Shedding Light on the Invisible Work of Peer Recovery Support Specialists through the Design of a Point of Care Technology

Abstract (144/150)

Peer Recovery Support Specialists (PRSSs) are certified professionals that provide social, informational, and logistical supportto people in recovery fromsubstance use disorders. In this study, we report the findings from design work completed with a teamof PRSSsto define workflowand informationalneeds for the developmentofa novel point of care software. Through our qualitative research we uncovered a significant amount of undocumented work that is necessary for a PRSS to provide individualized support to their clients. Thus, much PRSS work is not quantified and rendered “invisible.” We present technology design strategies that help quantify these tasks and illuminate the complexity of PRSS work.

Introduction

Substance Use Disorders (SUDs) are a global health crisis. Drug overdoses, especially involving synthetic opioids and stimulants, have consistently risen across the last decade¹. In addition to evidence-based medication assisted treatment, recovery support services are an integral part of long-termrecovery and a life-saving strategy for thosewho have overdosed². Recovery support takes many forms, but one approach that has gained popularity and has empirical validation^2,3 is peer support. Peer recovery support specialists (PRSSs) are certified personnel⁴ with lived experience that walk alongside individuals in their substance abuse recovery journey⁵. PRSSs help clients navigate their recovery journeys, which can include being a liaison between the recoveree and medical, legal, and community organizations, providingdirect emotional support and case management, and providingrecovery information5. By doingso, they aim to reduce the likelihood of relapse by engaging with clients on a regular basis and extending the reach of treatment beyond a clinical setting into the everyday lives of people seeking sustained recovery⁶.

SUD recovery is a complex, non-linear process⁷. Thus, the work of PRSSs who support clients in this journey is often overwhelming, as increasing workloads can be challenging to peers from both an emotional and time-burden perspective. Moreover, PRSSs navigate the demands of complex, stressful caseloads that may expose them to secondary trauma⁸. Their work tasks include those that we label “invisible” or “hidden”. These tasks include taxing, highly-skilled, embedded work to complete their job outside of or even within traditional job descriptions^9,10. Within a health context, this includes the work between encounters or across distributed care teams¹¹. Examples could include reviewing and combining client-specific data from a variety of sources, managing a repository of community-based resources, working around client constraints, and providing bereavement support. Invisible work is an important construct whenexamining workflowswithinhealthcare contexts^12,13 and lendsto a more comprehensive understanding of both routine and non-routine (e.g., due to unexpected events) tasks¹³.

Digital recovery support solutions are ripe for exploration, as previous research has found that PRSSs utilize a variety of technology-facilitated tools to enact their role, including telephone calls, social media, text messages, and video-conferencing¹⁴. There is also a need to quantify PRSS work so that it can be reimbursed through traditional medical billing mechanisms, such as Medicaid. The latest Medicaid guidelines specify that self-help/peer services delivered individually and face-to-face that provide “structured, scheduled activities” (p. 4) are the types of peer recovery support services that are reimbursable¹⁵. It is unclear, however, to what extent PRSS interactions are systematically documented in a way that would facilitate reimbursement. Moreover, it is currently unknown how much of the day-to-day PRSS work would qualify as a reimbursable activity under current Medicaid billing regulations.

An analysis of the extent to which actual PRSSs working in the field engage in these types of reimbursable activities as compared to other types of facilitation, coordination, and less-structured forms of support is a necessary step to determine how much “invisible work” will remain unreimbursed under these new funding mechanisms. In addition, understanding the holistic activities in which PRSSs engage can better support software designers in developing systems that best meet PRSSs’ needs for care delivery and documentation. This investigation with on-the-ground PRSSs is essential to designing practicable solutions to integrate ways to quantify and support visible and invisible work. In this paper, we utilize rapid ethnography and group participatory design with PRSSs in the field to shed light on this topic and propose technological solutions to address existing challenges related to care delivery and documentation for PRSSs. These techniques have proven effective in examining the interplay of social and technical dimensions of technology use¹⁶ and eliciting contextually rich understanding from complex and collaborative work environments¹⁷.

Methods

This research study is part of a larger initiative^18,19 focused on the development of a secure communication platform for PRSSs and their clients. This qualitative study has several components²⁰ including an embedded ethnography within the PRSS office and participatory design sessions with PRSSs to validate findings from the ethnography and collaboratively design and refine communication platform features. As this was part of their daily work product, participants were not compensated for their participation. Participant activities took place over a seven-week period between April and June 2023 with an estimated total of 20 hours engaging in study activities per participant. This study, funded by the Agency for Healthcare Research and Quality (AHRQ), was reviewed and approved by the Parkview Institutional Review Board. This research took place at Parkview Health, a not-for-profit health system in the Midwest United States serving a population of over 1,000,000 people.

Recruitment.

The research team met with the PRSS director who identified individuals who represented different recovery domains (e.g., maternal, emergency department, community coaches). Contact details for these individuals were made available to the research team who approached coaches, explained the study was completely voluntary, and then consented individuals agreeing to participate.

Data Collection.

We completed four weeks of rapid ethnography with five members of the PRSS team to understand the PRSS workflow and how our technology platform could add value and capacity to the PRSS work. We observed three active PRSSs, each specializing in one of the following areas: 1) maternal clients; 2) community-based clients (e.g. involved with drug court and police-based mobile response team); and 3) clients referred by the emergency department. Additionally, we observed their supervisor as well as the PRSS coordinator, who was responsible for matching new clients with PRSSs and managing the schedule. Ethnographic observations occurred with the three coaches at four sites. During each observation, two research team members shadowed PRSSs during their work shift for 6-8 hours per shift. During the observations, research team members asked PRSSs for in-situ clarifications, additional context on data types, their processes on documentation and follow-up, and approaches to managing their workload. A data collection form that included columns to identify the observation by observer, site, and time was used. The collection form was developed for a previous large-scale observation study, and organized to optimize findings by theme and whether the observation could be translated into a technical or policy point of opportunity²¹. Observable activities, such as interaction with the electronic health record (EHR) and other PRSS tasks, and verbalizations were recorded on the observation form. At the conclusion of the ethnographic observations, we conducted two participatory design sessions with three coaches to validate our themes and categories of work. Participatory design workshops ensured that the themes extracted from the ethnographic observations accurately captured the work done by peers and adequately accounted for their technological needs going forward.

Data Analysis.

At the conclusion of each observation, the two researchers engaged in consensus coding22 of their ethnographicfieldnotes, synthesizing observations, open questions, and notesthat warranted further discussion within the focus group.. Over iterative discussions, codes were developed collaboratively, using an inductive process to discern patterns in the data collected in the field notes. Any inconsistencies in themes were resolved by discussion, which resulted in a singular presentation of thematic findings in two parallel areas: workflow and categories of work. These findings were validated via the participatory design workshops, wherein PRSSs assisted in providing more contextual details and corrections to our initial frameworks. These validation notes were compared, and changes were made to create a refined workflow and categories of work frameworks. For the reporting of identified invisible work, we used the Abraham et al., approach of organizing tasks within routine and nonroutine dimensions¹³.

Field Site Context and Author Positionality.

Our PRSSs are embedded within a health system, which is different from a community organization setting in several key aspects. Specifically, in many community organizations, clients are identified and navigated within one point of care, wherein they are directed to other organizations and resources for recovery treatment and support. In contrast, with our embedded health care system model, PRSSs interact with clients indefinitely, using multiple touchpoints and modalities both within and outside the health care system to assist them through their recovery journey. As our PRSSs are employees of the health system, this allows them to have unique access to medical care and insurance referrals, as well as clients’ health records. This translates to a broader range of direct touchpoints within the health system but also greater responsibilities for client care privacy when handing off clients to other community agencies. It was therefore understood that the technology needs of these PRSSs may be more varied and/or greater than those of PRSSs in other community settings.

Moreover, since many of the research team members are also employees of the health system, care was taken during the ethnography to clearly define their role as observers of workflow for purposes of technology development rather than evaluators of work process for the health system. Defining this positionality at the onset was critical to the ethnographic process so that workflow could be observed as unadulterated as possible, without PRSS workers being afraid that their methods would be critiqued or reported back to the health system.

Results

Several themes emerged from the ethnographic observations that can be summarized into two main categories: workflow and types of work. In section 4.1, we discuss the issues with day-to-day workflow that were identified through both direct observation of workflows and the subsequent participatory design sessions. In section 4.2, we explore the categories of work in which the PRSSs engage, which includes the broad categorization of tasks as well as the weight and priorities assigned to these tasks by the PRSSs. Our PRSS team works closely with the local police department andcourtson severalcommunity-wide SUDinitiatives. One of theseincludes thelocal police department’s HART officers. HART is a mobile response team that seeks out and actively engages residents of Allen County, IN to support them on their road to recovery with addiction from opioids and stimulants. They partner with the PRSS team to follow-up on 911 overdose calls to provide information on the Peer Recovery program to that localized community. Additionally, the community corrections and drug courts are PRSS partners which are reflected in sections 4.1 and 4.2. Finally, in section 4.3 we highlight specific areas of work that were invisible or less visible and participant reflections.

4.1. Workflow.

Observations uncovered pointsof tensionfor PRSSs even at the onset of their interactions with clients. Specifically, there is a very wide network for referrals, which results in clients joining the PRSS program. Figure 1 highlights the different types of entry points. For each type of PRSS, they have different types of referral points that require investigation (e.g., manual chart reviews, gathering further context from referral source, etc.) prior to the first contact with the potential client. There were multiple modes of first contactacrossthese variouspointsof entry. PRSSs might first encounter clients in a face-to-face setting in an emergency department or during a HART mobile response team call. Alternatively, they might have their first contact on a phone call or in a text message, sometimes after referral from the hospital or a community agency. Due to the variety of intake mechanisms (both setting and mode of contact) and the extensive amount of research that takes place prior to an initial contact, documentation of this work presents complexities in both inputting the information and extracting the information for record-keeping and follow-up purposes.

Figure 1.

Simplified Observed PRSS Workflow. *Maternal Coaches Only. **Maternal clients will transition off the maternal coach panel after two years when they can transfer to a non-maternal coach or exit the program

The ethnographic observations exposed the need for a system that could support the complex PRSS workflows for recordingclient demographic data and tracking client engagement and progress over time. Based on their area of focus (e.g., maternal care vs. drug court vs. community-based client engagements), PRSSs have very different day-to-day workloads, community and health system contacts, and communication needs. For instance, those who are working with referrals from the drug courts visit the drug courts to make initial contact with potential clients. Once a client agrees to be part of the program, the PRSS connects the client with legal, medical, and community resources, the majority of whom do not have access to the health system’s electronic health records (EHR). Coordination across systems and communicating with different types of service providers poses a unique challenge, especially for workers in a health system who are bound by HIPAA constraints when communicating about individual needs. In contrast, those PRSSs specializing in maternal and fetal safety conduct much of their work in connection with existing health systemservices, helping pregnant mothers attend their prenatal appointments and receive medication assisted therapy. Although coordinating these activities may involve assisting mothers with transportation and childcare needs, PRSSs could also use the EHR to track whether appointments were missed and to make referrals and contacts. As some of their tracking and record-keeping already exists within the EHR, medical appointments and medical information were more easily accessed and monitored as compared to interactions between the client and outside community entities.

Finally, observers noted complexities in documenting the ongoing interactions between PRSSs and their clients. This complexity was due to a variety of factors, including the very large number of individuals who were referred to the program, the indefinite and unpredictable timeline of follow-up for those in recovery, and the different types of interactions (i.e., modes of communication and content) in which PRSSs and their clients engage. Clients from different referral points also have different logistical, relational, legal, and emotional needs, which require PRSSs to be prepared to be responsive to widely different circumstances. For example, maternal coaches shared that in addition to the more traditional PRSS services, they also support clients at their pre-natal appointments, secure infant-specific and maternal-specific resources, and provide gap-services (e.g., transportation to critical medical or legal appointments). Additionally, there are sometimes complex legal aspects that range from custody hearings to liaising with federal agencies. Technological solutions that support these complex workflows need to allow for flexibility capturing dynamic and client specific data, as everyone’s recovery journey has a different path.

4.2. Categories of Work.

The development of categories of PRSS work required the delineation of day-to-day tasks, analysis of the weight of the tasks from an administrative or functional perspective, and the PRSS’s perceived level of engagement in these tasks as they carried out their daily responsibilities. It was immediately apparent that the complexity of the workflow in terms of breadth of responsibilities was replicated in the categories of tasks in which the PRSSs engaged. As an example, as shown in Table 2, from an administrative and functional standpoint, administrative work and reports represented the most basic level of work but administration required a medium or high level of engagement to complete the tasks.

Table 2.

Invisible tasks grouped by the type of work it represents and mapped to the appropriate weighted category¹².

Based on our observations of day-to-day workflow and with PRSS input in the participatory design phase, we constructed a hierarchyof tasks based on its weight relative to job function. This hierarchy, shown in Table 1, provides a template for the categorical classification system that would exist within a technology to capture PRSS work from the most basic (Level 1) to the most complex (Level 6) daily work tasks.

Table 1.

Categorization of tasks and associated levels of effort

Of note, the ethnographic observations revealed that there was considerable intersection between tasks, as PRSSs shifted from communication with clients regarding referrals or their medical appointments to more general emotional support for issues the recoverees encountered in their journey (e.g., family member issues or issues related to other social determinants of health like housing or transportation). Because of geographic considerations, especially for those PRSSs working in the field or virtually, the PRSS could not simply leave the space to end an appointment as a clinician might do to demarcate the end of a session. It was noted that this might pose additional time, record-keeping, or task-switching burdens for PRSSs working in nonclinical settings. Further, some tasks revealed a significant task-switching burden even if tasks were within the same category of work. For example, within the complex care coordination category, communication across multiple entities can be challenging and time consuming for an individual client as too can providing elbow support for clients that are in crisis or are being affected by others in their networks that are in crisis.

4.3. Invisible Work.

We found examples of invisible work that could be considered routine and non-routine among PRSSs. We are operationalizing invisible routine work as “behind the scenes or work that is not part of the formal job description” and non-routine invisible work as “response mechanisms to get things back on track once unanticipated contingencies occur”¹³. The routine work that we identified can be classified as work behind the scenes that is necessary to provide holistic support to the client or aspects of a certain role that are not well understood unless you are directly in that role. The non-routine work included howcoaches responded to unanticipated challenges and issues not in their formal job description, but necessary to support their client(s). Table 2 highlights examples of these categories of work. While this list is not comprehensive, it provides insights into the diverse and complex nature of PRSS work that is invisible and hard to quantify due to the variance in the levels of effort it takes to complete these tasks.

Discussion and Conclusions

The complex nature of PRSS work makes it ripe for innovation in several key areas. The first area is documentation. Consideringthe ample evidencethat peer-based support is an extremelyeffective treatment modality^2,3, it is imperative that digital recovery support tools support easy documentation of data, including invisible work. This will allow for better outcomes-based research and evolution of peer recovery support services. A technical solution that is tailored to the tasks like those described in Table 2 will help amplify impact, alleviate administrative stressors around time accounting, and support appropriate funding and workload (e.g., client panel size) of PRSS teams. Furthermore, when technology platforms fully integrate more holistic assessments of the variety of PRSS activities, program leaders will have the depth of data needed to fully understand what is needed for program sustainability as securing ongoing funding for these types of programs can be challenging²³.

The second area is categorization. Current approaches like the MedSoc tool allow for medical social workers to document patient care information and track productivity which allowed for categorization of time²⁴. Likewise, Bryne et. al. recently developed a checklist for PRSS work that provides a way to document client engagements that are organized intosocioemotional, informational, and instrumental client supports²⁵. Integrating a process forcategorizing and classifying the complexity of work is central to presenting a more comprehensive picture of PRSS work and connecting to client outcomes. While there exist numerous tools to support the work of peers, none integrate the scaffolding needed to document the invisible work we identified in the field. For example, pre-engagement tasks are of critical importance to the PRSS to provide appropriate support to clients, especially for peers that are embedded within a health system and must take into consideration clinical contexts before approaching a client (e.g., a pregnant mom who just miscarried). These tasks may include manual chart reviews within the EHR and engaging with points of referral for contextual insights – none of this effort is easy to account for in digital recovery software systems.

The third and last area is technology implementation. Emerging literature highlights the challenges of capturing the breadth and depth of PRSS activity within the PRSS ecosystem, identifying how traditional methods are sufficient in tracking some aspects of the recovery process but do not account for a data-driven approach for understanding more nuanced aspects of the recovery journey like empowerment and self-efficacy²⁶. Thus, systems designed for PRSS work need to be agile enough to both support direct client engagement while also providing detailed evidence on the effectiveness of various implementation strategies²⁷. Noted implementation challenges include inconsistent role delineations and complexity of clinical boundaries² in addition to programmatic challenges like appropriate matching with clients²³. Designing for this type of complex milieu is not novel within the field of computer supported cooperative work^28–30. Thus, adapting and modifying strategies with emerging design best practices with the PRSS implementation research community is essential if the next class of digital tools are to maximize client support with the contextually rich data that is needed to appropriately evaluate the impacts of this model of care.

Through our PRSS engagements, we uncovered important workflow and task category complexities that lay a foundation for the design of a technology that can illuminate the invisible work of peer recovery support specialists. Further, the findings fromthis qualitativeresearch caninformtheimplementationof existingdigital recovery solutions that support the work of peers. Finally, the notion of quantifying the work of a PRSS and the connection to client outcomes may inform policy related to reimbursement of peer recovery services.