A Low-Intensity Behavioral Intervention for Depression in Older Adults Delivered by Lay Coaches: Proof-of-Concept Trial

Abstract

Objectives:

A broader workforce is necessary to expand U.S. geriatric mental health services. We examined 1) feasibility of training undergraduate students to deliver Do More, Feel Better (DMFB), an evidence-informed program for depression; and 2) feasibility, acceptability, and outcomes in a single-arm proof-of-concept trial.

Method:

In Study 1, we taught DMFB to 18 upper-level undergraduate students and assessed fidelity using role plays. In Study 2, four students delivered six weekly DMFB sessions to 12 community-dwelling older adults (M = 66.83 years old, SD = 10.39) with depression (PHQ ≥ 10). Patient outcomes were change in pre- to post-treatment depressive symptoms, disability, and the target mechanism of increased activity.

Results:

Fidelity was high in the course (Study 1; 82.4% of role plays rated as “passing”) and the trial (Study 2; 100% of 24 sessions rated as “passing”). The majority (83.3%) of patients were retained and evidenced statistically and clinically significant improvement in depressive symptoms (Hamilton Rating Scale for Depression [HAM-D]), disability (World Health Organization’s Disability Assessment Schedule 2.0 [WHODAS 2.0], and activity (Behavioral Activation for Depression Scale [BADS]).

Conclusion:

It is feasible to train bachelor’s-level students to deliver a brief, structured intervention for depression. Future research should consider implementation strategies and stakeholder feedback.

Although depressive disorders are a leading cause of disability, suffering, and healthcare expenditure (Greenberg et al., 2015; Liu et al., 2020), there are significant gaps in treatment availability and access across the United States (U.S.), particularly among adults 55 years and older. Most older adults with depressive disorders do not receive mental health treatment for their symptoms (Barry et al., 2012; Byers et al., 2012; Olfson et al., 2016). Untreated depression among older adults—even of a mild severity—is associated decreased quality of life (Jia & Lubetkin, 2017) and increased mortality (Eurelings et al., 2018; Saeed Mirza et al., 2018). When treated, older adults are usually treated with antidepressant medication and rarely receive psychotherapy for depression (Olfson et al., 2016), although psychotherapy is effective at improving depression (Cuijpers, 2017; Cuijpers et al., 2009) and preferred to pharmacological treatment by many older adults (McHugh et al., 2013; Raue et al., 2009).

New models of care are needed to address the unmet mental health needs of older adults. Given that the U.S. geriatric mental health workforce is too small to meet the needs of older adults, a landmark report from the Institute of Medicine (Institute of Medicine, 2012) identified nontraditional or non-specialist providers as viable solutions to help bridge the gap. Lay counselors may be a scalable approach to improve access to treatment for common mental health disorders, and lessons from low- and middle-income countries (LMICs) offer some insights to this approach by way of reverse innovation. The World Health Organization (WHO) is expanding access to mental health care in LMICs (Saxena et al., 2014) via task sharing—that is, by equipping individuals with little or no prior formal background in mental health to deliver brief, low-intensity psychological treatments (Singla et al., 2017). Lay counselors have been used to deliver low-intensity behavioral interventions to effectively treat (Singla et al., 2019) and prevent (Dias et al., 2019) depression in India and Nigeria (Gureje et al., 2019). However, lessons from LMICs may not apply to the United States because of differences in infrastructure across settings (Hoeft et al., 2018). Fortunately, evidence across settings, including high-income countries, is mounting. The largest demonstration is the English National Health Services’ Improving Access to Psychological Therapies (IAPT; (Clark, 2011; Wakefield et al., 2021) program. IAPT offers a stepped-care model to increase access to evidence-based psychotherapies across the country, in which patients receive the lowest intensity of services appropriate before “stepping up” to a higher level of intervention. For example, IAPT offers both low- and high-intensity services for mental health conditions. The low-intensity component offers brief structured services, including guided self-help, which include less clinician input than high-intensity services. Patients step up to high-intensity formal psychotherapy services (e.g., a full course of cognitive behavioral therapy [CBT]) if they do not respond to low-intensity services or have a disorder that is not appropriate for low-intensity services (e.g., high levels of functional impairment, higher severity of disorder). Perhaps most central to the success of this work are the Psychological Well-being Practitioners (PWP), who are non-specialists with some college or equivalent background who are trained in a structured intervention curriculum alongside supervised practice in low-intensity services. Such IAPT services have been beneficial to older and younger patients alike (Prina et al., 2014). While the empirical evidence for such a model in the United States is sparse, it too is growing. The first trial comparing lay providers to experts for anxiety or depression care in the United States found comparable outcomes between CBT delivered by bachelor’s-level lay counselors and PhD-level psychologists (Freshour et al., 2016; Stanley et al., 2014). More recently, older adults who received a brief behavioral intervention delivered by volunteer lay counselors at senior centers demonstrated improvement in depressive symptoms coupled with high levels of session attendance and satisfaction (Raue et al., 2019, 2021). Finally, behavioral activation (BA) delivered via telehealth by bachelor’s-level counselors resulted in statistically and clinically meaningful improvement in depressive symptoms among older adults (Choi et al., 2020).

Behavioral activation (BA) is an effective (Dimidjian et al., 2011) and widely used depression treatment that is appropriate for delivery by lay counselors (Richards et al., 2016). The theoretical underpinnings of this approach posit that depression arises and is maintained from a chronic reduction in positively reinforcing events (Martell et al., 2013). The resultant depressed mood leads to further avoidance and withdrawal from activities, creating a tailspin which increases inactivity, isolation, rumination, and low mood. Thus, treatment include a focus on activity monitoring and scheduling, thereby counteracting the depressive tendency to withdraw, be inactive, and isolate oneself (Lejuez et al., 2011). While retaining focus on the key therapeutic strategy of BA—activation—we simplified the intervention to better match the skill set of bachelor’s-level students and shortened the protocol from more than 10 to 20 sessions (Lejuez et al., 2011) to six brief (i.e., approximately 30-min) sessions. While effective intervention programs vary in length, we chose six sessions as consistent other similarly structured brief psychotherapies for depression in primary care and/or medical populations (Cully et al., 2017; Renn et al., 2018; Williams, Jr et al., 2000). We call this intervention “Do More, Feel Better” (DMFB). DMFB retains the following elements of BA for depression (Lejuez et al., 2011; Martell et al., 2013): 1) psychoeducation about depression; 2) introduction to treatment rationale; 3) selecting and planning enjoyable and important activities to engage with between sessions; 4) review and troubleshooting of weekly activities plan in subsequent sessions; and 5) relapse prevention. Other aspects of BA, including baseline self-monitoring, thorough review of values, mindfulness practice, and conceptualization of and therapeutic focus on behavioral avoidance (Lejuez et al., 2011; Martell et al., 2013) were not included in DMFB.

This streamlined intervention was piloted tested by trained senior center peer volunteers. It was effective and acceptable when delivered to older adults aged 60 years and older with depression (Raue et al., 2019, 2021). What is yet unknown is whether DMFB can be delivered by other lay counselors, such as bachelor’s-level interventionists, and whether such a task sharing approach is feasible, acceptable, and yields potential clinical improvement among older adults with depression. Further, there are uncertainties surrounding training and supervision needs and reliable methods to assure intervention fidelity and patient appropriateness and safety.

We present a training program and proof-of-concept trial inspired by the IAPT PWP workforce using undergraduate students as interventionists to deliver DMFB for community-dwelling older adults with depressive symptoms. The rationale for integrating such training into an undergraduate course is to leverage this potential workforce to improve access to services while diversifying the workforce pipeline. Psychology is one of the most popular undergraduate majors in the United States (Clay, 2017), but one that offers very few discipline-specific jobs at the bachelor’s level and often requires a graduate degree (Conroy et al., 2019). Such an undergraduate-level mental health workforce may be well positioned to expand services by offering low-intensity treatment in non-specialty settings such as primary care. Study 1 presents the feasibility of training undergraduate students to deliver this manualized, low-intensity intervention for depression, including recruitment, retention, and fidelity. Study 2 examined the feasibility, acceptability, and clinical targets of this program delivered to community-dwelling older adult participants with depressive symptoms.

Method

Study 1

Participants

This study was conducted in Seattle, WA. Participants were undergraduate students working toward their bachelor’s degree at the University of Washington who enrolled in a four-credit upper-level elective course on behavioral interventions for depression in primary care during the summer 2018 or 2019 quarters. The course was cross-listed in the psychology and social work departments. Students were 23.80 years old (SD = 5.71) on average and majority female (n = 13, 72.2%) and White (n = 11, 61.1%). Most were psychology majors (n = 16, 88.9%). See Table 1 for student characteristics.

Table 1.

Student Participant Demographics

	All Students (N = 18)	Coaches (N = 4)
Age, M (SD) years	23.80 (5.71)	25.25 (6.55)
Gender, n (%)
Female	13 (72.2)	4 (100)
Male	5 (27.8)	0
Ethnicity, n (%)
Hispanic/Latino	1 (0.06)	0 (0)
Race, n (%)
American Indian/Alaska Native	0 (0)	0 (0)
Asian/Asian American	2 (11.1)	0 (0)
Black/African American	2 (11.1)	1 (25.0)
White	11 (61.1)	2 (50.0)
Multiracial*	3 (16.7)	1 (25.0)
Major area of study, n (%)
Psychology	16 (88.9)	4 (100.0)
Other	2 (11.1)	0 (0)
Class rank, n (%)
Senior†	13 (72.2)	2 (50.0)
Junior‡	2 (11.1)	1 (25.0)
Other§	3 (16.7)	1 (25.0)

^*Students classified as “multiracial” encompassed racial identities of American Indian/Alaska Native, Asian/Asian American, Black/African American, and White. Coaches classified as “multiracial” identified as American Indian/Alaska Native, Asian/Asian American, and White.

^†Senior class standing is determined by having completed ≥ 135 college credits; in the United States, this is often the equivalent of a student in their final year (4^th or sometimes ≥ 5^th) of undergraduate studies.

^‡Junior class standing reflects the student having completed 90-134 credits; in the United States, this is reflective of an upperclass student and is often the equivalent of a student in the 3^rd of 4 years of undergraduate studies.

^§For all students, “other” class rank included sophomore standing (45-89 credits; typically, the equivalent of a 2^nd year university student); for coaches, this included a recent graduate.

Procedures

The 2018 course was team-taught by BNR, PAA, and PJR; the 2019 course taught by BNR and PJR. Each course consisted of nine weekly in-person class meetings, each approximately 2 hr in length. The course focused on an overview of depression, mental health treatment delivered in non-specialty service settings (i.e., integrated primary care models), basic clinical interviewing skills (e.g., empathic listening, structuring sessions), and training students in DMFB. Students typically spent the first hour in didactic classroom instruction and discussion and the second hour role-playing simulated patients played by either fellow students or instructors. Simulated cases were a mix of mock cases provided by the instructors and student-generated cases (e.g., based on the instruction to “Pretend to be a patient who presented to their primary care provider with depressive symptoms such as loss of enjoyment in usual activities, fatigue, and depressed mood, but no suicidal ideation.”).

Materials and Measures

Fidelity to DMFB was assessed in the seventh, eight, and ninth class sessions. As part of their homework, students were randomly paired with a different partner each week to conduct role plays of a full first DMFB session and the beginning of a follow-up session. Students took turns delivering DMFB and playing a simulated patient. These video-recorded role plays were submitted to course instructors for assessment. To rate student skill acquisition, instructors and trained graduate teaching assistants used a standardized fidelity tool that evaluated adherence to the DMFB protocol across seven specific skill domains and a final global score. Scores ranged from 0 (very poor) to 5 (very good); global scores of 3 (satisfactory) and higher were rated as passing.

Study 2

Participants

Student Coaches.

The first author invited previously trained students from the two classes in Study 1 to participate as coaches after the completion of class if they exemplified proficiency with the DMFB protocol, including interpersonal and communication skills, and expressed interest in pursuing postbaccalaureate mental health training. Students volunteered their time and were not compensated for their participation as coaches. The students who served as coaches were all female (n = 4, 100.0%) and identified as Black or African American (n = 1, 25.0%), White (n = 2, 50.0%), or multiracial (n = 1, 50.0%); see Table 1.

Patient Participants.

Electronic recruitment flyers were sent to senior centers, posted on community message groups, and directly emailed to research participant registries to recruit individuals in Seattle and the surrounding metropolitan areas. Eligibility requirements included the following: ≥ 55 years, English speaking, and PHQ-9 depression screener score of ≥ 10. Exclusion criteria included endorsement of passive or active suicidal ideation, presence of psychiatric diagnoses other than unipolar, non-psychotic depressive or anxiety disorders, severe or life-threatening medical illness (e.g., end stage organ failure), or concurrent receipt of psychotherapy. We did not exclude individuals concurrently receiving psychotropic medication for depression and/or anxiety. This study was approved by the University of Washington Institutional Review Board. Patient participants provided informed consent to participate in this study and were incentivized $25 for completion of the post-treatment assessment. Patient participants were on average 66.83 years old (SD = 10.39) and the majority identified as female (n = 11, 91.7%) and White (n = 8, 66.7%). Table 2 describes patient participant demographics.

Table 2.

Patient Participant Demographics

	Patient Participants (N = 12)
Age, M (SD) years	66.83 (10.39)
Gender, n (%)
Female	11 (91.7)
Male	1 (8.3)
Ethnicity, n (%)
Hispanic/Latino	1 (8.33)
Race, n (%)
American Indian/Alaska Native	1 (8.3)
Asian/Asian American	0 (0)
Black/African American	2 (16.7)
White	8 (66.7)
Multiracial*	1 (8.3)
Education, M (SD) years	15.45 (2.7)
Marital Status, n (%)
Never married	2 (16.7)
Married	5 (41.7)
Divorced	2 (16.7)
Widowed	3 (25.0)
Living Situation, n (%)
Alone	4 (33.3)
With others	8 (66.7)
Financial Status, n (%)
Can’t make ends meet	2 (16.7)
Have just enough to get along	3 (25.0)
Are comfortable	7 (58.3)
Psychotropic Medication Use†, n (%)
During trial	4 (33.3)
Ever	5 (41.7)

^*Multiracial patient participant identified as White and “other,” reporting Hispanic/Latino as racial identity in addition to ethnicity.

^†Including antidepressant medication, mood stabilizers, and anxiolytics.

Procedures

Training.

Prior to seeing patient participants, coaches met with the lead author for two 1-hr booster sessions to review the previously taught course material, orient to study procedures, and ask questions. During the study, coaches met for weekly 30-min DMFB supervision sessions with the lead author to review the previous session and course of treatment.

Assessment and Intervention.

This non-randomized, single-arm trial was conducted virtually during the initial months of the COVID-19 pandemic and resultant social distancing policies. Pre- and post-treatment assessments were conducted by trained research coordinators over the telephone or HIPAA-compliant Zoom teleconferencing software. The DMFB intervention entailed six weekly sessions conducted via Zoom teleconferencing or telephone. Session length was intended to be approximately 30-min to fit the typical constraints of non-specialty settings such as primary care; however, as this was a feasibility trial, sessions were allowed to flex between 30- to 45-min. Initial sessions often took longer (closer to 45-min) than follow-up sessions (typically 30-min) given the list of tasks to complete in the first meeting (details of session content in the following paragraph). Assessment and intervention sessions were conducted on weekdays based on patient and coach schedules. We allowed some flexibility (up to 3 months from Session 1) to complete the intervention to accommodate patient and interventionist schedules, preferences, illness, and unexpected life events.

After screening and baseline assessment, eligible patient participants were assigned to coaches based on coach and patient participant availability. A packet of materials including the consent form, coach and meeting details, and session forms were mailed to participants for reference prior to the first session. Each session followed a structured agenda to facilitate adherence to the protocol and time management. The agenda for the initial meeting consisted of (1) introducing the DMFB program and the coach’s role; (2) review of patient participant’s depression symptoms using the PHQ-9; (3) psychoeducation about depression and how the DMFB program works; (4) generating a list of pleasurable, meaningful, and/or rewarding activities, rated for their difficulty (easy, medium, hard); and (5) daily activity scheduling and action planning for the following week. Patient participants were provided with an action planning form and advised to keep a record of their activities and their satisfaction level each day (see supplemental materials for Form A, Activity List and Action Plan; for more information on session forms, refer to (Raue et al., 2019)). Sessions 2-5 focused on (1) review of depressive symptoms and any changes using the PHQ-9; (2) review of action plan and satisfaction level from the previous week, and (3) creation of a new action plan for the following week. The final session included a relapse prevention plan.

Measures

Feasibility.

Feasibility was assessed using guidelines from the National Institute of Health (National Center for Complementary and Integrative Health (NCCIH), 2021), including these targets: (a) can we recruit the target population?, (b) can we keep participants in the study?, (c) will participants do what they are asked to do?, and (d) can the treatment be delivered per protocol? Coach fidelity to DMFB delivery was evaluated at the first session for all patients and one randomly selected follow-up session per patient. Fidelity was rated according to the same standardized fidelity tool used in training; global scores ≥ 3 (“satisfactory” or above) were rated as passing. See supplemental materials for Form B, a sample fidelity rating form.

Acceptability.

We conceptualized acceptability of the DMFB program as satisfaction with the content, credibility, comfort, and delivery of services (Proctor et al., 2011). The Working Alliance Inventory—Short Revised (WAI-SR (Hatcher & Gillaspy, 2006)) assessed the patient participant’s perception of the strength of the relationship with their DMFB coach, including agreement on the goals and tasks of the program, and the development of a therapeutic bond. The 12 items (rated 1 = seldom; 5 = always) are summed for a total possible score between 12-60, with higher scores indicating stronger therapeutic alliance. Patient satisfaction was evaluated with three questions: “To what extent has the study therapist met your needs?” (1 = none of my needs have been met, 4 = almost all of my needs have been met); patient satisfaction with services received (1 = quite dissatisfied, 4 = very satisfied); and willingness to return to the same coach (1 = no, definitely not, 4 = yes, definitely).

Clinical Targets.

The Hamilton Rating Scale for Depression (HAM-D (Hamilton, 1960)) evaluated depressive symptoms at baseline and Week 6. The HAM-D was used to evaluate outcomes because using the same instrument for both eligibility and outcome may lead to early rapid decline in depression scores (Riordan & Avrumson, 2018). The HAM-D is a structured interview that considers both frequency and severity of psychological and somatic symptoms of depression. Scores are summed; increasing scores denote increased severity of symptoms (10-13 indicate mild depression, 14-17 moderate, and ≥17 moderate to severe depression). Treatment response was defined as a 50% reduction of the baseline HAM-D score. Remission was defined as the attainment of HAM-D ≤ 10 at the final assessment.

Disability was assessed at baseline and Week 6 using the 12-item World Health Organization’s Disability Assessment Schedule 2.0 (WHODAS 2.0). This is a generic health-status measure that assesses functioning across six domains: communication, mobility, self-care, interpersonal, life activities, and participation. Item scores (0 = no difficulty to 4 = extreme difficulty or cannot do) are summed, with higher scores representing greater disability. The WHODAS has been evaluated in the context of depressive disorders in primary care (Chwastiak & Von Korff, 2003) and used as an outcome in large-scale depression intervention trials (e.g., (Alexopoulos et al., 2020).

The target mechanism of behavioral activation—or, increased activity—was assessed at baseline and Week 6 using the Behavioral Activation for Depression Scale (BADS (Kanter et al., 2007)). The 25 items assess activation, avoidance or rumination, work/school impairment and social impairment. Responses range on a 7-point scale (0 = Not at all, 6 = Completely). For the total scale (possible score range: 0-150), higher scores represent increased activation.

Results

Study 1

Of the 20 students who enrolled over the two summer quarters, two dropped the class during the first week. A total of 18 upper-level undergraduate students completed the course. Of these, over half (62.5%) received a rating of “satisfactory” or better on their DMFB role plays by the seventh class. The majority (82.4%) were rated as competent (i.e., global rating of “satisfactory”) by the end of the course (ninth class session).

Study 2

Feasibility

Four students enrolled as coaches. We invited eight trained students to participate as coaches; three could not participate because of the study timeline (e.g., conflicted with upcoming study abroad, moving after graduation); one other declined for other scheduling and time commitment conflicts. All four coaches participated for the entirety of the study. They evidenced both 100% retention in the study and 100% attendance in weekly DMFB supervision meetings. Regarding coach fidelity to DMFB, 100% of sessions reviewed were rated as satisfactory, good, or very good. All patient participants were able to attend sessions via Zoom, either by using the telephone call-in option or videoconferencing. Only one patient participant elected to use the telephone-only option during the trial.

Acceptability

The majority (n = 10, 83.3%) of patient participants completed all six DMFB sessions (Figure 1). Each coach was assigned three patient participants (N = 12 patient participants total). From this sample, two patient participants withdrew: one after session 3 (declined to complete follow-up assessments) and one after session 4 (completed follow-up assessments). A total of 11 participants provided data for follow-up assessment. Patients reported a high level of working alliance, high satisfaction with the services received, and willingness to return to the same coach if they were to seek help again (see Table 3). Seven (63.7%) reported that “most” or “all” of their needs were met by the study therapist; the remainder (n = 4) reported “some” of their needs had been met.

Figure 1.

Patient Participant Flow through the Study

Table 3.

Patient Participant Outcomes

	M (SD)
Clinical Targets
Depressive symptoms (HAM-D)
Pre	18.09 (2.98)
Post	13.18 (4.64)
Disability (WHODAS 2.0)
Pre	16.91 (7.45)
Post	10.36 (8.13)
Behavioral activation (BADS)
Pre	73.18 (13.15)
Post	94.00 (23.77)
Acceptability
Working alliance (WAI-SR)	50.00 (8.14)
To what extent has the study therapist met your needs?	2.73 (0.65)
Patient satisfaction with services received	3.40 (0.84)
Willingness to return to the same coach	3.36 (1.03)

Note. Pre-treatment data for are reflective of the entire sample (N = 12). Outcome (“post”) data are based on the subset of 11 participants who completed follow-up assessment. HAM-D: Hamilton Rating Scale for Depression; higher scores denote increased severity of symptoms (10-13 indicate mild depression, 14-17 moderate, and ≥17 moderate to severe depression). WHODAS 2.0: World Health Organization Disability Assessment Schedule 2.0; higher scores represent greater disability. BADS: Behavioral Activation for Depression Scale; higher scores represent increased activation. These measures were collected at baseline (pre-treatment) and at Week 6 (post-treatment). Acceptability measures were collected post-treatment. Working alliance measured using the Working Alliance Inventory-Short Revised (WAI-SR) total score; possible ranges 12-60, with higher scores indicating stronger therapeutic alliance. To what extent has the study therapist met your needs?” (1 = none of my needs have been met, 4 = almost all of my needs have been met); patient satisfaction with services received (1 = quite dissatisfied, 4 = very satisfied); and willingness to return to the same coach (1 = no, definitely not, 4 = yes, definitely).

Clinical Targets

Patient participants demonstrated a 5-point reduction in HAM-D scores at 6-week follow-up (see Table 3). This improvement was statistically significant, t(10) = 3.53, p = .005, with a large effect size, Cohen’s d = 1.06. Of the 11 participants with follow-up assessment data, five (45.5%) responded to treatment (50% reduction of the baseline HAM-D score), while four (36.4%) achieved remission (HAM-D ≤ 10 at the final assessment).

Participants similarly evidenced improvement on disability, t(10) = 3.63, p = .005, with a large effect size, Cohen’s d = 1.09. We observed a significant increase in the target mechanism of increased activity over the 6-week treatment window as measured by the BADS, t(10) = −3.98, p = .003, Cohen’s d = 1.20.

Discussion

Our proof-of-concept demonstration provides preliminary evidence of feasibility for training bachelor’s-level coaches to deliver manualized, low-intensity treatment for depressive disorders. We successfully trained a majority of undergraduate students enrolled in a course focused on delivering our brief intervention. The coaches were highly engaged in the program and evidenced high fidelity to the DMFB protocol during the program, albeit with weekly individual supervision. We found acceptability of the intervention among our patient participants, and they exhibited clinically significant outcomes of improvement in depressive symptoms and functioning. Importantly, we noted an increase in our target mechanism of behavioral activation (i.e., increased activity) during the course of treatment, suggesting that DMFB may exert its influence on depressive outcomes through the intended pathway of increasing activation and reducing avoidance. This relationship between depression and the mechanism of increased activity has been documented in other research on behavioral activation delivered by professionals (Manos et al., 2010).

Despite a brief dose of treatment (six 30-to-45-min sessions), we had promising findings across acceptability and clinical targets. Almost half (45.5%) of our patient participants evidenced treatment response and 36.4% achieved remission. These findings compare favorably with remission other trials of lay counselor-delivered treatment for depression (Choi et al., 2020) as well as the remission rate across 51 trials of antidepressant medication vs placebo for older adults with depression (Kok et al., 2012). Our 5-point reduction in HAM-D scores over six sessions was statistically significant. A prior demonstration of DMFB delivered in person by trained volunteers to community-dwelling older adults in senior centers evidenced an 8-point reduction in HAM-D scores over 12 weekly sessions—twice the dose of our brief intervention (Raue et al., 2019). Further testing of volunteer-delivered DMFB in senior centers found similarly robust HAM-D improvements of approximately 7 points across 9 weeks of treatment (Raue et al., 2021). Strong support for brief models of treatment comes from ten years of practice-based evidence from the English IAPT program, in which the average treatment duration is 6.7 sessions and is associated with large effects for depression and anxiety outcomes (Wakefield et al., 2021). However, the durability of treatment effects from brief models of intervention over time is largely unknown. Our study did not assess long term outcomes among patient participants, and some data suggest that relapse after low-intensity treatment is common (Ali et al., 2017). Future research using fully powered randomized controlled trials is needed to establish the optimal dose of treatment, which may depend on setting.

These outcomes lend initial support for the potential of a non-specialist workforce to deliver brief, structured behavioral interventions for common mental health conditions such as depression and are responsive to the need for new models of care to expand access to geriatric mental health care. Larger scale work is needed to more robustly assess and address acceptability of both the brief intervention and the service delivery model to improve such access. Two patient participants discontinued participation in our trial (16.7% attrition), although one of these two participants agreed to follow-up assessment. This is consistent with the 19% attrition among older adults with anxiety assigned to receive CBT from a trained bachelor’s-level provider (Stanley et al., 2014), but higher than the attrition (5%) in a trial of BA delivered by lay counselors for homebound older adults (Choi et al., 2020). Four patient participants indicated that “some” of their needs were met by our brief, structured behaviorally-focused intervention. Structure of this program and other task sharing models will likely differ by setting (e.g., primary care versus community or specialty mental health); this study did not explore these differences. The English National Health Service supports IAPT across primary care and community settings. Given the majority of common mental health conditions in the United States are recognized and treated in primary care (Archer et al., 2012; Olfson et al., 2016), we designed DMFB for non-specialist delivery in such settings, rather than traditional mental health settings. Leveraging non-specialty settings such as primary care is intended to expand access to care in the very setting where most common mental health conditions are detected. Such integration into primary care also has the potential to reduce stigma associated with mental health treatment and maximize sparse specialist resources. Other U.S.-based studies have investigated models of brief, non-specialist-delivered interventions by training volunteers in senior centers (Raue et al., 2019, 2021) and case managers employed by aging services agencies (Choi et al., 2020). These are other potentially relevant models for improving access to services for older adults. We suggest brief lay-delivered intervention may be best considered as part of a stepped-care model, where low-intensity services such as DMFB are only one type of offering, alongside more longitudinal supports and other treatment options delivered by independently licensed professionals for more complex cases. Further work is needed to identify unique training and supervision needs to best support such a lay mental health workforce.

Our findings are notable given two significant historical events that occurred during the study—the global COVID-19 pandemic and the 2020 resurgence of the Black Lives Matter movement protests after the murder of George Floyd. Both were prominent sociocultural influences in Seattle during Spring and Summer 2020. Coaches pivoted in their use of DMFB to support patient participants in adapting to social distancing restrictions, managing anxieties, and navigating discussions of systemic racism, all in the context of planning and structuring activities. This highlights the flexibility of the DMFB protocol as a tool to support guided self-management of depressive symptoms and further suggests feasibility of such a program.

Limitations

Methodological limitations include a small sample and lack of comparison group; future studies will need to have adequate power and include control conditions to answer questions about treatment efficacy. Coaches received high intensity supervision (weekly individual meetings with a licensed clinical psychologist); it will be important to assess learning and supervision needs and test more scalable (e.g., group supervision) models in future iterations. One patient participant discontinued treatment after three sessions and was lost to follow-up; consequently, we do not have data on their reason for discontinuing and how their discontinuation relates to treatment feasibility or acceptability.

The composition of the study sample—for both coaches and patient participants—was predominately female. While this is consistent with other depression trials described herein (Choi et al., 2020; Raue et al., 2019), further work needs to attend to establishing feasibility and appropriateness of this treatment with men and gender non-binary individuals. However, we were able to recruit and retain a racially diverse sample of coach participants. Likewise, the racial and ethnic identity of patient participants this sample was reflective of the racial and ethnic composition of Seattle (67.3% White alone, 7.3% Black or African American, 6.9% multiracial, 6.7% Hispanic or Latino; (United States Census Bureau, 2021)), with the exception of an absence of Asian/Asian American participants. Other patient factors that may influence the results and limit generalizability include a majority with “comfortable” financial situations and attainment of a college degree, on average.

Our patient participants were required to have access to and be able to use a telephone or computer to conduct sessions via Zoom. The study was entirely remote given the need for social distancing during the COVID-19 pandemic. We took care to offer a telephone-only option to offset barriers associated with video conferencing (e.g., ownership of and ease using a computer, stable internet connection), which one patient participant utilized. Nonetheless, access to and comfort with technology may have posed a barrier and excluded individuals at greatest risk of not accessing mental health care and community services and supports.

Conclusion

Older adults encounter barriers to access mental health care that are compounded by the geriatric workforce shortage. The success of IAPT in England and elsewhere directly informed our vision to transform access to behavioral health care in the United States. Training bachelor’s-level students to deliver a brief behavioral intervention for depression is feasible. Low-intensity interventions such as DMFB may be effective at improving symptoms of depression when delivered by such non-specialists and may be one way to enhance the workforce equipped to provide evidence-based geriatric mental health services. However, restructuring mental health services is a complex endeavor and will require multilevel stakeholder commitment from policymakers, funders, healthcare system leaders, primary care organizations, guild and professional organizations, and other relevant service providers.