Sign Up
Login
Try for Free
Login
The C-SSRS (Columbia-Suicide Severity Rating Scale) is a widely adopted instrument that classifies suicidal ideation and behavior into defined categories. Hospitals, clinics, schools, and research programs use it to standardize screening and the response that follows. This page covers how to document C-SSRS administration and results in the chart, with a fictional, structure-only sample note.
Anyone may administer the healthcare versions per the steward; clinicians own interpretation, assessment, and follow-up
Treating teams and consultants, accreditation surveyors and tracers, quality and risk managers, attorneys reviewing records
4 to 10 chart lines for the screen, plus the assessment a positive result triggers · screener completion 1 to 3 minutes
Structured ideation and behavior classification; category results, not a summed score
Suicide risk screening at intake and transitions, ED triage, inpatient monitoring, VA Risk ID, clinical trial safety monitoring
Free for healthcare use and EHR embedding; copyright held by RFMH; major regulators require a validated process, not this tool by name
The Columbia-Suicide Severity Rating Scale (C-SSRS), also distributed as the Columbia Protocol, is a structured instrument for identifying and classifying suicidal ideation and behavior, developed by Kelly Posner and colleagues for an NIMH-funded adolescent treatment study and first validated at scale in 2011. It is disseminated by the Columbia Lighthouse Project under Columbia University auspices, and its copyright is held by the Research Foundation for Mental Hygiene, Inc. The licensing is bifurcated and widely misunderstood: healthcare and community use is free, including creating setting-specific triage steps and embedding the tool in an EHR, while pharmaceutical-funded clinical trials require a per-protocol license, with the electronic eC-SSRS routed through Clario. Training splits the same way: the steward recommends but does not require training for clinical use, while research raters must train and re-train every two years.
The version family is the first documentation trap, because materially different instruments share the name: brief screeners (Recent and Since Last Contact), full scales (Lifetime/Recent for baselines, Since Last Visit for follow-up), pediatric and very-young-child adaptations, a frequent-monitoring screener for shift-based settings, an inpatient-discharge screener, and self-report versions. "C-SSRS completed" identifies none of them. The second trap is the output: the instrument classifies ideation categories and behavior categories with recency, and it produces no summed total; the steward's own materials put "score" in quotation marks and map affirmative answers to example triage tiers. A defensible note therefore records the category-level result, the local triage decision, and the clinician's overall risk formulation as three different things, because they are.
Hospitals and behavioral health organizations document C-SSRS results inside accreditation-driven screening workflows; emergency departments and inpatient units use the setting-specific screeners at triage, during stays, and at discharge; the Veterans Health Administration mandates the screener internally through its Risk ID strategy, with a comprehensive evaluation required the same day in ambulatory care and within 24 hours in inpatient and emergency settings; schools, corrections, and crisis lines use the community versions; and clinical trials use licensed versions for prospective safety monitoring under FDA draft guidance. The documentation always sits inside a larger chain: a positive screen feeds a suicide risk assessment, mitigation lands in a safety plan, acute encounters generate a crisis note, and the generic screening-layer architecture lives in a risk screening note. This page owns the instrument-specific layer: version, categories, and the record that connects them to action.
No regulator prescribes a C-SSRS note format, but surveyors trace the record as a chain: covered patient, validated tool administered as designed, categorical result, response to a positive, assessment, overall risk with rationale, mitigation, and follow-up. The elements below make that chain visible. Each carries the pitfall that most often breaks it.
Exact version and administration circumstances. Name the instrument and version (Screener Recent, full Lifetime/Recent, Since Last Contact, and so on), the timeframe it covers, the mode, the date and time, the administrator's role, and any language, privacy, or participation limits. Pitfall: "C-SSRS done." Different versions measure different timeframes with different items, and the record cannot be traced, compared, or defended without the version.
The category-level result. Record the highest ideation category identified, the behavior category or "none identified," the relevant recency, and whether the result is positive or negative under the named local protocol. Pitfall: a bare number or color. "C-SSRS score 4" is ambiguous (item, category, intensity subtotal, or local tier?), and "red" without the underlying category and timeframe documents the wall chart, not the patient.
Branching fidelity. Note that all applicable branches were completed, and record anything that limited participation. Pitfall: partial administration read as negative. The instrument's branching is part of its validity, and accreditation guidance warns against altering validated wording or skipping the flow.
The screen-to-assessment chain. A positive screen leads to an evidence-based assessment covering ideation, plan, intent, behavior, risk factors, and protective factors, documented in a suicide risk assessment the same visit. Pitfall: the screen standing in for the assessment. Screening organizations that stop at the screener are among the most-cited survey findings, and a negative screen never closes an inquiry the clinical picture keeps open.
Overall risk level, with the reasoning. Where policy requires a categorical level, document it under the organization's rubric with the clinical rationale: current state against baseline, dynamic and historical factors, protective factors that are specific and currently accessible, collateral information, and foreseeable changes. Pitfall: the screener category pasted in as the overall level. The instrument result, the local triage tier, and the clinician's formulation are three outputs, and reviewers expect the third to carry reasoning, not a color.
Actions completed, not recommended. Document the safety plan work, means-safety counseling at a non-method-specific level, monitoring status, consultation and its outcome, confirmed referral or handoff with a named receiver, disposition, and who owns the next contact. Pitfall: a positive result with no visible response, or a reflex disposition. No authority makes hospitalization or one-to-one observation automatic; the steward's own experience is that 1 to 2% of people screened need more acute care, and the record should show why the chosen response is proportionate.
Reassessment cadence and triggers. Record when re-screening is due under written policy and what changes trigger immediate reassessment. Pitfall: invented universal intervals. "Every encounter" is a Zero Suicide convention, annual screening is a VHA rule inside VA care, and the frequent-monitoring screener exists precisely because cadence is a local policy decision.
C-SSRS DOCUMENTATION BLOCK Version: [Screener Recent / Since Last Contact / full Lifetime-Recent / Since Last Visit / other] Timeframe: [ ] Date/time: [ ] Administered by (role): [ ] Mode: [ ] All applicable branches completed: [yes / limits noted] RESULT (category level, never method detail) Highest ideation category: [ ] Behavior category: [ / none] Recency: [ ] Result under [named local protocol]: [pos / neg] IF POSITIVE -> same-visit evidence-based assessment [ideation, plan, intent, behavior, risk factors, protective factors; documented in the risk assessment note] OVERALL RISK (clinician formulation, org rubric) Level: [ ] Rationale: [state vs baseline; dynamic + historical factors; protective factors and their current accessibility; collateral; foreseeable changes] ACTIONS COMPLETED [safety plan updated / means-safety counseling / monitoring status / consultation + outcome / confirmed handoff to (name)] Disposition + why proportionate: [ ] FOLLOW-UP Next contact owner: [ ] Reassessment due: [per policy] Immediate-reassessment triggers: [status change / new ideation / new behavior / care transition] Clinician signature / credentials: Date:
Free to use and share, no signup. The PDF includes a one-page cheat sheet with element-by-element pitfalls and a pre-sign checklist; the DOCX is the blank documentation block, ready to adapt.
Scenario: an outpatient visit where the brief screener is positive for active ideation without intent, the clinician completes the assessment the same visit, and care continues safely outpatient. The sample shows the documentation structure only: it names categories and actions, and deliberately contains no method, means, or item content. All details are fictional.
Patient: J.L., 41 · Visit: Outpatient follow-up · Clinician: A. Reyes, LPC · Note date: 08/11/2026
Screen: C-SSRS Screener Basic, Recent version, administered verbally in private at 10:20 by this clinician; all applicable branches completed. Highest ideation category: active suicidal ideation without intent. No suicidal-behavior category identified in the screener's timeframe. Result is positive under the clinic's written screening protocol.
Same-visit assessment: An evidence-based suicide risk assessment was completed this visit and is documented in today's risk assessment note: current and historical ideation and behavior, intent, mental status, dynamic and historical risk factors, protective factors and their current accessibility, and collateral information from J.L.'s spouse with consent. No prior suicidal behavior by history or record review.
Overall risk and rationale: Elevated from J.L.'s baseline but managed outpatient under the clinic's formulation rubric. Reasoning: ideation is recent and active but without intent, plan, or preparatory behavior; presentation is engaged and future-oriented; protective factors (children at home, treatment alliance, involved spouse) are current and accessible; the identified stressor is acute and being addressed. The screener category informed but did not determine this formulation.
Actions completed today: Collaborative safety plan updated with J.L.; means-safety counseling provided and arrangements confirmed; supervising psychiatrist consulted, concurred with outpatient management; spouse engaged with J.L.'s consent; crisis and after-hours contacts reviewed, including 988. Disposition: continue outpatient care with increased contact, proportionate to the assessment above.
Follow-up: This clinician owns next contact: telephone check-in within 48 hours and appointment in 5 days. Re-screen at the policy interval, and immediately on any status change, new ideation, new behavior, or transition of care.
This sample is fictional and for educational purposes. It does not describe a real patient or record, shows documentation structure only, and intentionally contains no method or means content. If you or a client needs immediate support: call or text 988 (US), 9-8-8 (Canada), or Lifeline 13 11 14 (Australia).
Writing these after every session? BastionGPT drafts complete notes from bullets, dictation, or a transcript.
Generate a note from bulletsThe C-SSRS is widely adopted, and almost nothing mandates it by name: that distinction shapes defensible documentation. Joint Commission accreditation requires a validated suicide screening tool and names the C-SSRS only as an example, and its 2026 reorganization split the standard's location: hospitals and critical access hospitals now work under NPG.08.01.01 while behavioral health care programs remain under NPSG.15.01.01, with closely aligned elements: screening from age 12 for patients evaluated or treated primarily for behavioral health conditions, an evidence-based assessment after a positive screen covering ideation, plan, intent, behavior, risk factors, and protective factors, a documented overall risk level and mitigation plan, written policy on reassessment and monitoring, discharge counseling and follow-up, and quality monitoring. CMS is expressly tool-neutral: its March 2026 interpretive guidance states it does not endorse or require a particular instrument, while the underlying regulations require care in a safe setting and psychiatric records that justify the degree and intensity of treatment. Inside VA care, PAYER POLICY is specific: the Risk ID strategy uses the C-SSRS Screener with annual universal screening, and a positive screen requires the comprehensive suicide risk evaluation the same day in ambulatory care and within 24 hours in inpatient and emergency settings. LAW appears mostly at the state layer: California requires general acute-care hospitals to screen patients 12 and older under written policies, and a dozen states mandate suicide-specific training for defined professions; none of these makes the C-SSRS itself compulsory. Zero Suicide's every-encounter screening is a CONVENTION worth naming as one in policy documents.
Canada and Australia run the same closed loop with different labels, and one genuine divergence. Accreditation Canada applies Health Standards Organization's suicide prevention program standard as a required practice (the current edition uses Required Safety Practice terminology; older materials say Required Organizational Practice), covering routine screening, assessment, safety planning, transitions, and postvention, all tool-neutral. Australia's NSQHS Comprehensive Care actions 5.31 and 5.32 require systems that identify and respond to suicide and self-harm risk, and the Commission's June 2026 fact sheet, which supersedes the old AS18/14 advisory, explicitly calls screening without resulting action and carrying forward old risk ratings without reassessment low-value care. The divergence: US accreditation requires a documented overall risk level, while NSW Health policy discourages high/medium/low stratification as the basis for care decisions, citing evidence that most people categorized high-risk never die by suicide and about half of deaths arise from lower-risk groups. A defensible chart works in both worlds the same way: record the category-level result, then carry the weight in the formulation and the actions rather than in a tier label. On the instrument itself, keep the validated wording and branching intact, record refusals as refusals rather than negatives, and remember the prediction limits: screening organizes detection and response, and no instrument accurately predicts individual outcomes. If you or a client needs immediate support: call or text 988 (US), 9-8-8 (Canada), or Lifeline 13 11 14 (Australia).
The C-SSRS is disseminated by the Columbia Lighthouse Project under Columbia University auspices; its copyright is held by the Research Foundation for Mental Hygiene, Inc. Healthcare and community use, including EHR embedding, is free under the steward's terms, and clinical-trial use requires a per-protocol license. BastionGPT is not affiliated with, or endorsed by, the Columbia Lighthouse Project, Columbia University, or the Research Foundation for Mental Hygiene. This page reproduces no instrument items, stimuli, or scoring materials.
The evidence sets the tone: this instrument classifies and organizes, it does not predict. In a cohort of 18,684 psychiatric-emergency patients, the screener's one-week discrimination ran near 56.5% sensitivity and 75.6% specificity; across the risk-categorization literature, roughly 95% of patients categorized high-risk do not die by suicide while about half of deaths arise from lower-risk groups; and the first C-SSRS-specific meta-analysis (2026) found meaningful associations with later nonfatal attempts but could not estimate fatal-outcome prediction at all. The steward's own experience is that 1 to 2% of people screened need more acute care. The BastionGPT Clinical Advisory Board sees the same errors most often in C-SSRS documentation reviews:
BastionGPT is specifically trained, tuned, and clinically tested on behavioral health progress notes and screening documentation.
See how clinicians use it day to day on the AI therapy notes page.
Many BastionGPT users report saving more than 90 minutes per day on documentation.
HIPAA-compliant with a signed BAA on every plan. Your data is never used to train models. BastionGPT drafts, you review and sign.
Not in the PHQ-9 sense. The instrument classifies the highest ideation category reached and any behavior category with its recency; research protocols may code categories numerically and total a separate intensity component, but there is no validated summed clinical score, and the steward's own materials put "score" in quotation marks before mapping answers to example triage tiers. A chart entry like "C-SSRS score 4" is ambiguous four different ways. Document the category, the timeframe, and the result under your named local protocol instead.
Use and name the version that matches the clinical job: the full Lifetime/Recent scale when a comprehensive baseline history is needed, the brief Screener (Recent) for front-line identification, Since Last Contact or Since Last Visit forms after a prior administration, the frequent-monitoring screener in shift-based inpatient workflows, the inpatient-discharge screener at that transition, and pediatric or very-young-child adaptations where age requires them. The setting-specific triage layouts change the suggested next steps, not the questions. Whatever you choose, the note records the exact version and timeframe, because results from different versions are not interchangeable.
No. No steward, accreditor, or regulator reviewed for this page makes admission, involuntary treatment, or one-to-one observation the automatic consequence of a positive screen; the steward's published experience is that 1 to 2% of people screened need more acute care. A positive result triggers the locally defined response, normally a same-visit evidence-based assessment, and disposition follows the full formulation, legal criteria, available supports, and the least restrictive safe option, with the reasoning documented. What reviewers flag is not the disposition itself but a disposition with no visible rationale.
Generally no authority requires it by name. Joint Commission requires a validated screening tool and lists the C-SSRS as an example; since 2026, hospitals work under NPG.08.01.01 while behavioral health programs remain under NPSG.15.01.01. CMS states expressly that it does not endorse or require a particular tool. Canada's accreditation requirement and Australia's NSQHS framework are tool-neutral. The clear exception is internal: the VA's Risk ID strategy mandates the C-SSRS Screener within VHA care. So the defensible description is widely adopted to satisfy validated-process requirements, and the defensible chart names the local policy the result was interpreted under.
No, and collapsing the two is the most-cited survey finding in this territory. The screener detects and classifies; the assessment integrates ideation, plan, intent, behavior, risk factors, protective factors, collateral, and clinical judgment into a formulation with a documented overall risk level and mitigation plan. One process can satisfy both roles only when it genuinely covers all the assessment domains. A negative screen does not end the obligation either: when history, behavior, or presentation raises concern, the assessment proceeds anyway, and the note says why.
There is no universal interval. Joint Commission requires each organization's written policy to define reassessment frequency and triggers (status change, new ideation, new behavior); Zero Suicide's every-encounter screening is a convention its adopters choose; the VA's annual universal screening is internal VHA policy; and the steward publishes a frequent-monitoring screener precisely for settings that re-screen by shift or day under local rules. Document the cadence your policy sets, the next due date, and the triggers, and re-screen immediately at material transitions rather than carrying an old result forward; Australian guidance now explicitly calls stale carried-forward risk ratings low-value care.
For healthcare and community use, no license and no fee: the steward's terms allow use, setting-specific triage steps, and EHR embedding without permission, and training is recommended rather than required, with free modules in about 30 languages. The boundaries: pharmaceutical-funded clinical trials require a per-protocol license (copyright sits with the Research Foundation for Mental Hygiene, and the electronic eC-SSRS routes through Clario), research raters must train with re-training every two years, and your own organization or accreditor may mandate training as a competency requirement even though the steward does not.
A refusal is not a negative screen, and the chart should never convert one into the other. Document the exact version offered, that the patient declined or could not participate, any stated reason, the steps taken to make participation easier, the sources used instead (observation, records, collateral), the limits created by the missing self-report, the clinical assessment and safety actions taken anyway, and the reassessment plan. When concern remains despite refusal, follow local law and policy, and record the reasoning; the liability literature is consistent that over-reliance on a single declined self-report is the pattern that fails review.
Yes, for the documentation layer. Give it the category-level facts and it drafts the screen block, the assessment scaffold, the formulation-with-rationale prompts, and the completed-actions and follow-up sentences for your review; it can also check a finished note for a positive screen with no assessment, a level with no reasoning, or a missing version. The clinical decisions (risk formulation, disposition, safety planning) remain yours, and the tool is built to keep them that way. BastionGPT is HIPAA-compliant with a signed BAA on every plan, and your data is never used to train models.
The instrument facts and compliance claims on this page trace to these sources, last verified August 2026:
Educational content, not legal or billing advice. Sample notes are fictional. Follow your organization's policies and your board, payer, and jurisdiction requirements.