LIS Autoverification Rules — A Practical Guide

Autoverification turns approved laboratory policy into repeatable release decisions. When every configured business rule passes, eligible requisitions can release without manual intervention; anything that fails stays in the queue for human review. Poorly designed rules can release results the lab cannot defend in an inspection. This guide walks through how clinical labs structure LIS autoverification rules in production — what the rules actually check, how they interact with QC, where the regulatory frame sits, and what evidence the lab needs to keep.

What autoverification is

Autoverification is the automated release of a clinical-lab result without a technologist clicking “verify” — when, and only when, the result satisfies all defined criteria the lab has approved in advance. The LIS evaluates each result against a configured rule set; pass → release to the chart; fail → hold for manual review.

Three things have to be true for autoverification to be defensible:

1. The rules are written down, version-controlled, and approved by the laboratory director before they are activated.
2. Every result the LIS releases under autoverification carries an audit-trail record that names the rule version, the QC status of the run, and the result values that drove the decision.
3. The hold conditions are designed to catch the failure modes that matter — analytical, biological, and patient-specific — not just the obvious out-of-range values.

The promise of autoverification is throughput plus consistency. The risk is releasing a result that a technologist would have caught. The whole rule-design exercise is calibrating that trade-off.

How autoverification rules are structured

A typical autoverification rule has four kinds of checks, evaluated in order:

• QC status checks. The most recent QC run for the analyte must have passed the lab’s Westgard / Levey-Jennings rule set on the relevant control level. If the morning QC failed and was not corrected, autoverification is disabled for that analyte until QC is restored.
• Analytical range checks. The result is inside the analytical measurement range for the assay (above the detection limit, below the upper linear bound). Out-of-range → hold for dilution or repeat.
• Reference range and delta checks. The result is within the reference range for the patient’s demographics (age, sex, sometimes pregnancy status), OR if outside, the delta-check against the patient’s prior result is within an expected band. Large deltas → hold for confirmation.
• Patient-context checks. The patient is not flagged for special handling (e.g. dilutional studies, factor deficiencies, known interfering substances), and the specimen is not flagged (e.g. hemolyzed, lipemic, icteric).

When all four sets pass, the result auto-releases. When any one fails, the result holds with a coded reason the technologist sees on review.

The rule set is per-analyte and per-instrument, not global. The criteria for autoverifying a CBC differ from those for a comprehensive metabolic panel, which differ again from those for a coagulation panel.

CLIA and CAP framing

CLIA (42 CFR Part 493) does not prohibit autoverification — it requires that the laboratory director establish, approve, and periodically review the rules, and that the LIS produce defensible audit-trail evidence for each released result. The relevant subsection is 42 CFR 493.1281, which addresses test-report responsibilities; the CAP General Laboratory checklist (and the discipline-specific chemistry, hematology, and microbiology checklists) builds on this with explicit autoverification line items.

CAP inspectors typically ask three questions:

1. Where is the documented autoverification policy? A signed document by the laboratory director, listing analytes covered, rule structure, exceptions, and the review cadence.
2. How does the LIS prove autoverification fired correctly for any given result? Audit log entries that name the rule version, the QC state at the time of release, and the values that drove the decision.
3. When was the rule set last reviewed? A documented review at least annually, and after any major reagent lot, instrument, or assay change.

A clean answer to all three is the bar. Autoverification rules without those three pieces of evidence are operationally fine but inspection-fragile.

Westgard, Levey-Jennings, and the QC interplay

Autoverification depends on having QC running and passing. The Westgard multi-rule framework (1-2s, 1-3s, 2-2s, R-4s, 4-1s, 10-x and variants) is the standard for evaluating each QC run; Levey-Jennings charts are the visualization. The interaction with autoverification is:

• QC pass on the level relevant to the patient result → autoverification eligible for that analyte
• QC fail in any active rule → autoverification disabled for that analyte until corrective action is documented and a passing QC repeats
• QC overdue (the lab’s defined run cadence has elapsed without a new QC) → autoverification disabled

This is why “the QC failed but the patient results look fine” is not an acceptable basis for releasing under autoverification. The rule structure has to enforce the dependency programmatically.

Two practical pitfalls:

• QC ran but on the wrong level. Some analytes have multiple QC concentrations (low, normal, high). Autoverification should only fire when QC at the level relevant to the patient result has passed.
• QC failure that was overridden. Overrides happen — a manufacturer-issued lot-specific shift, for example. The override needs to be documented with a signed reason; the LIS should not auto-fire on a level whose last QC was overridden until the override expires or the next clean QC posts.

The QC LIS software feature page covers how LIMS IQ evaluates and documents QC, and the Westgard rules and failed-QC review post walks through Levey-Jennings charts, multirule evaluation, batch-review alerts, and SOP-controlled sample holds.

When manual review is required

Autoverification is the wrong tool for some scenarios. Standard hold-for-review categories include:

• Critical / panic values. Results outside life-threatening thresholds — typically held for explicit clinician notification, not auto-released
• Delta-check failures. Patient’s current result differs from their last result by more than the analyte’s defined delta band, where biological plausibility is in question
• Specimen integrity flags. Hemolyzed, lipemic, icteric, or otherwise compromised specimens
• First-time results for a patient. No delta-check baseline yet; many labs prefer manual review on the patient’s first encounter
• Specific analytes by lab policy. Some labs choose to manually review all results for high-stakes analytes (troponin, glucose under certain protocols, blood gases, etc.)
• Investigative cases. A clinician has requested a hold while a workup is in progress
• Outliers from rules-based logic. Reflex testing is in progress, or a calculated result depends on a held component

The right autoverification rule set explicitly hands these to a human. Trying to autoverify them is where labs get into trouble.

See the critical-value reporting workflow for the nine-stage notification, read-back, and escalation process that takes over once a critical value is held back from auto-release.

Implementation considerations

When a lab is building or revising an autoverification rule set, a few patterns recur:

• Start with the high-volume analytes that drive throughput. Common chemistry panels and CBCs are usually the first wins. Coagulation, immunoassay, and specialty molecular tests typically come later.
• Pilot in shadow mode first. Let the LIS evaluate the rules without actually releasing — the technologist still verifies — and compare. This surfaces edge cases before they cause a real-world release.
• Build the review report. A daily or weekly report that shows the autoverification rate, the hold rate, the override rate, and the rule version. If those numbers can’t be pulled cleanly, the rule set is not inspection-ready.
• Coordinate with reflex logic. Reflex testing (positive screen → confirmatory) needs to interlock with autoverification — the autoverification rule for the screen needs to know whether to hold pending reflex or release. See the molecular reflex testing guide for how the LIS generates the reflex order and binds it to the original specimen once the screen is verified, and the toxicology reflex testing guide for the presumptive-screen-to-LC-MS/MS-confirmation cascade.
• Document the exclusions. The list of analytes the lab explicitly does NOT autoverify is as important as the list it does. Maintain both.

Validation: an autoverification rule set is a system control under CLIA and CAP, so it follows the lab’s validation playbook for system changes. New rules → test data → expected vs actual → signed evidence → activation. See the LIS validation guide for the broader validation framework that autoverification sits inside.

Common pitfalls

• Autoverifying without delta-check. A first-pass rule set that doesn’t compare to prior results misses true patient-state shifts. Delta is one of the highest-yield hold criteria for catching real changes.
• Forgetting about reference-range edge populations. Pediatric, geriatric, and pregnancy reference ranges need to be in the rule set; using adult ranges across the board produces false holds and false releases.
• Letting rules drift away from the lab director’s approval. The director approves the rule set at activation, but ongoing tweaks happen on the bench. Every change needs the same approval trail — a quick “I just adjusted a delta threshold” message is not enough.
• No annual review. Reviews lapse silently. Inspectors notice.
• Mixing autoverification with manual override. A technologist’s “I override this hold” carries the same weight as autoverification firing — the audit trail needs to capture both with equal rigor.

Where LIMS IQ fits

LIMS IQ runs configurable rule-based autoverification across chemistry, hematology, microbiology, molecular, and toxicology workflows. The rule engine evaluates analytical range, reference range, delta-check, and patient/specimen flags before releasing, with full audit-trail capture per release event. Rule sets are version-controlled and tied to the laboratory director’s approval record. See the rules-based resulting and autoverification feature page for the platform detail, the QC LIS software feature page for automatic QC evaluation and failed-run review, the hematology LIS guide for how delta checks and reflex slide review apply to CBC and differential autoverification, the chemistry LIS guide for how result rules and documented QC review work together on the chemistry bench, and the immunology LIS guide for how immunoassay autoverification and ANA reflex rules apply to autoimmune and serology result release.

The broader context — accessioning, instrument integration, reflex logic, billing handoff — sits in the clinical LIS guide and the cloud LIS software overview.

Next step

If you are designing or revising an autoverification rule set for a clinical, toxicology, or molecular lab, the highest-leverage next step is a working session that maps your analytes, QC framework, and clinician expectations against a real rule structure. Request a demo and we will walk your team through how LIMS IQ handles autoverification across your test catalog and what the audit-trail evidence looks like for a CAP inspection.