This is a post in the Clinical Skills and Interventions series, part of the ongoing clinical application of the Primary Care Behavioral Health (PCBH) model. It is a companion to The Warm Handoff: Suicide Risk, which covers the acute warm handoff encounter—what the BHC does when suicidal ideation enters the room and the patient is still in the building. This post goes deeper: safety planning, means restriction, chronic suicidality, welfare checks, and the professional accountability structures that support good clinical practice over time.
A note before reading: This post discusses suicide, suicidal ideation, and clinical approaches to suicide risk assessment in detail. It is written primarily for behavioral health clinicians and primary care providers. If you are personally struggling with thoughts of suicide or self-harm, please reach out for support—resources are listed at the end of this post.
A note on scope: This post addresses clinical practice standards and professional accountability in suicide risk management. It is written for licensed behavioral health clinicians and primary care providers. References to legal standards—documentation requirements, duty to protect, involuntary hospitalization criteria—describe general principles derived from the clinical and professional literature. Legal requirements vary by jurisdiction, and nothing in this post constitutes legal advice. Clinicians with questions about their specific legal obligations should consult their state licensing board, a licensed attorney, or their organization’s risk management or legal counsel.
Charles Dickens opens Bleak House with fog. Not metaphorical fog—actual fog, described in obsessive detail, accumulating in the streets and settling most thickly at the heart of the High Court of Chancery, where the case of Jarndyce and Jarndyce has dragged on so long that no living person can say with certainty what it was originally about. The suit has become self-sustaining. It generates process: motions, hearings, delays, documentation that refers to other documentation. Everyone who enters it is changed by it. Some are destroyed. When the case finally reaches its conclusion late in the novel, the entire estate at its center has been consumed by legal costs, and there is nothing left for anyone. The machinery produced nothing but the machinery.
Against that institutional fog, Dickens sets Esther Summerson, one of the novel’s two narrators—a housekeeper, a caretaker, the person who actually shows up and does the work that needs doing. She does not resolve Jarndyce and Jarndyce. She cannot. But she keeps her household keys. She tends to the people the fog has damaged. She is present, particular, and persistent in a way the Court is constitutionally incapable of being.
The warm handoff post establishes the acute scenario: a PHQ-9 screen flags item 9, or a patient discloses unprompted, and the BHC is in the room within minutes. That encounter has its own logic—brief, assessment-focused, aimed at a disposition decision before the patient leaves the building. But the clinical work of suicide risk in primary care does not end there. What follows the acute encounter—safety planning, means restriction, the management of patients who carry chronic suicidality, and the professional practices that keep clinicians competent and accountable over a career—is the subject of this post.
The fog of Chancery has its analogs here. Institutional risk-management culture, EHR automation that mistakes documentation for care, defensive clinical practice designed to satisfy systemic processes rather than serve the person—these are the forces that generate the clinical equivalent of Jarndyce and Jarndyce: encounters that close loops without opening them, charts that document questions without capturing answers, systems that protect themselves while the patient sits in the waiting room. Esther’s alternative is also available. These are not advanced topics reserved for specialists. They are the standard of care for any licensed clinician working with suicidal patients in any setting, and primary care is no exception.
The Safety Planning Intervention
The evidence-based response to moderate suicide risk—the clinical territory where hospitalization is neither clearly indicated nor clearly inappropriate—is the Safety Planning Intervention (SPI), developed by Stanley and Brown . The SPI is not a no-harm contract, and this distinction is not semantic.
No-harm contracts—in which the patient agrees not to harm themselves before the next appointment—have no demonstrated efficacy as a suicide prevention strategy and carry documented risks: they create a false sense of security, shift accountability from clinician to patient, and can damage the therapeutic alliance when the patient experiences suicidal urges and feels they have “broken” a commitment rather than encountered a clinical situation that requires support. Bryan and Rudd observe that patients frequently sign these contracts to appease the provider while having no intention of honoring them—and that providers often know this—creating a dynamic in which the contract functions as ritual rather than treatment. Jobes notes that no-harm contracting persists in practice despite professional guidelines recommending against it, which is itself a testament to how reliably anxiety drives clinicians toward rituals of reassurance. The no-harm contract is the clinical equivalent of Jarndyce paperwork: it looks like action, satisfies an institutional need, and does nothing for the person whose life is at stake.
The SPI is built around six collaboratively developed components:
- Warning signs: The patient’s individualized early indicators that a suicidal crisis is approaching—specific thoughts, images, feelings, or situations.
- Internal coping strategies: Things the patient can do alone to reduce suicidal ideation without contacting anyone else.
- Social contacts for distraction: People and settings that provide distraction and support, not necessarily people the patient will disclose their distress to.
- People to ask for help: Contacts the patient will explicitly tell they are struggling.
- Professional and crisis contacts: The BHC, PCP, a crisis line, an emergency department. The 988 Suicide and Crisis Lifeline belongs on every safety plan.
- Making the environment safer: Means restriction steps the patient agrees to take.
What distinguishes the SPI from every administrative process that resembles it is the same thing that distinguishes Esther Summerson from the Court of Chancery: it is particular. It is built around this patient’s warning signs and this patient’s contacts. It cannot be generated from a template, cannot be completed by proxy, and cannot function if the patient did not genuinely participate in building it. A safety plan that reads “call a friend” under social contacts and lists “988” under professional resources is better than nothing in the same way that any documentation is better than none—but experienced BHCs will recognize the difference between a plan the patient helped construct and one they acquiesced to in order to leave the room. The first is a clinical instrument. The second is Jarndyce paperwork.
The PCBH model’s emphasis on brief, population-level encounters is not a ceiling on clinical care, and treating it as one when a patient is in suicidal crisis is a choice that serves the schedule, not the patient. A safety planning encounter that requires forty-five or sixty minutes to complete well is clinically appropriate—any primary care provider would extend an appointment to manage a medical emergency that presented during a routine visit, and suicidal crisis is a behavioral health emergency. The 30-minute norm describes the typical encounter, not the upper limit of what the patient deserves. The BHC who cuts a safety planning encounter short because the next patient is waiting has made the same category of error as the one who initiates a welfare check to resolve their own anxiety, or who sends a patient to the emergency department because the encounter produced more uncertainty than they could tolerate: a decision organized around the clinician’s comfort rather than the patient’s clinical picture.
That said, staging SPI completion across visits is sometimes the right clinical decision—and the distinction between deliberate staging and running out of time is not trivial. A first encounter with a patient the BHC does not yet know may be better spent establishing alliance and completing a thorough assessment than rushing through a six-component plan the patient has not had time to invest in. A safety plan built on a thin alliance is fragile in precisely the moments it needs to hold. What is not acceptable is leaving a plan incomplete because the slot expired. The BHC who stages completion deliberately, schedules explicit follow-up before the patient leaves the building, and documents what was initiated and what remains is doing clinical work. The one who initiates a plan, runs out of time, and hands the patient a half-filled form is producing the appearance of safety planning without its substance—which is, again, Jarndyce paperwork (GATHER: Generalist; 4 Cs: Comprehensiveness—the BHC addresses suicidal crisis as part of whole-person primary care, not as a simple referral out of the system ; .).
Part of what makes the SPI effective in primary care is the spirit in which it is offered. Dialectical Behavior Therapy‘s foundational emphasis on the therapeutic relationship as a treatment mechanism offers a useful frame here: a safety plan built collaboratively communicates something beyond its procedural content. It communicates that the patient’s continued presence matters, that the clinician will show up to the next appointment, and that the work they are doing together is worth protecting. A plan built in that spirit is a different clinical object than a form completed to document compliance. The former reflects genuine engagement. The latter reflects the clinician’s need to move on to the next patient.
The completed safety plan belongs in three places: in the EHR where the care team and any covering provider can access it, as a patient copy through the after-visit summary or patient portal, and in the clinical note as a documented summary of what was completed and what remains in progress. A covering provider who encounters this patient on a day the BHC is out should be able to find the plan, understand its contents, and use it. If they cannot, the documentation failed in its purpose regardless of how thoroughly it was completed (GATHER: Team-based; 4 Cs: Coordination—the safety plan is a shared care team document, not a record belonging to one provider.).
Means Restriction Counseling
Means restriction counseling deserves its own section because it has the strongest evidence base of any single suicide prevention intervention and is, relative to that evidence, underemphasized in most clinical training.
The logic is direct: suicidal crises are often acute and time-limited. Among patients who survived near-lethal attempts, research has found that the interval between deciding to act and making the attempt can be measured in minutes for a substantial proportion . A patient who cannot immediately access a lethal means during that interval is more likely to survive it. Restricting access extends the interval; extending the interval saves lives. This is not a complicated argument, and the evidence supporting it—drawn from studies of bridge barriers, medication packaging regulations, and firearm storage practices across multiple countries and decades—is among the most consistent in the suicide prevention literature .
The clinical translation is equally direct. Means restriction is not about eliminating all theoretical access to harm; it is about reducing access to the specific means a patient is most likely to use during the most acute phase of a crisis, when impulsivity is highest and the capacity for deliberate decision-making is most impaired. The conversation is brief, targeted, and clinically necessary. Avoiding it because it feels intrusive or because the clinician anticipates resistance is a choice that privileges the clinician’s comfort over the patient’s safety.
In primary care, this conversation takes three primary forms.
Firearms: Firearms account for a disproportionate share of suicide deaths nationally, and they do so because self-inflicted gunshot wound carries a far higher case fatality rate than most other methods . This is a clinical conversation, not a political one—the question is not about gun ownership but about temporary storage options during a period of elevated risk. Providing patients with information about gun locks, temporary transfer to a trusted person, and local safe storage programs reduces access without requiring permanent surrender of the firearm. Bryan and Rudd note that resistance from gun-owning patients is best met not by arguing but by returning to the shared clinical goal: reducing suffering and keeping options open while the crisis resolves. Framing the conversation around the patient’s own values—their reasons to live, their stated desire to get better, their relationships—is more effective than framing it around the clinician’s concern. The patient who declines temporary storage after a direct, respectful conversation has still heard the conversation. Document it. If risk escalates and there is a subsequent encounter, the topic has already been raised, and the second conversation is easier.
Family members and household contacts are often the most effective means restriction partners for firearms. When a patient is willing but reports that a firearm belongs to a spouse or partner who is not present, the BHC should explicitly invite that conversation into the clinical picture—either through a follow-up visit that includes the partner, or by helping the patient identify how to raise the subject at home. The firearm conversation that ends with the patient saying “my husband won’t agree to it” is not a completed means restriction conversation. It is the beginning of one.
Medications: Prescribing smaller quantities of potentially lethal medications—not just psychotropics but analgesics, over-the-counter medications stored in quantity, and other drugs with overdose potential—is a standard risk management step. The BHC’s role includes flagging this recommendation explicitly to the PCP during the staffing conversation so that prescribing decisions reflect the clinical picture. When suicidal risk is identified during a scheduled BHC visit or in another clinical context—a dental exam, for example, where a patient may disclose to a provider they feel less guarded with—a reverse handoff to the PCP is the right clinical decision so that prescribing can be addressed in the same encounter. Identifying a patient’s elevated risk and then allowing them to leave with a 90-day supply of a medication with significant overdose lethality is a failure of clinical coordination, not a documentation problem (GATHER: Team-based; 4 Cs: Coordination—means restriction requires the BHC and PCP to act on shared clinical information within the same encounter.).
Pill storage is a related but underaddressed point. A patient who agrees to transfer a firearm to a neighbor’s safe may not have considered that they have three months of sertraline, two months of trazodone, and a bottle of acetaminophen in the same medicine cabinet. The BHC who asks about firearms but not about medication storage has completed half a means restriction conversation.
Other means: The assessment should address whatever means the patient has contemplated. A patient who described a specific method during the risk assessment warrants a specific means restriction conversation targeting that method. This is not a secondary concern. A patient who has a plan involving a specific means has, by definition, already done some of the cognitive work required to act on it. The clinical task is to increase the distance between that plan and its execution.
The means restriction conversation is almost never comfortable and almost always necessary. Evidence consistently shows that patients do not resent being asked—many report that direct questioning communicates that the clinician is taking their situation seriously . Clinicians who avoid it because they fear offending the patient or provoking pushback are protecting their own comfort at the patient’s expense. That is a choice worth naming clearly, because it is easy to mistake it for clinical sensitivity.
Chronic Suicidality and the EHR Automation Problem
Here is a clinical reality that tends to be handled quietly in training programs and encountered loudly in practice: some patients carry suicidal ideation as a chronic feature of their presentation. Patients with borderline personality organization—a term from Kernberg’s object relations tradition describing the structural features of borderline presentations rather than a DSM diagnosis—treatment-resistant depression, chronic pain, and several other complex presentations may endorse suicidal ideation with a consistency that, if treated as equivalent to acute risk on every encounter, produces clinical distortion rather than clinical care.
It is worth stating directly what this does not mean: chronic suicidal ideation is not “attention-seeking,” and treating it as such is both clinically inaccurate and harmful to the therapeutic alliance. A patient who reports suicidal ideation at every encounter is reporting something real about their experience, even when the function of that reporting is partly relational. The clinical task is to take the ideation seriously while also assessing accurately—not to dismiss it, and not to treat every expression as equivalent to acute crisis.
Bryan and Rudd frame this through fluid vulnerability theory, which distinguishes baseline risk—a patient’s chronic set point, elevated in multiple attempters by definition—from acute exacerbation, when a triggering stressor activates the full suicidal crisis. For chronically suicidal patients, the clinical goal is not elimination of all suicidal ideation, which is unrealistic, but monitoring for change from baseline and containing acute exacerbations as quickly as possible. This is consistent with how primary care manages other chronic conditions: the patient with treatment-resistant depression and chronic suicidal ideation is managed more like a patient with poorly controlled hypertension than like a patient presenting with a first suicidal crisis.
The relevant clinical question for a patient with chronic suicidal ideation is not “are you thinking about suicide?”—the answer to which is predictably yes and differentiates nothing—but rather “is this different from what you usually experience?” Change from established baseline, not presence or absence of ideation, is the clinically significant variable . That distinction requires the clinician to know the baseline, which requires that the baseline be explicitly documented—in language a covering provider or an EHR prompt will encounter and understand.
This is where the problem compounds. In clinics that have adopted Zero Suicide frameworks or similar protocols, EHR systems may flag any patient with prior documented suicidal ideation and require completion of a formal risk assessment at every subsequent visit. The intention is sound: ensuring that high-risk patients are not lost to follow-up. The execution, at its worst, produces something Dickens would have recognized immediately. The alert fires. The provider answers the alert. The documentation closes the loop. The patient—who has reported passive suicidal ideation at every appointment for three years as reliably as they report their blood pressure—watches the provider work through a structured prompt that was designed for someone presenting with suicidal ideation for the first time, answers the same questions they answered at the last six visits, and leaves having had an encounter that resembled clinical care without quite being it.
This is Chancery logic applied to behavioral health: process that generates more process, documentation that refers to prior documentation, a machinery that keeps itself running while the person at its center waits. The fog settles most thickly not at the court’s entrance but at the most complex and chronic cases, precisely where clinical judgment is most needed and most difficult to automate.
The corrective is not to abandon structured risk assessment for patients with chronic presentations—it is to document the baseline explicitly so that any deviation from it registers as meaningful. A chart note that reads “patient endorses passive SI, consistent with chronic presentation; no change from baseline; safety plan reviewed and updated, patient verbalized continued ability to use plan” communicates something a covering provider can act on. A note that reads “patient denied SI” records a question that was answered in a way that closed the alert without engaging the patient. The first note is Esther’s work. The second is Jarndyce paperwork.
Establishing that documented baseline requires a specific clinical conversation, ideally early in the treatment relationship with a chronically suicidal patient. What does their suicidality usually look like? Passive ideation without intent? Fleeting thoughts of self-harm that do not organize into plans? What has changed in the past when things got worse? What were the precipitants? What pulled them back? The answers to those questions are the baseline. They belong in the chart in language clear enough that any provider who covers for the BHC during a vacation week can read them and understand immediately whether what they are encountering is business as usual or a departure from it. A baseline that lives only in the treating BHC’s memory is not a baseline. It is a liability—for the patient, and for everyone on the care team (GATHER: Team-based; 4 Cs: Continuity—documented baselines ensure that any covering provider can assess change from baseline, not just recognize the presence or absence of ideation.).
Autonomy, Ethics, and the Legal Frame
The tension between respecting patient autonomy and the clinician’s ethical and legal obligation to protect life is not resolvable by formula. It is navigated case by case, informed by the patient’s actual risk level rather than the clinician’s discomfort.
A useful orienting principle: involuntary hospitalization is the most restrictive available intervention, carries significant risks of its own—loss of employment, insurance implications, trauma responses in patients with prior involuntary holds—and is clinically appropriate when risk is imminent, safety cannot be maintained in a less restrictive environment, and voluntary engagement with safety planning has failed or is not feasible. It is not appropriate as a default response to suicidal ideation, as a mechanism for managing clinician anxiety, or as a substitute for a completed assessment. The clinician who sends a patient to the emergency department because the encounter produced more uncertainty than they could tolerate has made a risk management decision, not a clinical one. Those are not always the same thing.
Jobes argues persuasively that the collaborative approach—engaging the patient as a partner in understanding and addressing their own suicidality rather than positioning the clinician as the expert who controls the outcome—is not only ethically preferable but clinically more effective. Patients who participate in building their own safety plans are more likely to use them. Patients who experience the assessment as collaborative rather than evaluative are more likely to disclose accurately. Clinicians who relinquish the fantasy of controlling the patient’s behavior in favor of reducing the patient’s suffering alongside them are better positioned to make clear clinical decisions, because their anxiety is no longer distorting the clinical picture.
The documentation question deserves direct attention here, because it is where the Chancery temptation is strongest. The legal standard in most jurisdictions is not that the clinician correctly predicted the patient’s behavior—foreseeability is not certainty—but that the clinician met the standard of care: structured assessment, reasonable clinical judgment, documented rationale, and appropriate follow-up. The note that exists to satisfy that standard is a Jarndyce document: it records the process, closes the loop, and tells the story the institution needs to hear. The note that exists to communicate clinical reasoning to the next provider who cares for this patient is something different. It documents the domains assessed, the specific risk and protective factors identified, the rationale for disposition, the safety plan completed, the means restriction discussion, and the team staffing. In the event of a bad outcome, the difference between those two notes is not trivial. But more importantly, one of them actually serves the patient who is still alive.
The defensive note and the clinical note often look similar from the outside. The difference is in what they are organized around. A note organized around protecting the clinician from liability documents whatever would be most useful in a deposition. A note organized around the patient’s clinical picture documents whatever would be most useful to the provider seeing this patient next week. These goals are not in opposition—a genuinely clinical note is also legally defensible—but they are not identical, and clinicians who conflate them tend to produce notes that serve neither purpose well.
A concrete example clarifies the distinction. Consider two notes for the same patient—a 34-year-old with a history of major depression and one prior attempt, presenting following a recent job loss:
Note A: “Patient presented following job loss. Denies active suicidal ideation. Safety plan reviewed. Follow-up scheduled.”
Note B: “Patient presents with significant acute stressor (recent job loss). Assessed for suicidal ideation given history of MDD and prior attempt in 2019. Patient reports passive ideation (‘sometimes wish I wouldn’t wake up’) without active intent or plan; this represents a change from baseline (typically denies SI at scheduled visits). Identified precipitant: financial stress and perceived failure. Protective factors: two children at home, strong relationship with PCP, previously responded to safety planning. Safety plan reviewed and updated; patient verbalized understanding of warning signs and identified two new contacts. Discussed medication storage; patient agreed to have spouse hold medications this week. PCP staffed; 48-hour phone follow-up planned. Will reassess at next scheduled visit.”
Note A closes the loop. Note B opens a clinical picture. In the event of a bad outcome, any reviewer will know immediately which clinician was doing clinical work and which was doing paperwork. More to the point: a covering provider seeing this patient two days later, working from Note A, knows nothing. Working from Note B, they know nearly everything they need to care for this patient safely.
A Note on Welfare Checks
The welfare check—requesting that law enforcement visit a patient at home to assess their safety—is a tool that deserves its own careful treatment, because it is one of the most commonly anxiety-driven interventions in suicide risk management.
A welfare check is clinically indicated when a patient has expressed imminent suicidal intent, contact has been lost, and there is reason to believe they are in immediate danger. It is not indicated as a reflexive response whenever a clinician cannot reach a patient, cannot resolve their own uncertainty, or simply feels that something must be done. Sending police to a patient’s home because the clinician is anxious—rather than because the patient’s risk level warrants it—can cause significant harm: for patients with prior trauma involving law enforcement, for patients from communities with historically adversarial relationships with police, and for any patient whose experience of a wellness visit is coercive rather than caring. Drustrup and colleagues argue that welfare checks in particular carry far greater potential risk than benefit for many patients—and that those risks are not distributed equally. Data from the National Suicide Prevention Lifeline found that in over 99% of cases, the 911 system was unnecessary for a successful rescue when callers were engaged collaboratively. People with mental health concerns represent a disproportionate share of police-involved deaths—roughly 25%, against a general population prevalence of approximately 4%—which means that dispatching police to the home of a patient in psychiatric distress is not a risk-neutral intervention regardless of the clinician’s intent. For patients from communities with documented histories of adversarial or lethal encounters with law enforcement, a welfare check initiated primarily to manage clinician anxiety is a decision whose costs fall on the patient. A welfare check can irreparably damage the therapeutic alliance and, in the worst cases, escalate rather than resolve a crisis.
The most important clinical practice here is prospective transparency. Whenever there is any realistic possibility that a welfare check might become necessary—for a patient with a history of suicidal crises, for a patient whose follow-up plan includes a check-in call, for anyone whose risk level warrants monitoring between appointments—the possibility should be raised directly with the patient before the encounter ends. The conversation might sound something like: “Part of how I’ll follow up with you is to try to reach you by phone in the next day or two. If I try to reach you and can’t make contact, and I’m concerned about your safety, I may ask someone to check on you at home. I want you to know that ahead of time so it’s not a surprise.”
That transparency is not a bureaucratic disclaimer. It preserves patient autonomy, it keeps the therapeutic alliance intact, it reduces the likelihood that a welfare check will feel like a betrayal, and it gives the patient the opportunity to make a different choice about their safety. It is good clinical practice, and it takes about thirty seconds to deliver.
The same logic applies to mandatory reporting obligations more broadly. Patients who learn for the first time that the clinician is required by law to take certain actions—when those actions are already underway—experience that disclosure as a violation. Patients who learn about the limits of confidentiality before they disclose, when the information is hypothetical rather than immediately consequential, have the opportunity to make genuinely informed decisions about what to share and how. This is not a loophole. It is basic informed consent, and it belongs at the beginning of the clinical relationship rather than at the moment when it is most inconvenient.
Continuing Education, Peer Review, and Professional Accountability
Suicide risk assessment is one of the few clinical competencies for which the research on training gaps is genuinely alarming. Studies examining graduate training in psychology, social work, and counseling have consistently found that formal instruction in suicide risk assessment and management is sparse—often amounting to a handful of hours across an entire doctoral program . Licensure does not correct this. A clinician can enter independent practice having managed very few supervised cases involving suicidal patients and accumulate decades of clinical work without ever formally revisiting what they learned in a single graduate seminar.
This gap is not primarily a knowledge problem, though it is partly that. It is a cultural problem. Suicide risk encounters are frightening. Clinicians who feel undertrained tend to avoid the discomfort by relying on habit, ritual, and the no-harm contracts their supervisors modeled for them—because those things feel like action even when they function as avoidance. The clinician who has never taken a continuing education course on suicide risk assessment because their licensing board does not require it and their supervisor never mentioned it is not, by any visible measure, deficient. The deficiency is invisible until it is not. This is precisely the structure of Jarndyce and Jarndyce: a system that sustains itself through incremental inaction, in which the absence of a bad outcome is mistaken for the presence of competent practice.
The corrective is straightforward: recurring continuing education in suicide risk assessment should be part of every licensed clinician’s routine professional development, whether or not the licensing board requires it. Some boards do require it; many do not. The requirement should not be the motivation. Applied Suicide Intervention Skills Training , the Collaborative Assessment and Management of Suicidality (CAMS) framework , and Zero Suicide workforce training curricula all offer evidence-informed approaches to skill development that extend meaningfully beyond what most graduate programs cover. Engaging with this literature on a renewal-cycle schedule—rather than as a checkbox completed once and forgotten—is the professional standard. The clinician who last engaged seriously with suicide risk literature a decade ago is practicing at a decade-old standard, regardless of how many patients they have seen since.
Supervision is the second accountability structure worth naming directly, particularly for early-career clinicians whose graduate training left gaps. The BHC who has a supervisor, consultant, or senior colleague they can call when a case feels difficult is in a structurally better position than the one who practices in isolation. That is not a sign of weakness or inadequate training. It is a sign of professional maturity. The capacity to recognize the limits of one’s own judgment in a given moment and to seek input before acting is one of the most clinically protective habits a clinician can develop. The supervisor who hears “I saw a patient today who scared me, and I’m not sure I made the right call” is getting the supervision conversation that actually develops competence. The one who hears only polished case presentations is not.
Peer review—the structured process by which cases are reviewed by colleagues for quality assurance and learning—is a third accountability mechanism worth naming directly. Peer review exists in many healthcare organizations as a formal process, typically convened after a significant adverse event. What is less commonly practiced is voluntary peer review: a clinician identifying a case in which they made an error—or where the outcome was good but the reasoning was flawed—and bringing it forward for collegial review without being required to do so. The culture that makes peer review most valuable depends on clinicians being willing to say, without prompting: I’m not sure I handled this well and I want to talk about it. A case that ended without a bad outcome is still worth reviewing if the reasoning was driven more by the clinician’s discomfort than by the patient’s clinical picture. That is often the case from which the most learning is available.
Chus Arrojo, writing in response to the companion warm handoff post, identifies a related failure mode that peer review is particularly well-positioned to address: what she calls aligned error under distributed responsibility. In a team-based encounter, no single clinician may make an unreasonable decision, yet the gaps in the system briefly line up—each provider trusting that someone else has conducted the harder assessment, each acting on assumptions that go unexamined precisely because the team structure creates the appearance of redundancy. The Swiss cheese model of safety assumes that multiple layers catch what individual judgment misses; Arrojo’s observation is that the layers can also align in the wrong direction. The corrective is not less teamwork but more explicit ownership: at the close of any suicide risk encounter involving multiple providers, someone should be able to name clearly who conducted the assessment, what it found, and who holds the follow-up. If that question cannot be answered without ambiguity, the encounter is not complete regardless of how many people were in the room.
Esther Summerson kept her household keys. She inventoried what she had. She checked what she did not know. She asked for help when the situation exceeded her skill, and she returned to the work the next morning. She was not immune to the fog of Chancery—it touched her directly, as it touches everyone in the novel—but she did not wait for the fog to clear before doing the work that needed doing. That is precisely the professional posture that continuing education, supervision, and peer review are designed to support: not the elimination of uncertainty, which is not achievable, but the development of competence sufficient to act well inside it.
Zero Suicide as Organizational Context
Individual clinical competence is necessary but not sufficient for suicide prevention at the population level. The Zero Suicide framework situates the individual clinical encounter within a system-level approach that includes workforce training, care transitions protocols, means restriction counseling policies, and continuous quality improvement. BHCs working in systems that have adopted Zero Suicide will recognize many of the clinical practices described here as standard components of that framework; those in systems that have not may find it a useful reference for program development conversations.
The BHC who cares about this work at a systems level—who wants the clinic to be the kind of place where suicidal patients are reliably identified, carefully assessed, and effectively engaged—is in a position to contribute to that culture. That is a different conversation than the one happening in any single exam room. It is also a necessary one. Zero Suicide as a framework makes explicit what good clinical practice implies: that the safety of suicidal patients is a system responsibility, not a clinician-by-clinician lottery that depends on whether the right person happened to be available on the right day (GATHER: Highly Productive; 4 Cs: Comprehensiveness—suicide prevention functions as a population-level responsibility distributed across the entire care system, not a task assigned to whichever clinician happens to be present.).
The EHR automation problem discussed above in the context of chronic suicidality applies at the systems level as well. A Zero Suicide implementation that installs structured prompts without providing the clinical training to use them thoughtfully has built more Chancery machinery. The prompts are not the intervention. The training, the consultation structures, the culture of voluntary peer review, the willingness of the care team to have difficult conversations with difficult patients—these are the intervention. The prompts are the documentation. Mistaking documentation for care is the single most reliable way to believe that a system is doing something while it is doing nothing.
Closing
Kay Redfield Jamison, writing about her own suicide attempt in Night Falls Fast , describes the experience with a precision that no clinical formulation quite captures: it was the end of what she could bear, the last afternoon of having to imagine waking up the next morning only to start all over again. Nothing warm could make its way through. She survived. The account belongs in the clinical literature not because it is unusual but because it is precise—a description of what suicidal crisis feels like from inside it, which is exactly the perspective that risk assessment tools cannot fully capture and that the therapeutic relationship, when it is working, is designed to reach.
Many patients in primary care waiting rooms are living inside some version of that sentence. They have come for a refill, for a physical, for a score that ticked up on a questionnaire. The clinic they have walked into is either Chancery or it is something else. If it is Chancery, it will document their visit, close the relevant alerts, and send them home with paperwork that protects the institution. If it is something else, it will have built what Dickens gave Esther: someone with keys, someone who keeps the household in order, someone who notices when a person has stopped eating. Someone who is present, particular, and persistent in a way that the fog cannot replicate and the machinery cannot replace.
The clinical practices described in this post and its companion are not a burden added to an already demanding role. They are the reason the role exists.
If you or someone you know is in crisis: The 988 Suicide and Crisis Lifeline is available 24 hours a day, 7 days a week.
- Call or text: 988
- Chat: 988lifeline.org
- For veterans: Press 1 after dialing 988, or text 838255
- For Spanish speakers: Press 2 after dialing 988
If there is immediate danger, call 911 or go to your nearest emergency room.
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
Cite this article as:
Robert Allred, "Suicide Risk Assessment in Primary Care: A Clinical Reference," Allred Consulting, June 4, 2026, https://allred.consulting/2026/06/suicide-risk-assessment-in-primary-care-a-clinical-reference/.
or
APA Style, 7th Edition:
Allred, R. (June 4, 2026). Suicide Risk Assessment in Primary Care: A Clinical Reference. Allred Consulting. https://allred.consulting/2026/06/suicide-risk-assessment-in-primary-care-a-clinical-reference/
Bluesky Discussion
View on BlueskyNo replies yet. Be the first to comment on Bluesky!
Discover more from Allred Consulting
Subscribe to get the latest posts sent to your email.
