Most People Get Better Without Us

This post contains Amazon affiliate links. If you purchase through them, I earn a small commission at no extra cost to you.

This is the first post in the Addiction and Recovery in Primary Care series. For related reading across the blog, see the Warm Handoffs and Trauma in Primary Care series.

---

The Tyrone family spends a single August day in 1912 cataloguing one another’s failures. James blames his wife Mary’s morphine dependence on her weak character—on a softness he has spent decades privately resenting. Their older son Jamie blames James, who hired the cheapest doctor available when Mary was in labor, a physician who knew one remedy for pain and used it without restraint. Edmund, the younger son and Eugene O’Neill’s autobiographical stand-in, understands his mother through grief—through the accumulated weight of a family that has never quite recovered from itself. Mary, for her part, oscillates between understanding her morphine use as something that happened to her and understanding it as evidence of something broken inside her that no amount of recovery can repair.

Long Day’s Journey into Night is not a play about addiction so much as it is a play about the inadequacy of explanations. Each character holds a theory of what Mary’s dependence is and how it came to be, and each theory illuminates something real while foreclosing something else. None of them can agree. And in the space between their competing explanations—moral failure, medical negligence, grief, constitutional weakness—Mary gets worse. The fog rolls in off Long Island Sound, and the family retreats, as it always does, into its separate fictions.

What the Tyrones could not do, primary care behavioral health consultants (BHCs) are asked to do every day: hold multiple frameworks simultaneously, remain curious about which one is illuminating something the others are missing, and keep the patient in view rather than the model. That is harder than it sounds, because the models we inherit about addiction are not neutral. They determine what we see, what questions we ask, and—most consequentially—what we think our job is. Before any of the clinical skills that subsequent posts in this series will address, the BHC needs a clear accounting of what those models are, where they come from, and what the empirical literature actually says about how people with substance use disorders get better.

The short version: most of them get better without us.

---

The CFHA Post and This One

Earlier this year I published a piece in the Collaborative Family Healthcare Association (CFHA) blog titled When the Chart Says Substance Use and the Room Says Something Else. That post described the clinical experience of walking into a warm handoff in which the chart announces a problem and the patient in front of you tells a more complicated story. It was a post about clinical attunement—about learning to listen for what the room is saying when the record has already decided what it thinks.

This post is the theoretical ground beneath that observation. If the room so often tells a different story than the chart, it is partly because the chart is written inside a particular model of addiction, and that model may not be the most clinically useful one. What follows is an attempt to name those models honestly, to trace what each of them costs, and to situate the BHC’s role against the most important and least-discussed finding in the addiction literature.

---

The Models We Inherit

Every clinician working with substance use carries an implicit framework—usually more than one, in uncomfortable coexistence. These frameworks are not chosen deliberately in graduate training so much as absorbed, from supervisors, from institutional culture, from the language embedded in diagnostic systems, from whatever the clinic already believes. Making them explicit is the first step toward using them well.

The Moral Model

The oldest framework, and the one that clinical training most explicitly rejects, is also the one that most reliably survives in clinical practice. The moral model understands addiction as a failure of character: a problem of will, of virtue, of the kind of person the patient has chosen to be. It generates the language that follows patients through the medical system—”drug-seeking,” “non-compliant,” “manipulative,” “difficult”—and that language, once charted, shapes every subsequent encounter. A patient who arrives at the clinic carrying that chart history is not meeting the care team fresh. He is meeting the judgment that has preceded him.

The moral model is not without empirical anchoring. Volitional elements are genuinely present in substance use behavior, and acknowledging that patients make choices—including choices that can change—is not the same as blaming them for those choices. The problem is not that the moral model recognizes agency; it is that it assigns moral weight to that agency in a way that produces shame, and shame is among the most reliable predictors of treatment avoidance and dropout. The BHC who has internalized even a residual version of the moral model will find it difficult to sustain genuine curiosity in encounters with patients whose use patterns have caused harm—to themselves, to their families, to people who trusted them. Moral models produce moral reactions, and moral reactions are poor clinical instruments.

The Disease Model

The dominant contemporary framework in American addiction treatment understands addiction as a chronic brain disorder—a condition characterized by compulsive substance use despite adverse consequences, driven by neuroadaptive changes that alter reward processing, decision-making, and stress reactivity .

The disease model’s clinical contributions are substantial. It reduces blame, frames relapse as a feature of chronic illness rather than a moral failure, and opens the door to medication-based treatment—a door that has been transformative for opioid use disorder in particular. Medication for opioid use disorder (MOUD)—buprenorphine, methadone, and naltrexone—is now routinely delivered in primary care settings, where the PCBH model and MOUD programs are increasingly integrated . Low-threshold buprenorphine models—emphasizing same-day treatment entry, flexible policies, harm reduction orientation, and availability in unconventional settings—represent an emerging operationalization of this approach that is designed specifically to reach patients who would not engage with traditional specialty addiction treatment . When a primary care team understands addiction as a disease, patients are more likely to be treated with the same matter-of-fact competence that the team brings to diabetes or hypertension.

But the disease model carries its own limitations, and primary care is the setting where those limitations are most visible. The model was developed and validated largely in treatment-seeking populations—people who had already arrived at specialty addiction care, often after years of unsuccessful attempts to stop on their own. That population is not representative of the patients a BHC meets in primary care, most of whom have never sought addiction treatment and never will. The disease model also sits in persistent tension with one of the most replicated findings in the addiction epidemiology literature, which is that the majority of people who meet diagnostic criteria for a substance use disorder at some point in their lives resolve it without formal treatment. If addiction is a progressive brain disease, that finding is difficult to explain. If it is a condition shaped by context, reinforcement history, developmental stage, and social circumstance—as well as by neurological changes—the finding makes considerably more sense. (GATHER: Accessible—the BHC who is present in the primary care setting, without referral barriers, is positioned to reach patients who would never seek addiction treatment through any other door .)

The Biopsychosocial Model

The biopsychosocial framework, which George Engel developed as a challenge to biomedical reductionism, offers the most theoretically complete account of addiction: it is biological and psychological and social, all at once, with each domain influencing the others in ways that resist simple hierarchical ordering. The framework is accurate. It is also, on its own, clinically underspecified. Knowing that a patient’s alcohol use disorder involves neuroadaptive changes, a trauma history, a social environment saturated with heavy drinking norms, and a set of maladaptive coping beliefs does not, by itself, tell the BHC what to do first, or how to do it in fifteen minutes.

The biopsychosocial model is most useful as an orientation—a reminder that no single-axis account of addiction will capture the patient in front of you—and as a corrective to frameworks that reduce the clinical picture to any one dimension. It is the foundation on which more clinically actionable models need to be built.

The Social Learning and Behavioral Model

Albert Bandura’s social learning theory offers something the disease and biopsychosocial models do not: a direct path from theoretical framework to clinical target. If substance use is learned behavior, shaped by reinforcement contingencies, outcome expectancies, modeling, and the social contexts in which use occurs, then the points of clinical intervention are identifiable. Expectancies can be examined and modified. Alternative reinforcers can be identified and strengthened. The social context can be assessed and, where possible, changed.

Central to the social learning account is self-efficacy—the patient’s belief in her capacity to manage her use, to tolerate cravings, to navigate high-risk situations without returning to the substance. Low self-efficacy is among the strongest predictors of relapse and treatment dropout, and building it is among the most important things a BHC can do in any encounter, including the first one. The self-efficacy post elsewhere on this blog covers that clinical target in detail; the connection to substance use is direct and worth naming explicitly here.

The social learning model also provides the theoretical grounding for Motivational Interviewing, which William Miller first described specifically in the context of problem drinkers—and which remains, four decades later, the most evidence-supported brief approach to substance use in primary care. Miller’s original paper explicitly grounded MI in self-efficacy and attribution theory, which is to say it was a social learning intervention from the beginning, whether or not it was always described that way.

The Contextual Model

The model that carries the most clinical weight for the population primary care BHCs actually encounter is rooted in functional contextualism—a philosophical framework associated most closely with Steven Hayes and colleagues. Functional contextualism understands behavior not as a symptom of underlying pathology but as an event occurring in context, shaped by history and current circumstances, and best analyzed by asking what function it serves. The goal of clinical analysis is not formal diagnosis but prediction and influence: what is this behavior doing, under what conditions, and what would need to change for something different to become possible?

This framework has generated multiple clinical implementations, two of which are particularly relevant to primary care. Acceptance and Commitment Therapy (ACT)—and its primary care adaptation, Focused ACT —applies functional contextualism to psychological flexibility: helping patients make contact with their values, defuse from unhelpful cognitions, and move toward behavior change without requiring that distress be eliminated first. ACT and FACT have had an enormous influence on how PCBH conceptualizes brief intervention, and the FACT framework is woven throughout the clinical approach described on this blog.

The implementation that carries the most weight in my own clinical practice, however, is Steven Gold’s Contextual Trauma Therapy (CTT). That emphasis is not a theoretical preference so much as a response to the population. The patients I have worked with most extensively have been in Federally Qualified Health Center (FQHC) settings, where the prevalence of complex trauma histories—often rooted in poverty, structural racism, adverse childhood experiences, and chronic community-level stress—is high enough that a trauma-informed lens is not a specialty add-on but a baseline clinical assumption. In those settings, Gold’s framework fits because it was built for exactly this population: people whose substance use is not a primary disorder so much as a functional adaptation to unresolved traumatization. CTT frames substance use disorders not as co-occurring conditions that happen to accompany trauma histories but as functional responses to affective states, intrusive memories, and physiological hyperarousal that the patient has not yet developed other means of tolerating. In this reading, the substance is doing something. It is working, in the most limited sense of the word, even as it destroys.

The empirical substrate for the CTT argument is the Adverse Childhood Experiences (ACE) study, which established a graded dose-response relationship between the breadth of childhood exposure to abuse and household dysfunction and nearly every category of adult health risk behavior, including alcoholism, drug abuse, and smoking . The relationship is not incidental. It is one of the most robust findings in behavioral medicine, and it suggests that for a substantial portion of the primary care patients with substance use concerns, the clinical question is not “what is wrong with this person’s relationship with alcohol” but “what happened to this person, and what has alcohol been helping them survive.”

This is the model that animated the CFHA post. It is also the model that most directly shapes the BHC’s clinical posture in the encounter—curious, contextually grounded, alert to the function the substance is serving, and slow to intervene in a way that removes coping before anything is in place to replace it. The relationship between trauma and substance use will receive its full treatment in the next post in this series.

---

The Fact That Changes Everything

Four decades ago, Barry Tuchfeld published a study of adults who had resolved significant alcohol problems without formal treatment. The finding was not new—clinicians had been observing spontaneous remission for decades—but Tuchfeld’s systematic documentation of the phenomenon contributed to a growing body of evidence that the specialty addiction treatment system was studying a highly selected sample and drawing conclusions that did not generalize to the broader population of people with substance use disorders.

The Sobells’ research on self-change sharpened that argument considerably. Their work demonstrated not only that people with significant alcohol use disorders could and did change without formal treatment, but that many of them achieved outcomes comparable to those of treated populations—and that the change process often looked less like a dramatic conversion experience than like a gradual reappraisal of costs and benefits, frequently catalyzed by a life event, a relationship, a shift in circumstances. The Transtheoretical Model that James Prochaska and Carlo DiClemente developed from their study of self-changers provided a map of that process: people move through stages of readiness—precontemplation, contemplation, preparation, action, maintenance—and they do so whether or not a clinician is present.

The population-level data confirmed what these clinical studies suggested. Large epidemiological surveys consistently find that the majority of people who meet lifetime criteria for a substance use disorder recover without seeking formal treatment . A 2016 systematic review and meta-analysis of remission from substance use disorders found that between 35 and 54 percent of individuals with a diagnosable disorder achieved remission, typically after a mean follow-up period of seventeen years, and that treatment intensity was not the primary determinant of outcome . This is not an argument against treatment—for people with severe, long-standing disorders, particularly opioid use disorder, treatment is often essential and life-saving. It is an argument about who the primary care BHC is actually seeing.

The patients who walk into a primary care exam room are not, as a population, the patients who populate specialty addiction treatment programs. They are the full distribution—people at every stage of readiness, most of them ambivalent, many of them not yet ready to call what they are doing a problem, some of them already well into a change process they have not disclosed to anyone. The specialty treatment literature, which dominates addiction training, was built on the more severely affected, treatment-seeking end of that distribution. Applying its assumptions to primary care patients is a category error.

What natural recovery research suggests is that the BHC’s job in primary care is not to be the intervention. It is to be a person in the clinical relationship at a moment that may or may not “matter”—to reduce shame, to build self-efficacy, to provide accurate information, to keep the door open, and to not inadvertently make things worse by confronting ambivalence before the patient is ready to move. Edwards and colleagues demonstrated in their landmark trial that a single session of advice—not treatment, not structured therapy, not a specialty referral—produced outcomes for problem drinkers that were comparable to intensive inpatient treatment. The finding was controversial and contested, running counter to clinical intuition and the assumptions of service delivery; but as Jim Orford noted on the study’s fortieth anniversary, it was ahead of its time—anticipating decades of subsequent evidence showing that brief interventions can be as effective as more intensive modalities for many drinkers, and that more treatment does not reliably produce better outcomes . That finding has never fully dislodged the specialty treatment industry’s assumptions about what adequate care requires. In primary care, it is the premise from which the BHC’s work should begin.

---

What This Means for the BHC

None of the above is an argument for therapeutic nihilism. The BHC who decides that most people get better without us and is therefore off the hook has misread the finding. The natural recovery literature consistently identifies facilitating conditions—social support, reduced access to the substance, life events that shift the cost-benefit calculus, encounters with trusted people who respond with curiosity rather than judgment. Primary care is an extraordinary site for providing those conditions, precisely because patients are not there to talk about their substance use. They are there for a diabetes check, a well-child visit, a sore throat. The BHC who shows up in that context, trauma-informed and non-judgmental and genuinely curious, is positioned to be one of those facilitating conditions—not because they perform a specialized addiction intervention, but because they are present, in a trusted setting, at a moment when the door may briefly be open.

This is also an argument about what the BHC brings to the primary care team. PCPs are trained in a medical culture that is deeply influenced by the disease model—and they should be, because for OUD specifically the disease model’s treatment implications are life-saving. But that model, applied without modification across the full spectrum of substance use presentations, produces clinical nihilism (“there’s nothing we can do until they’re ready”) and moral frustration (“she keeps coming back and nothing changes”) in roughly equal measure. The BHC who can articulate a broader framework—who can explain why the patient who presents ambivalent today is not failing treatment but inhabiting a predictable stage of a change process that primary care is uniquely positioned to support—is doing something the PCP cannot easily do alone. (4 Cs: Comprehensive—the BHC expands the team’s conceptual framework for substance use, not just its clinical capacity .)

Mary Tyrone, by the play’s end, has retreated so far into her morphine and her memories that the other three Tyrones can no longer reach her. The fog has come in, and she is lost inside it. What the play cannot give us—what it was never designed to give us—is an account of the thousands of people for whom the fog eventually lifted, not because anyone found the right model in time, but because something in their lives shifted and the cost of staying became higher than the cost of changing. They got better without the Tyrones noticing. They get better without us all the time.

Knowing that is not a reason to stop showing up. It is a reason to show up differently.

---

References