How to Triage Collaborative Care Patients for Specialty Psychiatry
How should clinicians triage primary care collaborative care patients with depression or anxiety who may need specialty psychiatric services?
Patients entering primary care collaborative care for depression or anxiety often can remain managed in that setting, but some will ultimately need escalation to specialty psychiatry. This guide applies to pharmacologic collaborative care programs like antidepressant monitoring in which intake symptom burden and selected comorbidities can help identify patients who may warrant closer follow-up or earlier referral consideration.
-
Confirm the patient is appropriate for primary care collaborative care
Use this workflow for patients with depression and/or anxiety being treated exclusively in primary care rather than specialty mental health clinics. In the study program, patients with severe mental illness such as bipolar affective disorder or psychotic disorders were not included because they required a higher level of service from the outset.
-
Collect baseline measurement-based symptom data at enrollment
At intake, measure depressive symptoms with the PHQ-9 and anxiety symptoms with the GAD-7. In this program, these instruments were part of routine collaborative care monitoring, and higher baseline PHQ-9 score was independently associated with later referral to specialty psychiatric services, with an odds ratio of 1.067 per point.
-
Screen specifically for sleep problems and substance use history
At enrollment, document whether sleep impairment is present and whether there is a current or past alcohol use disorder or cannabis use disorder. These baseline features remained independently associated with referral in the multivariable model: sleep issues OR 1.752, alcohol use disorder OR 1.702, and cannabis use disorder OR 2.070.
-
Flag higher-risk patients for closer stepped-care review
Treat patients with greater baseline depressive burden, sleep issues, alcohol use disorder, or cannabis use disorder as more likely to need escalation during collaborative care. The article supports using these baseline characteristics to inform treatment flow algorithms and to identify patients for closer monitoring, although it does not provide a validated cutoff or formal triage score.
-
Follow patients with scheduled collaborative care contacts
In this model, RN collaborative care managers contacted patients by telephone every 3 to 4 weeks and collected PHQ-9, GAD-7, medication adherence, and tolerability data. Patients were followed for a goal minimum of 6 months, with longer follow-up when medication changes or other clinical indications required additional monitoring.
-
Escalate to specialty psychiatry when improvement remains inadequate
Referral to a higher level of service can be made at any time during collaborative care at the discretion of the treating or consulting physicians. In this cohort, referral was driven overwhelmingly by persistent symptoms, with 121 of 131 referred patients, or 92.37%, still having at least moderate depression and/or anxiety symptoms at the time of referral.
Clinical Considerations
- These findings come from a retrospective secondary analysis and should not be treated as a validated prospective referral algorithm.
- The cohort consisted entirely of veterans in a VA pharmacologic collaborative care program, so generalizability to non-VA populations may be limited.
- Baseline anxiety severity differed between referred and nonreferred groups but did not remain significant in the multivariable model.
- Referral in this stepped-care model should not be interpreted as treatment failure, because escalation is part of appropriate population-based care.
Bottom Line
In primary care collaborative care for depression and anxiety, higher baseline PHQ-9 burden, sleep problems, alcohol use disorder, and cannabis use disorder should prompt closer monitoring and earlier consideration of specialty psychiatric escalation.