measures

2.6.1. Primary and secondary outcomes assessed in the full sample

The primary outcome is a Veteran-reported measure of whether a reproductive health discussion involving SDM occurred during their primary care visit. Veterans will be asked whether they discussed reproductive health topics at their appointment. If they did, SDM is measured using the 3-item CollaboRATE scale (scales and instruments available in Appendix 2), [33] with SDM defined using a top box approach. [34] Additional outcomes measured in the full sample include whether a reproductive health discussion occurred at their visit (with or without SDM), communication self-efficacy (Perceived Efficacy in Patient-Provider Interactions or PEPPI scale), [35] and reproductive health and contraceptive knowledge. The CollaboRATE and PEPPI scales have each been validated within varied patient populations and have been translated and validated in multiple languages [36–40]. The family planning knowledge scale was developed by our research group, both drawing from questions used in previous studies of contraceptive interventions [23,41] and developing new items to capture prepregnancy health considerations. In pilot testing of the intervention, the knowledge scale demonstrated reliability with a Cronbach alpha of 0.7, although it has not been formally validated. Provider-patient communication quality will be further assessed among Veterans who had a reproductive health discussion, with measures based on topics discussed in the visit; the Patient-Centered Contraceptive Counseling scale [42,43] will be used to assess contraceptive counseling quality; and study team-developed items will assess quality of reproductive goals and prepregnancy health counseling.

2.6.2. Secondary and exploratory outcomes assessed in subgroups

Certain contraceptive or prepregnancy health outcomes will be relevant only for subsets of the full sample. Two subgroups of participants will therefore be defined at baseline: participants at risk of unwanted pregnancy (defined as those who report being sexually active with a person assigned male sex at birth who has not had a vasectomy, and who report they are not seeking or open to pregnancy in the next year) and participants for whom prepregnancy counseling may be relevant (defined as those who report that they desire, would be okay with, or are unsure if they desire pregnancy in the next two years). Participants may belong to one subgroup only, both subgroups (e.g., people who have been sexually active with a male partner and desire pregnancy one to two years from now), or neither subgroup (e.g., people who have not been sexually active with a male partner in the past year and do not desire pregnancy in the next two years).

Post-visit secondary outcomes assessed only in the subgroup at risk of unwanted pregnancy include decision conflict about their contraceptive choice (Decision Conflict Scale), [44] whether they feel their current contraceptive method is right for them (measured using a single Likert scale item), [45] and contraceptive method chosen. Additional outcomes will be assessed at three- and six-month follow-up and include current contraceptive method at six months, continuous contraceptive use over six months, consistent contraceptive use at six months, contraceptive satisfaction at six months, and incidence of unplanned pregnancy over six months (London Measure of Unplanned Pregnancy) [46].

Exploratory outcomes will be collected among the subgroup considering or open to pregnancy. We will assess self-reported prepregnancy health risks at baseline, then measure outcomes at follow-up relating to each patient’s specific risks. Risk factors include chronic medical or mental health conditions; body mass index >30; lack of folic acid supplementation; tobacco, marijuana, alcohol or drug use; housing instability; and food insecurity. Outcomes are measured over six months and include occurrence of healthcare discussions, receipt of services or treatments, and changes in behavior relating to personal prepregnancy health risks.

2.6.3. Exploratory goals-concordant outcomes

In keeping with convention in contraceptive use trials, contraceptive and prepregnancy health outcomes at 6-month follow-up will be assessed according to baseline pregnancy intentions. For example, a participant who does not desire pregnancy at enrollment will be asked about contraceptive use regardless of whether their orientation toward pregnancy changes during the follow-up period. This convention helps preserve the benefits of randomization; use of MyPath could conceivably influence a change in pregnancy intention in the intervention group differentially. However, attitudes toward pregnancy commonly fluctuate [47] and may do so in both study arms; thus novel measures that accommodate changes in contraceptive needs over time are needed. In addition to our conventional contraceptive outcomes, we will assess three exploratory outcomes measuring reproductive behaviors concordant with participants’ stated pregnancy goals as captured over time (Table 1). Veteran behavior will be considered “goals-concordant” if their contraceptive utilization and/or folic acid supplementation aligns with their stated desires for pregnancy at the current time.

Table 1.

Exploratory goals-concordant reproductive health measure definitions.

Patient-reported pregnancy intentions at 6-month follow-up	Goals-concordant continuous contraceptive use	Goals-concordant consistent contraceptive use	Goals-concordant reproductive behaviors
Desires pregnancy now	Not using contraception continuously	Not using contraception consistently	Taking folic acid
Not trying, but would be okay with pregnancy	Either continuous or non-continuous contraceptive use	Either consistent or non-consistent contraceptive use	Consistent contraceptive use and/or taking folic acid
Desires pregnancy later but not now or never desires pregnancy	Continuous contraceptive use	Consistent contraceptive use	Consistent contraceptive use
Unsure	Continuous contraceptive use	Consistent contraceptive use	Consistent contraceptive use and/or taking folic acid

2.7. Sample size justification

Sample size and power calculations are based on our primary outcome and rely on formulas specified for use in cluster randomized trials accounting for intervention non-adherence and attrition. Calculations assume a baseline prevalence of a reproductive health discussion with SDM of 50% in the usual care arm [25,48], 30% non-adherence in the intervention arm [49], and 5% attrition between enrollment and the post-visit survey. Assuming a kappa coefficient of 0.01, a sample of 456 patients will yield 80% power in 2-sided tests with a type-1 error rate of 5% to detect a 20-percentage point increase in the occurrence of a reproductive health discussion with SDM to 70%, which is clinically meaningful and consistent with the magnitude of change we observed in pilot testing. [25]

2.8. Data collection

Surveys will be performed via telephone. Providers will be surveyed about demographics and practice characteristics at enrollment. Veterans will be surveyed at enrollment about demographics, health factors, and pregnancy goals. Three follow-up surveys with Veterans will assess outcomes post-visit (within three weeks after their primary care visit) and at three and six months.

2.9. Analyses

Primary analyses of treatment effects are intent-to-treat, with participants analyzed in the randomization groups to which they were originally assigned irrespective of whether they used the intervention. The primary outcome will be analyzed using a mixed effects logistic regression model testing whether the odds of reproductive health discussion with SDM are significantly different in the intervention versus control arm. Dichotomous secondary outcomes will use the same model and test whether the odds of a positive response (e.g., satisfaction with contraceptive method) are significantly different in the intervention versus control arm. Continuous secondary outcomes will use a mixed effects model testing whether average outcomes differ between intervention and control groups.

2.9.1. Analysis of interaction effects

We will assess modification of the effect of assignment to the intervention on the primary outcome by age, race/ethnicity, visit modality (virtual/telephone versus in-person) and clinic type (WHC yes/no).

2.9.2. Missingness

For key variables that have 15% or more missing values, we will analyze factors that are associated with missingness and perform a sensitivity analysis using multiple imputation to allow an unbiased analysis for all subjects. Because informative missingness would violate the covariate-dependent missing at random (CDMAR) assumption of the standard multiple imputation procedure, we will conduct sensitivity analyses in which we multiply impute missing data under plausible informative missing data mechanisms – in particular, lower contraceptive utilization rates among participants with missing values of this outcome. In all analyses, we will check for departures from model assumptions.

2.9.3. Implementation barriers and facilitators

Quantitative and qualitative data will be collected throughout the study to explore potential barriers and facilitators to implementing the intervention in VA clinical settings. These efforts are guided by the RE-AIM (Reach, Effectiveness, Adoption, Implementation and Maintenance) framework, a widely used implementation evaluation framework incorporating both individual- and organizational-level elements. [50,51] Findings will be summarized into recommendations to inform adaptations and implementation strategies.

Quantitative and qualitative assessments of each RE-AIM construct are summarized in Table 2. Quantitative outcomes will be assessed using data collected from the trial. Intervention cost assessment will be conducted using methods described by Liu et al. [52]. Semi-structured qualitative interviews will be conducted among a sample of approximately 24 Veterans in the intervention group, purposively sampling to achieve thematic saturation among Veterans who did and did not use the tool. [53] We will also sample to achieve variability in race/ethnicity and sexual orientation. Interviews will occur after Veterans complete their final survey. Additionally, we will invite intervention arm providers and women’s health and clinic leaders at intervention sites to participate in interviews after all Veterans have been enrolled. Interviews will be conducted by telephone or virtually. Interview guides address the five RE-AIM constructs, [51] using open-ended questions to elicit perceived barriers and facilitators to MyPath implementation. Interviews will be audio recorded and analyzed using rapid qualitative analysis, a pragmatic approach to analyzing structured qualitative data in implementation research. [54]

Table 2.

Quantitative and qualitative assessment of MyPath implementation.

RE-AIM Element	Quantitative Assessment	Qualitative Assessment
Reach: Proportion of individuals who participate in the intervention	Proportion of Veterans who	Veterans’ reasons for use or non-use of the tool and strategies to improve and increase use
	Were interested in information about birth control, fertility, or preparing for pregnancy among those screened Were excluded because they have opted-out of receiving VA text messages Used the MyPath tool among those in the intervention arm
	Comparison of demographic characteristics of intervention arm participants who did or did not use the MyPath tool
	Average time spent using the MyPath tool; sections of the MyPath tool most frequently visited
Effectiveness/ Efficacy: Impact of the intervention on outcomes	Assessed in primary and secondary study analyses	Veterans’ perspectives regarding
		whether and how the intervention influenced their discussion with their health care provider or their health care decisions what they learned (if anything) from using the MyPath tool the most and least helpful aspects of intervention
		Providers’ perspectives regarding
		whether and how the MyPath tool affected their counseling the most and least helpful aspects of intervention
Adoption: Proportion of settings and intervention agents who are willing to initiate the intervention		Providers’ and clinic leaders’ attitudes toward adoption and factors that could influence uptake of the implementation in routine practice
Implementation: Extent to which intervention is implemented as intended at the setting level, and costs of the intervention	Implementation costs:	Providers’ and clinic leaders’ perspectives on barriers and facilitators to intervention delivery and fidelity
	Cost of maintaining and hosting the MyPath website Time spent engaging VA leadership and clinical stakeholders Time spent training and supporting enrolled providers Time spent administering text messages
Maintenance: Extent to which a program becomes part of routine practices		Providers’ and clinic leaders’ perspectives on factors that may influence whether the intervention can be sustained as-is or modified, or should be discontinued, and why

3. Discussion

The MyPath trial tests a novel and scalable web-based intervention developed to address gaps in delivery and quality of reproductive health services in VA primary care settings nationally. While MyPath demonstrated positive results in pilot testing, additional evaluation is needed to establish effectiveness and to inform wider use if it is found to be effective. The MyPath trial is innovative in several ways, including the approach of the intervention itself, aspects of trial design, and considerations for implementation.

First, while web-based tools that support discrete contraceptive decisions have been tested in family planning clinics, [23,55] few studies evalute web-based tools designed to address a broader range of reproductive needs in primary care settings. Interventions such as MyPath that aim to improve the quality and person-centeredness of clinical encounters are needed given evidence of low rates of primary care-based reproductive counseling as well as negative experiences among marginalized populations. [56–59] MyPath is designed to support reproductive autonomy and connect users with information and services that can meet the full spectrum of their reproductive needs, including healthy pregnancies when desired. [60] This approach is consistent with the increasing focus on providing reproductive health care in a manner aligned with reproductive justice, a rights-based framework defined by Black women activists that points to the need to honor and support reproductive autonomy across the range of reproductive experiences, particularly among individuals who have faced discrimination or devaluation of their fertility in clinical care and society more broadly. [61,62] Our assessment of modification of the intervention’s effect by race/ethnicity will provide preliminary data to assess the intervention’s effect on equity as well as quality.

Because the intervention’s scope extends beyond contraception, the MyPath trial includes populations who have traditionally been excluded from contraceptive or reproductive health studies, including people who do not have sperm-producing partners. These individuals have reproductive health desires and needs but are often excluded from family planning-related studies, which typically focus on contraceptive care. [63] By defining two analytic subgroups, one at risk of unwanted pregnancy and one open to pregnancy, the MyPath trial is able to evaluate clinical outcomes relevant to these subgroups but also evaluate patient-centered outcomes that apply to a broader group of people who endorse a full range of pregnancy desires. Including these participants for assessment of the primary outcome strengthens the study’s representation of the diverse reproductive health needs reflected in primary care settings.

Our study is also innovative in its use of patient-centered outcomes. This choice reflects increased awareness that, given the realities of fluctuating reproductive desires and the complexity of reproductive decision making, conventional clinical measures such as contraceptive continuation or unintended pregnancy often fail to accurately assess what is meaningful to individuals. [64,65] Patient-centered outcomes are increasingly recognized by health care systems and funders as essential for evaluating interventions based on what matters most to patients. [66] Our outcomes, selected based on qualitative work with Veterans [56,67] and review of the literature, assess the extent to which Veterans discuss their needs in a person-centered conversation as well as dimensions of decision quality and communication. We also include novel exploratory outcomes that account for fluctuating pregnancy intentions. Collecting these data will enable us to draw comparisons with traditional measures and advance discussion in the field regarding use of patient-centered measures to evaluate reproductive health interventions.

Lastly, the study is innovative in its pragmatic strategy for integrating patient-facing web-based tools into clinical practice. The challenge of delivering web-based tools efficiently within clinical workflows often prevents their adoption in health care systems. Furthermore, the COVID-19 pandemic has increased demand for interventions that are compatible with virtual care. The MyPath trial adds to the existing body of literature about implementing decision-support tools by using a low-touch dissemination strategy – one-way text messages with the potential for automation – that allows individuals to use the tool at a time and place convenient to them. Intervention delivery requires little to no clinic staff time and does not rely on extensive provider training. We also collect data on potential barriers and facilitators to implementation and cost to facilitate real-world uptake within the VA and future adaptation for other health care systems.

The MyPath trial also has several limitations. Provider eligibility is limited to those with sufficient volume of women Veterans to enable us to achieve recruitment goals; our findings may therefore not be generalizable to providers who care for very few women Veterans. Additionally, Veteran eligibility was limited to those who receive VA text messages. While only a small proportion (4%) of Veterans opt-out of VA text messages, [68] this nevertheless limits the generalizability of our results to patients who receive messages. No published data describe the characteristics of VA users who do not participate in the VA text messaging program and how they might differ from those who do; however, during the course of this trial, we will be able to measure and describe differences between our study population and those Veterans who are excluded due to non-participation. In addition, Veteran eligibility was limited to those identified as women in the medical record. As discussed earlier, due to current limitations of VA medical record data, we are not able to ensure the consistent inclusion of transgender or gender-diverse Veterans who can become pregnant and their reproductive health experiences may not be fully represented within this study. In addition, several limitations of our outcome measures deserve mention. First, the PEPPI scale was designed and validated to inquire about interactions with a participant’s “doctor,” although some Veterans receive counseling or care from a nurse practitioner or physician assistant; additional research is needed to validate broader versions of this measure. Second, the availability of some contraceptive methods that require specialized training such as IUDs and implants may vary by VA facility, and our contraceptive use outcomes do not account for this variability. A national VA assessment of contraceptive access is currently ongoing; results from this assessment could eventually help to contextualize any potential site-level effects on contraception method mix and use among participants in this trial.

4. Conclusion

Providing person-centered high-quality care for the rapidly growing population of Veterans capable of pregnancy is a high priority for VA. The recently passed Protecting Moms Who Served Act (H.R.958/S.796), which will commission a comprehensive study on the scope of America’s maternal health crisis among women Veterans, highlights the growing national spotlight on these issues among Veterans. [69] Discussions about reproductive goals that ensure individuals receive the services they need to avoid undesired pregnancy and achieve healthy ones are often absent in primary care, representing important missed opportunities to improve both Veterans’ experiences of care and their reproductive health outcomes. Pragmatic clinical trials of novel and scalable interventions, such as MyPath, can facilitate the timely translation of research into practice and identify strategies to create sustained impact both within and beyond VA.

Appendices

Appendix 1. MyPath PRECIS-2 domain figure and ranking explanations

PRECIS-2 rankings are on a scale from one to five, with higher numbers indicating more pragmatic decisions and lower number indicating more explanatory decisions. The figure below is followed by a table with explanations of our own self-rankings.

PRECIS-2 domain	MyPath ranking	Explanation for ranking
Eligibility	4	This trial does not have strict eligibility criteria creating an “ideal” set of participants; however, some eligibility criteria do make our study populations less pragmatic, including:
		Provider eligibility was limited to those who saw ≥ 30 Veterans of reproductive age identified as women in the medical record in the past year for practical purposes (without this criterion we could not ensure sufficient eligible patients for recruitment). Provider eligibility was limited to non-trainees for practical purposes (without this criterion we could not avoid contamination of intervention vs. usual care provider communication practices between trainees and attending physicians). Patient eligibility is limited to those who are not currently pregnant, have reproductive capacity (exclusion of those who do not have a uterus), have decision making capacity (excluding those with diagnosed dementia and cognitive degeneration), and who endorse having interest in receiving information about reproductive health topics. These eligibility criteria make the study more person-centered and limit the delivery of the intervention to patients for whom reproductive health decision support is applicable (analogous to studies which limit inclusion to those with a health condition the intervention is intended to treat). These criteria do not limit the pragmatism of the study; however, limiting inclusion to patients who participate in the VA text message program does limit the pragmatism of the trial. We recognize that these criteria limit the study sample to those who could be more receptive to and likely use a web-based intervention than the broad population of all Veterans who need reproductive health decision support.
Recruitment	5	The trial recruitment strategy includes all types of VA primary care settings, and recruits patient participants who have regularly scheduled primary care appointments. Patient participants were recruited regardless of appointment modality, including in-person, videoconferencing, or telephone visits.
Setting	5	This trial tests the intervention in the same healthcare system (VA) it was designed for and may be implemented in, if found to be effective. Provider participants were recruited from all types of VA primary care settings (hospital-based primary care clinics, designated women’s health clinics, and community-based outpatient clinics) and in VA health care systems distributed geographically throughout the U.S.
Organization	5	The intervention is integrated fully into the flow of usual care without any additional clinical staff time or clinic resources. Intervention delivery does not depend on clinical staff or providers and is accomplished using existing one-way text messaging infrastructure. This is a model which would be highly feasible and no- or low-touch to implement in a clinical setting.
Flexibility (delivery)	5	The intervention is delivered remotely/electronically via existing text-message infrastructure, so there is no effort to promote uniform delivery or to measure whether the delivery remains true to design. It is possible some patient participants may not receive the text message or may not notice the text message, which matches what would happen in usual care.
Flexibility (adherence)	4	Patient enrollment happens prior to intervention delivery and patients are encouraged to use the intervention if sent a MyPath weblink via text message. This would not happen in usual care. However, the study allows for patients to elect how to interact with the intervention (including the possibility to not interact with it at all, and to engage with it as deeply as they prefer) because the intervention is delivered remotely via text message to patients’ phones prior to visits. We limit our sample to patients who report being interested in information about reproductive health topics and who receive VA text messages, but there is no mechanism ensuring patients use the intervention or enforcing how or when they do so.
Follow-up	3	Study follow-up is more frequent/rigorous than what would be expected in usual care for purposes of outcome assessment; however, follow-up is done remotely for patient-reported outcomes and on a standard schedule for all participants (with no additional follow-up triggered by outcome-related events).
Primary Outcome	4	The study’s primary outcome (occurrence of reproductive health discussions involving shared decision making during a primary care visit) is a patient-centered and patient-reported outcome of great importance both to patients and to healthcare systems. This outcome is not routinely collected in healthcare settings and use of administrative/medical record data is therefore not possible.
Primary Analysis	5	The analysis is intent-to-treat for primary and secondary outcomes.

Appendix 2. Scales utilized to measure outcomes

Scale	Questions	Response Options
CollaboRATE Scale (Shared Decision Making)	Thinking about the conversation you had with your provider about pregnancy and/or birth control:	No effort was made A little effort was made Some effort was made A lot of effort was made Every effort was made Decline to answer
	How much effort was made to help you understand your health issues or choices?
	How much effort was made to listen to the things that matter most to you?
	How much effort was made to include what matters most to you in choosing what to do next?
Perceived Efficacy in Patient-Physician Interactions (PEPPI) Scale	On a scale of 1 to 5, where 1 is “not at all confident” and 5 is “very confident,” how confident are you in your ability to:	1 = Not at all confident 2 3 4 5 = Very confident Decline to answer
	Know what questions to ask a doctor
	Get a doctor to answer all of your questions
	Make the most of your visit with the doctor
	Get a doctor to take your main reproductive health concerns seriously
	Get a doctor to do something about your main reproductive health concern
Family Planning Knowledge Scale1	When during a person’s menstrual cycle are they most likely to become pregnant?	During her period 3 days after her period ends Two weeks before her next period starts 3 days before she gets her period I don’t know Decline to answer True False I don’t know Decline to answer As soon as she gets a positive pregnancy test After she has her first prenatal visit 3 months before she gets pregnant One month before her due date I don’t know Decline to answer
	Is this statement true or false?
	Someone who skips periods or has irregular periods can still become pregnant.
	When is the best time for a someone to start taking folic acid to prevent birth defects?
	Which of the following conditions can result in complications during pregnancy? Please choose all that apply.
	Medical problems like diabetes or high blood pressure Taking prenatal vitamins Being overweight Being underweight Mental health conditions like depression Moderate exercise Physical or mental abuse by a partner	Yes (It can result in complications) No (It cannot result in complications) I don’t know Decline to answer
	If someone might become pregnant, what should they do if they are taking prescription or over the counter medications?	Stop all medications Make an appointment to talk to her provider as soon as she has a positive pregnancy test Make an appointment to talk with her provider before she is pregnant I don’t know Decline to answer 3 months 6 months 18 months 5 years I don’t know Decline to answer
	What is the safest length of time for someone to wait to get pregnant again after giving birth?
	For each of the following pairs of birth control methods, please choose which one you think is better at preventing pregnancy, or if you think they are equally good.
	Birth control pills or condoms	Birth control pills are better Condoms are better Both are equally good Don’t know Decline to answer Birth control pills are better An IUD is better Both are equally good Don’t know Decline to answer Condoms are better The shot is better Both are equally good Don’t know Decline to answer True False I don’t know Decline to answer True False I don’t know Decline to answer Yes, it is safe No, it is not safe I don’t know Decline to answer Birth control pill Intrauterine device, or IUD Contraceptive injection, also known as Depo-Provera All of the above None of the above Don’t know Decline to answer 1 day 3 days 5 days 20 days Don’t know Decline to answer
	Birth control pills or an intrauterine device (IUD)
	Condoms or the birth control shot, Depo Provera
	Is this statement true or false?
	Young people can use IUDs, even if they have never given birth.
	Is this statement true or false?
	The implant or an IUD cannot be removed early, even if someone changes their mind and wants to get pregnant.
	Is it safe to use a birth control method that stops your period?
	It is often hard to become pregnant soon after stopping which of the following birth control methods?
	Up to how many days after unprotected sex is the “morning after pill” (or Plan B) effective?
Decision Conflict Scale (DCS)	Please indicate how much you agree or disagree with the following statements about your birth control decisions.	Strongly disagree Disagree Neutral Agree Strongly Agree Decline to answer
	I know which birth control options are available to me.
	I know the benefits of each option.
	I know the risks and side effects of each option.
	I am clear about which benefits matter most to me.
	I am clear about which risks and side effects matter most to me.
	I am clear about which is more important to me (the benefits or the risks and side effects).
	I have enough support from others to make a choice.
	I am choosing without pressure from others.
	I have enough advice to make a choice.
	I am clear about the best choice for me.
	I feel sure about what birth control to choose.
	The decision is easy for me to make.
	I feel I have made an informed choice.
	My decision shows what is important to me.
	I expect to stick with my decision.
	I am satisfied with my decision.
London Measure of Unplanned Pregnancy (LMUP)	Below are some questions that ask about your circumstances and feelings around the time that you became pregnant. Please think of your current (or most recent) pregnancy when answering the questions below.
	Please choose the statement that most applies to you.	I/we were not using contraception I/we were using contraception, but not on every occasion I/we always used contraception, but knew that the method failed (i.e., broke, moved, came off, came out, not worked, etc.) at least once I/we always used contraception Decline to answer Right time OK, but not quite right time Wrong time Decline to answer I intended to get pregnant My intentions kept changing I did not intend to get pregnant Decline to answer I wanted to have a baby I had mixed feelings about having a baby I did not want to have a baby Decline to answer My partner and I had agreed that we would like me to be pregnant My partner and I had discussed having children together, but hadn’t agreed for me to get pregnant We never discussed having children together Decline to answer Took folic acid Stopped or cut down smoking Stopped or cut down drinking alcohol Ate more healthily Sought medical/health advice Took some other action [fill in blank] I did not do any of the above for my pregnancy Decline to answer
	In the month that I became pregnant…
	Please choose the statement that most applies to you.
	In terms of becoming a mother (first time or again), I feel that my pregnancy happened at the…
	Please choose the statement that most applies to you.
	Just before I became pregnant…
	Please choose the statement that most applies to you.
	Just before I became pregnant…
	In the next question we ask about your partner. That might be (or have been) your husband, a partner, a partner you live with, a boyfriend, or someone you’ve had sex with once or twice.
	Please choose the statement that most applies to you.
	Before I became pregnant…
	Before you became pregnant, did you do anything to improve your health in preparation for pregnancy? (Please choose all that apply)

¹The Family Planning Knowledge Scale was originally developed using gendered language (e.g., “when during a woman’s cycle is she most likely to become pregnant?”). As part of our efforts to improve the inclusivity of our study and its measures, we adapted and altered this measure to be gender neutral.

Data availability

No data was used for the research described in the article.