: Impact of a Change in Testing Method for Clostridium difficile Infection (CDI) on Rate of Diagnosis, Physician Practices and Costs

Background: The laboratory diagnosis of CDI is imprecise and more efficient means of diagnosis are needed. Objective: Evaluate if a change in the assay for CDI results in an alteration in the diagnosis of CDI, physician ordering practices, or cost. Methods: In mid-July of 2009, the Wampole C. diff Quik Chek Complete (Inverness Medical, Princeton, NJ), a combined assay detecting glutamate dehydrogenase (GDH) and toxins A and B, replaced the Premier Toxin A+B  Assay (Meridian Bioscience, Cincinatti, OH).  Communications to personnel described the GDH assay test characteristics and recommended usage.  C. difficile testing of inpatients over two 2 month intervals, one before introduction of the new assay and one after, were reviewed and analyzed. Results: During May-June 2009 (Period 1), 49 of 702 toxin assays for CDI, performed on 323 patients (mean 2.2 assays/pt), were positive at a total cost of $175,500.  37 episodes (separated by at least 2 weeks) of CDI were diagnosed in 35 patients.  During August-September 2009 (Period 2), 110 of 683 GDH assays completed on 332 patients (mean 2.1 assays/pt) were positive at a cost of $170,750.  78 episodes of CDI (either GDH (+) or toxin (+)) in 75 patients were diagnosed.  The rate of CDI increased from 1.50 cases/1000 inpatient days during period 1 to 3.07 cases/1000 inpatient days in period 2 as did the rate of healthcare facility-associated CDI (0.84 to 1.35 cases/1000 days). 

182 (56.3%) patients during period 1 underwent 196 episodes of repeat testing (test repeated within 7 days of a negative test or 14 days of a positive test) and 151 (45.5%) patients during period 2 underwent 177 episodes of repeat testing.  Patients underwent an average of 2.66 tests/repeat testing episode (range 2-7) during period 1 and 2.65 tests/repeat episode during period 2 (range 2-6).  Episodes of repeat testing rarely resulted in initially negative assays becoming positive(5.0% of episodes during period 1, 7.4% during period 2) and frequently a negative test followed a positive test (73.3% of episodes during period 1, 41.4% during period 2).  The cost of repeat testing during period 1 was $81,500 and $73,000 during period 2.  Only 25 of 78 (32.1%) episodes of CDI diagnosed during period 2 had a positive toxin assy.  Repeat testing in patients with GDH positive stools resulted in only 1 of 40 (2.5%) tests being positive for toxin.

Conclusions: The change in assay was associated with an increase in the diagnosis of CDI.  Physician practice did not change despite multiple communications and the availability of institutional guidelines.  Repeated tests for C. difficile rarely result in relevant diagnostic information and are associated with substantial economic costs.  The use of the GDH assay increased the diagnosis of CDI, but the clinical significance of a positive GDH assay is unknown.