Trauma in Primary Care: When to Refer, When to Stay, and How to Step Up

This is part four in the Trauma in Primary Care series.

There is a particular moment that most behavioral health consultants recognize. A patient begins to disclose something significant—a history of childhood abuse, a recent assault, something they have not said aloud before—and the BHC feels the floor shift. The visit is already twelve minutes in. There is no promise of a next appointment. The clinical instinct that fires in that moment, for many of us, is some version of this belongs somewhere else.

That instinct deserves scrutiny, because what it usually encodes is not a clinical judgment but a question about whose discomfort is being managed. A referral is not a neutral act. For many patients, it is not a transfer of care—it is the end of care, because the system they are being sent to will not hold them. The referral question—when to stay in primary care with a trauma patient, when to step up to specialty care, and what staying actually looks like when a referral is in motion—is the subject of this fourth post in the series. The first three posts established the clinical foundation: why most trauma-history patients do not require specialty referral the wrong tool for the wound, how to build safety and relational ground in a first contact first contact and safety, and how stabilization skills prepare patients for the unorganized exposure they are already experiencing in daily life stabilization as phase-one trauma treatment. This post takes on what those earlier posts deferred.

Specialty Mental Health Is Different Care, Not Better Care

The assumption built into most referral decisions is hierarchical. The thinking runs: this patient is complex, primary care handles mild to moderate concerns, therefore she needs something more than I can offer. That assumption about mild to moderate presentations has been challenged directly in the integrated care literature. Bridget Beachy and David Bauman have named it explicitly as a myth—one that mischaracterizes both the scope of primary care generally and the capacity of BHCs specifically, and that licenses a kind of pre-emptive abdication before any clinical work has been attempted. Reiter, Dobmeyer, and Hunter are explicit in the PCBH model’s own definition: BHCs engage with mental health and substance use concerns “of any severity and chronicity,” and patients are not screened out on an a priori basis based on presentation complexity.

The same principle governs the rest of primary care, and the ethical and risk-management literature makes it explicit. The AMA Code of Medical Ethics holds that referral is adjunctive—the PCP retains fiduciary responsibility for the patient’s care regardless of specialist involvement, and may not discharge a patient simply because their condition has become more severe or more complex . The threshold for changing that arrangement is competence and safety: whether the clinician can still provide guideline-concordant care within their scope, not whether the patient’s numbers look alarming . A PCP managing a patient with an A1c of eleven does not transfer that patient to endocrinology and step back. She intensifies management, involves the specialist when specific clinical questions require it, and continues coordinating care. The behavioral health argument follows exactly the same logic, and Neftali Serrano , writing from a decade of building PCBH programs that serve populations with serious mental illness, arrives at the same conclusion: the question of what is too severe for primary care has one answer, and it is nothing. The more important question, Serrano argues, is one every BHC should ask before placing a referral: who is this referral truly for—the patient, or the clinician?

The more defensible frame is that specialty mental health is a different kind of care, not a higher one . There is no standardized definition of clinical complexity, and the definition varies so widely across clinicians, clinics, and systems that it functions more as a rationalization than a clinical criterion, producing wide variation in care and the systematic denial of access to patients who are told, in one form or another, that they are too much for the setting they came to . Primary care has genuine clinical advantages that specialty settings often lack: a multidisciplinary team, longitudinal knowledge of the patient across years of medical contact, access to family and social context, and the structural capacity to re-engage episodically without the formality of a new intake. The BHC who has seen a patient through a diabetes diagnosis, a pregnancy, and a period of job loss knows something about him that a specialty therapist meeting him for the first time does not.

What specialty care offers in exchange is genuinely worth describing accurately, because a post arguing that primary care can handle more than it is typically given credit for does not serve anyone if it minimizes what specialty care actually does well. The dedicated mental health appointment—an hour uncontested by medication refills, lab results, or three other active problems—creates conditions for a kind of sustained focus that primary care visits rarely achieve. For patients who are ready for it and will actually attend, longer-duration work creates space for processing that fifteen to thirty-minute visits, however skillful, cannot fully replicate. Some SMH settings bring psychiatry, case management, peer support specialists, and structured group programming under one roof in ways that primary care cannot match. Highly structured, multi-session protocols—full courses of Cognitive Processing Therapy or Prolonged Exposure that unfold across eight to twelve planned appointments—require a treatment architecture that the episodic PCBH visit is not designed to provide. And for some patients, the separateness of a specialty mental health setting carries its own therapeutic meaning: going to that appointment is itself an act of prioritizing their mental health, distinct from everything else that happens at the primary care clinic .

These are real advantages. The problem is not that specialty care offers them. The problem is the assumption that these advantages are what most patients referred from primary care actually need and will actually receive. The referral does not give the patient access to something categorically different. It gives her access to a different structure for receiving largely similar treatment, in a setting she may not engage with.

Reiter documents the follow-through problem in detail. The majority of patients referred from primary care to specialty mental health never attend a single appointment—only 32% of referred adults and 18% of referred adolescents complete even a first visit . Of those who do engage, many drop out before completing a meaningful course of treatment, and the net effectiveness of empirically supported treatments—accounting for refusals, dropouts, and non-responders—is far lower than their trial efficacy suggests. Serrano describes the structural cruelty of this from the patient’s perspective: the very symptoms that make specialty care feel necessary—instability, unpredictable life circumstances, difficulty following through—are the same symptoms that get patients closed out of specialty systems for non-attendance. The referral, for patients with the most complex presentations, is often least likely to result in care. For the individual patient sitting across from a BHC, the realistic alternative to primary care engagement is often not specialty treatment. It is no treatment.

Research on PCBH outcomes makes the clinical case for staying. For example, Cigrang and colleagues conducted a randomized controlled trial of brief Prolonged Exposure for PTSD delivered by a BHC in primary care in four thirty-minute visits; treatment outperformed a waitlist control and benefits held at six months. Large outcome studies of PCBH in military primary care settings found that severely distressed patients improved more, and faster, than patients presenting with mild concerns . Patients with serious mental illness showed better mental health and other outcomes in well-supported primary care than in usual care supplemented by specialty services . Complexity, in other words, is not a contraindication for primary care. It is often exactly what primary care is equipped to address.

(GATHER: Generalist and Routine—stepped care depends on BHCs functioning as generalists who treat first and step up on evidence of non-response, rather than triaging by perceived severity before the intervention begins . 4 Cs: Comprehensive and First Contact—primary care’s scope encompasses the full spectrum of behavioral health presentations, and BHCs, like PCPs, are the first step of care for most concerns regardless of severity .)

The Referral Impulse Is Not Always a Clinical Signal

Kazuo Ishiguro‘s The Remains of the Day offers a frame for understanding what happens when the professional self overrides the clinical one. Stevens, the novel’s narrator and protagonist, is a consummate butler—precise, self-effacing, devoted to his role. He is also a man who has organized his entire inner life around the suppression of anything that might require him to act as a full human being. When the moment calls for genuine engagement—with Miss Kenton, with the moral questions his employer’s politics raise, with the life he has not lived—Stevens retreats into function. He is simply doing his job. The dignity of the withdrawal is, from his perspective, the point.

What Ishiguro renders so exactly is that Stevens never asks what the withdrawal is actually for. He experiences it as professionalism. It is, in fact, a long-practiced avoidance of anything that might cost him something.

The diagnostic question for BHCs facing the referral impulse is the same one Stevens never poses: What is this for? A referral grounded in patient safety, scope-of-practice limits, structural constraints, or genuine evidence that a different level of care is indicated is not flight—it is clinical judgment, and the distinction matters. But when the impulse is driven by the BHC’s discomfort with strong affect, fear of opening something she cannot close in twenty-eight minutes, or a tacit professional belief that trauma content belongs in a different building, the parallel to Stevens holds. That is the dignified withdrawal dressed as clinical reasoning.

Bauman and Beachy have described this pattern in BHC training. Novice clinicians sometimes pre-announce the limits of what they will discuss in a visit—not to orient the patient but to protect themselves from going where the patient might take them. Reiter identifies the same pattern as a systemic problem in blended-model settings, where the presence of a specialty track subtly reframes the BHC’s work as preliminary screening before the real treatment: the BHC becomes, in effect, a triage function rather than a clinician, and both the BHC and the patient receive the message that serious engagement belongs somewhere else and that primary care is incapable of providing deep work.

Jeff Reiter puts the clinical principle plainly: referral decisions should be patient-centered, not clinician-centered. The BHC’s discomfort is not a clinical criterion. Neither is the BHC’s belief, however sincerely held, that the patient’s presentation exceeds what primary care can address—because that belief, without evidence of actual non-response, is a prediction. And we cannot reliably make that prediction before the work has begun.

We Cannot Predict Who Will Improve

One of the most counterintuitive and clinically important findings in the psychotherapy literature is that we cannot predict which patients will respond to briefer or less intensive intervention. Most change in therapies planned for extended durations happens early, often in ways that were not anticipated at intake. A substantial proportion of treatment responders experience rapid, unexpected improvements—what researchers call sudden gains—for which there are no reliable clinical predictors . Strosahl, Robinson, and Gustavsson have argued that many patients with complicated and chronic presentations will improve in response to less intensive intervention, and that defaulting to high-intensity specialty care for complex presentations is an inefficient use of a scarce resource.

The stepped-care framework exists precisely because individual prognosis is uncertain . In primary care, stepped care means attempting to help first, monitoring response with some consistency, and stepping up when there is actual evidence of non-response—not when there is an expectation of it based on presentation severity or diagnostic complexity. This logic is not unique to behavioral health—it is how primary care manages chronic disease across the board. The American Diabetes Association’s Standards of Care describe a treatment sequence for type 2 diabetes that begins with lifestyle counseling and self-management education, progresses through oral agents, advances to injectables, and escalates to insulin and more intensive regimens as needed, with referral to endocrinology reserved for specific clinical indications . No one refers every patient with an A1c of ten directly to endocrinology at diagnosis because the number looks severe. Severity prompts intensification of care, not removal from the setting. The same logic, applied to behavioral health, produces the stepped-care model. A BHC who concludes at intake that a patient’s trauma history makes her unlikely to benefit from primary care behavioral health is making a judgment that her medical colleagues would not make about any other chronic condition she walked in with.

When Referral Is the Right Move

The argument above is not an argument against referral. Stepped care is not the same as refusing to step. There are situations in which specialty mental health is clearly indicated, and a BHC who holds onto patients beyond those thresholds is not serving them.

Robinson and Reiter , drawing on consensus across the field, identify the conditions that reliably warrant referral: the patient is at acute or imminent risk of harm to self or others; the patient requests a referral; the PCP requests a referral; the patient needs a service the BHC does not provide; or the patient is not improving after several visits . In the trauma context, each of these deserves some elaboration.

Non-response after three to four visits is the most common stepped-care threshold, and it is worth being precise about what non-response means. In trauma work, it does not mean that the patient remains symptomatic—many trauma-history patients will carry symptoms indefinitely, and symptom persistence alone is not evidence that primary care is failing. Non-response means that functioning is not improving: the patient is not more able to do what she values, her distress tolerance has not increased, she is not more able to manage the triggers she encounters in daily life. A patient who is stabilizing, building skills, and re-engaging with meaningful activity is responding to primary care treatment, even if her scores on a standardized measure have not crossed a threshold. The measurement guides the conversation; it does not make the decision.

Serrano offers a reframe of the referral criterion that is worth adopting: the question is not how severely ill the patient is, but whether treatment intensity matches her motivation and capacity. A patient who is motivated for a more intensive level of care and has the practical capacity to engage with it—transportation, schedule, stability enough to attend regularly—is a good candidate for specialty referral even before non-response is established. A patient with significant complexity but little motivation for a different treatment structure, or whose life circumstances make specialty attendance unlikely, is better served by continued primary care engagement regardless of symptom severity. That is a patient-centered criterion. Severity alone is not.

The three-to-four visit threshold also requires a clarification about what it is not. In some settings, clinic policy limits all patients to three or four behavioral health visits per episode, after which a referral is required or services end. This is not stepped care—it is an administrative ceiling, and it is no more consistent with how primary care manages chronic conditions than a policy requiring referral to endocrinology after four diabetic visits regardless of trajectory. The threshold is a signal to consider referral when a patient is not improving, not a timer that runs regardless of progress. A patient with complex trauma who is stabilizing, building distress tolerance, and re-engaging with valued activities should continue to be seen, provided that doing so remains consistent with the model—which means the BHC must also maintain availability for warm handoffs and same-day access to new referrals . A BHC carrying a caseload heavy enough to compromise access is drifting toward functioning as a specialty clinic housed in a primary care building, which undermines the model’s capacity to serve the whole population. When a patient has made genuine progress and is functioning well, the clinically and operationally sound move is often to shift to a self-management frame: practice the skills you have been working on, follow up if symptoms worsen, re-engage if functioning declines. That is not abandonment. It is exactly how the rest of primary care manages patients with chronic but stable conditions—and it sets up the re-engagement pathway that is one of primary care’s most distinctive structural advantages.

That advantage is worth naming explicitly. A patient who transitions out of active BHC contact does not disappear from the system. Primary care is built on longitudinal relationships—patients return for physicals, prescription renewals, lab work, chronic disease management, well-child visits . The team is there, the chart is there, and the BHC is there. If access is being maintained and warm handoff systems are in place, a future exacerbation—a return of avoidance, a new stressor, a decline in functioning—can get her back to the BHC within days, without a new intake, without a waitlist, and with a clinician who already knows her history and her strengths. That kind of frictionless re-entry is not something specialty care can offer. It is one of the things primary care does uniquely well, and it is only available if the BHC has managed the panel in a way that preserves that availability.

The PCP’s request for referral deserves particular attention. Taking it seriously is not the same as acting on it automatically. Serrano frames the BHC’s role with the medical team as helping providers say “yes, and”—empowering them to engage with complexity rather than routing around it, while supporting them to set appropriate limits when a situation genuinely exceeds the team’s capacity for safe and effective care. That framing positions the BHC as a consultant who helps the PCP think, not an order-taker who executes whatever the PCP requests. Part of that consultative function is helping the team read the clinical picture accurately . A PCP who has known a patient for fifteen years and believes she is deteriorating in ways the BHC has not yet seen is offering clinical information the BHC should take seriously. A PCP who becomes anxious whenever a patient mentions trauma history, and who reflexively requests referral whenever that history surfaces, is offering something different: their own discomfort, dressed in clinical language. The BHC who cannot tell the difference, or who defaults to compliance in either case, is not functioning as a consultant. A brief conversation with the PCP about what specifically concerns them, what they have observed, and what they are hoping specialty care will accomplish often clarifies which situation the BHC is actually in.

It is also worth noting that the consultative dynamic runs in both directions. In mature PCBH settings, the BHC may have a fifteen-year history with a patient that the PCP does not. A recently graduated family medicine resident, a PCP new to the clinic, or a provider whose prior experience was in a setting that operated on the mild-to-moderate myth may not yet know what integrated behavioral health care looks like when it is functioning well. They may default to referral not because the clinical picture warrants it but because referral is what they have seen done with complex presentations. The BHC’s role in this case is not to override the PCP but to educate and reassure: to describe what the BHC can offer, what the stepped-care pathway actually looks like, what the data say about outcomes in primary care for complex presentations, and what resources exist when the clinical picture does eventually exceed what primary care can provide—whether that is a consult with psychiatry, a connection to a collaborative care manager, or a supported warm handoff to specialty services. A PCP who understands what the BHC team can actually do is a different clinical partner than one who is making referral decisions in the dark.

(GATHER: Team-based—the BHC’s consultative function includes helping PCPs understand the full range of what integrated behavioral health offers, so that referral decisions reflect team capacity rather than individual clinician anxiety. 4 Cs: Coordinated—primary care coordinates the patient’s care across settings, including specialty involvement, rather than transferring responsibility.)

When a referral is appropriate, how it is framed to the patient carries clinical weight. Language that positions the referral as a response to the patient’s complexity—this is beyond what primary care can handle—communicates something about the patient that is probably false and is almost certainly harmful: you are too broken. The framing that Reiter recommends is more accurate and more useful: the current approach has not produced the hoped-for results, and a different approach is worth trying. That is a description of a clinical process, not a judgment about the patient.

Staying in the Room

When a referral is made, primary care does not end. This is the feature of the PCBH model that most sharply distinguishes it from specialty care, and one that BHCs who trained in traditional mental health settings sometimes find genuinely difficult to internalize. As Bauman and Beachy observe, the concept of discharge simply does not map onto primary care. PCPs do not discharge their outpatients. The relationship continues, episodically, for as long as the patient is in the practice—and the BHC relationship follows the same logic.

For the BHC, this means a referral is not a handoff. It is adding a resource while maintaining the existing relationship. Robinson and Reiter describe this as bridging care: continuing to see the patient with active clinical intent until she is genuinely engaged in specialty services or until she begins to improve, whichever comes first. Bridging visits are not administrative check-ins to see whether the patient has called the therapist yet. They are clinical visits, with real intervention. The skills from earlier in this series—grounding, distress tolerance, values-based goal work, psychoeducation—remain available and appropriate throughout. The referral changes the structure of care; it does not suspend it.

(GATHER: Routine—the BHC remains a routine part of the patient’s care throughout the referral process and beyond it, just as the PCP does. Primary care does not terminate; it continues.)

For patients who decline specialty referral, the recommendation should be renewed at each visit and documented, but the work continues. This is not merely a PCBH convention—it reflects a well-established principle in medical ethics and risk management. The AMA Code of Medical Ethics states that physicians have a fiduciary duty to recommend consultation or referral when clinically indicated, and that a patient’s refusal of a recommendation is not grounds for terminating the care relationship . Risk management guidance from MedPro and MICA is explicit: when a patient refuses a clinically indicated referral, the clinician continues to provide care within their competence, explores the reasons for refusal, offers alternatives where possible, renews the recommendation as clinically appropriate, and documents the recommendation and the patient’s informed decision at each visit . A patient who refuses a cardiology referral is not discharged from primary care. The PCP documents the recommendation, continues offering whatever she can, and keeps trying to help the patient engage with what she has declined. The BHC operates on exactly the same logic. The framing Reiter offers is simple and worth keeping close: avoid telling patients there is nothing more primary care can do for them. The more accurate version is: I’d like you to try this other approach, and we’ll still be here. That second clause is not a courtesy. It is a clinical commitment, and it changes what the patient understands to be true about the care relationship.

Stevens’s tragedy is not that he made the wrong choices. It is that he never examined them. He experienced his withdrawals as dignity; he understood his avoidances as professionalism; he mistook self-erasure for excellence. The novel’s long final drive, in which he reconstructs the meaning of his life, is the reckoning he never permitted himself in the moment.

The BHC who pauses at that floor-shift moment—who asks what the referral impulse is for, whose needs it serves, what the patient is likely to experience on the other end of it—is already doing the thing Stevens never could. The question is not whether to refer. It is whether the decision is being made from clinical ground.

Claude Sonnet 4.6 (Anthropic) was used to convert Zotero citation keys to ZotPress shortcodes for publication. All writing, research, and clinical content are the author’s own.

References