Depression often remains undetected despite its high prevalence. From 2005-2008 Henry Ford developed, tested, and refined a model for integrating depression screening and treatment into primary care clinics. Pilot study outcomes resulted in a 23% detection rate for depression (220/974 patients). Clinical action to positive screenings (treatment or referral) was 90%, and patient refusal for screening was only 1%. The practice model utilized a nurse practitioner experienced in Behavioral Health to serve as coach to the primary care clinic team for implementing depression screening processes and tools. Medical assistants determined patients’ screening eligibility and administered screening tests, and physicians used test results to drive discussion with patients for diagnosis and treatment plans. The Behavioral Health nurse served to reinforce clinical screening and treatment guidelines and also provide psychiatric treatment. Henry Ford integrated two depression screening tools, as well as evidence-based treatment guidelines, into the electronic medical record system for availability to all Henry Ford clinicians.
Integrating Depression Screening and Treatment in Primary Care Clinics
To improve the detection and appropriate treatment of clinical depression in Primary Care using an integrated, collaborative depression care model.
In the first month we tracked weekly the number of eligible patients screened (chart audit of all filled appointments).
- Number of patients screening positive for depression (DST > 10).
- Disposition of those screening positive (#/% receiving antidepressant treatment versus referred to Behavioral Health Services versus patient refusal, etc.)
- Clinical outcomes of antidepressant treatment (mean DST change score, % with 50% reduction in DST score, % with DST < 5).
- Impact on comorbid chronic disease clinical parameters (e.g., pre-post change in LDL and HgA1C).
Embedded a nurse practitioner experienced in Behavioral Health in the Primary Care clinic to serve as a depression coach/mentor in a team approach.
Trained the team how to use a standardized screening tool for depression (e.g., DST adapted from the PHQ-9) and evidenced-based treatment guidelines. Convened a team that included representation from all clinic roles (e.g., clinic service representative, medical assistant, RN, MD) to help develop a clinic process for depression care that would be sustainable.
Developed screening criteria for high-risk patients.
Moved from a paper process to the DST embedded within the electronic medical record (EMR) system.
Moved to a two-step screening process, using the PHQ-2 followed by the full DST when the PHQ-2 was positive.
Collaborated with Information Technology to build two new note types in the EMR to document the PHQ-2 screening, and worked to build electronic reports to track monthly depression screening rates.
Pilot results indicated that approximately 50% of eligible patients were screened although initial screening criteria were too broad, resulting in over 300 patients being screened in the first month with a 21% detection rate. This led to the process being amended in two ways: 1) screening criteria were streamlined to patients with either CAD, HF, DM, or complaints of depression; and 2) the PHQ-2 was added, with only those positive on the two-question survey going on to complete the full DST. After adding the PHQ-2, detection with the DST rose dramatically to approximately 60% while significantly reducing the time burden on clinical staff. The integrated model led to Primary Care physicians addressing detected depression > 90% of the time, with initiation of an antidepressant in about 70% of the cases.
Clinical outcomes were impressive, with a group mean change in DST score from 15 at baseline to 7 at 24-week follow-up (50% reduction indicates treatment response and a DST score of < 5 indicates remission).
Further subanalysis of patients with comorbid DM showed that for patients whose depression was successfully treated (DST reduction by at least 50%), 65% showed improvement in HgA1C with a group mean pre-post drop in HgA1c of 1.01 (a 1 point drop in HgA1c has been shown in the literature to significantly reduce diabetes complications).
Next Steps/Lessons Learned
- Produce electronic monthly reports to track PHQ-2 frequency by site and provider; can assess usage trends over time; email these reports to administrators at the pilot sites monthly.
- Booster (training) sessions at the initial pilot sites at set time intervals to re-review the process.
- Periodic case conference at sites that have implemented this process to promote continuous learning and collaboration with Behavioral Health Services.
- Use of the DICE tool with the design team at the beginning was critical to assess the perceived importance of the initiative and likelihood of project success.
- Involving the clinical staff (end users) was critical to help build the process.
- Use of multiple PDCA cycles to test the processes developed, and to revise what was not working well, helped the team to make changes quickly as needed.
- Giving clinical staff frequent data reports helped each to know how they were performing and what positive impact the new process (and their actions) had on patients.
- Plan to roll out the program across the Henry Ford Medical Group over time.
Contact Information: Terri Robertson, PhD, Project Manager, Center for Clinical Care Design, Henry Ford Health System, 313-874-2495.
Keywords: access, behavioral health, care team, chronic care, Detroit, depression, evidence-based medicine, Henry Ford Medical Center, health information technology, health screening, Michigan, MyHealth web portal, patient support, patient web portal, safety, treatment, value
FOR MORE INFORMATION: