Psychiatry and Primary Care: A Clinician’s Guide

Prioritize the Collaborative Care Model (CoCM): team-driven, measurement-guided integration that consistently improves depression and anxiety outcomes while controlling costs compared with standard referral pathways (see here). If you take one thing from this guide, take that: CoCM has the deepest evidence base of any psychiatry and primary care integration model, and it scales from a two-physician clinic to a large health system without needing to reinvent the wheel each time.

Here’s why it works when other integration efforts stall out: CoCM pairs a primary care provider with a behavioral health care manager and a psychiatric consultant who reviews a caseload registry on a set schedule, rather than seeing every patient face to face. That structure is what makes it affordable and sustainable, not just clinically sound.

Three things to do this week if you’re considering this path:

• Talk to your medical director or practice manager about designating a care manager, even part time, before you build anything else.
• Start measuring with a single validated tool. PHQ-9 for depression, GAD-7 for anxiety, repeated every visit.
• Check your billing codes. CoCM has its own CPT codes, and most practices leave revenue on the table by not using them.

A working CoCM team typically includes:

• A primary care provider who owns the treatment relationship and prescribes
• A behavioral health care manager who tracks the registry and checks in with patients
• A psychiatric consultant who reviews cases weekly, usually without seeing patients directly
• A shared electronic record so nothing lives in someone’s head or a sticky note

Key Takeaways

The Collaborative Care Model works because it pairs routine measurement with a defined team structure, and it outperforms looser integration efforts on both outcomes and cost.

Table of Contents

• Psychiatry and Primary Care Integration: The Main Models Explained
• What the Evidence Actually Shows for Collaborative Care Psychiatry
• Who Manages What: PCPs, Care Managers, and Psychiatric Consultants
• How Warm Handoffs and Case Review Actually Work Day to Day
• Screening Tools and Measurement-Based Care: PHQ-9 and GAD-7 in Practice
• Building a Collaborative Care Program: Staffing, Timeline, and Billing
• Where Integration Efforts Break Down, and What Actually Fixes It
• A Collaborative Care Vignette: Following One Patient Through the Model
• How Nortexpsychiatry Applies These Same Principles
• What We’ve Learned About Making Integration Actually Work
• Sources
• FAQ

Psychiatry and Primary Care Integration: The Main Models Explained

Not every “integrated” clinic is actually integrated. We’ve walked into practices that call themselves collaborative because a therapist has an office down the hall, and we’ve walked into practices with no on-site behavioral health staff at all that run tighter, more effective psychiatric support than the first group. The model matters less than the structure behind it.

Co-location means a behavioral health clinician and a primary care provider share a physical space, sometimes a building, sometimes a suite. Patients can be walked down the hall for an introduction. That’s it. Co-location does not require a shared chart, a shared treatment plan, or any routine communication between providers. It’s a real step up from cold referral, but on its own, co-location does not equal integration.

Consult-liaison psychiatry is the model most primary care providers already know from hospital settings, adapted for outpatient use. A psychiatrist is available for case-by-case consultation, usually initiated by the PCP for a specific question. There’s no registry, no population management, and no ongoing psychiatric oversight of a caseload. It works well for occasional complex cases but doesn’t scale to manage a clinic’s full mental health burden.

Integrated behavioral health (IBH) is a broader umbrella term. It describes any model where behavioral health services are woven into the primary care visit itself, sometimes through co-location, sometimes through embedded therapists, sometimes through structured psychiatric consultation. IBH is the category; CoCM is the specific, most rigorously tested version of it.

Collaborative Care Model (CoCM) is the one with the strongest research backing. It’s team-driven, population-focused, measurement-guided, and evidence-based, with a defined care manager role and a psychiatric consultant who works from a registry rather than a full patient panel. This is what lets one psychiatric consultant effectively support hundreds of patients across several primary care sites.

If you have limited behavioral health staffing and a Federally Qualified Health Center’s budget constraints, co-location plus a strong warm handoff protocol might be your realistic first step. If you have the ability to hire even a part-time care manager, CoCM outperforms every alternative on the outcomes that matter most.

What the Evidence Actually Shows for Collaborative Care Psychiatry

We get asked this constantly by colleagues weighing whether to invest in a care manager position: is the evidence for CoCM real, or is it consultant marketing dressed up as science? It’s real, and it’s been real for over two decades.

The IMPACT trial, one of the earliest and largest randomized controlled trials of collaborative care for depression in older adults, established the template most modern CoCM programs still follow: a care manager, a psychiatric consultant working from a registry, and measurement-guided treatment adjustments. That early work seeded decades of replication across different populations and settings.

A recent international systematic review of collaborative care programs found consistent symptom improvement for depression and anxiety across the majority of included studies, spanning work published between 2008 and 2024. The review also flagged something worth taking seriously: evidence for PTSD outcomes under collaborative care is far less consistent, and implementation features like stable funding, staff training, and registry use were the strongest predictors of which programs actually sustained their gains.

The strongest evidence sits with:

• Adult primary care populations with depression and anxiety
• Rural and VA settings, where psychiatric access is otherwise thin
• Programs that maintained fidelity to the four core CoCM elements over time

Where the evidence thins out:

• PTSD and more complex trauma presentations
• Severe mental illness requiring specialist-level ongoing management
• Programs that adopted the “collaborative care” label without the registry or measurement infrastructure behind it

The American Psychiatric Association’s own dissemination report treats caseload-focused psychiatric consultation and measurement-driven registries as implementation essentials, not optional extras. Additional randomized trials and meta-analyses indexed on PubMed reinforce the same pattern across independent research groups, which matters when you’re building a case to a skeptical CFO or board.

Who Manages What: PCPs, Care Managers, and Psychiatric Consultants

The most common integration failure we see isn’t a bad model. It’s ambiguous ownership. Someone assumes someone else is following up, and a patient falls through a gap that nobody built on purpose.

Here’s how the roles typically break down in a functioning collaborative care psychiatry program:

• Primary care provider: owns the overall treatment relationship, prescribes and adjusts medication with consultant input, and remains the patient’s first point of contact for most concerns.
• Behavioral health care manager: tracks the registry, checks in with patients on a defined schedule (often every 1-2 weeks initially), administers PHQ-9/GAD-7, and flags cases that aren’t improving for consultant review.
• Psychiatric consultant: reviews the caseload registry on a set cadence, usually weekly, recommends medication or treatment plan changes, and directly evaluates only the small subset of patients who need it.
• Behavioral health clinician (therapist): provides brief, structured psychotherapy where indicated, often working from the same registry as the care manager.

Referral thresholds matter more than most clinics realize, because unclear thresholds are what cause both over-referral (burning limited psychiatric capacity on cases primary care could manage) and under-referral (patients who needed specialist care getting stuck in a stepped-care loop too long). Escalate to direct psychiatric evaluation when you see:

1. Active suicidality or a recent suicide attempt
2. Psychotic symptoms of any kind, new onset or worsening
3. Treatment resistance after two adequate medication trials
4. Diagnostic uncertainty, particularly bipolar spectrum features masquerading as depression
5. Complex polypharmacy or a medical comorbidity complicating medication choice

Shared-care decisions, meaning cases where a PCP continues prescribing based on consultant recommendation, should be documented explicitly in the shared chart, including who initiated the change and why. That documentation habit does double duty: it protects patients, and it protects you medico-legally if a case is ever reviewed after an adverse outcome. Consent for information sharing between the primary care team and psychiatric consultant should be discussed with the patient at the outset, not assumed.

How Warm Handoffs and Case Review Actually Work Day to Day

A warm handoff is not just a friendly phrase for “we told the patient to call someone.” It’s a specific, scheduled action: introducing the patient to the behavioral health team member, or booking the follow-up appointment, before the patient leaves the building or ends the telehealth visit. Practices that operationalize scheduling at the point of care retain dramatically more patients than those relying on the patient to make a follow-up call from home.

A simple warm handoff script sounds something like this: “Your PHQ-9 today suggests moderate depression. I’d like you to meet Sarah, our care manager, right now for about ten minutes before you head out, and she’ll get you scheduled for a follow-up call next week.” No ambiguity, no homework assigned to a person who’s already struggling to get through the day.

The registry is the backbone underneath all of this. Someone, usually the care manager, has to own populating it every time a patient screens positive or enters the program, and outreach cadence needs a rule, not a hope. Weekly outreach for new or unstable patients, tapering to monthly once symptoms stabilize, is a reasonable default.

Case review meetings are where the psychiatric consultant earns their keep without seeing every patient. A workable structure:

• Weekly, 30 to 60 minutes depending on caseload size
• Care manager presents new patients and anyone not improving on the standard timeline
• Triage rule: patients with no improvement after two consecutive PHQ-9 or GAD-7 checks get prioritized for discussion first
• Consultant documents recommendations directly in the shared chart during the meeting, not after

Pro Tip: Keep case review meetings strictly time-boxed and cap the number of cases discussed per session. The moment these meetings run long or become open-ended clinical debates, care managers stop bringing cases forward, and that’s when patients quietly slip through.

Screening Tools and Measurement-Based Care: PHQ-9 and GAD-7 in Practice

You cannot manage what you don’t measure, and this is where a lot of well-intentioned integration efforts quietly underperform. The PHQ-9 takes about two minutes for a patient to complete and screens for depression severity across nine items tied directly to DSM criteria. The GAD-7 does the same for anxiety in seven items. Both are validated, brief, and reliable enough to use as the backbone of routine tracking rather than a one-time intake formality.

GAD-7 uses a similar structure, with scores of 10 or higher generally warranting active treatment and closer follow-up. Re-screening cadence should track severity: every two weeks for patients starting treatment or in crisis, stretching to every four to six weeks once someone stabilizes. Baking automated PHQ-9/GAD-7 prompts into your EHR intake workflow, tied to a nursing or medical assistant task rather than relying on the physician to remember, is one of the highest-yield, lowest-effort changes a clinic can make.

Building a Collaborative Care Program: Staffing, Timeline, and Billing

Most clinics overestimate what it takes to start and underestimate what it takes to sustain. A pragmatic staffing model for a mid-sized primary care practice looks like one full-time equivalent care manager for roughly 100 to 150 active patients in the registry, alongside a psychiatric consultant working a fraction of a day per week, often two to four hours, dedicated specifically to caseload review rather than direct patient visits.

A realistic rollout runs three to twelve months depending on your starting infrastructure, and clinics that skip the pilot phase to “go live” everywhere at once tend to stumble on exactly the workflow kinks a small pilot would have surfaced cheaply.

Telepsychiatry deserves specific mention here because it changes the math for smaller or rural practices. A psychiatric consultant doesn’t need to be on-site, or even in the same state in many cases, to review a registry and join a weekly case conference by video. This has let telepsychiatry consultation and mentoring programs extend psychiatric expertise into primary care sites that could never otherwise recruit a consultant, and it’s worth building your telehealth workflow before you assume you need to hire locally.

On billing: CoCM has its own dedicated CPT codes designed specifically to reimburse the care manager and consultant time that doesn’t fit a traditional office visit. Documentation needs to capture time spent by the care manager, the psychiatric consultant’s caseload review, and the specific patient’s inclusion in the registry, since payers generally require evidence the four core CoCM elements are actually in place, not just claimed. Understanding your state’s Medicaid spending priorities and payer mix early in planning will save you from building a program your payer landscape can’t sustain financially. National utilization data on mental health visit volume can also help estimate expected registry size before you commit to a staffing ratio.

Where Integration Efforts Break Down, and What Actually Fixes It

The barriers are predictable, and so are the fixes, if you’re honest about them early.

• No time for screening: build PHQ-9/GAD-7 into the medical assistant intake, not the physician’s already-full visit.
• EHR doesn’t support a registry: many EHRs have a basic registry function that goes unused simply because nobody configured it; ask your IT team before assuming you need new software.
• Reimbursement uncertainty: start billing CoCM codes on a small pilot cohort first, so any documentation gaps surface before you scale.
• Clinician skepticism: bring one or two PCPs into the pilot design early rather than mandating the workflow from above.
• Staffing an FTE feels expensive: a part-time care manager covering a smaller patient panel is a legitimate starting point, not a compromise.

Pro Tip: If you can only fix one thing first, fix the registry. Everything else, warm handoffs, case review, measurement cadence, depends on someone being able to see who’s in the program and where they stand.

Cultural competence deserves a direct mention rather than an afterthought: care managers and consultants working across diverse patient populations need training on how symptom presentation, stigma, and help-seeking behavior vary by community, and registries should be reviewed periodically for disparities in who’s engaging with treatment versus who’s screening positive and disappearing from follow-up.

A Collaborative Care Vignette: Following One Patient Through the Model

A 42-year-old patient comes in for a routine physical and screens positive on the PHQ-9 with a score of 14, moderate depression, something the PCP would have easily missed in a fifteen-minute visit focused on her diabetes management.

• Day 0: PCP reviews the score, does a warm handoff to the care manager on-site, who schedules a follow-up call for the next week.
• Week 2: Care manager checks in by phone, reviews symptoms, confirms the PCP started a low-dose SSRI, logs the case in the registry.
• Week 6: PHQ-9 re-screen shows minimal improvement, score still at 12. Care manager flags the case for the psychiatric consultant’s weekly review.
• Month 3: Consultant recommends increasing the SSRI dose rather than switching medications, given partial response, and suggests adding brief structured psychotherapy through the practice’s behavioral health clinician.

By month four, her PHQ-9 drops to 6. She never once needed a direct psychiatric evaluation, and her PCP managed the entire episode with consultant guidance in the background.

How Nortexpsychiatry Applies These Same Principles

We didn’t invent collaborative care, but we’ve built our own workflow around the same core elements the evidence points to. At Nortexpsychiatry, that means routine measurement at nearly every visit, not just intake, so medication decisions are grounded in tracked scores rather than a general impression of “doing better.” It means telehealth isn’t an afterthought bolted onto in-person care. It’s a full option for patients across Allen, Frisco, McKinney, and Plano who need psychiatric follow-up without taking half a day off work.

When a patient’s depression or anxiety score isn’t moving after a reasonable trial, that’s the signal to reconsider the plan, whether that means adjusting medication, adding a therapy referral, or exploring options like TMS or Spravato for cases that haven’t responded to standard approaches. Waiting too long to reassess is one of the more common ways good treatment plans lose momentum.

A few practical habits we’d recommend borrowing regardless of your setting:

• Schedule the follow-up before the patient leaves the visit, not after
• Use telehealth triage to reserve in-person slots for patients who genuinely need them
• Give patients plain-language education about what their screening scores actually mean

If you’re evaluating whether it’s time for a more thorough psychiatric evaluation for yourself or a family member, our guide on when to seek psychiatric care walks through the practical signs worth acting on.

What We’ve Learned About Making Integration Actually Work

If we had to boil down what separates a collaborative care program that lasts from one that quietly fades, it comes down to three things: measurement, communication, and leadership buy-in. Measurement gives you an honest signal instead of a guess. Communication, meaning an actual shared chart and a real case review rhythm, keeps that signal from getting lost between providers. And leadership buy-in is what protects the care manager position when budget season gets tight, because it will.

The pitfalls we’d flag first: skipping the pilot phase, treating the registry as optional, and assuming co-location alone counts as integration. Track two outcomes from day one, PHQ-9/GAD-7 trends and time-to-follow-up after a positive screen, and you’ll know within a few months whether your model is actually working or just looks good on paper.

Start smaller than feels ambitious. A focused pilot with fifteen patients teaches you more than a program-wide rollout ever will.

Sources

• Learn About the Collaborative Care Model
• Healing the Whole: An International Review of the Collaborative Care Model between Primary Care and Psychiatry

FAQ

Can a psychiatrist be a primary care doctor?

A psychiatrist is a physician who can technically manage some general medical concerns, but psychiatry is a distinct specialty focused on mental health diagnosis and treatment, not a substitute for primary care. Most patients need both a primary care provider and, when indicated, psychiatric care coordinated between the two.

Do I need a primary care doctor to see a psychiatrist?

No, you don’t strictly need a referral from a primary care doctor to see a psychiatrist in most settings, though many insurance plans and integrated care models work best when the two providers communicate. Within a Collaborative Care Model, the primary care visit is often where screening happens first, which naturally leads to psychiatric involvement when needed.

What is the 3 month rule in mental health?

There’s no single universally recognized “3 month rule” in mental health diagnosis; the concept usually refers to symptom duration thresholds used in specific diagnostic criteria, such as adjustment disorders resolving within roughly six months of the stressor ending. Definitions vary by diagnosis, so it’s best to discuss the specific criteria relevant to your situation with a psychiatric provider.

What are the most serious mental disorders?

Conditions generally considered the most severe include schizophrenia, bipolar I disorder, major depressive disorder with psychotic features, and severe treatment-resistant depression, largely because of their impact on functioning and elevated safety risks. These are exactly the presentations where primary care should escalate quickly to direct psychiatric evaluation rather than manage through stepped care alone.

When should primary care escalate to psychiatry instead of managing treatment directly?

Escalate for active suicidality, new or worsening psychotic symptoms, treatment resistance after two adequate medication trials, or significant diagnostic uncertainty. Stable, mild to moderate depression and anxiety generally respond well to primary care management with psychiatric consultant support through a collaborative care model.

Recommended

• Why Seek Psychiatric Care: a Guide for Adults
• Starting psychiatric treatment: a step-by-step guide for adults