Quality Assurance, Validation, and Laboratory Regulatory Standards
Frameworks
Know the regulatory landscape: CLIA '88 (CMS) governs clinical testing; FDA regulates test kits and, historically, laboratory-developed tests (LDTs); CAP provides accreditation and checklists aligned with CLIA. Understand test categories (waived, moderate, high complexity), personnel qualifications, and the requirement that an LDT be validated by the performing laboratory before clinical use. Know the role of the medical director and the quality management system (document control, competency, proficiency testing, corrective action).
Validation vocabulary
Be precise: accuracy (closeness to truth), precision (repeatability/reproducibility), analytical sensitivity (LoD), analytical specificity (cross-reactivity/interference), reportable range, reference interval, and clinical sensitivity/specificity. Validation establishes performance; verification confirms a manufacturer's claims. Understand the elements required for a new assay: specimen types, extraction, controls, LoD studies (often ≥20 replicates at low level), precision across operators/days, interference (hemolysis, bilirubin, drugs), and bioinformatic pipeline validation for NGS.
Traps
Items blur validation vs verification, confuse proficiency testing (external, mandated) with internal quality control, and misstate that CAP replaces CLIA (it does not; CAP is deemed-status accreditation). Watch for questions about when a modification requires revalidation and about the distinction between a reference standard and a calibrator.
Rule of thumb
If a laboratory changes a reagent lot, primer, platform, or bioinformatic pipeline in a way that could affect performance, treat it as a modification requiring documented verification. For NGS, validate both the wet lab and the dry lab, including variant calling accuracy against orthogonal methods.
Worked example
A lab introduces a new somatic variant caller. Validation should include a truth set (e.g., well-characterized cell lines or reference materials) spanning variant types, allele fractions near the LoD, and difficult contexts (homopolymers, low GC). Report sensitivity, specificity, PPV, and reproducibility, and document the minimum allele fraction reliably detected.