CHSPS Measures
Standardized measure specifications that let accredited organizations measure quality, safety, integrity and reliability with consistent definitions and auditable data.
Self-reported data are not automatically CHSPS-validated data. A result earns the “validated” label only by passing independent validation — see Validated Performance Data.
What every CHSPS measure must demonstrate
Importance & clarity
Addresses a meaningful quality, safety, integrity or reliability issue, with a precisely defined construct.
Reliability & validity
Consistent enough for its intended use, with evidence it reflects what it claims to measure.
Feasibility & fairness
Reasonable data burden, equity-aware design, and resistance to gaming through classification or exclusion.
Auditability
Every result traceable to objective source data an independent validator can verify.
The measure lifecycle
Every measure moves through a controlled lifecycle, and its current status is always displayed so a developmental or reporting-only measure cannot be mistaken for a mature comparative indicator.
| Stage | What happens |
|---|---|
| Concept & review | Anyone may propose a measure; the Measures Advisory Function advances, revises, holds or rejects it. |
| Specification | Exact denominator, numerator, exclusions, timing, data sources, stratification, scoring direction and data-quality rules. |
| Feasibility & field testing | Data availability, burden, reliability, validity, distribution, exclusion behavior, fairness and gaming potential. |
| Public comment & approval | Rationale, specification and testing summary published for comment; an independent governance body approves final use. |
| Confidential first cycle | New measures ordinarily complete at least one confidential reporting cycle before any public comparative use. |
| Validated public reporting & benchmarks | Only after evidence supports the intended use — and benchmarks only with adequate validated volume and meaningful variation. |
| Maintenance & retirement | Reevaluated on a cycle; retired when topped out, invalid, burdensome or obsolete. Versions are never overwritten. |
Measure status labels
Intended-use categories are equally explicit: gating indicator, monitoring, public performance, benchmark, recognition — and diagnostic, for measures like near-miss reporting and appeal overturns that must never be read as simply higher- or lower-is-better.
The flagship: CPDS
CPDS — The Credential Performance Data Set
The core cross-pathway measure set with fixed definitions, one annual reporting season, independent validation and disciplined publication — a published comparability framework for credentialing. Twelve CPD measures are published in Specification 2026.1 with controlled definitions, calculation rules and VPD validation requirements.
The dashboard demonstrates the published reporting structure. Validated public results and empirical benchmarks are governed separately by the VPD and benchmark methodologies.
Read the CPDS conceptFramework measure families 2026.1
Each specialty accreditation standard includes a defined 2026.1 measure family with controlled calculation and validation requirements. Empirical benchmarks are published only after sufficient validated reporting volume exists.
AIM · Assessment Integrity
Published measures for identity verification, security events, technology interruptions, accommodations, incident response, appeals, proctor competence and automated-system oversight.
REM · Remote Examination
Published measures for session continuity, identity exceptions, proctor response, platform disruptions, recovery, accommodations and remote-delivery incident handling.
WCM · Workforce Competency
Published measures for credential currency, initial competency before independent assignment, high-risk skill currency, remediation, temporary-workforce equivalence and record accuracy.
MSM · Medication Safety
Published measures for reconciliation, high-risk safeguards, medication-technology conformance, preventable harm, event learning, recall response and downtime recovery.
CEM · Clinical Education
Published measures for objective-assessment alignment, content currency, faculty disclosure, completion-record accuracy, accommodation fulfillment, remediation and transfer to practice.
AIG · AI Governance
Published measures for AI-system inventory, validation currency, human-oversight conformance, change control, incident review, performance drift and critical vendor governance.
Performance recognition — the maturity ladder
Comparative recognition arrives in deliberate stages: accreditation only → validated measure profiles → aggregate benchmarks → performance distinction → comparative rating — each step gated on data volume, validity and independent governance approval. CHSPS will not introduce star ratings until sufficient validated data support fair and meaningful comparison, and a fee can never purchase recognition.
Published operating bands support consistent interpretation.
CHSPS uses a three-band display model: Preferred operating band, Watch band and Review trigger. These are controlled CHSPS Reference Criteria, distinct from empirical benchmarks and not automatic pass/fail thresholds unless a normative clause expressly incorporates them.