C. difficile Positivity Doubles After a Testing Change
A 280-bed hospital's C. difficile testing moved from a two-step algorithm (GDH antigen plus toxin EIA, with NAAT only for discordant results) to standalone NAAT on November 1. Over the next quarter the NAAT positivity rate rose from 7% to 15%, and the laboratory reports 41 positive tests per month versus 19 previously. Chart review of 30 consecutive positive results shows that 13 patients had received laxatives within 24 hours, 6 had formed stools at the time of collection (accepted by the laboratory and tested), 5 had been tested within 7 days of a prior negative result, and 4 were asymptomatic on admission and were screened as part of an admission order set. The CMO wants a C. diff outbreak task force convened.
1. Is this an outbreak? How do you separate increased detection from increased transmission?
Read the worked answer
Not on this evidence. The timing is the tell: positivity doubled immediately after a test-method change on November 1. NAAT detects toxigenic C. difficile DNA at a far lower limit of detection than toxin EIA, and a two-step algorithm deliberately filters out toxin-negative/GDH-positive specimens. Some of that increase is real disease that the old algorithm missed - but a large share is colonization and testing artifact.
Build the right comparison before calling a task force:
- Plot tests performed per 1,000 patient-days, not just positives. If test volume rose alongside positivity, ordering behavior changed.
- Plot positivity separately; a rising numerator with a rising denominator is a different problem from a rising numerator with a flat denominator.
- Compare severe outcomes: ICU transfer for CDI, colectomy, CDI-attributable mortality, and 30-day readmission. Transmission with unchanged severity suggests detection bias.
- Check clinical case review on a sample: does the patient have diarrhea, leukocytosis, fever? Was an alternative explanation present?
- Look for true clustering: new cases in the same unit and overlapping dates, not just an aggregate rise.
Only the combination of rising rates plus rising severity plus unit-level clustering supports a transmission investigation. What you have here is a diagnostic stewardship failure with a possible real-disease component - and it needs fixing either way.
2. Which of these 30 positives count as NHSN laboratory-identified (LabID) CDI events, and what are the reporting implications?
Read the worked answer
LabID CDI is a laboratory-result-based definition. It counts a positive C. difficile toxin or molecular assay on an unformed stool specimen; clinical symptoms are not required, and the definition does not depend on whether the treating clinician agrees the patient has disease. Key rules:
- Only the first positive per patient per 14 days counts; a second positive within 14 days of the first is not a new LabID event.
- Present-on-admission (POA) applies if the specimen was collected on hospital day 1-3; hospital-onset (HO) if collected on day 4 or later. The 4 asymptomatic patients screened on admission would be counted as POA, not HO - they inflate the POA rate, not the HO rate the CMO is reacting to.
- Formed stools should be rejected by the laboratory, not reported. If the lab tested and reported a formed stool as positive, that is a laboratory policy failure, and once reported it generally must be counted - you cannot retroactively exclude it.
- The 5 repeat tests within 7 days would not generate additional LabID events if they follow an earlier positive within 14 days.
Implications: your HO-CDI standardized infection ratio may rise for reasons unrelated to care quality. That must be communicated to leadership and to the publicly reported data - but do not manipulate the definition. Fix the ordering and laboratory rejection criteria going forward, and document the method change as a confounder in your internal trend analysis.
3. Design the intervention package and the metrics that will tell you whether it worked.
Read the worked answer
Interventions - make the right test hard to order:
- Laboratory rejection criteria, hard-coded: reject formed stools, reject specimens from patients on laxatives within 48 hours, reject repeat testing within 7 days of a prior result, reject rectal swabs, and require the specimen to meet the Bristol 5-7 (unformed) standard.
- Order-set gating: C. difficile NAAT requires at least 3 unformed stools in 24 hours and no laxative in the prior 48 hours, attested by the ordering clinician or captured by nursing flowsheet data.
- Nurse-driven triage protocol so the test is ordered appropriately at the bedside rather than reflexively at 2 a.m. for any loose stool.
- Reconsider the algorithm. A two-step approach (GDH plus toxin, NAAT only for discordant results) reduces overdiagnosis of colonized patients. This is a laboratory-medical staff decision with stewardship input; document the rationale.
- Remove C. difficile from the admission order set unless a specific indication exists.
- Antimicrobial stewardship review of high-risk agents, since antibiotic exposure is the dominant modifiable driver of true disease.
Metrics, measured monthly with a run chart:
- Tests per 1,000 patient-days (should fall).
- Percent of tests meeting appropriateness criteria on audit (should rise).
- Positivity rate (will change when the denominator is cleaned up - interpret alongside volume).
- HO-CDI LabID rate per 10,000 patient-days and the standardized infection ratio.
- Severe CDI events, colectomy, and CDI-attributable mortality - the safety counterbalance that proves you did not suppress necessary testing.
Report back to the CMO with the test-volume chart first; it usually resolves the 'outbreak' question without a task force.