Test plan review as a quality practice — peer review of test procedures before execution is standard in established SDLC processes.
Class B applicability added by A1:2015. Terminology shifted from 'verify for correctness' to 'evaluate for adequacy' — broadening the evaluation beyond binary correctness to sufficiency of coverage. [Class B, C]
Documented adequacy evaluation record dated before test execution, with coverage analysis against 5.6.4 content areas and an explicit adequacy conclusion.
Maps to
IEC 62304: §5.6.5 Evaluate software integration test procedures
ISO 13485: §7.3.6 Design and development verification
Pre-QMSR Part 820 (legacy QSR): §820.30(f) Design verification.
Requirement text
The manufacturer shall evaluate the integration test procedures for adequacy. [Class B, C]
Why this clause exists
Test procedure adequacy evaluation is a meta-verification step: it asks whether the tests themselves are good enough to detect the failures they are designed to find. Without clause 5.6.5, a manufacturer could execute a complete set of integration tests that pass while the test procedures are so weak that they would not detect actual integration failures. For safety-critical software, where integration testing is the mechanism for verifying that risk controls function correctly in the combined software item, inadequate test procedures represent a systematic gap — not a random quality problem. The adequacy evaluation forces explicit documentation of why the test procedures are sufficient before relying on their results as compliance evidence.
What changed
The clause title in A1:2015 changed from 'Verify software integration test procedures' (IEC 62304:2006) to 'Evaluate software integration test procedures' (A1:2015), with the verb in the requirement body changing from 'verify' to 'EVALUATE ... for adequacy.' Amendment 1 updated the term from 'correctness' to 'adequacy' throughout the testing clauses, reflecting a shift from binary correctness to a broader judgment of sufficiency. Class B applicability was added.
Common gaps (what we see in audits)
- Test procedure adequacy evaluation not conducted as a distinct step — Integration tests are written and executed without a separate evaluation of procedure adequacy. The test plan approval (if any) confirms process compliance but does not evaluate whether test procedures are adequate to detect relevant integration failures.
- Adequacy evaluation performed by the test author without independent review — The developer who writes the integration test procedures also evaluates their adequacy, creating a conflict where the same person assesses the quality of their own work. For safety-critical systems, independent evaluation provides a meaningful adequacy check.