Care Managers and Role Ambiguity: The Challenges of Supporting the Mental Health Needs of Patients with Chronic Conditions

Abstract

As U.S. healthcare organizations transition to value-based healthcare, they are increasingly focusing on supporting patients who have difficulties managing chronic care, including mental health, through the growing role of care managers (CMs). CMs communicate with patients, provide access to resources, and coach them toward healthy behaviors. CMs also coordinate patient-related issues internally with healthcare practitioners and externally with community organizations and insurance providers. While there have been many interaction design studies regarding the work of clinical and non-clinical healthcare providers and how best to design support systems for them, we know little about the work of CMs. In this study, we examine the role of CMs, particularly focusing on their work to support patient mental health, through interviews with 11 CMs who are part of a large Midwestern U.S. health system. Workflow observations were conducted to supplement the interview data. We describe the role of CMs and identify challenges that they face in supporting patient mental health. A key challenge is a high degree of role ambiguity in this professional role. We discuss sociotechnical implications to better support care delivery processes and technologies for the delivery of mental health services by CMs.

1. INTRODUCTION

An estimated 53.4% of adults in the United States have one or more chronic physical or mental health conditions (Merikangas et al., 2007). For healthcare organizations providing direct assistance to patients for managing these long-term illnesses, there is a growing emphasis on treating mental health needs alongside physical health needs, as they are often interrelated (Moussavi et al., 2007). Because of the focus on long-term patient health within the new value-based healthcare model in the U.S. (Petersen et al., 2016), coupled with the expense of untreated mental health needs (Greenberg et al., 2015), providing mental health treatment and ongoing management has more organizational support than ever before. However, at the primary care level, providers often struggle to adequately treat mental health needs in tandem with physical health needs. Therefore, to meet the growing need for coordinated health services in the U.S., new provider roles and resources need to be created.

One new role created by healthcare organizations to support these needs are Care Managers (CMs). CMs have a complex multifaceted role to help patients to navigate the healthcare system and support their ongoing health behavior change (e.g, quit smoking; exercise more frequently). CMs also work toward healthcare system goals to reduce system-wide costs by reducing unnecessary patient Emergency Department visits (Coleman et al., 2017). CMs typically manage a caseload of patients who not only have physical health conditions such as diabetes, but who also have a high rate of comorbid mental health problems such as depression and anxiety. They do so by working with patients from 3 months up to one year to improve the patient’s health status. They do this as part of a larger team of CMs who share information and best practices with each other. Although physical health needs has been the primary focus of the CM’s work activities, there is a growing focus in US healthcare systems to also address a patient’s mental health needs due to recognition of the value of holistic healthcare and mind-body relationships. Therefore, CMs are being asked to tackle mental-health specific complexities such as social stigma and lack of professional resources (particularly in rural areas). Yet, while a CM’s role in supporting physical health is well-defined, their role and expectations in dealing with mental health issues are not as well-defined by these health systems. Therefore, we need to better understand their roles in addressing mental health issues order to support their needs as they strive to improve the mental and physical health of their patients.

These issues are of particular interest to CSCW researchers. CSCW has long been interested in better understanding work and collaboration in healthcare organizations and more generally, in the delivery of healthcare services (Doherty et al., 2012). The primary focus of this work has been to address the physical condition of the patient (Mentis et al., 2010; Tentori et al., 2012). More recently, there has also been growing interest in CSCW in the mental health space particularly in how individuals navigate mental health in online communities (Kanera et al., 2016; Park et al., 2015; Pater and Mynatt, 2017). An increasing variety of mental health support technologies for use by individuals have been designed and developed including peer-to-peer chat tools (O’Leary et al., 2017; O’Leary et al., 2018), self-management support through smartphone apps (Huang et al., 2015; Mohr et al., 2017) and internet-based psychotherapeutic programs (Andersson and Cuijpers, 2009). However, there has been little work in CSCW (and HCI) so far exploring how healthcare providers such as CMs conceptualize and provide mental health support. Consequently, in order to better understand the broader context of mental health support options and current delivery challenges to overcome, we conducted a qualitative study of a care management team in a large healthcare system in the U.S. Midwest, as the system was expanding its value-based care services.

Similar to other CSCW studies of healthcare work (Fitzpatrick and Ellingsen, 2013), we undertook this study to better understand the challenges that CMs face in providing mental health support, how they attempted to address those challenges, and opportunities for design to support their needs. We build upon a number of recent CSCW studies investigating the role of patient navigators (Bonner et al., 2012; Jacobs et al., 2014; Jacobs et al., 2015; Okeke et al., 2019), showing their important, but often underresourced roles in healthcare systems. We contribute to this literature by focusing on the specific challenges of supporting patient mental health needs. In this paper, we are interested in addressing the following question: what role do CMs play in supporting patients with mental health concerns and how do they accomplish these activities? Our primary goal here is to identify the practices and challenges that these CMs face in addressing patient mental health needs. This study is part of a larger research collaboration with the healthcare system to design, develop, and sustainably implement digital mental health tools within the care management team that will help CMs better support the mental health needs of their patients.

Through our study, we found that CMs faced a number of challenges in their work to support patients living with mental health problems. In particular, because of the recent creation of this role in the organization, CMs’ role ambiguity exacerbated challenges and led to varying levels of support for patient mental health issues. Critically, this meant that traditional methods of intra-organizational knowledge-sharing and collaboration, such as asking experts, creating ad-hoc teams, disseminating information via email (Spence and Reddy, 2012) or having smaller best-practice routines to fall back on in times of uncertainty (Ackerman and Halverson, 2004) described in previous CSCW studies, were not observed in our research. This paper makes three major contributions. First, we describe the work practices of a growing role in the healthcare system – CMs. Second, we highlight the different perspectives on mental health management among CMs and their challenges of delivering mental health services in health systems where there is often little support for doing so. Third, we discuss potential sociotechnical approaches to better support CMs’ engagement with patients around mental health.

2. RELATED WORK

We provide an overview of the literature focusing on three related issues. First, we review studies examining collaboration in healthcare organizations. We then summarize the challenges of delivering mental health services in the shifting U.S. healthcare system. Finally, we describe the focus of collaborative technology research in the mental health space.

2.1. Collaboration in Healthcare Organizations

CSCW researchers have long been interested in understanding the role that healthcare providers and other members of the healthcare organizations play in providing patient care (Wolf and Karat, 1997). Researchers have examined a variety of roles, activities, and goals for providers and staff working in healthcare organizations. In particular, clinical staff – physicians, nurses, pharmacists, and other providers – have been the focus of research interest for a number of years. This has particularly been the case as researchers try to design and implement health information technologies to better support clinical staff members. Fitzpatrick and Ellingson (2013) highlight the focus on hospitals and electronic health records that has driven much of the previous health-focused research in CSCW. For example, there have been numerous studies of clinical staff and how to support their work across a variety of settings including intensive care units (Reddy and Spence, 2008), emergency departments (Kulp et al., 2017; Park et al., 2012), inpatient care (Zhou et al., 2009), and dental care offices (Cederman-Haysom and Brereton, 2006). CSCW researchers are also starting to investigate the important roles that healthcare providers can play in mental health management from the patient perspective, e.g. (Murnane et al., 2018), yet we still need to learn more about this context from the healthcare provider perspective.

2.1.1. Patient Navigation Within Healthcare Systems

CSCW and HCI researchers have also focused on understanding the work of supporting patients as they navigate the healthcare system. Recent studies have begun to investigate healthcare system roles, beyond traditional physician and nurse roles, which focus on patient navigation and ongoing treatment. These include Home Health Aides (HHAs) who provide home-based care, often for individuals managing chronic conditions (Okeke et al., 2019) and patient navigators who assist patients throughout the breast cancer diagnosis and treatment process (Jacobs et al., 2014; Jacobs et al., 2015). These studies elucidate the complex and challenging work of patient navigators helping patients through often very difficult health conditions. Healthcare workers in these roles must collaborate with and support patients in their illness trajectories, often within organizations with limited resources, in addition to coordinating patient care with other healthcare providers. However, while these studies describe the complex and shifting nature of patient navigation work, there is less focus on provider experiences and needs regarding supporting the mental health needs of their patients. In contrast with discussions about physical health conditions, mental health is more likely to be stigmatized and is often a more sensitive topic for individuals (Byers et al., 2012). Therefore, understanding the practices and challenges of supporting patient navigation in the mental health context will extend our current understandings of the diversity and depth of the work of healthcare providers, particularly in the new role context of care management.

Other literature regarding navigation focuses on how patients, caregivers, and community organizations support patient navigation needs. These needs range from finding needed information to dealing with the hospital bureaucracy (Gui et al., 2018; Robinson-White et al., 2010). For example, Mishra and colleagues (2016) described how patients and their caregivers act as active participants in managing the patient’s care during hospitalization. Similarly, Miller and colleagues (2016) studied caregivers, noting their role in navigation to help guide patient’s decisiosn and help them understand their condition. The authors identified design considerations for building tools to better enable them to support their loved ones during a hospital stay. Gui and colleagues (2018), introduce the term “navigational competence,” describing how parents of young children’s navigation practices “included not only information practices where interviewees sought, analyzed, and shared information, but also coordination and negotiation with organizations just to make things work.” Finally, Bonner and colleagues (2012) describe the work of Child Life Specialists in helping children in a hospital learn about their condition as well as distracting them from the daily challenges of being in the hospital. Thus, through deep investigation of the work of individuals in healthcare settings, CSCW researchers create essential understanding to guide the design of tools to better support their activities and goals.

To the best of our knowledge, this is the first CSCW paper investigating the role of care managers in the mental health context. While there is growing literature about the work of clinical providers and of how patients navigate the health system (often with the help of family and friends), we still know relatively little about the CM role and ways to best support their needs as they navigate the health system alongside their patients. Understanding the communication and coordination needs of CMs, who play a complex role in the health system, will be particularly valuable.

2.2. Models of Mental Health Service

The role that CMs play is multifaceted. CMs work with many patients who have co-morbid mental health problems, as mental health problems are common and undertreated (Otte et al., 2016). In the general population, the 12 month prevalence rate of common problems such major depressive disorder is 6.7% and 28.1% for anxiety disorders (Kessler et al., 2005). These rates are even higher in subpopulations of patients with chronic medical illnesses (Kunik et al., 2005; Moussavi et al., 2007), who are often the recipients of care management services.

There are a number of evidence-based models for providing mental health services, primarily focused in pharmacotherapy (Arroll et al., 2009) and in psychotherapy (Ijaz et al., 2018). These services typically require repeated effort from patients, including multiple follow-up appointments with a specialist either for medication adjustments or for psychotherapy sessions. There exist a large number of pharmacologic treatments for mental health concerns (Magni et al., 2013) which typically involve a daily oral medication. Similarly, a large number of types of psychotherapy exist, ranging from cognitive-behavioral to interpersonal to mindfulness-based treatments, with robust evidence for the effectiveness of these treatment programs (de Mello et al., 2005; Hunot et al., 2013).

Despite the high prevalence rate of mood and anxiety disorders and the number of evidence-based treatments for such disorders, only 19.6% of individuals with mood disorders and 12.7% of individuals with anxiety disorders receive minimally adequate treatment (Wang et al., 2005). There are a number of barriers for people to receive mental health treatment including stigma, lack of motivation, time constraints, and difficulties with transportation (Mohr et al., 2010). On top of those barriers, there are not enough mental health professionals to provide standard care for the approximately 50 million adults who will experience common mental health problems each year in the U.S. (Kazdin & Blase, 2011). Thus, even if the treatment being provided were adequate and patients were ready and willing to engage in treatment, the current U.S. care system is not capable of treating all people with common mental health problems.

One approach that has been taken in an attempt to improve the delivery of mental health care is the Collaborative Care Model (Gilbody et al., 2006). In the Collaborative Care Model, behavioral health professionals are embedded within a primary care practice and collaborate with the primary care providers to develop mental health care plans for distressed patients. The care team (comprising of primary care provider, behavioral health care manager, and psychiatric consultant roles) shares responsibility for tracking this defined group of patients. Specifically, these patients’ mental health outcomes are routinely tracked. The embedded behavioral health professionals provide evidence-based behavioral health treatment (e.g., problem solving therapy, behavioral activation, cognitive behavioral therapy) that have demonstrated effectiveness in primary care settings. While a Collaborative Care Model team does include a behavioral health care manager, this is a different role than the care managers and the care management service observed in this study (Bower et al., 2006). Unlike in the Collaborative Care Model where care managers focus solely on mental health issues, in our care management setting, the CMs need to focus on both physical and mental health issues.

2.3. Mental Health and Collaborative Technology

Mental health is a growing research topic within CSCW and the broader HCI community (Arnrich et al., 2013; Calvo et al., 2018; Huang et al., 2015; Pater and Mynatt, 2017). There is a significant body of research focused on better understanding the mental health issues and needs of individuals, primarily within social media and other online settings (Zhang et al., 2018). There has been interest in peer support tools for mental health (O’Leary et al., 2017), and developing models for identifying mental illness in social media data (Feuston and Piper, 2018; Sadeque et al., 2018; Tsugawa et al., 2015). There has also been work done on identifying factors that affect the disclosures of mental health issues in social media (De Choudhury et al., 2017).

Within the psychological and medical research literature, there are a wealth of studies examining how people manage their mental health (Lawn et al., 2007; Leamy et al., 2011; Schrank et al., 2012), and a number of technology-enabled services to provide mental health treatment have been tested in efficacy trials (Andersson et al., 2014; Cuijpers et al., 2013). More recently, there has been growing interest among CSCW and HCI researchers to understand how people manage mental health issues and to design technologies that can support this management. For example, Murnane and colleagues (2018) examined social relations among individuals with bipolar disorder in regards to how these people played roles in symptom management, and provided an understanding of how personal informatics mediate these relations. Yamashita and colleagues (2013) examined how family caregivers dealt with the challenges of taking care of depressed family members and found they utilized a variety of strategies. Walsh and Richards (2017) examined experience and engagement with the design features of an internet-delivered treatment program for generalized anxiety, and found that having a flexible program structure, being able to visualize improvement, and receiving help and guidance from an online supporter were key for program engagement.

Most of these studies have focused on the end user (i.e. patients, caregivers), rather than on supporting individuals within the healthcare system, such as care managers and most interventions have been tested in research studies focused on feasibility and efficacy, rather than on implementation in large healthcare organizations. While there has been some early work in automating training for and evaluation of psychotherapists (Hirsch et al., 2017; Hirsch et al., 2018) and in examining challenges to designing care (e.g., need for permanent technology champions & integration into hospital care routines; Thieme et al., 2016), we are unaware of any studies in HCI that have examined the delivery of mental health services in large healthcare organizations. This study starts to address this gap by examining the challenges that CMs face in delivering remote mental health services in the context of a healthcare organization.

Because care managers provide remote-delivered services to patients, technology-enabled mental health services may be particularly relevant to their work. Technology-enabled mental health services have been developed to address ongoing mental health treatment gaps, and are largely based on evidence-based psychotherapy principles – most commonly, cognitive-behavioral therapy. These programs can treat mental health problems using digital technologies such as web-based and mobile applications coupled with support from a CM or coach. Indeed, a large number of RCTs over the past 15 years have consistently shown that digital treatments, when coupled with low intensity coaching, can be highly effective (Andersson et al., 2014; Cuijpers et al., 2009; Donker et al., 2013; Richards and Richardson, 2012). While there has been great hope that these technologies can extend care, many attempts to implement commercial technology-based interventions in healthcare organizations (e.g. Kaiser Permanente, Priority Health) have failed and real-world engagement with these interventions has varied widely with many programs reporting low engagement (Fleming et al., 2018). Some studies have been successful, for example, an internet-based cognitive behavioral therapy program for generalized anxiety demonstrated effectiveness in a recent trial of college students (Richards et al., 2016).

However, in a large implementation trial of the most widely used technology-based treatments for depression, deployed in primary care in England’s National Health Service, showed no significant effect (Gilbody et al., 2015). Patient- and service-level variables that predict a positive response to treatment have recently been examined and can be used to inform ongoing service implementation strategies (Catarino et al., 2018).

Attempts at implementing technology-enabled mental health services have failed because neither patients nor providers in real world settings use them. Mohr and colleagues (2017) argue that these technologies have been designed “top down” with little input from the users of these technologies. Many interrelated factors affect the adoption of technologies that support the delivery of mental healthcare. For instance, some current key limitations relate to such tools’ inability to sustain engagement; challenges in evaluating their appropriateness and effectiveness across highly variable patient circumstances; providers’ ability to integrate new duties and tools into their workflow; and ongoing issues related to ethics and bias implicated in mental health support resources and interventions. Therefore, to design technology-enabled services that will be used by CMs to support the ongoing mental health needs of their patients, we need to better understand the needs, limitations, and workflows of CMs.

3. METHOD

To address our research goals, we utilized semi-structured interviews with CMs and conducted observations of CM workflows. We used this methodological approach because it allowed us to better surface and understand the challenges that CMs face in helping patients address their mental health issues. We collected and analyzed the data in an iterative manner that allowed us to progressively refine our conceptual understanding of the collected data (Lincoln and Guba, 1985; Maxwell, 2012).

3.1. Site and Participants

We studied a CM team that is part of a large healthcare system in a state within the US Midwest region. Like most healthcare systems in the U.S. (Frank and Glied, 2006), at our site, there are often long waitlists for mental health providers within this system and patients often raise mental health-related concerns to providers who do not specialize in mental health. The CM team is tasked with facilitating care with a focus on supporting patients who struggle to manage their own health conditions. Therefore, they collaborate with a patient’s primary care provider(s) and help patients coordinate services with other providers within the wider healthcare system (e.g. chronic disease specialists, mental health providers). Patients are referred to care management through a variety of mechanisms, ranging from direct referral by their primary care provider to patients being automatically identified and referred through metrics listed in their health records (e.g. consistently high A1C levels, indicating that diabetes is not well-controlled). Specific CMs specialize in different patient populations, some of which are defined by medical status (i.e. one CM specializes in perinatal and postpartum care), while other patient populations are defined by demographic status (i.e. one CM works with Spanish-speaking patients, one CM works with older adults in a certain geographic region).

3.2. Data Collection

We interviewed 11 of the 12 CMs, who were located in two offices (Table 1); the 12th non-participating CM was on leave during the study period. Five participants worked in Location 1 – a large single story office building with a primarily open floor plan. The CMs all had cubicles. Because of the open format of this office, sound transferred easily and words spoken during phone calls could intermittently be heard throughout the office space depending on the current level of ambient noise. The remaining six participants worked in Location 2, a single floor of an office building in a different town than Location 1. In Location 2, CMs shared private offices with each other. Words spoken during the phone calls could not be heard directly outside of the rooms. Both Location 1 and 2 are located in suburbs of a large city. Location 1 was approximately 20 miles from the city in a densely populated area, and Location 2 was approximately 45 miles from the city in a more sparsely populated area. The CMs at Location 2 served a larger ratio of rural residents.

Table 1:

Study participants

Pseudonym	Observed	Location
Nancy	No	1
Jane	No	1
Sarah	Yes	1
Katie	Yes	1
Mallory	No	1
Gina	No	2
Dawn	Yes	2
Francesca	Yes	2
Mary	No	2
Jennifer	Yes	2
Emma	Yes	2

The authors initially met with the care management team during a regularly scheduled team meeting to introduce themselves and the research study. During the introductory meeting, and at each subsequent contact, the authors emphasized to the CMs that participation in the study was voluntary and not a requirement of their job. Thus, the CMs were recruited through contact with their manager who allowed the researchers to visit the offices. The CMs were notified that the researchers would be visiting and that their participation was voluntary. Participants were largely receptive to participation, and no CMs declined to participate. The study was approved by the university IRB.

All CMs were female and most were nurses by training (10/11). The remaining CM was trained as a social worker. Two of the nurse CMs had formal psychiatric training. To protect anonymity of the participants, we have removed any explicit mentions of their training so that they cannot be linked to their pseudonyms. In Table 1 below, we note which CMs also agreed to an observation in addition to an interview, and in which of the two office locations they were conducting their work.

We conducted semi-structured interviews that lasted approximately 30–45 minutes. The 1^st and 2^nd authors met participants individually in their offices or in a private conference room. The interviews focused on CM work responsibilities and practices (e.g. How do you communicate with patients about their symptoms, their follow-up with referrals, and their progress toward recovery?), experiences in working with patients with common mental health problems (e.g. What resources or tools do you currently use for patient treatment of depression and/or anxiety?), and the potential for incorporating technology enabled mental health services into the care management program (e.g. How do you envision a technology enabled care system working in this clinic? What concerns might you have about delivering mental health services remotely?).

While the interviews were the primary method of data collection, we also conducted workflow observations of seven CMs to further contextualize and understand their work practices. The observations lasted approximately 20–30 minutes, were conducted after the interviews, and focused on understanding the workflow of the individual CMs. During these observations, the authors observed CMs making and receiving phone calls, organizing and distributing resources for patient provision, and filling out patient chart information in their electronic medical record system. The CMs who participated in the observations appeared to provide a realistic view of their behavior during observations. CMs who opted out of observations primarily did so due to time constraints, such as leaving the office for lunch or to go to the hospital.

All participants were given $5 gift cards as a small token of appreciation for participating in the study.

3.3. Data Analysis

The interview transcripts served as the primary data source and were analyzed by the authors using a thematic analysis approach (Braun, 2006). This approach enabled the coders to become familiar with the data as they systematically organized individual codes into broader final themes.

Interviews were transcribed. One participant did not wish to be audio recorded, and therefore we typed up our field notes for that participant. We used open coding to start the process of theme identification. The 1^st and 2^nd authors met frequently to discuss the generation of open codes, and to identify areas of overlap and discrepancy in these codes. After a list of 112 open codes was generated, they engaged in an affinity diagramming exercise to create axial codes (O’Leary et al., 2017). The 14 axial code categories that emerged during this exercise were used to create a code book and served as the set of codes for use in the second round of coding. Axial coding was conducted through a cloud-based qualitative coding software package (NVivo). The results of the axial coding were then organized into the themes described in this paper. The authors met regularly throughout the analytic process to discuss these codes and ensure validity.

Along with interview coding, the workflow observation data was also analyzed using thematic analysis. We primarly utilized the observational data as means to triangulate our interview data (Reeves, Kuper, & Hodges, 2008). In particular, we used the observational data to better understand the workflow processes that the CMs described in their interviews.

4. FINDINGS

In this section, we first highlight the overall role and work practices of CMs (4.1). We then focus in on the experiences and challenges the CMs face in delivering mental health support (4.2). We conclude by detailing the variety of ways in which CMs try to address these challenges (4.3).

4.1. The Role of Care Managers

Our participants expressed that CMs have a multifaceted role within the healthcare system. On one hand, they communicate directly with patients to support their health needs. To do so, they assess patients to identify health facilitators such as adequate health insurance, good mental health, and abundant social support, as well as barriers such as low-income levels or comorbid depression. After assessment, CMs connect patients with community-based and online resources, encourage them to make relevant health goals, and coach them toward healthy behaviors through ongoing check-ins. Mary describes the breadth of her activities to help patients follow their plan of care:

‘My role is to provide support, assistance, health coaching, and whatever it is that is needed to help [patients] to follow a plan of care for them to maintain their health.’

- Mary

Beyond their work with each individual patient, CMs coordinate with other providers both internal and external to their healthcare organization. For example, they connect internally with healthcare practitioners such as their patients’ primary care physicians, as well as externally with community organizations and insurance providers to try to ensure that patients receive the appropriate care. Within their care management team, they also engage in a number of individual and team-based processes to accomplish their tasks (detailed in 4.1.1). As an essential part of their internal and external collaboration and ongoing support of their patients, CMs also widely use the Electronic Health Record (EHR).

Unlike other healthcare providers who have a structured day of patient appointments, a CM’s workday is largely unstructured. This is partly because of the largely telephone call-based contact they have with patients. While some CMs set up specific times to call with patients, the majority of CMs have to contact patients when patients are willing to be contacted. Thus they often call and leave messages for patients, and then have to wait for a response. Patients also may call CMs with questions or problems at any time. Consequently, the day-to-day flow of CM work is often unpredictable.

A major aspect of care management for this team is to encourage patients to develop customized health solutions for their own lives using methods common in health coaching, such as Motivational Interviewing (Rollnick et al., 2008). Prompting patients to create and follow-through with their own goals is viewed as a key learning goal of care management. CMs are also tasked with reducing unnecessary use of the healthcare system such as reducing the number of patient trips to the Emergency Department (ED). This can be particularly important to do sensitively for individuals managing mental health needs. Gina describes her partial success in re-training a patient to utilize healthcare resources appropriate for his level of need:

‘He quite often goes to the emergency department for medical issues, but it’s because his anxiety gets so high and he obsesses about his problems. We’ve kind of broken him of going to the ED a lot and we try to make sure that he has access to his providers for appointments for non-emergencies.’

– Gina

Gina presents a view of the multiple, often complex processes and goals that CMs are expected to accomplish in their work supporting their patients. To highlight the workflow and responsibilities of a care manager, we present a brief description of a typical workday for a CM, constructed from observations and interviews with severals CMs.

4.1.1. A Day in the Life

As we highlighted above, the CM’s primary task is to help patients manage their medical conditions. Below, we provide a brief overview of the daily activities of CMs highlighting the interplay of technologies (phone calls, EHR records), adaptable work structure, and collaborative problem-solving inherent in the daily work of CMs. The following is a composite of data collected during interviews and observations.

“Christina” arrives at work, turns on her computer and logs into the EHR. She notices that 3 new patients have been assigned to her, and makes a note to herself to make outreach phone calls to these patients today, in addition to 5 ongoing patients who are due for outreach calls. Christina next stops by a co-worker’s desk to consult about a patient. Christina and her CM co-worker discuss one of Christina’s patients who has been having difficulty controlling his diabetes and has seemed withdrawn recently. They discuss the patient’s history of depression and how it appears to impact his ability to engage with treatment, and they identify a combination of community resources and medical recommendations for the patient.

Using the EHR, Christina spends the next hour researching her patients’ recent medical activity to prepare for her phone calls, looking to see whether they have been in to see their primary care physicians or other specialists, and if they have had any recent hospital stays. Then, she calls the 3 new patients assigned to her. She reaches one patient who says he isn’t sure if he wants to enroll, but is agreeable to her sending him more information via mail. Christina is unable to get a hold of the other 2 new patients assigned to her, and leaves messages for both of them. After trying to contact these new patients, Christina begins making phone calls to the 5 ongoing patients she has identified as being due for a call. She reaches the first patient immediately, and spends 20 minutes on the phone with this patient to check in on her current health, follow-up with recent referrals, and medication adherence. The patient sounds stressed and ambivalent about taking her currently prescribed medication, so Christina utilizes Motivational Interviewing techniques (Miller and Rollnick, 2012) to prompt the patient to come up with her own goals for being more adherent to her medical care plan.

Christina then leaves messages for the other 4 patients. After every outreach call that day, she records the call attempt and any relevant information in the EHR. As she is documenting one of these outreach attempts, a patient calls her back. This patient had previously scored high on her PHQ-9 depression assessment. Based on the high score, Christina had recommended that the patient see her primary care doctor for a diagnosis and potential treatment. Now, a few weeks later, Christina talks with the patient, who is now diagnosed with major depression disorder in addition to her previous diabetes diagnosis. The patient is looking for psychiatrist referrals in-network for her health insurance.

After the call ends, Christina spends approximately 20 minutes looking up resources online for the patient. Christina notices that her patient does not have adequate insurance coverage for a behavioral health treatment such as talk therapy. Christina calls the patient back to let her know that there is a 5-month wait to see a psychiatrist and encourages her to get on the appropriate waitlists. She reports that she has also located some appropriate community-based referral resources, and the patient states that she is currently driving and requests the Christina mail the waitlist information, and organization brochures to her house.

As described here, Christina’s workflow can be unpredictable due to the remote nature of her work and the availability of her patients. The example also highlights that while her patients are primarily being seen for physical health issues, many of them have mental health issues that can interfere with their engagement in addressing physical issues.

In the following section, we describe in greater detail CMs’ conceptualizations of their role in handling mental health and their ensuing perspectives about and activities to address the mental health concerns of their patients.

4.1.2. Managing Mental Health Services.

Patients are referred to to care management primarily because of physical health conditions. Thus, many of the CMs have expertise in chronic physical disease management and the protocols and practices in place by the division of Care Management were well-suited to this type of work. However, among the common physical chronic health conditions which makes a patient a candidate for care management (e.g., diabetes; chronic kidney disease) there is a high rate of comorbidity with mental health issues. Thus many patients also have a mental health issue (e.g. depression, anxiety) when they are assigned to a CM.

‘generally [patients] are not referred for mental health issues…probably almost every one of them has mental health issues…at some level whether they have situational depression, maybe they’ve lost a spouse and they are down (they are not suicidal or homicidal) or they have a history of depression or anxiety and they are in treatment and currently on medications.’

– Gina

CMs estimated that half or more of their patients actively manage comorbid mental health conditions in addition to a chronic physical health condition. CMs are encouraged to conduct mental health screening assessments as part of their patient intake process, and on an ongoing basis as needed.

However, while all of the CMs felt comfortable discussing and managing patient’s physical health issues, some CMs were uncomfortable discussing the topic of mental health with patients because of what they viewed as a lack of sufficient mental health training. However, there was little formal guidance available to them on how their role would differ for physical health issues compared to mental health issues. Therefore, CMs took individual approaches to dealing with these issues. These individual approaches were based on each CM’s conceptualization of their role in handling mental health issues, how it was connected to physical health, and their understanding of the organizational policies. Consequently, mental health services were provided unevenly across the CMs on the team. Thus not every patient received the same standard of care for mental health.

First, some CMs felt comfortable discussing mental health with their patients and viewed it as an important part of their practice. For example: “you never know in your conversation with them as you’re guiding them through better health…better eating, better diet, better, you know, [mental health] should be included.” – Jennifer. CMs who felt comfortable with mental health discussions believed that mental health and physical health were equal, or near equal contributors to well-being. For example, Mary described her focus on the “whole person”.

‘I feel that [mental health is] part of the whole person and so from that standpoint it’s making sure to include it, to factor that into the overall understanding of, well, if you’re depressed you really aren’t going to be motivated.’

– Mary

CMs with this perspective also viewed mental health and physical health as interrelated. Others described addressing mental health because of the sheer amount of patients with these needs.

‘I mean if they’re just on some antidepressants and they’re not getting counseling or doing anything like that I feel comfortable talking to them because like in our world now who isn’t on an antidepressant.’

– Sarah

Therefore, these CMs believed that as part of their role, they were responsible for addressing not only the physical health but also the mental health of their patients.

However, this perspective was not shared by all the CMs. For other CMs, the chronic physical health condition was of primary importance. Mental health issues are viewed by this group of CMs either as the purview of other healthcare practitioners or as a hindrance to be “managed” or “worked through” in order to address the physical health condition. This approach represents a different conceptualization of their role in mental health management.

‘my focus is medical only, not behavioral so if there are any behavioral issues…I’ll call and make sure they [are] connected with whatever community resources they were referred to [and] to make sure they followed up so but that’s the extent to it.’

- Mallory

In general, across perspectives, CMs almost uniformly described being comfortable talking to patients with “managed” mental health needs where, for example, a patient had a history of depression and currently had a therapist whom they visited regularly. However, many of the CMs were much less comfortable dealing with more severe mental health needs, like schizophrenia or bipolar disorder, or for serious and urgent mental health situations such as suicidality. Regardless of their perspectives, the CMs all identified a variety of challenges that they had to deal with in regards to mental health support.

4.2. Mental Health Support Challenges

The CMs faced three major challenges in trying to provide mental health support. These were (1) initiating mental health discussions, (2) having a clear process for supporting patient mental health, and (3) documenting mental health information within the electronic health record (EHR). We describe each of these in greater detail below.

4.2.1. Initiating Mental Health Discussions.

Many CMs reported very little or no prior training in mental health, making them wary of engaging with this topic within patient discussions. The majority of CMs had a nursing background and their training did not include formal mental health support education beyond suicide prevention.

‘We don’t manage pure behavioral health cases because I am not trained as that type of care manager. There are care managers or case managers that are specifically for behavioral health that is not what we do that is not my expertise.’

– Gina

A few of our participants described that they had some background in mental health support from mental health training from past jobs and continuing education workshops and webinars for mental health services. However, the majority of CMs did not receive any mental health training specific to their current job. This lack of training made CMs concerned about making a mistake during sensitive mental health discussions. They noted that rapport with patients was an important component of care management that developed over time. Thus, the fear of saying something wrong and ruining rapport with patients regarding mental health issues was an issue that troubled CMs. As Katie noted,

‘I’m not very comfortable…because of my lack of experience obviously, but I know from personal experience how sensitive the subject [of mental health] can be to other people and I don’t want to cross that line, and if I am forming relations with somebody already I don’t want to ruin that relationship. I’m concerned about damaging their trust.’

Care management is an elective service and so in order to successfully retain patients, CMs need patients to be willing to work with them. If a CM damages the relationship in any way, the patient may choose to discontinue the service. Knowing that they lacked sufficient background training, CMs were concerned about damaging the relationship with patients by “saying the wrong thing” in mental health discussions with patients.

Beyond considerations about rapport, the fear of potentially saying the wrong thing and causing a patient distress was foremost on CMs’ minds when discussing mental health issues. As Sarah stated,

‘I know if I say the wrong thing and the patient that night committed suicide, oh my god, the guilt would be ridiculous. I don’t feel comfortable with that.’

Multiple CMs stated that to feel more comfortable engaging with the topic of patient mental health, they wanted additional training around mental health issues and mental health management. As Jennifer stated,

“Any training is better than what I have now.”

Thus a lack of comfort among several of the CMs led to a lack of engagement with the real and ongoing mental health needs of their patients.

4.2.2. Having A Clear Process to Support Patient Mental Health.

In addition to difficulties with initating discussions about mental health with patients, the process by which support should be delivered over time was not clear. There was a stark contrast between the number of protocols and guidelines to help CMs assist patients in addressing physical health issues, and the guidelines for supporting mental health. For example, CMs described multiple collaborative behavior change support activities for physical health needs, including making patients aware of relevant resources, collaboratively setting personal health goals, and conducting gentle check-ins regarding progress toward those goals. However, encouraging patient behavior change in the mental health context was more difficult because of stigma and socioeconomic barriers. The social stigma around mental health prevented some patients from active engagement with CMs on this topic. Francesca described this pushback:

“Mental health has stigma and I think patients are sometimes reluctant to admit that they’re having problems in that way.”

Furthermore, multiple CMs discussed that socioeconomic barriers including housing and transportation had to be addressed before they could start to work with an individual on health-related goal development. However, there was no clear guidance on how the CMs could help patients address these issues.

CMs also only had a single protocol in mental health, where they had to screen new patients for depression using a common depression screening measure, the PHQ-9 (Kroenke, Spitzer, & Williams, 2001). As Sarah notes, “our case management assessment isn’t really very in-depth it just does the depression screening tool.” If the patient scored a 10 or above (scale of 0–27), CMs were instructed to recommend that the patient see her primary care doctor for a mental health evaluation. However, beyond this assessment, other aspects of supporting patient mental health over time were unclear, as Emma noted:

‘I think [a protocol] would be very helpful because of my lack of not knowing what to do, where to go…How to approach [mental health problems].’

Lacking knowledge of how to approach mental health problems can create issues because patients expect to receive help from the CMs regarding these topics. For instance, if a patient did go to see their primary care physician and received a diagnosis of depression, she would likely return to the CM for help in identifying support resources. However, for the CM to be able to do this effectively, they would need protocols and guidelines on how to discuss these issues with the patient and how to respond to patient mental health needs that may emerge during the resource seeking process. They had no clear processes for either set of activities.

4.2.3. Electronic Health Record Use.

Similar to other studies that highlight challenges that clinical staff face in using the EHR (Collins et al., 2012; Fitzpatrick and Ellingsen, 2013), we also found that the EHR was not designed to appropriatly support CMs. In particular, CMs faced two main challenges, (1) accessing and (2) documenting mental health information, with using the EHR. These challenges led to additional gaps in care. First, if a patient had any specific mental health records in the EHR, those records are stored separately from the patient’s general health records. Consequently, as non-mental health professionals, CMs cannot easily access the patient’s mental health records in the EHR. The CMs call this practice storing information “behind the glass.” Some CMs discussed “breaking the glass” to access information other providers have documented because mental health information is important to holistic understanding of the patient situation. Other CMs explicitly avoid information behind the glass, due to belief that breaking the glass is not a permissible part of their job and/or that the information is private and not immediately relevant to CM work.

A second and related problem that they encounter was how to digitally document their own notes about a patient’s mental health status, whether that information is about the mood of the patient or a patient’s potential for depression. Since CMs are not considered mental health professionals, their notes do not have the added “behind the glass” protection as described. Thus, some CMs are worried about including information about any mental health/mood challenges. Therefore, they often do not document this information.

‘I know sometime in the future we’re supposed to get rid of the glass and everybody is supposed to be able to see everything but until that happens I am not comfortable putting [mental health] information in the chart’

– Dawn

Several CMs further described the ability of patients to ask for a full print-out of their EHR, and that this ability resulted in CMs fearing that any of their mental health notes, viewed unfavorably by the patient, could ruin hard-won rapport. Due to these documentation challenges, many CMs avoid documenting mental health information within the medical record. Further, some CMs avoided asking about mental health related issues because their formal documentation system did not highlight or prompt mental health as a topic for routine discussion.

4.3. Responding to Challenges

Given the lack of protocols within the system and relative lack of personal mental health training, CMs often turned to their colleagues to share information. CMs with more mental health training often became a resource for other CMs when dealing with mental health issues, and those CMs working within the same office spaces typically touched base with one another throughout the day to discuss patient needs:

‘I’ll reach out to [another CM with mental health training] to see if maybe she has anything under her hat that she might be able to offer’

– Mallory

Similarly, Emma mentioned

‘I will just ask one of our behavioral health nurses for support if she has any suggestions.’

Experienced CMs would thus provide ideas for supporting the mental health needs of patients, both process-based (i.e. ways to talk with a patient about a topic) and more tangibly (i.e. a specific resource to offer the patient). Although not often done, CMs could also move a patient to another CM who had the requisite mental health expertise. However, this raises other problems because transferring patients puts extra burden on the CMs with mental health training.

CMs have also developed a set of work-arounds to address the problems with EHR documentation. First, some CMs used a paper-based notation system in addition to the required EHR documentation, in which they wrote patient mental health information in their private paper files. These files are shredded at the end of management with a patient client. For instance, one CM described having a personal notation code for when a patient is angry (e.g. “the client vehemently declined services.”) Second, another CM who only documented a limited amount of mental health information in the EHR system, documented more freely in the secure insurance-provider system because she knew that patients could gain access to their EHR notes but did not have ready access to the insurance-provider system. These workarounds have all developed due to the lack of protected space for CMs to document mental health status in the main EHR system.

In summary, the CMs responded to the challenges that they faced in supporting patient mental health needs by intra-team collaboration, recommending community and online resources, and and developing alternate documentation workarounds.

5. DISCUSSION

In this section, we first discuss the context of a healthcare system in transition, using the concept of role ambiguity from the organizational literature to unpack the ways in which CMs struggled in their mental health delivery, and why even though they collaboratively supported each other, they ran into ongoing challenges. We then discuss sociotechnical design implications for supporting CMs in their future mental health work.

5.1. Care Managers and Role Ambiguity

In the United States, an increasing number of healthcare systems are moving from the traditional fee-for-service model to a value-based care model where they are paid to manage populations of patients and are financially incentivized to be more proactive in keeping patients healthy by meeting quality metrics for both medical and mental health (Petersen et al., 2016). Care management plays a key role in supporting the goals of value-based care systems such as improving service quality and reducing inappropriate system costs from patient over- and under-utilization (Fortini, 2015) by increasing the level of individual patient attention.

This focus on improving the quality of care and reducing costs puts the CMs at the interface of the patient and the healthcare system. Therefore, similar to the role of patient navigators (Phillips et al., 2018), CMs spend much of their time helping the patient to move through the healthcare system, for instance organizing appointments and making appropriate choices based on level of health insurance coverage. However, unlike patient navigators, CMs are also responsible for ensuring that patients are not over-utilizing system resources inappropriately to keep organizational costs low. The management of these dual roles can be challenging. Consequently, CMs depended on organizational protocols and policies to guide them. However, as a new role in the healthcare system, there was not a clear set of policies and guidelinces regarding supporting the mental health needs of patients. The lack of policies and guidelines created role ambiguity challenges for CMs.

Role ambiguity occurs when “shared specifications set for an expected role are incomplete” (Nugent, 2018). Furthermore, a lack of clear and consistent information related to the actions required for the particular role (Kahn et al., 1964) may lead to unpredictable consequences within the workplace (Pearce, 1981). In healthcare, role ambiguity can have deleterious effects, including emotional exhaustion (Glasberg et al., 2007), burnout (Tunc and Kutanis, 2009), and loss of job satisfaction (Siefert et al., 1991). It also has been identified as a leading cause of stress and uncertainty in many organizational domains including education (Rizzo et al., 1970) and social work (Postle, 2002).

For CMs, role ambiguity raised significant issues. First, as Postle notes in her study of CMs in social work, role ambiguity exposes ‘staff to conflicting demands,’ making it difficult to know how to prioritize tasks (Postle, 2002). The CMs in her study were organizationally driven to make quick and structured assessment (to fit into the IT system) but their clients required more in-depth and nuanced assessment. So, they were caught between the organization’s demands and what their clients actually needed. Similarly, in this setting, CMs had to deal with a variety of sometimes conflicting demands. They had to balance organizational goals (e.g., encouraging appropriate use of services, reducing spending) with supporting patient needs (e.g., validating patient concerns, spending additional time with patients). The respective goals and needs do not always align and CMs may potentially be caught between supporting the organization or the patient, given the breadth of their job responsibilities.

5.1.1. Role Ambiguity and Mental Health Support

In the context of mental health support for their patients, CMs’ role ambiguity led to uncertainty about how to best provide support. Therefore, many followed their own ideas of best practices of mental health care, leading to differing levels of mental health support depending on the particular CM. Additionally, many CMs were placed in an uncomfortable position of trying to ascertain whether to respond to a patient’s mental health needs or focus on their physical health needs. We discuss these issues below.

First, the need to balance the variety of role demands along with the lack of policies and procedures around supporting mental health made CMs uncertain and cautious about how to handle mental health issues. This complicated the interaction between the CMs and their patients, as worries about stigma, “saying the wrong thing,” and ruining rapport made CMs reluctant to help patients manage their mental health. In addition, as described in the findings, many CMs were reluctant to initiate mental health discussions. Consequently, CMs often had limited knowledge of patient’s symptoms. This lack of knowledge affected their ability to address how mental health symptoms were interfering with patients’ ability to manage their other health conditions and reducing the overall effectiveness of their care.

Second, role ambiguity led each CM to follow their own ideas of best practices of care for mental health issues. Individual CMs determined how much they wanted to engage in providing mental health support for patients. This is a particular concern in healthcare where standardized care attempts to limit or avoid quality of care being dependent on the preferences and idiosyncracies of the care provider. Standardized care is generally viewed as desirable because patients can anticipate receiving the same level of care from care providers in the same job type. Furthermore, healthcare systems can anticipate that new providers will be able to manage patient health similarly to providers with a longer work history (LeGros and Pinkall, 2002).

Finally, role ambiguity left CMs in the uncomfortable position of trying to ascertain whether to respond, and if responding, how to best meet patient concerns about mental health needs. Their approach to addressing these situations of uncertainty was, as described in the findings, to turn to other CMs who might have greater expertise in dealing with mental health issues. Addressing knowledge gaps by turning to other members of the team has long been a focus of team collaboration (Paul and Reddy, 2010). However, as many of the CMs noted, this did not directly address the issues surrounding role ambiguity that they had to deal with. Without guidance from supervisors on how to help patients manage their mental health, CMs were left to their own conceptualizations of mental health management. The lack of role and task clarity prevented the CMs from effectively doing their own work or collaborating with others on these issues.

Rogers and Molnar (1976) argue that the most effective way to reduce role ambiguity is at the organization level. Organizationally, CMs require more clearly identified goals and robust processes and appropriate tools for addressing mental health issues. In the next section, we discuss ways that we can better support CMs to clarify their processes and in turn reduce role ambiguity.

5.2. A Sociotechnical Perspective

Taking a sociotechnical perspective focuses us on both the technology and the social system surrounding it (Reddy et al., 2003). Through this lens, we see the need to improve both technologies and processes in tandem to better support the efforts of CMs. We present three areas of focus: (1) Improving engagement with patients; (2) Supporting knowledge sharing; and (3) Documenting mental health information.

5.2.1. Improving Engagement with Patients

CMs are uniquely well-positioned to support patient mental health, within healthcare systems that currently provide little support in this area. However, as described in the findings, CMs were concerned about the lack of training and related discomfort in discussing mental health concerns with patients, and the impact this lack of training and discomfort had on their ability to connect with and engage patients in their health management. We are not suggesting that CMs take a primary role in treating the mental health concerns of their patients, but rather, we are suggesting that the work of CMs would benefit from all CMs routinely screening for mental health concerns, providing referrals and resources to address identified concerns, and following up with patients on their progress with identified concerns, similar to how they already manage physical health problems.

At a basic level, integrating technology supported training programs (e.g., additional webinars, online courses) into CMs’ expected work could help increase their comfort in speaking about mental health issues. As noted in the results, most CMs had little training in mental health management and saw this as an area for professional growth. Indeed, past research provides support for the use of computer-based training for nursing practice (Hart et al., 2008) and it allows nurses to complete standardized trainings during times that are convenient for them. Being able to complete trainings at personally convenient times is particularly important for CMs, due to the largely unstructured nature of their daily work. While recent literature regarding home health aides calls for broadly similar knowledge-sharing applications, Okeke and colleagues (2019) push for ‘just in time’ smartphone-based education for aides. In contrast, for care managers who interact with a diversity of patients via telephone on an ongoing basis, we advocate for more robust and ongoing training both during onboarding and as continuing medical education. By building confidence and knowledge long before patient interactions, care managers can be prepared to broach the topic of mental health and importantly, understand the roles depression, anxiety, and other mental illnesses can play in patient’s larger adherence practices to chronic disease self-management.

Second, one way to promote CM interactions with patients could be to develop technology-enabled service protocols to guide their patient interactions (Carman et al., 2013). For example, after a CM completes a mental health screening measure (e.g., PHQ-9) with a patient, a decision support tool could prompt her with follow-up questions and information to help guide the discussion with the patients. These tools would embody protocol decisions to show next steps (Morris, 2000) for example, regarding what will occur when a patient scores high on a depression screening, and address CM concerns about “saying the wrong thing.” Making these next-step processes clear to CMs (in a topic area where they were previously hesitant and uncertain of how to proceed) so they can guide the patient is important especially in the sensitive context of mental health, and could be vital in engaging patients in their mental health management. For instance, encouraging patient follow-up with their primary care provider for a diagnosis if there is concern that the patient has a high likelihood of experiencing depression or anxiety based on the PHQ-9 scores. However, to avoid presenting new barriers to care delivery, design of such protocols will need to be sensitive to varying levels of digital comfort and digital literacy among CMs.

Beyond the initial conversation, workflow technologies such as reminder systems (Shea et al., 1996) could also prompt CMs to check in periodically with people who already have a diagnosis of depression or anxiety, so that the CM can ensure they are receiving the support they need from their therapist, psychiatrist, or other mental health professionals. Crucially, technology-enhanced workflows should fit within the unstructured setup of the CM’s day so that CMs can respond in real time to patient needs and ongoing illnesses. We see these technology solutions operating at a (more granular) task level within the EHR systems that are used to document telephone-based encounters with patients. As described in our findings, CMs are confident in their ability to carry out basic tasks for chronic disease management, including: (1) coaching patients using the motivational interviewing process, (2) answering medical-related questions, and (3) helping participants find resources and navigate the healthcare system. CMs however struggled to accomplish these activities regarding the mental health needs of their patients. Our process and design recommendations, therefore, seek to support CM confidence and competence in carrying out these task-level aspects of their work within the mental health support context. In discussing these solutions with the administrative leadership of the team, these approaches appear to be feasible. Their leadership acknowledges that they are focused on improving quality of care and are invested in providing the staff with the training opportunities needed to help the CMs feel comfortable and competent in handling patient mental health concerns. They are currently embarking on a year-long quality improvement initiative with a psychologist on our team to increase depression screening rates, and the hospital leadership is sufficiently invested in these potential solutions to allow CMs to be trained during standard working hours. Given the diversity of background training for CMs that is likely to continue at least until institutional programs start providing care manager-specific training, we are primarily suggesting trainings that would occur once CMs are hired

5.2.2. Supporting CM Knowledge Sharing

Similar to teams in settings such as clinical care (Reddy and Dourish, 2002) or air traffic control (Bentley et al., 1992), CM team members need to share knowledge and collaborate to solve patient-related problems. Effective knowledge sharing is crucial for the successful functioning of the team (Staples and Webster, 2008). However, currently, there are few mechanisms that are available to CMs for them to share their information with other CMs except through word of mouth. This paucity of mechanisms limits both the ability for widely sharing the information and the ability to pass on the information as turnover occurs in the team.

One approach that could start to address these issues is the use of an organizational memory system (Steinand and Zwass, 1995). These systems serve a variety of purposes including identifying who has a particular expertise (e.g., so a CM could ask them questions) and allowing for the storage and retrieval of relevant information (e.g., in digital files or data stores). When designed appropriately, these types of systems can be accessible to employees with varying levels of digital comfort and can minimize the likelihood of information overload. Ackerman (Ackerman, 1998; Ackerman and McDonald, 1996) previously described the development of an organizational memory system, Answer Garden, which supports the retrieval of relevant organizational information.

As noted in the findings, some CMs had more expertise in mental health than others, and this created some strain to continually support the CMs without expertise. Attending to worker strain is particularly important because nursing and related ambiguous and complex roles are prone to high levels of burnout (Tunc and Kutanis, 2009). An organizational memory system could minimize the burden on CMs with mental health expertise and provide ongoing support to those CMs who are less comfortable discussing mental health. Having a repository of key questions that continue to be asked, for instance, how often patients should be screened for mental illness, what the recommended treatment path is for depression or anxiety, basic ways to enter into and exit a discussion about mental health, and the procedure to follow if a patient is having a mental health crisis. Some aspects of an organizational memory system are certainly connected to automated decision support systems described above, but critically, organizational memory information (either in documentation form, or in knowing who to ask questions about what issue) should ideally be understood before a CM encounters a patient need.

Additionally, because of the limited level of current mental health support knowledge in the CM organization, an organizational management system could also connect with other parts of the larger healthcare system. For instance, for certain questions, CMs may like to hear advice from and connect with other behavioral health practitioners in the healthcare system (e.g., psychiatrists, psychologists, counselors, and other behavioral health care practitioners). Building skills in approaching and respectfully engaging with patients about their mental health will not happen overnight, but as designers we can think more broadly about how to pull from the knowledge of the larger healthcare system to inform, encourage, and support CMs to feel comfortable doing so.

However, while organizational memory systems might be useful for understanding proper procedures and care plans, as well as the expertise of certain individuals, there are still challenges in how they can support the work of the CMs. Organizational memory systems can take many forms, however, workplace chat applications such as those described by McGregor and colleagues (2019), are unlikely to be an effective fit for this context where patient data is highly protected. In addition, the CMs are physically co-located, therefore they are more likely to discuss best practices in-person (as described in our findings), than to seek information via chat. For instance, CMs in our study discussed a practice of having two CMs sit side-by-side while calling a patient who was having a particularly rough time. Given this current practice and the sensitive nature of sharing patient information, designers creating organizational memory systems for these types of organizations would need to consider how to build upon these routines without overburdening CMs existing work practices.

Some of the best practices that would need to be included in these systems (and in the protocol and documentation systems which are discussed in the preceding and following sections, respectively) are not currently clear. Therefore, in order to design effective memory systems for this organization, there needs to be further needs finding across the different levels of management and the entire CM team to surface expertise, common questions, and issues. Furthermore, the organization needs to create new policies and norms about how to access others who have expertise, with an explicit focus on avoiding overburdening them in this task (e.g., having ‘office hours’ time, or periodic team Q&A times). Therefore, it is important to first create proper procedures, as described in 5.2.1, and then to facilitate ways of accessing these procedures over time for use in-context as CMs interact with patients.

5.2.3. Documenting Mental Health Information

Documenting mental health information within the electronic health record (EHR) was a challenge to the CMs. While the EHR has supported collaboration amongst a variety of providers (Reddy et al., 2001), the design of these systems has frequently been cited as a challenge for many types of healthcare providers (Hartswood et al., 2003). Of particular importance, were the restrictions around who can and who cannot document and access mental health information stored “behind the glass” in the EHR system. Many of the CMs were uncertain about how to balance patient privacy with ease of provider communication, and found themselves at a loss for an appropriate method of doing so.

While the first two subsections focused on potential technical opportunities, here, the challenge is not the technology itself but rather the policies and processes in place to support the CM use of the EHR. As researchers have highlighted (Berg, 2001; Miller and Sim, 2004), a mismatch between organizational policies and technologies will lead to challenges in using the system. In this case, the CMs were not allowed to enter or access mental health information in the EHR when they clearly had a need to enter and access that information. Consequently, the processes and policies need to be re-designed to better support the CM use of the EHR for mental health. For example, CMs may be granted access to documenting and accessing mental health information “behind the glass” or they may be provided with approved templates in which to document mental health information in the existing system. These processes and policies must be created with ethical considerations in mind for maintaining patient privacy, and thus if CMs are given broad access to information “behind the glass,” they must be trained on when and how to access that information for necessary patient care, and best practices for maintaining confidentiality and privacy for their patients

In summary, our findings resonate with prior work describing how mismatches between policies and technologies can have a negative effect on patient care (Murphy and Reddy, 2017), as well as negative effects on the well-being of healthcare providers (Duquette et al., 1994; Lloyd et al., 2002). Consequently, as CSCW researchers and designers in these spaces of healthcare organizations in transition, there is a strong need for organizational policies and processes to be well-aligned with the design of future mental health technologies. Together, this would reduce the pervasive challenges of role ambiguity, more effectively support CM activities, and improve essential patient mental health care.

6. LIMITATIONS

Since we studied a single team of CMs in a particular healthcare system, the findings may not generalize to CM teams in other organizations that have different policies and guidelines in place for interacting with patients about mental health. Further, the observations were conducted at single time points, and thus we reported on a limited subset of information. Observing the CMs in their work practices over a longer period of time may have led to richer insights into challenges faced, such as specific difficulties in connecting patients with mental health care and how different CMs’ beliefs about mental health played out in their work. Finally, this study was conducted in the U.S. This is important to consider when interpreting the findings, not only because the structure of care systems can vary considerably across the world, but also because attitudes and beliefs about mental health vary significantly by culture. Despite these limitations, the study provides us insights into the role that CMs play and the challenges that they face in providing mental health support.

7. CONCLUSION

We present one of the first studies examining the role of CMs in supporting mental health services in a healthcare system. The CMs shared their experiences of delivering mental health support services together with other physical health focused care services, and discussed opportunities and challenges for mental health support technologies in this clinical role. There is a growing gap between the need for mental health services and the availability of these services in the United States. However, there are opportunities to address some of this gap through roles such as CMs. Providing CMs with the appropriate tools and processes to effectively support a patient’s mental health needs would improve the quality of mental health services delivered through this expanding healthcare delivery service. Finally, technology alone will likely fail to make significant improvements in the absence of change in the organizational culture and policies (e.g. Doherty et al., 2012; Orlikowski, 1995). For CMs to effectively support patients’ mental health needs, healthcare organizations need to explicitly acknowledge the role that they play in in this area. Organizational policies and processes need to be aligned with technology design, to reduce role ambiguity and best support the work of CMs in supporting patient mental health.