Delta-Check Removal to Improve Autoverification
A 600-bed academic medical center's core laboratory processes 4,200 chemistry specimens daily. Six weeks after go-live of a new middleware autoverification platform, the autoverification rate is 78% against a 90% target. The laboratory medical director proposes removing delta checks for sodium and potassium, which generate 60% of holds but have a 0.4% true-positive rate, to raise the rate. Two recent cases involved pseudohyponatremia in patients with paraproteinemia; a potassium delta flag previously caught a specimen drawn above an IV line. The chemistry supervisor supports removal. The hospitalist group has not been consulted. The LIS cannot yet apply autoverification rules by ordering location, and the vendor says that enhancement is 9 months away.
1. How would you evaluate the proposal to remove the sodium and potassium delta checks? Specify the analyses, stakeholders, and safety considerations.
Read the worked answer
Frame the problem. Delta checks are a post-analytic safety net for specimen integrity and transcription errors; the 0.4% figure is an average that can hide high-yield subgroups. Removing both analytes to hit a throughput target trades a measurable risk for an operational metric.
Analyses to request
- Stratify holds by analyte, ordering location, patient type (inpatient, ED, outpatient, dialysis, oncology), and collection time.
- Define a true positive explicitly: confirmed specimen error, corrected report, recollection, or clinical action. Calculate PPV per 1,000 results and per 1,000 holds, not just overall.
- Review the two pseudohyponatremia cases and all missed or near-miss events; check whether sodium or potassium flags would have caught them.
- Audit a random sample of held and overridden results for harm and delayed treatment.
- Compare with peer laboratories, but do not rely on published rates alone.
Stakeholders and governance. The laboratory director cannot change a safety rule unilaterally. Involve hospital medicine, ED, nephrology, oncology, nursing, patient safety, and the medical executive committee. Document a formal risk assessment and change-control record.
Safety options. Keep at least one flag, use analyte-specific rules, reference change values, and reflex testing (e.g., osmolarity or direct ion-selective electrode) instead of simple removal. Route ambiguous results to a technologist rather than auto-releasing.
2. What decision and monitoring plan would you recommend, including how to address the LIS limitation?
Read the worked answer
Recommendation: do not remove both delta checks. Replace the current rules with a tiered, location-aware set while fixing the LIS limitation.
Interim rule design
- Potassium: retain the delta check for inpatient and ED draws where IV-line contamination is plausible; auto-release small changes within a short interval.
- Sodium: retain a flag for oncology and paraprotein patients and use age- and sex-specific reference change values; add a comment or reflex osmolarity when flagged.
- Route only high-risk flags to manual review; auto-release low-risk changes with an audit trail.
Address the LIS gap. Do not wait nine months. Use middleware rules keyed to collection location or order priority if available; otherwise build a worklist by unit and manually review. Ask the vendor for a written timeline and consider contractual remedies or an interface workaround.
Monitoring plan
- Primary: autoverification rate overall and by location.
- Safety: corrected reports, specimen recollections, missed-result events, delta-check true positives.
- Process: hold volume, technologist review time, turnaround time, override reasons.
- Report monthly to the laboratory and patient safety committees.
If the rate stays below target, examine pre-analytic causes (hemolysis, clots), analyzer flags, rule logic, and staffing before weakening safety rules. The target should not be met by lowering the safety bar.
3. How would you communicate and govern this change, and what would you do if clinicians or the laboratory director disagreed?
Read the worked answer
Communication
- Present the stratified data, not just the overall rate, to the laboratory and hospitalist groups.
- Explain the specific failure modes the flags catch: IV contamination, paraprotein interference, and wrong-patient specimens.
- Describe the interim rules and the plan to restore location-specific logic.
- Set a review date and publish the safety and throughput metrics.
Governance
- Take the proposal to the laboratory utilization or patient safety committee, with the medical executive committee as the final arbiter for a safety-relevant change.
- Document the decision, dissent, and rationale in the change-control record.
- Assign owners for rule build, LIS vendor escalation, and monitoring.
If disagreement persists
- If the director wants removal despite the analysis, ask for a written risk acceptance and escalate to the medical executive committee and patient safety officer.
- If clinicians resist any flags, offer a pilot on selected units with predefined stop rules.
- If the vendor misses the location-specific deadline, consider alternative middleware, a manual worklist, or contractual penalties.
- Re-evaluate after three months with the same metrics; do not let the target override the safety analysis.