Out of Specification (OOS)

OOS Investigation Procedure for GMP Labs

An OOS investigation in a GMP lab follows a defined two-phase procedure set by FDA's 2006 guidance and reinforced by the United States v. Barr Laboratories decision. The original failing result must be investigated before any retesting is considered, the procedure must be documented end-to-end, and batch disposition must be supported by evidence — not by repeated testing until a passing value appears.

At a high level, the procedure runs: lock the original result, execute Phase 1 (laboratory investigation), execute Phase 2 (full-scale investigation) if no assignable lab cause is identified, document a scientifically justified retest hypothesis when retesting is appropriate, and close with batch disposition and CAPA linkage for any systemic root cause.

See the In Practice tab for the full phase-by-phase steps, the Key Controls tab for the controls that keep the procedure inspection-ready, and the Regulatory Context tab for the cited FDA, EU GMP, USP, and MHRA references.

What Out of Specification means

An Out of Specification (OOS) result is a test result that falls outside the specifications, acceptance criteria, or control limits established in a registered product specification, compendial monograph, or validated analytical method.

OOS is distinct from Out of Trend (OOT) — a result inside specification but trending in a concerning direction. Both require investigation, but OOS carries higher urgency and regulatory scrutiny.

Why OOS is so heavily regulated

The 1993 United States v. Barr Laboratories decision fundamentally shaped how pharmaceutical labs must handle failing results. The court ruled that retesting without investigation to dismiss an OOS is not permissible. FDA's 2006 guidance codified the expected two-phase investigation process.

OOS mismanagement — testing into compliance, inadequate investigation, late reporting — is one of FDA's most common data integrity findings. Warning letters frequently cite OOS handling as evidence of broader quality failure.

Where OOS is regulated

OOS handling expectations appear in:

• FDA Guidance for Industry: Investigating OOS Test Results (2006) — the primary framework
• 21 CFR §211.192 — Investigation of unexplained discrepancies
• 21 CFR §211.194 — Laboratory records requirements
• EU GMP Chapter 6 — Quality control laboratory operations
• USP <1010> — Analytical data interpretation and treatment
• MHRA Data Integrity Guidance — ALCOA+ applied to laboratory records

The two-phase OOS investigation

FDA's 2006 OOS guidance defines a two-phase investigation process:

• Phase 1 — Laboratory Investigation: analyst reviews calculations, equipment, method execution, sample prep, and any obvious lab error. Performed immediately by the analyst and supervisor.
• Phase 2 — Full-Scale Investigation: if no assignable lab cause is found, expand to the full quality system — manufacturing, raw materials, equipment qualification, environmental controls.
• Hypothesis & Retesting: only if a scientifically justified hypothesis is established, additional testing can be performed. Retesting requires a pre-approved protocol.
• Batch Disposition: based on investigation findings — release, reject, rework, or reprocess (if permitted).
• CAPA linkage: systemic root causes trigger CAPAs across the quality system.

What good OOS controls look like

Robust OOS programs have:

• Immediate lock of the original result — no overwrite in the LIMS
• Mandatory phase 1 investigation within a set time (commonly 24 hours)
• Controlled retesting protocols — QA pre-approval required
• Formal hypothesis documentation before any retest
• Batch hold automatic pending closure
• Trend analysis linking OOS results across batches, methods, analysts
• Full ALCOA+ data integrity over all raw data and calculations

How Complere handles OOS

OOS results are ingested as a dedicated Deviation sub-type in Complere, with structured phase 1 / phase 2 templates aligned to FDA 2006 guidance.

• Immutable original result — captured with full ALCOA+ metadata
• Phase 1 and phase 2 templates drive consistent investigation depth
• Retesting protocol approval flow with e-signatures
• Batch hold propagates automatically across dependent records
• Trending dashboards by product, method, analyst, and root cause
• CAPA escalation with one-click linkage
• Complete audit trail suitable for 483 response and inspection evidence

Related terms

OOS is part of a broader set of laboratory and quality investigation terms:

• Deviation — broader category including OOS events
• CAPA — systemic follow-up to OOS root causes
• Root Cause Analysis — method used inside phase 2 investigation
• Audit Trail — record of every test, retest, and calculation
• Validation — method validation baseline for interpreting results