Comparative effectiveness of clinician versus peer supported problem-solving therapy for rural older adults with depression

Abstract

Objectives

Self-guided and peer-supported treatments for depression may address barriers to treatment common among rural older adults. This pilot study compared the effect on depression of peer-supported, Self-Guided Problem-Solving Therapy (SG-PST) to Case Management with Problem Solving Therapy (CM-PST) among older adults in rural California.

Methods

105 older adults with depression received an introductory PST session with a clinician followed by 11 sessions of CM-PST with a clinician (n=85) or SG-PST with a peer counselor (n=20).

Results

Both interventions resulted in clinically significant improvement in depression by week 12. Depression in the CM-PST group dropped by 4 points more than the SG-PST group between baseline and 12 weeks (4.1 point difference, 95% CI 0.99 to 7.22 t(332)=2.59, p<0.001, Hedges’ g=1.08).

Conclusions

Our results suggest that peer-supported SG-PST is a viable, acceptable option for rural older adults with depression as a second line of treatment if access to clinicians is limited.

Introduction

Evidence-based psychotherapies for depression are underutilized by older adults (1)(2). Older adults comprise about 20% of rural populations (3), where barriers to accessing care include a lack of mental health professionals, transportation, and stigma (4, 5). Self-guided and peer-assisted adaptations of evidence-based treatments have been developed to overcome some of these barriers. The Improving Access to Psychological Therapies program, a lay-delivered behavioral health intervention in the UK, has shown efficacy and acceptability (6). Similar studies in the US are limited in scale and evidence-base (7–10), and none have been conducted in rural settings.

Problem-Solving Therapy (PST) is an evidence-based psychotherapy that can potentially be adapted to a self-guided, peer-assisted format to address access problems and stigma in rural communities. A straightforward therapy approach that teaches people to generate realistic solutions for life problems that contribute to depression, PST has demonstrated effectiveness in older adults across a range of settings and clinical presentations and can be successfully administered by a variety of community providers (e.g., social workers, case managers, and promotoras) (11). Although not yet evaluated as a form of bibliotherapy, a short booklet called the Participant Guide to Using PST has been successfully utilized by clinicians (12). Case management (CM) is the process of connecting older adults with needed social services and assistance, such as health insurance and home-delivered meals. CM has been combined with clinician-delivered PST to offer support for a fuller range of problems experienced by older adults (13).

Self-guided therapy (also referred to as self-help, self-instruction, self-care, or psychoeducation) offers a structured model for treatment that is adopted and adhered to by the individual with little or no clinical contact (9). Self-guided psychotherapy treatment in older adults has positive effects on depression, especially when combined with some form of support (9).

Growing literature supports the use of non-traditional, non-specialist providers in delivering low-level behavioral health interventions and/or providing support for self-guided interventions (6–8, 10). One type of such providers, Senior Peer Counselors (SPCs), offer a viable and effective alternative to services provided by mental health professionals (8, 10), and are an essential resource within rural communities. SPCs are typically members of the patient’s own community and have personal experience with mental illness. Peer counseling is similar to “befriending”, as both offer companionship and reciprocal sharing of personal information (14), However, SPCs are also trained to provide supports to other older adults with depression.

The Senior Peer Alliance for Rural Research on Wellness (SPARROW) project (13) included a pilot intervention which adapted the Participant Guide to Using PST to serve as the basis for a peer-supported, self-guided intervention. The aim of this pilot was to compare treatment outcomes and adherence to treatment of an SPC-supported, Self-Guided PST (SG-PST) to a clinician-supported CM with PST (CM-PST). The objectives of the pilot were to establish the feasibility of SG-PST and provide direction for future research and implementation on a larger scale.

Methods

The SPARROW Project was a collaboration between UCSF and Catholic Charities Diocese of Stockton (CCDS), a community agency providing behavioral health services to residents of three rural counties in California: Tuolomne, Calaveras, and Stanislaus. CCDS recruited and trained clinicians in CM-PST, as well as volunteers aged 60 and over to become SPCs and learn to support SG-PST.

Clinicians were master’s-level therapists or social workers, either licensed or in the process of licensure. Training for clinicians in the CM-PST model consisted of an 8-hour workshop, six 1.5-hour sessions of group role play practice, and weekly hour-long case review sessions with the study’s clinical supervisor (RC). Full details of CM-PST training have been described elsewhere (13).

SPARROW SPC were chosen based on observed compassion, empathy, respect, and leadership ability. As with their traditional SPC role, SPARROW SPCs were not compensated for their time, but could be reimbursed for mileage. CCDS training for their SPC program consisted of 20 hours of training on basic counseling interventions; issues of depression, loneliness, grief, and suicide; substance use and medication abuse issues; mental health issues in older adults; elder abuse and neglect; confidentiality and mandatory reporting; and available community resources. Additional training in the SG-PST model consisted of a 4-hour workshop followed by 2 hours of individual role-play. After training, SPCs attended weekly case review meetings with the study’s clinical supervisor, and an annual booster training to assure continued adherence to the SG-PST protocol.

SPCs were trained in using guided self-help methods to support participants in applying the PST model, including: reinforcing the rationale for PST, identifying key problems and goals, encouraging the use of the Participantś Guide to Using PST, supporting efforts to change, and using assessment and outcome measures to review progress and assess the need for further help from the clinician. All SPCs were supervised by a clinical case manager and supervisor, who were responsible for monitoring the participant’s wellbeing and response to treatment. SPCs participated in role-plays with the study’s clinical supervisor to practice supporting SG-PST when participants forgot to use the materials, had trouble solving problems, failed to enact action plans, or displayed increasing cognitive impairment or suicidal ideation. Over the course of the study, 9 trained SPCs served 1–2 clients each. SPCs were mostly women (7/9). All sessions were recorded for treatment fidelity monitoring by the study supervisor and an external reviewer.

This study used a randomized experimental design. Participants were community dwelling rural older adults (60+) with depression (Hamilton Depression Rating Scale (HAM-D) score ≥ 20). Individuals were deemed ineligible if they had any of the following: psychotic depression, high suicide risk, presence of any Axis I psychiatric disorder or substance abuse other than unipolar major depression, Telephone Cognitive Screen (T-CogS) score below 24, acute and/or severe medical illness, or use of drugs causing depression. Participants could receive co-occurring treatment, but not if they had begun psychotherapy or changed antidepressant medication or dosage within 3 months. Further details of recruitment and inclusion/exclusion criteria have been described elsewhere (13).

Participants were recruited through community organizations, and received an introductory session with a clinician followed by 11 sessions of either CM-PST or SG-PST delivered in their home. Screenings and assessments were conducted via telephone by a researcher blinded to their treatment group. Participants received $20 gift cards for each comprehensive assessment (baseline, 12, and 24 weeks). Study procedures were approved by UCSF’s institutional review board and participants provided written informed consent. The HAM-D was used to assess depression at baseline and at 3, 6, 9, 12, and 24 weeks (15). Participants’ data was eligible for analysis only if they had attended at least 10 of the 12 planned sessions.

Mixed effects linear regression with random intercepts and slopes (and an unstructured covariance matrix) was used for all analyses with degrees of freedom determined by the Kenward-Roger method. For the analysis of CM-PST alone, the primary comparisons were changes over baseline. For the analysis of CM-PST and SG-PST, the primary comparisons are differences between the groups in change over baseline (tested using the interaction of group assignment and time). Estimated values and differences are based on least squares means and comparisons between least squares means; we report a small-sample-size-corrected Hedge’s g for our primary comparison of interest.

Results

SPARROW screened 205 individuals and consented and randomized 105 participants that met study criteria. Using 4:1 randomization procedure, 85 were randomized to CM-PST and 20 to SG-PST, given the limited capacity of the SPC program. Of these, 77 (91%) and 12 (60%) completed 12-week assessment, respectively. Participants were on average 70 years old with 14 years of education. Most participants were female (n=84, 80%), non-Hispanic (n=96, 91%), and Caucasian (n=98, 93%). Intervention groups did not differ significantly in age, gender, ethnicity, race, or baseline depression scores.

Depression scores in both CM-PST and SG-PST groups declined significantly over the course of the intervention (baseline-Week 12) and were maintained until Week 24 follow-up (Figure 1). The reduction in depression scores from baseline to Week 12 was significantly more pronounced in the CM-PST vs SG-PST group (4.1 point larger drop in CM-PST, 95% CI 0.99 to 7.22 t(332)=2.59, p<0.001, Hedges’ g=1.08), suggesting an advantage of CM-PST over SG-PST. However, between Week 12 and Week 24, depression scores in the CM-PST group increased slightly, whereas scores in the SG-PST group continued to decline.

Figure 1.

Hamilton depression rating scale (HAM-D) outcomes for SPARROW CM-PST participants vs. SPARROW SG-PST participants

Possible scores range from 0 to 50, with higher scores indicating more severe depression. The scoring ranges are 0–7 no depressive symptoms, 8–16 mild depression, 17–23 moderate depression, and over 24 severe depression.

Response was defined as a 50% or more decrease in depression scores between baseline and week 12. Remission was defined as a depression score $\le$10 at week 12. Both response and remission rates were higher in the CM-PST compared to the SG-PST groups. In the CM-PST group, 69.4% (59/85) of participants had more than 50% decrease in depression scores at week 12 compared to 45% (9/20) in the SG-PST group, although this was not a statistically significant difference. By Week 12, mean depression score in the CM-PST group was 7, compared to 11 in the SG-PST group.

Discussion

While improvements in depression were significant after 12 weeks in both treatment groups, improvements were more pronounced at Week 12 and Week 24 in the CM-PST group. Consequently, CM-PST with a clinician should be considered a preferred treatment option when such resources are available. However, this study demonstrated that SG-PST supported by SPCs may be an appropriate and acceptable alternative to CM-PST (or other evidence-based interventions) for rural older adults facing barriers to access as a result of stigma, poor transportation, or a lack of clinicians/services.

Regarding the sustained impact of treatment, our results show that at 24-weeks, SG-PST participants reported continued improvement in depression compared to the slightly higher rates of depression reported by CM-PST participants. This could imply a longer-lasting impact of the SG-PST intervention given its focus on empowering the patient to address their own problems. It could also be a result of the low sample size or completion rate in this arm of the intervention.

Given the lower completion rate in the SG-PST group, we recommend further research to determine if and why SG-PST may be less acceptable than CM-PST. Further replication of SG-PST will be dependent on the capacity of existing SPC programs. Although CCDS provides services to all residents across religious categories, the generalizability of our results could be limited given their perceived religious affiliation. Further research is needed to establish an informed triaging of clients into CM-PST or SG-PST. Naturally, we would recommend a higher level of supervision and/or intervention for older adults with more severe symptoms of depression or risk factors for self-harm.

Conclusion

Our results suggest that peer-supported SG-PST is a viable, acceptable option for rural older adults with depression, especially when supervised by professional clinicians. Further research could combine clinical and peer-supported interventions, and triage to the appropriate level of care. Effective use of SPCs to support evidence-based treatments has the potential to maintain the independence of rural older adults by improving access to mental health services using existing infrastructure and minimal community resources.

Highlights:

• PST can be adapted into a peer-supported, self-guided treatment model (SG-PST).

• SG-PST was effective at reducing depression after 12 weeks of treatment, but not as effective as case management with PST (CM-PST) with a clinician.

• SG-PST with a volunteer peer counselor may be an effective treatment option for older adults with depression in rural communities, where workforce and other resources can be scarce.