The VA is supporting development of collaborative care models for improving care for Veterans with mental health (MH) needs in primary care (PC). These models enhance collaboration between Veterans, primary care and mental health providers. However, better strategies are needed for sites to select, adapt, and implement an optimal set of these models.
Our objective was to use a collaborative process to implement the TIDES/PTSD Care Management (TPCM) model in two PC clinics utilizing Patient Aligned Care Teams (PACT), and further develop methods for implementing Veteran-centered care management that would facilitate spread of the VA's national tool kit for integration of MH services into PC.
An evidence-based, collaborative quality improvement process was used to update existing care management models used by PC-MH Integration (PC-MHI) teams operational in two different PACT clinics. Collaborative implementation activities included engaging clinician and administrative stakeholders from relevant services at multiple organization levels. Updates included modifying the existing TIDES Care Management (CM) model to be applicable for depressive disorders and PTSD. TPCM updates were made using VA software allowing for exportation to other VA sites. Evidence-based changes to this CM model included incorporating measurement-based care, improved screening, education, supportive care, and referral management. Specific types of activities included: (a) an Expert Panel to adapt pilot study findings, (b) clinic-specific local tailoring; (c) care manager and PACT staff training; and (d) ongoing site implementation support.
Evaluation questions addressed the feasibility of implementing the TPCM model, contextual factors influencing the implementation, and model acceptability. Potential influence on access to care, process of care, and clinical outcomes was explored. A site-level, comparative case study design was used. Data sources included clinical data collected in the usual process of care, implementation field notes, and stakeholder's interviews. A descriptive matrix analysis was used to synthesize quantitative/qualitative data on Veteran characteristics, extent of model use, process of care, clinical outcomes, staff experiences with implementation, and staff and patient satisfaction. The case studies provided formative feedback for adjusting the implementation and support for preparation of products.
Cross-site implementation activities found important for implementation included the Expert Panel, weekly Care Manager Training Calls, weekly Implementation Support Calls, and ongoing engagement with multidisciplinary stakeholders.
TCPM templates, tools and protocols in CPRS for Initial and Follow-up visits were comprehensively updated at one site (E). The other site (D) chose an incremental approach, adding a few PTSD-specific elements to the existing templates. Care Manager roles and activities differed locally. At Site E, the Care Managers maintained their previous, consultative role within the PC-MHI Team, remaining physically separate. At Site D, their role evolved along with the establishment of a PC-MHI Team within PACT, where they played a multi-faceted role in communication and collaboration. Relevant policies and procedures were adapted to fit each site's specific needs. Each team set up outreach and education activities specific to the existing clinical team meetings at their site, e.g. PACT, PC-MHI, Social Work, and Women's Clinic teams.
Quantitative data indicate that 85 patients (92% male) were engaged in ongoing TPCM. Of those, 83 completed the PHQ-9 with, 87% scoring > 10 indicating a likely depressive disorder. The mean PHQ-9 was 15.7 and the mean change in the PHQ-9 was a decrease of 3.71. The mean number of TPCM sessions per patients was 3.89. For those that completed 6 months of TPCM or had at least 6 care manager sessions (10 patients), 50% dropped their PHQ-9 score by > 50% and 50% achieved a PHQ-9 score of < 10 indicating a clinical response.
Of the 85 patients, 63 completed the PCL. From this sample 31 patients had a PCL score > 50 indicating a likely comorbid diagnosis of PTSD and 23 patients had a PCL score between 35 and 50 indicating sub-threshold PTSD. The mean number of TPCM sessions per patient was 3.82. For the course of TPCM, those patients with both a depressive disorder (PHQ-9 > 10) and a PCL > 35 (49 patients), 10 patients (20%) had a PCL score drop of at least 5 points indicating a significant clinical response; in addition, 9 patients (18%) had a drop of at least 10 points.
Qualitative data indicate that clinic staff and patients generally perceived TPCM to be valuable as part of a complementary set of care options. They noted that telephone contact improves access for patients who have various barriers to in-person visits. They also described a number of examples of care managers collaborating with providers for medication management, monitoring and adjustment of treatments, and providing patient education and support that helped improve outcomes.
Facilitators and barriers to implementation described were common to both sites and site-specific. Potential influences on local tailoring appeared related to individual leaders/champions, staffing and other resources, and local PC-MHI and PACT structures and processes. In general, interdisciplinary collaboration and partnerships for implementation helped tailor TPCM for the local PC-MHI contexts and overcome barriers.
Limitations: These results reflect the feasibility of the TPCM model at these sites, further work is needed to understand the transferability and impact on outcomes at other sites.
Veterans with depressive disorders and comorbid PTSD often present in PACT where treatment is challenging. This project evaluated the implementation of the TPCM model into PACT. Results demonstrated strategies for improving the delivery of evidence-based treatment, educating PACT providers about these disorders, and improving the collaboration between patients and PACT providers. Specific tools were developed for both patients and PACT providers to facilitate the delivery of TPCM. These tools and future reports will assist other sites in implementing strategies to care for this patient population. In addition, lessons learned will better inform the implementation of other treatment models into PACT.
- Hinton L, Apesoa-Varano EC, Unützer J, Dwight-Johnson M, Park M, Barker JC. A descriptive qualitative study of the roles of family members in older men's depression treatment from the perspectives of older men and primary care providers. International journal of geriatric psychiatry. 2015 May 1; 30(5):514-22.
- Dwight Johnson M, Apesoa-Varano C, Hay J, Unutzer J, Hinton L. Depression treatment preferences of older white and Mexican origin men. General hospital psychiatry. 2013 Jan 1; 35(1):59-65.
- Tuepker A, Allen E, Johnson M. Integrating Resident Education into PACT. [Cyberseminar]. 2013 Jul 17.
- Lam CA, Sherbourne CD, Rubenstein LV, Huynh AK, Lee ML, Chu K, Fickel JJ, Gelberg L, Farmer Coste MM. Outcomes of Minimal Depression Treatment and Follow-up Measures for Women Veterans. Poster session presented at: VA HSR&D / QUERI National Meeting; 2015 Jul 8; Philadelphia, PA.
- Fickel J, Bonner L, Chaney E, Davis T, Felker B, Jakupcak M. Stakeholder Partnerships in Quality Improvement of VA PC-MH Collaborative Care Management. Paper presented at: Southern California Dissemination, Implementation, and Improvement Science Annual Symposium; 2015 Mar 24; Los Angeles, Ca.
- Oishi SM, Garovoy N, Hyun JK, Canelo I, McCutcheon SJ, Strauss J. Primary care-mental health integration: Tailoring care arrangements for women. Poster session presented at: VA Women's Health Research Conference; 2014 Jul 31; Arlington, VA.
- Chang E, Wells K, Johnson MD, Fickel JJ, Jou K, Rubenstein LV, Young AS, Stockdale S. Achieving communication between primary care and mental health: why is it so difficult, even in the VA? A QI approach. Paper presented at: AcademyHealth Annual Research Meeting; 2013 Jun 23; Baltimore, MD.