CAP Inspection Checklist for Clinical Laboratories

A CAP inspection rarely fails on the science. It fails on the paper trail: a competency that was never documented, a maintenance log with a gap, a corrective action that was closed without evidence it worked. This checklist is organized the way inspectors actually work through a lab — by category, not by checklist question number — so you can walk your own floor before they do.

1. Personnel and competency records

• Every tested employee has an initial competency assessment on file, dated before they went live on that test.
• Competency is re-assessed at defined intervals (typically annual, more often in the first year) using all six required methods where applicable: direct observation, monitoring of QC/PT, blind sample review, review of records, problem-solving assessment, and instrument function checks.
• Roster changes — new hires, terminations, role changes — are reflected in the training file within days, not months.
• Job descriptions match what the person is actually doing at the bench.

Common finding: competency exists for the test method but not for the specific instrument or platform the tech is running today.

2. Quality control records

• QC is run at the required frequency for every active test, with acceptable ranges defined and reviewed before patient results are released.
• Out-of-range QC has a documented corrective action, not just a re-run that happened to pass.
• Lot-to-lot QC verification is on file whenever reagent lots change.
• QC records are traceable to the specific instrument, operator, and result they gated.

Common finding: QC was reviewed and signed, but the corrective action for an earlier failure was never written down — only remembered.

3. Instrument maintenance and calibration

• Preventive maintenance (PM) is performed on the manufacturer's schedule, and every PM event has a completed, dated, signed record.
• Calibration verification is performed at the required frequency (and after major maintenance, reagent lot changes, or QC trending out of range).
• Instrument downtime and repair events are logged, including what was done before the instrument was returned to service.

Common finding: the PM schedule exists as a spreadsheet, but the actual completed logs live in a binder that's missing two months.

4. Document control

• Procedures (SOPs) match what's actually being done at the bench — not last year's version.
• Every SOP has a review date, an approver, and a defined review cycle (most commonly biennial).
• Superseded versions are archived, not just deleted, with a clear effective-date history.

Common finding: the SOP on the shared drive doesn't match the printed copy taped inside the instrument's cabinet.

5. Proficiency testing (PT)

• PT is enrolled for every regulated analyte, submitted on time, and reviewed by the lab director or designee with a signature and date.
• Unsuccessful or unsatisfactory PT events triggered a documented investigation and corrective action — including a check for a pattern across events, not just a one-off explanation.
• PT results are compared against peer group and flagged if they trend toward the edge of acceptable range before a failure occurs.

6. CAPA (corrective and preventive action) history

• Every CAPA has a defined problem statement, root cause, corrective action, and — critically — effectiveness check performed after enough time has passed to know if it worked.
• CAPAs that recur (same root cause, different date) are flagged and escalated rather than closed identically each time.
• CAPA records are retrievable by test, instrument, or date range in under a few minutes — inspectors will ask for this live.

7. Safety and general lab operations

• Chemical hygiene plan, safety data sheets, and biohazard handling procedures are current and accessible at the bench, not just in a binder in the office.
• Incident reports (needlesticks, spills, exposures) are logged with follow-up documented.

A realistic pre-inspection countdown

30 days out — Run a self-audit against this list. Pull a random sample of competency files, QC logs, and CAPA records and confirm they're complete and consistent with each other.

7 days out — Confirm every SOP review date is current. Close any open CAPA that has enough data to support an effectiveness check.

1 day out — Make sure whoever answers the door can pull competency, QC, and CAPA records for any test the inspector names, without needing to ask three other people first.

The pattern behind most findings

Almost every finding above traces back to the same root cause: the evidence exists, but it lives in four different places — a binder, a shared drive, an instrument's local log, someone's memory — and nobody can reassemble the story fast enough when asked. The fix isn't more paperwork. It's one system where competency, QC, instrument maintenance, and CAPA all reference the same roster, the same instrument records, and the same audit trail, so "show me" takes minutes instead of a scramble.