Self-Efficacy in Primary Care—Why It Matters and How to Build It

I came to psychology by way of English literature, which means I spent several years before graduate school learning to read carefully and argue from evidence before I had any idea what a confidence interval was. The skills transferred better than I expected. What I did not expect was how often I would find, buried in the novels and poems I had studied as an undergraduate, precise illustrations of concepts I was only later given clinical names for.

This post is something of an indulgence in that direction. Self-efficacy—Albert Bandura‘s construct for a person’s belief in their capacity to act effectively in a given domain—turns out to be one of those concepts that writers understood long before psychologists formalized it. Baum knew something about the difference between capacity and belief in capacity when he sent three companions down a road they were already equipped to travel. Twain knew something about how belief revision works when he put Huck on a raft with Jim. Salinger knew something about why good advice fails when the relationship cannot hold it. Woolf knew something about the attributional errors people make when circumstances are hard and they mistake the hardness for a fact about themselves. None of them were writing psychology. They were doing what writers do: attending carefully to how people actually behave, and why.

I want to shift tone slightly from the last few posts in this series, which have been heavier going—trauma, suicide risk, the relational work of first contact with patients who have every reason not to trust the encounter. This post is still clinical, and the argument it makes matters practically. But self-efficacy is, as clinical targets go, one of the more hopeful ones. It is a belief, which means it is revisable. And the literature, as it turns out, has a fair amount to say about how that revision happens.

There is a particular kind of stuck that does not look like resistance.

The patient agrees with everything. She nods when you explain that walking thirty minutes three times a week would help her blood sugar. She says yes, she knows she should quit smoking. She fills out the wellness plan, takes the handout, and does not do any of it—not because she does not understand, not because she disagrees, and not because she does not want to change. She does not do it because she does not believe she can. That belief is not laziness and it is not ambivalence, though it can look like both. It is low self-efficacy, and if you design an intervention that does not account for it, you will spend the next six visits wondering why nothing is working.

Self-efficacy is a proximate determinant of behavior, one that sits between a person’s knowledge, their motivation, and what they actually do. People who believe they can change are more likely to attempt change, persist through difficulty, and recover from setbacks . People who do not believe they can change generally do not try, or they try in ways that are already organized around anticipated failure.

That second profile appears in primary care with considerable regularity. It is not confined to any single presentation. It shows up in the patient with type 2 diabetes who has been counseled on diet and exercise at every visit for three years. It shows up in the patient with CTr history who has tried therapy before, found it unhelpful or actively harmful, and now presents as cooperative but untouched by clinical contact. It shows up in the patient managing depression with a substance use habit that everyone in the room agrees is not helping, who nonetheless cannot imagine the version of himself who does not need it. In each case, the surface presentation differs; the underlying structure is the same. The patient does not believe that the thing being asked of him is something he is capable of doing.

What Low Self-Efficacy Looks Like in the Room

The BHC who has not looked for low self-efficacy as a clinical variable tends to interpret its presentations through other frameworks—motivation, ambivalence, insight, character. That substitution costs something. The patient who says “I’ve tried everything” and means it is not making an excuse; she is reporting a learned conclusion about her own efficacy. The patient who agrees readily and then does nothing has not, in most cases, agreed at all. They have produced the social performance of agreement while experiencing something more like resignation. The patient who frames all behavioral suggestions as things that “work for other people” is telling you, quite directly, that they do not count themselves among those other people.

Distinguishing low self-efficacy from motivational ambivalence matters because the two warrant different clinical responses. The ambivalent patient is weighing competing values and needs help articulating and resolving that tension—which is the work motivational interviewing is specifically built to do, and which I will take up at length in a later post. The patient with low self-efficacy around a specific domain has largely resolved the motivational question; he has decided he wants to change and concluded that he cannot. What he needs is not help clarifying his values. He needs evidence that the conclusion is wrong.

The Warm Handoff Is Already an Intervention

I want to center something: by the time a well-conducted warm handoff ends, the BHC has already begun addressing self-efficacy, whether or not that framing was ever explicit.

Bandura identified four sources from which self-efficacy beliefs are built and revised: mastery experiences, vicarious learning, verbal persuasion, and physiological and affective state . The first of these is the most powerful, and the warm handoff—designed well—provides it before any formal skill is introduced.

I wrote at length in the second post of the trauma series about the warm handoff as a learning event, drawing on Bandura’s social learning theory. The argument there was about relational learning—the patient with CTr history has learned through direct experience that helpers cannot be trusted, that disclosure leads to harm, and that managing distress alone is safer than asking for help. The warm handoff that goes well provides counter-evidence to each of those beliefs. It is, structurally, a corrective experience.

The self-efficacy dimension of that argument deserves explicit development here. When the BHC invites the patient into collaborative problem definition—”What would be most useful to work on today?” rather than “Here is what your chart says about you”—the patient is positioned as someone whose perspective on their own situation is clinically relevant. When the visit ends with a plan the patient helped to shape, rather than a plan handed down from expertise, the patient has engaged in something they have already done successfully: they have participated in their own care. That is a mastery experience in the most basic sense. It may be a small one. But for a patient who has spent years experiencing the healthcare system as something that happens to them, it is not nothing (GATHER: Accessible; 4 Cs: First Contact—the warm handoff is where the first mastery experience is made available, before any formal skill is introduced .).

Self-Efficacy Across Presentations

The reason this post exists as a cross-cutting piece rather than as a section within one of the series is that low self-efficacy around recovery is not a presentation-specific phenomenon. It appears consistently enough across CTr, chronic disease, depression, and substance use that it functions as a clinical target in its own right, one that should be assessed and addressed regardless of what brought the patient in.

Chronic disease. In chronic disease management—diabetes being the clearest example, and the one I have written about most directly elsewhere in this series—low self-efficacy is often the proximate barrier to behavior change. The patient who does not understand the treatment plan is a familiar clinical problem; the patient who understands it perfectly, agrees with all of it, and does none of it is the one who generates genuine frustration. Lorig and Holman’s foundational work on self-management in chronic disease established that self-efficacy is not merely correlated with self-management behavior but is a primary mechanism through which self-management education produces its effects.

CTr presentations. Among patients with complex traumatization histories, low self-efficacy around recovery has a specific texture that distinguishes it from the chronic disease version. The patient with diabetes who does not believe she can change her diet is contending with habit, environment, and competing demands. The CTr patient who does not believe she can get better has usually tried. She has attempted therapy, perhaps multiple times. She has disclosed things she had never told anyone and found herself retraumatized, dismissed, or simply unchanged. She has concluded, on the basis of actual experience, that recovery is not available to her. That conclusion is not irrational. It is an accurate generalization from the data her life has provided, and it will not yield to psychoeducation about the effectiveness of evidence-based treatment. What it will yield to, slowly, is new experience—which returns us to the learning-event framing I developed in Post 2.

Depression. The relationship between depression and self-efficacy runs in both directions: low mood erodes confidence in one’s capacity to act, and low confidence reduces the motivation to try, which deepens the low mood. Behavioral activation—the subject of its own post in this series—is partly a self-efficacy intervention by another name. When the clinician helps a depressed patient identify one small, concrete, achievable activity and that patient does it, the resulting shift is not only behavioral. The patient has evidence that he was able to act. That evidence does something to the story he is telling himself about what is possible.

Substance use. Self-efficacy in the context of substance use takes a form that clinicians working in this area will recognize immediately: the distinction between wanting to stop and believing one can. These are related but not identical. A patient can have substantial motivation to change his relationship with alcohol and near-zero confidence that he will be able to sustain it, particularly given prior attempts. The efficacy belief, not the motivational state, is often the more relevant clinical target in those cases. This is one of the reasons that SBIRT alone—the population-level screening and brief intervention model—has limited traction with patients who have tried and failed before. Brief feedback and advice, however well-delivered, do not address the efficacy question.

Building Self-Efficacy Across Visits

Bandura’s four sources provide the theoretical map, and I want to use them as backdrop rather than framework—a structure to be aware of, not a checklist to march through.

Using them well requires knowing which one is actually relevant for this patient, in this visit, today. That knowledge does not come from the presenting problem or the diagnostic category. It comes from context—the kind of context the Contextual Interview is designed to surface and that careful listening across visits continues to refine. A patient who has never attempted the target behavior needs a different intervention than one who has attempted it repeatedly and failed; a patient who has no credible model in their social world for what recovery looks like needs something different than one who simply cannot hear encouragement because the therapeutic relationship is not yet stable enough to carry it. The four sources are not interchangeable, and deploying the wrong one is not merely inefficient. It can confirm exactly the belief the BHC is trying to revise—another well-intentioned clinical encounter that did not help, which the patient files alongside all the others as evidence that help is not available to them. What the patient brings into the room today matters as much as what the chart says about the last six months. Efficacy beliefs are not static, and neither is the clinical question of how to address them.

Mastery experience is the most powerful of the four, which means the BHC’s central job is engineering the conditions under which those experiences are possible. Consider the Scarecrow, the Tin Man, and the Cowardly Lion. The Scarecrow reasons carefully on every page of The Wonderful Wizard of Oz ; the Tin Man weeps for the insects he accidentally crushes underfoot; the Lion, when it matters, acts with courage. None of them lacks the capacity they believe they lack. What they lack is a belief in that capacity, and what the Wizard offers at the end of the journey—the diploma, the heart-shaped clock, the medal—is not what changes them. The journey changes them. They did something hard, repeatedly, under pressure, together, and the doing of it constituted the evidence. By the time they reach the Emerald City, the case has already been made. The Wizard is a fraud; the intervention that worked was the road.

The clinical implication is a matter of goal calibration. A goal that the patient has no realistic chance of meeting is a learning experience—just not the one the BHC intended. Another failure confirms rather than challenges the existing efficacy belief, and the patient has learned, again, that change is not available to them. The goal that the patient meets, even partially, does something different. This is the logic behind SMART goals, and it is also the argument for the iterative structure of Primary Care Behavioral Health (PCBH) follow-up visits: each visit is an opportunity to review what happened and locate evidence of capacity. “You said you’d try to walk once this week, and you did it twice” is not praise. It is data. The BHC who helps the patient see that data—who resists the temptation to immediately escalate the goal before the patient has internalized what they just learned about themselves—is doing self-efficacy work.

Bandura’s second source—vicarious learning—is efficacy built through watching someone sufficiently like oneself succeed at a comparable challenge. The clinical key is that word sufficiently. Watching someone exceptional navigate a difficulty does not raise efficacy beliefs; it confirms the distance between that person and oneself. The model has to be credible, which means ordinary, which means recognizably human in the way that matters to the observer.

Twain showed this. What teaches Huck Finn is not a sermon and not an argument. It is the river. Over the weeks he spends on the raft with Jim in Adventures of Huckleberry Finn, Huck watches a man keep watch so he can sleep, grieve his children with a grief that does not perform itself for anyone’s benefit, and act with consistency and dignity under conditions that would excuse the opposite. None of this is heroic in any conventional sense. Jim does not transcend his circumstances; he simply inhabits them with more humanity than Huck had been told was there. That accumulated observation is what revises Huck’s beliefs—not about Jim’s capacity in the abstract, but about what is actually possible for a person in a situation like Jim’s. That is vicarious learning. The model was credible precisely because he was not extraordinary.

Peer support programs, group medical visits, and structured psychoeducation in a group format all exploit this mechanism. So does the clinical narrative, used carefully: the BHC who can describe, without violating confidentiality, the general arc of how other patients with similar presentations have moved through a process is offering vicarious evidence. That arc does not need to be triumphant. It needs to be believable.

The third source, verbal persuasion, is the one clinicians are most likely to reach for first and rely on too heavily. Telling a patient they can do something has a modest effect on efficacy beliefs under narrow conditions: specifically, when the patient already has some reason to credit the speaker’s judgment and some experience of their own on which to build. In the absence of those conditions, encouragement functions mainly as noise—and the literature has a precise case study in what that failure looks like. Mr. Antolini delivers what is arguably the most earnest and accurate piece of advice in The Catcher in the Rye . He is right about Holden, he cares about Holden, and he has more credibility than almost anyone else in the novel. It does not work. Holden wakes in the night, misreads Antolini’s motives, and flees. The persuasion failed not because the content was wrong but because the relational conditions were not stable enough to carry it. The BHC who tells a patient “I know you can do this” before the relationship has enough foundation to make that credible is in the same position as Antolini: right, caring, and unheard.

Physiological and affective state is the fourth source, and it operates in a direction patients rarely anticipate. Anxiety, physical tension, and somatic distress tend to lower efficacy beliefs—not because they impair performance directly, but because the patient interprets his own nervous system as evidence of incapacity. The internal experience of attempting something difficult feels, from the inside, like proof that he cannot do it.

Virginia Woolf made this argument in a different register in A Room of One’s Own, though she was writing about women writers rather than behavioral change. Her point was that the woman who sits down to write in a cold room, underfed, subject to constant interruption, and finds the work difficult has not discovered something true about her abilities. She has discovered something true about her conditions. The difficulty is not evidence of incapacity. It is the experience of trying to do something real while badly resourced. The error—and Woolf was precise about this—is attributional: the woman mistakes the hardness of the circumstances for a fact about herself.

Patients make this error routinely. The patient who attempts a behavioral change, finds it harder than they expected, feels anxious and uncomfortable in the attempt, and concludes they were right that they could not do it—has misread their own data. The discomfort is not a verdict. It is what it feels like to try. Stabilization skills—breathing, grounding, distress tolerance—address this source of efficacy erosion not by eliminating the somatic experience but by giving the patient enough regulation to interpret it more accurately. That is why the stabilization work developed in the trauma series belongs here as well: it is not only phase-appropriate trauma care. It is part of the mechanism by which change becomes believable.

The Relationship to Motivational Interviewing

The overlap between self-efficacy building and motivational interviewing is substantial enough that it is worth naming directly, even though the MI post in this series will develop that argument in full. MI is built on a model of change readiness in which self-efficacy—specifically, confidence in one’s ability to change—is one of two central variables, the other being the patient’s perception of the importance of change . The clinical conversation that explores and strengthens a patient’s confidence in their ability to make a specific change is doing MI and self-efficacy work simultaneously. The distinction is a matter of emphasis rather than kind.

What MI adds to the self-efficacy framework is the attention to ambivalence—the possibility that the patient’s difficulty with change reflects unresolved conflict between competing values rather than, or in addition to, a belief in their own incapacity. The BHC who can hold both questions at once—”What does this person believe about their capacity, and what do they believe about the value of change?”—is operating at the intersection of the two frameworks, which is where the most clinically precise work gets done.

What This Does Not Mean

Self-efficacy is one clinical target among several. It is not the explanation for all treatment failure, and the BHC who filters every stuck patient through an efficacy lens will miss the cases where the explanation is something else entirely: a goal that was wrong for the patient’s actual situation, a therapeutic relationship that has not established enough safety for honest engagement, a social context that makes the target behavior genuinely impossible, or a presentation that exceeds what PCBH can appropriately contain.

Nor is self-efficacy a fixed attribute. Efficacy beliefs are domain-specific and situation-specific . The patient with low self-efficacy around alcohol use may have high self-efficacy around managing his diabetes; the patient who cannot imagine maintaining an exercise routine may be entirely confident in his ability to manage his medications. Domain specificity matters clinically because it means the BHC is not in the business of rebuilding a person’s global sense of self. The work is building evidence in a specific domain, one experience at a time, through visits deliberately structured to make that evidence available.

None of this requires explicit self-efficacy language. None of it requires a protocol. What it requires is knowing what you are doing and why—which is, in the end, the argument this whole post has been making. The work of revising a person’s beliefs about their own capacity, conducted across brief visits in a primary care clinic, is not a pale substitute for something more intensive happening somewhere else. It is its own thing: specific, targeted, and—when the BHC understands the mechanism—capable of producing changes that outlast the visit and accumulate across the relationship.

This post was written by the author. Claude (Anthropic) assisted with verifying current‑event details, refining the structure of the series, supporting revision and editing, and generating ZotPress and WordPress shortcode syntax for citations and formatting. All positions, interpretations, and personal statements are the author’s own.

References