Using human‐centered design to map the perinatal ecosystem for pregnant people with substance use disorders

Abstract

Objective

Given the profound impacts of substance use in pregnancy on the health and well‐being of parents and newborns, improving the organization and delivery of perinatal services for this population is a priority. The objective of this study was to characterize the perinatal ecosystem for pregnant people with substance use disorders (SUDs) and identify operational gaps and opportunities to strengthen care delivery.

Methods

We used process mapping, a human‐centered design (HCD) methodology, to characterize the steps involved with care delivery across the perinatal care continuum. Clinical and community‐based staff within a safety‐net healthcare system participated in group and individual process mapping sessions in which they responded to 39 structured questions spanning prenatal through postpartum care. Following an HCD approach to qualitative analysis, we identified themes and generated opportunities.

Results

Forty‐seven stakeholders participated in process mapping. Analysis revealed three major themes: (1) divergent understandings of policies and procedures among providers, (2) challenges with care coordination, and (3) lack of standardized procedures for engaging Child Protective Services. These themes led to the identification of 13 actionable opportunities to strengthen perinatal care delivery.

Conclusions

Process mapping is a valuable methodology for improving systems of care for pregnant people with SUDs. In our ecosystem, process mapping revealed a lack of clear policies and challenges with care coordination. Strengthening healthcare systems is essential to improving consistency, quality, and outcomes for families affected by SUDs in pregnancy.

1. INTRODUCTION

Substance use in pregnancy is common, with more than 1 in 20 individuals across the United States reporting non‐medical drug use during pregnancy [1]. Over the past two decades, the prevalence of opioid use disorder among pregnant patients presenting to labor and delivery units has quadrupled [2, 3]. More recently, methamphetamine use during pregnancy has increased, surpassing the rates of opioid use in some areas of the western United States [4]. Rising rates of substance use disorders (SUDs) during pregnancy have contributed significantly to the US maternal mortality crisis, with drug‐related deaths accounting for more than 1 in 10 maternal deaths nationally [5]. Given the impacts of perinatal substance use on the health and well‐being of both birthing people and neonates, optimizing perinatal care delivery systems for families affected by SUDs is a public health priority.

Providing perinatal care for families affected by SUD is operationally complex, reflecting the intersection of healthcare delivery, legal systems, and social service provision. Stigma and punitive legal frameworks surrounding substance use in pregnancy shape clinical workflows and provider decision‐making, contributing to inconsistent implementation of evidence‐based care [6, 7, 8, 9]. Anticipation of judgment or legal consequences may limit disclosure of substance use, impacting clinicians’ ability to deliver appropriate care and coordinate indicated referrals [6, 7, 8, 9]. Care delivery is further complicated by mandated reporting requirements and engagement with Child Protective Services (CPS), as parental fears around custody loss influence care‐seeking behaviors and interactions with providers [10, 11, 12].

In addition, perinatal SUDs frequently co‐occur with other complex social needs, such as housing instability, mental illness, and interpersonal violence, increasing the number of referrals and cross‐sector coordination required to deliver appropriate care [13, 14, 15]. Clinical and social support services are often fragmented across multiple clinical sites and community‐based programs, which complicates care provision [16, 17]. As a result, care delivery for pregnant patients with SUDs often relies on coordination between multiple providers, departments, and agencies, increasing the risk of fragmented workflows and loss of continuity [17, 18, 19]. Collectively, these operational challenges contribute to missed opportunities for timely, coordinated perinatal care and are associated with lower utilization of prenatal services and negative patient care experiences among pregnant individuals impacted by SUDs [20].

Given the complexity of care delivery systems that support families affected by SUD during pregnancy, innovative methodologies are needed to understand how care is operationalized across clinical and community settings. Improving outcomes for this population requires insight into how policies, workflows, and inter‐organizational coordination shape care delivery. Designing novel solutions to address care fragmentation in the perinatal ecosystem therefore requires rigorous examination of existing workflows and engagement with the broad network of stakeholders involved in patient care. Human‐centered design (HCD) is a well‐established, solution‐oriented methodology for understanding complex systems, such as care delivery systems, by engaging frontline stakeholders and identifying barriers to care and opportunities for system improvement [21]. Drawing on ethnographic research methods, HCD prioritizes the experiences of users (e.g. clinicians, staff, and/or patients) to elucidate unmet needs and the contextual factors that shape care delivery within complex systems. By centering analysis on human behavior and decision making, HCD enables a nuanced understanding of how work is performed rather than how it is prescribed. This human focus is especially valuable in complex care environments where workflows are dynamic and spread across multiple roles, teams, and settings. In the context of complex care delivery systems, HCD offers a distinct advantage by making workflows explicit, visualizing variation across roles and settings, and revealing misalignment between policies and practice.

In this study, we applied HCD methodologies to map the perinatal care ecosystem for pregnant individuals with SUDs within a safety‐net healthcare system. Leveraging process mapping—a core HCD method—we engaged clinical and community stakeholders to document existing workflows, identify barriers to care, and generate actionable opportunities for system‐level improvement.

2. METHODS

2.1. Study design and setting

We conducted a prospective observational study from January 2019 to March 2020 using HCD methodologies to explore the perinatal care journeys of pregnant patients with SUDs in San Francisco. This work was carried out at San Francisco's only safety‐net hospital through a collaboration between The Better Lab, an HCD research laboratory; Solid Start, a multi‐agency partnership focused on improving the health and well‐being of pregnant people and families with young children in the safety‐net system; and the Plan of Safe Care Community Collaborative, a previously established coalition of clinical and community stakeholders working to identify and address gaps in care for families affected by SUDs in San Francisco County [16]. The Community Collaborative included more than 50 members representing a broad range of professional roles across the perinatal care continuum, including obstetricians, pediatricians, child protective services staff, nurses, social workers, community health workers, home visiting providers, public health representatives, and community‐based organizations (CBOs), reflecting the Collaborative's emphasis on cross‐sector care coordination and transitions beyond the clinical obstetric setting. Of note, in this safety‐net setting, nurse midwives provide a substantial proportion of attending‐level obstetric care, and their inclusion both in the Collaborative and as research participants ensured representation of frontline obstetric decision‐making within routine perinatal workflows.

Importantly, multiple obstetricians in this safety‐net system are trained in addiction medicine and routinely initiate and manage medications for opioid use disorder (MOUD), including buprenorphine, in both inpatient and outpatient prenatal care settings. Methadone treatment is co‐managed in collaboration with opioid treatment centers or inpatient addiction consult services.

The University of California San Francisco's Institutional Review Board reviewed and exempted this study. Verbal consent was obtained from participants, and all methods were conducted in accordance with relevant ethical guidelines and regulations.

2.2. HCD methodology

HCD uses qualitative, ethnographic research methods to understand users’ needs and identify opportunities to improve systems and services. Given the complex and intersecting challenges of providing care to families affected by SUDs, we applied HCD methodologies to examine current systems of perinatal care for this population. In this study, the primary “users” examined through our HCD methods were clinical and community‐based stakeholders responsible for delivering and coordinating perinatal care. Accordingly, process mapping was designed to capture internal workflows and coordination processes, rather than patient experiences directly.

Our HCD process consisted of three phases: inspiration, ideation, and implementation (Figure 1). This study focuses on the inspiration phase. The goal of the inspiration phase is to develop a comprehensive and contextually grounded understanding of how care is delivered and experienced by stakeholders. By making workflows explicit and visualizing variation across roles, teams, and settings, this phase enables identification of system‐level misalignments between policies, procedures, and daily care delivery. During the inspiration phase, qualitative data are collected and then systematically synthesized into “insight statements,” concise summaries that capture recurring challenges, unmet needs, or tensions identified across participants’ responses. By centering analysis on human behavior, decision‐making, and lived experience, the identification of thematic patterns and development of insight statements support the articulation of opportunity statements—clearly defined areas in which changes to workflows, policies, or processes could improve care delivery [21]. In the subsequent ideation phase, these opportunity statements are used to generate and refine potential solutions to identified workflow challenges. The implementation phase then focuses on prototyping and testing these solutions to assess feasibility and potential for sustainable integration into practice.

FIGURE 1.

The three phases of the human‐centered design process and associated activities.

To understand how perinatal services for pregnant people with SUDs are delivered, we conducted process mapping exercises with key stakeholders working within an existing safety‐net system. Process mapping is a commonly used HCD methodology that documents and visualizes the sequence of activities, decision points, roles and transitions involved in delivering a service, with the goal of identifying operational gaps and opportunities for workflow improvement [22, 23]. The purpose of these exercises was to generate insight into how services are delivered and experienced in practice and identify opportunities to improve value for both patients and providers. Process mapping has been applied across a variety of healthcare settings to examine how teams deliver care and to identify potential workflow improvements that enhance team performance and patient outcomes [24, 25, 26, 27].

Our process mapping approach was based on a worksheet adapted from Children and Family Futures (CFF), which outlines a series of key process steps relevant to the ecosystem of care for patients affected by SUDs [28]. Framed as probing questions, our adapted exercise included 39 steps spanning the perinatal journey, from the initiation of prenatal care to postpartum management (Appendix S1). These 39 prompts asked participants to reflect on how care is typically carried out at each step of the existing care pathways, including where workflows varied by role or setting, where challenges or uncertainties arose, and where participants perceived opportunities for improvement.

Given that the objective of this study was to map internal care delivery workflows and coordination across sectors, process mapping exercises were conducted with staff directly involved in providing and coordinating care for perinatal patients with SUDs. In‐person process mapping workshops were conducted with groups of 7–10 stakeholders, including resident and attending physicians, medical students, social workers, midwives, public health nurses, and representatives from five CBOs. During these workshops, we arranged large sheets of paper around the room to represent three time periods in the perinatal journey—prenatal, intrapartum, and postpartum—and the 39 structured prompts were displayed on sequential notecards. Participants were asked to write free‐text responses to each prompt and place them at the corresponding point in the timeline (Appendix S2). A facilitator trained in HCD methodologies guided participants through the exercise, clarifying responses as needed and supporting discussion of current workflow challenges. Participants were asked to describe their role in providing care at each of the 39 steps and how their daily workflows aligned with, diverged from, or complicated the mapped process. Participants were also invited to share suggestions for how to improve care provision at each of the specific process steps.

For participants who were unable to attend in‐person sessions, we developed a virtual process mapping exercise using an online Qualtrics survey. This survey mirrored the in‐person process mapping structure and prompted participants to respond to the same 39 questions in sequential order using open‐ended free‐text responses. Process mapping surveys were distributed to existing groups of perinatal stakeholders, including public health nurses and physicians. For stakeholders unable to participate in either the in‐person or virtual formats, we conducted semi‐structured interviews using the same set of prompting questions employed in both the in‐person workshops and the online survey.

2.3. Analysis and synthesis

Our analysis followed three key steps aligned with HCD methodology: (1) extraction of data from written responses to the process mapping exercises, (2) identification of themes, and (3) translation of themes into actionable opportunities. Using a general inductive approach to thematic analysis, four researchers with expertise in HCD independently reviewed participants’ responses to process mapping prompts. Each researcher identified recurring themes that emerged from participants’ descriptions of their roles, workflows, and challenges across the care delivery pathway. Although the process mapping exercise included 39 detailed prompts spanning the perinatal journey, analysis focused on a smaller set of key care delivery phases representing major stages of perinatal care delivery and coordination that emerged during the analysis. The research team tracked the number of stakeholders who contributed responses or expressed perspectives related to each of these care delivery phases, allowing assessment of the breadth of engagement across roles and care settings throughout the process map. The research team then convened to review and discuss emerging themes, refining and consolidating them until consensus was reached.

Final major themes and subthemes were reviewed with a multidisciplinary group of physicians, nurses, social workers, public health nurses, child welfare workers, CBO staff, and members of our Community Advisory Board (current and former perinatal patients, including patients with past SUDs, who advise on patient‐centered initiatives within the studied safety‐net system) prior to finalization. The Community Advisory Board was engaged during the synthesis stage to assess whether the themes resonated with patient experiences, to identify potential gaps or misinterpretations, and to provide contextual insight from a patient perspective. Feedback from the Community Advisory Board was incorporated into the refinement of final themes and the development of opportunity statements. Additional patient‐centered qualitative work was conducted separately and is outside the scope of the present analysis.

Thematically organized data were systematically translated into targeted opportunities for systems improvement. In HCD, the goal of the initial inspiration phase is to uncover unmet needs and systemic barriers through deep ethnographic inquiry and inductive thematic analysis. These insights serve as the foundation for identifying actionable opportunities and inform the development of prototyped solutions to be evaluated in subsequent phases of the design research process [29]. To develop these opportunities, we facilitated a collaborative ideation session involving our core team of HCD researchers (n = 4), representatives from the Plan of Safe Care Community Collaborative (n = 3), and members of Solid Start (n = 2). Through this collective process, actionable opportunity statements were generated and rigorously refined. These opportunities were then reviewed by the research team and a multidisciplinary group of stakeholders from the Community Collaborative to ensure strong consensus and alignment.

3. RESULTS

A total of 47 stakeholders participated in the process mapping exercise, including 26 participants across three workshop sessions, 18 participants via the online survey, and 3 participants via semi‐structured interviews. Participants included attending physicians, nurse‐midwives, residents, a nurse, medical students, social workers, a community health worker, public health nurses, CPS case workers, a health educator, a coordinator from a SUD treatment center, and staff from CBOs (Table 1).

TABLE 1.

Participant demographics.

Role	Count	Department/Role
Attending	9	Family medicine (4), pediatrics (3), psychiatry (2)
Nurse midwife	3	Labor and delivery (3)
Resident	4	Obstetrics (2), pediatrics (1), family medicine (1)
Nurse	1	Labor and delivery (1)
Medical student	2	Not applicable
Social worker	5	Community clinic (1), clinic‐based outreach team (1), labor and delivery (1), community‐based organization (1), other (1)
Community health worker	1	Not applicable
Public health nurse	16	Not applicable
Child protective services	2	Case worker (2)
Health educator	1	Not applicable
Medication assisted treatment center	1	Coordinator (1)
Community‐based organization	2	Manager (1), director (1)

Participants responded to prompts in the process map based on their professional experiences supporting perinatal care for patients affected by SUDs. Responses were organized according to care delivery phase across the perinatal continuum. In the prenatal period, participants described verbal screening and/or toxicology for SUDs in prenatal care settings (n = 18), referrals to treatment or supportive services (n = 24), planning for prenatal care (n = 7), sharing information between care teams (n = 22), and developing Plans of Safe care (n = 16). In the intrapartum period, participants discussed screening/toxicology testing for SUDs on hospital admission (n = 9), sharing of prenatal information with the hospital‐based birth team (n = 18), and initiating or continuing MOUD in the hospital (n = 9). Finally, in the postpartum period, participants commented on toxicology testing infants for substance exposure (n = 5), notifying CPS of a new case (n = 9), sharing information from providers to CPS (n = 9), sharing information from CPS to providers (n = 4), communicating a history of prenatal SUD to pediatricians (n = 8), and referring families to postpartum services (n = 7). Figure 2 presents the final perinatal process map organized by care delivery phases. Table 2 summarizes the number of participants who contributed perspectives related to each care delivery phase.

FIGURE 2.

Perinatal process map organized by care delivery phases, with derived major themes overlaid across phases of care. CPS, Child Protective Services; MOUD, medications for opioid use disorder; OB, obstetrics; SUD, substance use disorder.

TABLE 2.

Number of participants commenting on care delivery phases in the perinatal process map.

Prompts	Count of participants
Prenatal
Screening (verbal) and/or toxicology testing for SUD in prenatal care settings	18
Referring to treatment or supportive services	24
Planning for prenatal care and addiction treatment	7
Sharing information between care teams	22
Creating a POSC	16
Intrapartum
Screening (verbal) and/or toxicology testing for SUD on hospital admission	9
Sharing prenatal information with the hospital‐based birthing team	18
Initiating or continuing medication‐assisted treatment in the hospital	9
Postpartum
Testing infants for substance exposure	5
Notifying CPS of a new case	9
Sharing information from providers to CPS	9
Sharing information from CPS to providers	4
Communicating history of substance exposure to pediatricians	8
Referring families to postpartum to services	7

Note: Counts reflect the number of participants who provided substantive comments related to each care delivery phase across process mapping activities.

Abbreviations: CPS, Child Protective Services; POSC, Plan of Safe Care; SUD, substance use disorder.

3.1. Themes and actionable opportunities

From the data collected through our process mapping activities and synthesis sessions, we identified three major themes pertaining to the current workflow for pregnant patients affected by SUDs. For each major theme, we identified multiple associated actionable opportunities (Figure 2, Table 3).

TABLE 3.

Insights and design opportunities identified through a synthesis of process mapping activities.

Major themes	Opportunities
Divergent understanding of policies and procedures among providers
(1) When staff have variable understandings of substance use services, some patients do not receive referrals. “When a pregnant person has a positive drug screen, ideally the provider gives a list of resources, makes a referral and appointment with [substance use] treatment at the prenatal visit. Currently, some providers don't know the referral resources.”—Health educator	Create shared knowledge among providers around available services to support families affected by SUDs Generate patient‐facing materials to standardize information provided to patients
(2) Policies governing urine toxicology testing are unclear and their use is inconsistent. “I have to go and look at the [urine toxicology] policy… because it changed, and I need to make sure.”—Resident	Clarify and publicize hospital policies specifying the indications for urine toxicology testing
Challenges with care coordination
(3) Many care team members provide similar services. “Screenings for substance use and other social needs and referrals to substance use treatment centers or public health nurses are examples of services provided by multiple providers.”	Identify and document the roles of each member of the care team Identify and document the services being provided to patient at various clinics and community organizations
(4) Without a release‐of‐information (ROI), providers at different sites are unable to share patient information. “We're not reporting things to OBGYN. If that's what you want to do, [it is] up to client. The client needs to release information via an ROI”‐ Substance use treatment worker “There are big firewalls between sharing information from substance use treatment and other medical providers”—Attending	Provide a way to sign ROIs earlier in pregnancy, "pre‐emptively,” and remotely Use ROIs to clarify each team members’ role on the care team Leverage ROIs to increase patient autonomy and control over how their information is shared (and with whom)
(5) Prenatal care information is often lost when patients arrive at labor and delivery. “Sometimes information is shared in the note, sometimes we just dig in prenatal charts.”—Attending “The plans for mothers happen in multidisciplinary rounds where we hear about a flagged patient with a history of domestic violence substance use, homeless, who is admitted to L&D. Patients will flagged because of history… the doctors don't know she is working with [a social worker] and is stable.”—Social worker	Standardize the transfer of information from interdisciplinary care coordination meetings to labor and delivery staff
(6) Many patients who would benefit from care coordination are not discussed at interdisciplinary rounds. “At interdisciplinary rounds [social work] leads and talks about transitions out of the hospital and new incoming patients. Patients in programs are discussed, but there is no way currently for other prenatal care providers to flag a patient.”—Social Worker	Standardize a referral pathway to interdisciplinary care coordination meetings.
Lack of standardized procedures for engaging CPS
(7) A lack of clear guidelines on when to call CPS results in provider‐level bias. “[CPS calls are] variable. There are no clear guidelines and a risk of [racial] bias as a result.”—Nurse (8) CPS is rarely referred to the team member who knows the patient's situation the best. “CPS interviews are sometimes team members, but are not necessarily those most involved in the case”—Medical student Regarding contacting CPS about a well‐known client: “I have had many communication challenges with CPS. If you keep calling, maybe you'll get some information. Occasionally CPS will call me during an open investigating to ask what services I am providing to the patient.”—Social worker	Clearly document and disseminate hospital policies specifying when providers should call CPS Create a “time‐out” in which team members gather to discuss a patient's case before making a referral Develop a new workflow so that when a patient is referred to CPS, the name and contact of the most pertinent care team members are given to the assigned child welfare worker.

Abbreviations: CPS, Child Protective Services; SUD, substance use disorder.

The first major theme reflected the varied understanding of policies and procedures among providers. This theme spanned portions of the prenatal, intrapartum, and postpartum periods and encompassed several subthemes. The first subtheme was: “When staff have variable understandings of substance use services, some patients do not receive referrals.” The second subtheme was: “Policies governing urine toxicology testing are unclear, and their use is inconsistent.” These findings underscore the need to develop standardized practices to ensure that families affected by SUDs receive equitable and consistent care. Opportunities emerging from this theme included fostering shared knowledge among providers about available services for families affected by SUDs, developing patient‐facing materials to standardize the information patients receive, and clarifying institutional policies regarding the clinical indications for urine toxicology testing.

Our second major theme reflected the challenges of care coordination among team members and spanned all stages of the patient's perinatal care journey. The first subtheme that emerged was: “Many care team members provide similar services,” which led to confusion and inefficiencies. Other subthemes were: “Without a release of information, providers at different sites are unable to share information,” and “Prenatal care information is often lost when patients arrive at labor and delivery.” A final subtheme highlighted that: “Many patients who would benefit from care coordination are not discussed at interdisciplinary rounds.” These findings underscored the need to establish systems that clearly define team roles and support timely communication across providers. Opportunities emerging from this theme included developing a formal release‐of‐information process to clarify care team responsibilities, enabling appropriate information sharing across settings, and empowering patients to have a voice in how their information is shared with team members.

Our final major theme reflected the lack of standardized procedures for engaging with CPS, which primarily arose during the postpartum period of the perinatal care journey. One subtheme that emerged was: “A lack of clear guidelines on when to call CPS results in provider‐level bias,” which disproportionately affects families of color. Another subtheme was: “CPS is rarely referred to the team member who knows the patient's situation the best.” These findings underscored the need to establish clear, equitable protocols for CPS engagement. Opportunities emerging from this theme included developing and disseminating hospital policies that specify when CPS should be contacted, implementing “time‐out” procedures to support team‐based decision‐making prior to a CPS referral, and creating workflows to ensure that CPS is provided with the information of the most appropriate care team member when a referral does occur.

4. DISCUSSION

Delivering perinatal care to pregnant people with SUDs involves a complex web of clinical, public‐sector, and community‐based stakeholders. Through process mapping methodologies, we identified three major themes describing challenges shaping current workflows within our hospital system: (1) divergent understandings of policies and procedures, (2) challenges with care coordination, and (3) a lack of standardized procedures for engaging CPS. Through our HCD approach, we identified over a dozen opportunities to strengthen care delivery—ranging from changes to health system policies (e.g. disseminating a formalized urine toxicology testing protocol) to designing processes to facilitate care coordination between clinical and community‐based teams (e.g., systems to document care team members across care settings).

Pregnant patients with SUD frequently receive care from multiple clinical teams, public‐service programs, and CBOs, making coordination and improvement efforts particularly challenging. Findings from our process mapping analysis demonstrated that challenges with care coordination are not only driven by the number of providers involved, but by overlapping service roles, inconsistent policy interpretation, and fragmented communication across care settings. Similar dynamics have been well described in safety‐net health systems [30], where siloed services result in breakdowns in continuity across transitions of care [18, 19, 31]. Our findings reflect these dynamics in the context of perinatal SUD care, where informational continuity was lost across transitions such as between prenatal care and admission to Labor and Delivery or between outpatient care teams due to important confidentiality protections and release‐of‐information documents that, in practice, became an undue burden and inhibited collaborative care planning.

Participants also described substantial ambiguity around when CPS involvement or urine toxicology testing was indicated, creating space for subjective interpretation and bias in high‐stake clinical scenarios. This finding is particularly concerning given well‐documented racial disparities in both toxicology screening and child welfare referrals that lead to lasting impacts on families [32, 33, 34, 35]. Prior work has demonstrated that fear of child separation contributes to “system avoidance,” resulting in decreased use of social services and the strategic concealment of information from providers and staff responsible for resource referral [36]. When CPS is perceived primarily as surveillance rather than support, inconsistent policy interpretation may exacerbate mistrust, undermine patient‐provider relationships, and reduce opportunities to connect families with supportive services. Our findings highlight how divergent understandings of CPS‐related policies within a single safety‐net system risk reinforcing these dynamics, underscoring the need for clearer, consistently communicated workflows that prioritize equity, transparency, and family‐centered care.

Our selection of an HCD approach was intentional and addressed key methodological challenges inherent to studying perinatal care systems for pregnant people with SUDs. Systems of care for pregnant people with SUDs are complex, complicating efforts to improve quality and access. By adopting an ecosystem perspective, this study leveraged HCD process mapping to generate a deeper, systems‐level understanding of complex patient workflows.

HCD methods also allowed us to translate qualitative themes into actionable opportunities to improve care delivery. While traditional research methods are commonly used to describe barriers to care and evaluate interventions, HCD extends this process by incorporating stakeholder‐driven co‐design of potential solutions [37]. HCD is increasingly recognized as a valuable methodology for centering participant experiences in the development of new care models and interventions, including within the perinatal care literature [38]. In our study, identified opportunities varied widely. Some aligned with larger national priorities, such as “[clarifying] and [publicizing] hospital policies specifying the indications for urine toxicology testing,” which reflects calls from the American College of Obstetricians and Gynecologists to limit toxicology testing to scenarios in which there is a clear medical indication for testing [39, 40]. Others were more system‐specific, such as “standardizing a referral pathway to interdisciplinary care coordination meetings,” referring to a unique multi‐departmental care meeting at our safety‐net hospital. Together, these examples illustrate how our thematic findings were translated into actionable, contextually grounded opportunities through the HCD analytic process.

Finally, our study relied on early and consistent engagement with stakeholders throughout each step of our research process, enhancing rigor and relevance. In alignment with HCD principles, we recruited a diverse set of participants with a range of professional backgrounds to elicit a variety of relevant perspectives. We collaborated closely with participants as partners in the discovery and solution‐generation phases, incorporating their goals and values into the final themes and opportunities [21]. This study illustrates how HCD can expand the frameworks and strategies used in traditional quality improvement initiatives by facilitating participatory data collection and placing participants at the center of solution development.

Next steps for this work include developing and testing prototyped solutions derived from our opportunities. A key challenge in translating research into actionable systems change is stakeholder buy‐in; successful development and implementation of the prototyped solutions arising from this work will ultimately depend on effective collaboration with the broad array of providers and staff involved in the care system [41]. As other institutions look to improve systems of care for families with SUDs, these methods offer a template for successful engagement of pivotal stakeholders.

4.1. Limitations

This study has several limitations. First, it was conducted within a single county healthcare system. San Francisco is relatively well‐resourced in terms of clinical and community‐based programs, which may limit the generalizability of our findings to settings with different levels of resource availability. Second, the sample size was small, consistent with other HCD studies. While a smaller sample may limit generalizability, our approach intentionally prioritized diversity of stakeholder perspectives across a complex healthcare system, encompassing a wide range of roles and responsibilities. Additionally, we did not formally audio record and transcribe our process mapping sessions, and as a result some data may not have been fully captured. Finally, while we used HCD methodologies to define opportunity statements derived from our insights, we do not describe the development, implementation, or assessment of prototyped solutions. The viability of these opportunities will be evaluated in the next phase of this work, which may ultimately lead to refinement of both the insights and the proposed interventions.

Importantly, this paper focuses on process mapping completed with providers operating within existing clinical workflows that serve pregnant patients with SUDs, and patients were not direct participants in the mapping activities. This approach was intentional, given the study's objective to examine internal care delivery workflows and coordination practices across clinical and community settings. Patients with lived experience were engaged in the design and review of this work through our Community Advisory Board and participated in other HCD‐centered activities (described elsewhere). Nevertheless, the absence of patient participation in our process mapping activities represents a key limitation of the current study, as patient perspectives are critical to understanding how care delivery processes are experienced and for identifying system‐level gaps and opportunities for improvement that may not be visible to providers alone.

5. CONCLUSION

This study demonstrates the use of process mapping as a unique methodology for improving systems of care for pregnant people with SUDs. In our perinatal ecosystem, process mapping revealed divergent understandings of policies and procedures, challenges with care coordination, and lack of standardized procedures for engaging CPS. The opportunities identified through this work highlighted multiple strategies to improve care coordination for families with SUDs, ranging from strengthening workflows and procedures to improving care coordination across clinical and community‐based services. As the burden of SUD continues to rise among pregnant populations, it is increasingly important that our healthcare systems adapt to the shifting needs of families affected by SUDs. Strengthening healthcare systems to better serve this patient population is central to improving clinical outcomes, fostering respectful care environments, and supporting patients in achieving their parenting goals.

AUTHOR CONTRIBUTIONS

Conceptualization: Benjamin S. Alpers, Devika Patel, Lara Chehab, Melanie Thomas, Neeti Doshi, Heather Briscoe, Marcy Spaulding, Liliana Ocegueda, and Amanda Sammann. Methodology: Christina N. Schmidt, Benjamin S. Alpers, Devika Patel, Lara Chehab, Melanie Thomas, and Amanda Sammann. Data collection: Benjamin S. Alpers and Devika Patel. Analysis: Christina N. Schmidt, Benjamin S. Alpers, Devika Patel, Lara Chehab, Melanie Thomas, Neeti Doshi, Heather Briscoe, Marcy Spaulding, Liliana Ocegueda, and Amanda Sammann. Visualization: Christina N. Schmidt, Benjamin S. Alpers, Devika Patel, and Lara Chehab. Writing—original draft: Christina N. Schmidt. Writing—review and editing: Christina N. Schmidt, Benjamin S. Alpers, Devika Patel, Lara Chehab, Melanie Thomas, Neeti Doshi, Heather Briscoe, Marcy Spaulding, Liliana Ocegueda, and Amanda Sammann. Supervision: Devika Patel, Lara Chehab, Melanie Thomas, and Amanda Sammann.

CONFLICT OF INTEREST STATEMENT

Amanda Sammann, Devika Patel, and Lara Chehab are design consultants for The Empathy Studio.

ETHICS STATEMENT

Human subjects approval was granted by the institutional review board of the University of California San Francisco.

Supporting information

Supporting Information

Supporting Information

DATA AVAILABILITY STATEMENT

Qualitative data may be available upon request.