Worked Example — CHAI Risk Categorization Tool (v3)

Subject: CKD Care Orchestration Framework (portfolio component 01). A completed pre-deployment risk categorization across both CHAI risk domains, with each rating tied to the CRRF gate that mitigates it at run-time — demonstrating the two-tier gating model end-to-end.
Ratings are made honestly against the framework as designed; where the honest answer is "not yet knowable," the CHAI "Need More Info" affordance is used rather than a false Low.
Use case: Longitudinal, multi-agent CKD→ESKD care orchestration; clinician-facing decision support
Primary users: Nephrology + clinical informatics; human confirms every action
Decision posture: Assistive — "not an autonomous prescriber" (Framework §24)
Data status: Runs on a fully synthetic 18-patient cohort; not validated for clinical use
Deployment scope (assumed for scoring): Nephrology service line, pre-production pilot
SaMD status: Not asserted; if it qualifies as SaMD, FDA guidance governs and supersedes

Result: 5 modifiers rated High → a rigorous risk assessment is triggered

Per the Tool's rule, any single High warrants a detailed hazard/harm/probability assessment before deployment. The Highs cluster around two honest themes: clinical proximity + consequence (the system sits in the care-planning path for a high-stakes disease), and the not-yet-built operational layer (runtime monitoring / audit trail). Neither is a reason not to proceed; both define the mitigation and build agenda.

High: LPS-1 Distance From Patient · LPS-3 Consequences of Failure · LPS-5 Monitoring Difficulty · LPS-9 Population/Disparity · T&D-9 Monitoring & Incident Detection.   Deliberate Lows (design strengths): LPS-2 Decision Autonomy · T&D-1 Sensitive Data (synthetic) · T&D-10 Detection & Traceability.

Domain 1 · Life & Patient Safety

#Risk modifierResponseRationale / evidenceCRRF mitigation (Tier 2)
1Distance From PatientHigh The framework sits inside clinical reasoning — trajectory placement, gap detection, next-best-action for nephrology care. This is "semi-direct involvement in patient care… used by a healthcare professional as part of broader clinical judgment," the CHAI High definition. Gate 7 — authority is earned per output, never presumed by proximity
2Decision AutonomyLow Assistive by explicit design: "It should not behave like an autonomous prescriber" (§24). AI informs/recommends; the human reviews and decides; risk rests on the human reviewer. A deliberate, documented choice. Gate 7 + six-tier ladder caps outputs at advisory (T4) absent clean passage
3Consequences of FailureHigh Worst-case severity in CKD/ESKD is high: a missed sustained stage-4 designation, an unsafe renal dose carried forward, or delayed access planning can contribute to permanent harm. Human review lowers probability, not worst-case severity. Gates 4–5 — interpret before use; abstain if insufficient
4Use Context & ComplexityMedium Use context is outpatient nephrology follow-up (non-emergency) — argues Low — but the population is medically complex and often unstable (multi-comorbid CKD, MGUS, HF, progressing stage 4+). The complexity axis pulls it to Medium. Gate 1 — declare the question and the patient's context up front
5Level of Difficulty Monitoring OutputHigh No real-time monitoring exists; longitudinal narrative output is hard to monitor at scale. Framework §21 lists "human override and audit trail" as a known gap. Per CHAI, monitoring "needs to be developed before implementation" = High. Gate 9 — learn under governance (principle present; tooling is the build item)
6React TimeMedium Most outputs feed longitudinal planning reviewed at 3–4-month cadence, giving reaction time (argues Low). But some surfaced signals (e.g., critical hyperkalemia, acute deterioration) would need faster response — Medium is the honest blend. Gate 8 — outputs ship with reversal conditions the clinician can act on
7Breadth of Potential HarmMedium Affects a moderate CKD population at pilot scope. Note: breadth scales with deployment — enterprise-wide rollout would move this toward High and should be re-scored at that decision point. Gate 6 — cross-stratum coherence before any population-level claim
8Cross-System Propagation RiskMedium The External Data Reconciliation Agent ingests ADT, discharge summaries, outside labs, and claims, so the system is integrated — but outputs are advisory and clinician-gated, which limits automated cascade. Integrated, limited connections = Medium. Gate 3 — provenance/currency checks on every ingested feed
9Population Sensitivity / Disparity RiskHigh CKD carries documented equity hazards: transplant-referral disparities, historical eGFR race adjustment, and unequal access to modality planning. A tool touching candidacy and access could reinforce inequities if not equity-gated. Significant disparity potential = High. Gate 6 — subgroups must cohere; Gate 8 exposes disparity-relevant assumptions

Domain 2 · Technology & Data

Several Technology & Data modifiers are architecture-dependent. Because the framework is a pre-production demonstration on synthetic data, the honest current rating is Low, with an explicit Need More Info flag marking where the score must be re-taken against the production architecture — using CHAI's own abstention affordance rather than asserting a Low that the production system has not earned.

#Risk modifierResponseRationale / evidenceCRRF mitigation (Tier 2)
1Use of Sensitive DataLow→ High at production Runs entirely on a synthetic cohort with documented ground truth — CHAI's Low definition ("synthetic, nonconfidential"). On real PII/PHI this becomes High. The synthetic-first design is a deliberate risk-lowering choice worth surfacing to governance. Gate 3 — sensitive-data handling is part of "what we have"
2Accuracy of DataMedium Designed to consume messy real-world coding (it explicitly harmonizes heterogeneous proteinuria measures and labels med-list confidence). Synthetic data is controlled today, but the operational target is Medium-accuracy source data. Gate 4 — a code is not a phenotype until validated
3Completeness of DataMedium Longitudinal CKD data is notoriously fragmented across systems. The framework's premise is to detect and flag gaps rather than paper over them — a mitigant that keeps this at Medium rather than High. Gate 3 — completeness/gaps vs. what the question expects
4Veracity of DataMedium External feeds (claims, outside labs, ADT) have variable provenance and refresh cadence. The data-confidence model (Confirmed/Probable/Conflicting/Missing/Outdated) is the internal control. Gate 3 — provenance, currency, trustworthiness
5Data TransparencyLow→ Med at production The synthetic cohort is fully documented (canonical Patients repository + ground truth), so provenance is currently complete. Production data-source documentation would need to be assembled — re-score then. Gate 8 — provenance is shown, not assumed
6Sufficiency & RepresentativenessMedium 18 synthetic patients are deliberately diverse but small and not a population sample; adequate for demonstration, not for generalization or model training. Representativeness of the operational data is the real test. Gate 2 + Gate 6 — declare needed data; check cohort coherence
7AI Model Security VulnerabilitiesLow→ Med at production Today a self-contained, application-bound HTML demo with no external interfaces (CHAI Low). An EHR-integrated, internal-facing production deployment would be Medium — a decision-point re-score. Gate 9 — governed deployment surface
8AI Model Lifecycle ManagementMedium Strong version control (git repo, changelog, versioned specs) — but no formal retraining/validation/rollback governance for a deployed model. The system is largely rule/heuristic-based, lowering drift risk but not removing the need for governed updates. Gate 9 — versioned, auditable knowledge layer
9AI Monitoring, Incident Detection & ResponseHigh No runtime monitoring, anomaly detection, or clinician-override audit trail is built (§21 gap). CHAI High: "monitoring is minimal or absent… no clear separation between operational incidents and AI safety incidents." This is the single most actionable pre-deployment gap. Gate 9 — the gate whose tooling must be built to clear this
10AI Detection & TraceabilityLow Traceability is a design default: rationale visibility on hover/click, explicit data-confidence labels, and separation of observation / interpretation / action. AI influence is legible to the clinician end-to-end — CHAI Low. Gate 8 — "show the work" is the framework's core output contract
Reading the result. The categorization is not a verdict against the framework — it is the map of where oversight must concentrate. Two Highs (Distance, Consequences) are intrinsic to doing anything useful in CKD care and are answered by the Assistive posture (LPS-2 Low) plus the CRRF authority ladder. One High (Population/Disparity) names a real equity obligation that Gate 6 must enforce with subgroup checks on the synthetic cohort. Two Highs (LPS-5, T&D-9) are the same finding from two domains — the monitoring/audit layer is not yet built — and become the top pre-deployment build item. Honoring the portfolio ethos, the categorization states plainly what is not yet known (the production-dependent Need-More-Info flags) rather than assigning a Low the production system has not earned.
Licensing & scope. Structured on the CHAI Risk Categorization Tool (v3), CC BY-NC-ND 4.0, Coalition for Health AI. This is an independent, internal application of the Tool to a portfolio framework — not a reproduction or modification of CHAI's document. Ratings reflect the framework as designed for demonstration and are a governance aid, not clinical, legal, or regulatory determinations. If the solution qualifies as Software as a Medical Device, FDA regulation governs.