Autoverification Change Control and a Missed Hemoglobin Delta Check
During a CAP inspection of a 420-bed hospital laboratory that performs 2.1 million tests annually, the inspector reviews autoverification. Seventy-eight percent of chemistry results are autoverified. Three months earlier, middleware rules were revised; the hemoglobin delta-check rule now applies only to inpatients. An outpatient had Hb 14.2 g/dL at 08:10 and 6.8 g/dL at 11:00 on the same day; both results were autoverified and released, and the patient was transfused in the emergency department. The inspector also finds that two technologists' competency records for autoverification were not updated after the change. Proficiency testing for hemoglobin is acceptable. The laboratory director must respond to the inspection findings and prevent recurrence. The vendor says the rule change was a 'minor configuration update' and did not require validation.
1. What is the most likely root cause, and what immediate patient-safety actions should be taken?
Read the worked answer
The root cause is a change-control failure, not a failure of the hemoglobin assay. The middleware rule was narrowed to inpatients, so an outpatient with a physiologically implausible fall from 14.2 to 6.8 g/dL in under three hours was autoverified without delta-check review. The vendor's 'minor configuration' label is irrelevant to patient risk; any change that alters autoverification logic requires risk assessment and validation.
Immediately, I would suspend autoverification for outpatient hemoglobin or for the affected rule set until the logic is corrected. I would perform a look-back over all outpatient CBCs since the go-live date, identify other missed delta checks, and review whether any result was acted on. The clinician caring for this patient should be notified that the result was released without delta-check review, and the specimen should be investigated for wrong-patient labeling, dilution, clot, or analyzer error. If the low result was analytically valid, the transfusion may have been appropriate; the quality problem is that the system did not prompt verification.
2. Design a corrective and preventive action plan, including validation of the revised rule.
Read the worked answer
The CAPA should correct the rule, validate it, and prevent recurrence. Correction: restore the hemoglobin delta check for all patients, with separate limits for inpatients and outpatients if clinically justified. Validation should use a retrospective dataset of paired CBCs that includes outpatients, known analytic errors, and true physiologic changes; I would define the number of pairs in advance, measure the rule's sensitivity for known errors and its false-positive review rate, and document acceptance criteria. A vendor configuration change should be treated as a major change with version control, approval, and revalidation.
Prevention: revise the change-control SOP to require risk assessment for middleware rules, add autoverification to the competency program, train the two technologists, and audit outpatient delta-check performance monthly for at least three months. Proficiency testing is necessary but not sufficient because it tests the assay, not the middleware logic. The CAPA should name an owner and a completion date and should include a follow-up audit.
3. At the inspection, how should the laboratory characterize and respond to the finding?
Read the worked answer
I would characterize it as a deficiency in autoverification validation and personnel competency, with a potential for patient harm, but not automatically as immediate jeopardy unless the unsafe rule remains active during the inspection. The response should be factual: describe the middleware change, the missed delta check, the look-back results, and the immediate suspension. The laboratory should provide evidence of the corrected rule, validation data, training records, and the audit plan.
I would not argue that acceptable proficiency testing exonerates the laboratory; PT does not cover autoverification logic. The response should also address disclosure and risk management for the affected patient. If the look-back identifies other missed results, the laboratory should report that scope and any patient notifications. The goal is to show systemic correction, not just a promise to be more careful.