Written and maintained by CASRAI Editorial Board
Last updated
General information for laboratory and clinic staff, not medical advice and not a substitute for your own validated procedures.
Short answer
Autoverification is the automatic release of laboratory results, with no human review, when a result passes a set of pre-defined computer rules. Results that fail any rule are held for a person to review. The rules and their limits are set and validated by each laboratory; the published CLSI guidance covers how to design and validate the algorithm, not what the clinical limits should be.
Definition and scope
One peer-reviewed hospital study defines it as “a process of automatic release of laboratory results with no need for human intervention” (Feitosa et al., 2016). The logic usually runs in the instrument, a middleware layer or the laboratory information system; see LIMS vs LIS for how those systems differ.
Two CLSI documents frame the topic:
- AUTO10-A, Autoverification of Clinical Laboratory Test Results (Approved Guideline, first edition, 31 October 2006), specifies recommendations for the design, building, implementation, validation and compliance of autoverification algorithms. It is aimed at information-system vendors, laboratories and instrument manufacturers.
- AUTO15, Autoverification of Medical Laboratory Results for Specific Disciplines (first edition, 11 September 2019; reaffirmed June 2025), gives recommendations for designing algorithms for specific disciplines and types of testing, including guidance on human intervention for automated and manually entered results and on building scalable algorithms from simple to complex criteria.
How autoverification rules work in practice
A result is released automatically only if it clears every rule; any failure routes it to manual review. The worked example below comes from a single hospital laboratory (Feitosa et al., J Bras Patol Med Lab 52(3), 2016). It illustrates the rule types, not a standard, and the limits themselves are local choices.
| Rule type | What it checks (as described in the study) | What happened on failure in that laboratory |
|---|---|---|
| Delta check | Current result against the same patient’s previous result, with limits calculated using a reference change value formula | Above the 99% significance limit: result retained. Between 95% and 99%: released with a supervisor flag |
| Critical values | Used as limits where the automatic release interval would otherwise extend past them | Result handled outside automatic release |
| Instrument flags | Lipemia, turbidity, hemolysis, equipment errors | New sample ordered when an interferent was flagged |
| Linearity limits | Result inside the analytical measuring range | Outside linearity: retained for dilution or investigation |
| Quality control | QC status of the run | QC outside limits: middleware blocks the test |
| Automatic release interval | Result inside the range allowed for automatic release | Unreleased results follow a manual-release flowchart |
Related reading on the quantities these rules use: reference intervals and the critical value reporting policy.
AUTO10-A and AUTO15 compared
| AUTO10-A | AUTO15 | |
|---|---|---|
| Published | 31 October 2006, first edition | 11 September 2019, first edition; reaffirmed June 2025 |
| Focus | Design, building, implementation, validation and compliance of algorithms | Algorithms for specific disciplines and types of testing; human intervention; scalable simple-to-complex criteria |
| Does not cover | Test parameter rules and limits for clinical practice; hardware, interfaces, connectivity, security | Transfusion medicine, microbiology, molecular medicine, anatomic pathology, point-of-care testing; gives no programming language, vendor-specific implementations or analyte-specific algorithms |
| Status on CLSI’s page (checked 1 November 2026) | Archived, no longer reviewed through CLSI’s consensus process; listed as FDA-Recognized | Reviewed and confirmed as suitable to remain published without revision (June 2025) |
Neither document supplies analyte limits, so a laboratory cannot copy limits from a standard; it has to derive and validate its own.
Where it matters for labs, clinics and supply decisions
Autoverification affects what a laboratory needs from its analyzers, middleware and LIS, and which results a person still reviews. A checklist of questions to put to a vendor or to your own informatics team, built from the rule types and scope statements above:
- Which rule types can the system apply: delta checks, critical-value limits, instrument flags, linearity limits and QC status?
- Can the limits be set and changed locally, and is each change recorded?
- Where does a failed result go, and what does the person reviewing it see?
- How will the algorithm be validated before go-live, using the design and validation recommendations in AUTO10-A?
- Does the discipline you test in fall within AUTO15’s scope, or is it one of the excluded areas such as point-of-care testing?
Accreditation and quality context: CLIA quality control requirements and ISO 15189 accreditation.
Common misunderstandings
Autoverification is not a clinical judgment
It applies pre-set rules to a result. It does not interpret the result for a patient.
The limits are not set by the standard
AUTO10-A excludes test parameter rules and limits for clinical practice, and AUTO15 gives no analyte-specific algorithms. Limits come from the laboratory’s own validation.
It does not apply the same way to every discipline
AUTO15 states it is not applicable to transfusion medicine, microbiology, molecular medicine, anatomic pathology or point-of-care testing.
“Automatic” does not mean “no human”
Failed results still go to a person, and AUTO15 includes guidance on human intervention.
FAQ
What is autoverification in a clinical laboratory?
The automatic release of results that pass pre-defined rules, without a person reviewing each one; results that fail are held for manual review.
Which CLSI standards cover it?
AUTO10-A (2006) for algorithm design, building, implementation, validation and compliance, and AUTO15 (2019, reaffirmed June 2025) for specific disciplines and types of testing.
Does the standard tell me which limits to use?
No. AUTO10-A excludes clinical test parameter rules and limits, and AUTO15 gives no analyte-specific algorithms.
Sources
- CLSI, AUTO10-A, Autoverification of Clinical Laboratory Test Results: clsi.org/shop/standards/auto10 (checked 1 November 2026).
- CLSI, AUTO15, Autoverification of Medical Laboratory Results for Specific Disciplines: clsi.org/shop/standards/auto15 (checked 1 November 2026).
- Feitosa et al., J Bras Patol Med Lab 52(3), 2016, DOI 10.5935/1676-2444.20160026: SciELO (checked 1 November 2026).
