Characteristics and outcomes of infection after a change in the diagnostic testing algorithm.

BACKGROUND

Polymerase chain reaction (PCR) testing for the detection of is a highly sensitive test. Some clinical laboratories have included a 2-step testing algorithm utilizing PCR plus toxin enzyme immunoassays (EIAs) to increase specificity.

OBJECTIVE

To determine the risk factors and outcomes of PCR-positive/toxin-positive encounters compared to PCR-positive/toxin-negative encounters.

DESIGN

Retrospective study.

SETTING

A Veterans' Affairs hospital.

METHODS

A retrospective case-control study of patient encounters with a positive test by PCR and either a toxin EIA-positive assay (ie, cases) or toxin EIA-negative assay (ie, controls). Clinically relevant exposures and risk factors were determined to assess CDI recurrence at 30 days. Available encounter stool specimens were cultured for and were subjected to restriction endonuclease analysis (REA) strain typing.

RESULTS

Among 130 PCR-positive patient encounters, 80 (61.5%) were toxin EIA negative and 50 (38.5%) were toxin EIA positive. Encounters that were toxin positive were more frequently treated (96.0%) compared to toxin-negative encounters (71.3%; < .01). A multivariable logistic regression model revealed that toxin-negative encounters were less likely to suffer a recurrent CDI episode within 30 days (odds ratio [OR], 0.20, 95% confidence interval [CI], 0.05-0.83). Additionally, a higher PCR cycle threshold predicted a lower risk of CDI recurrence at 30 days. (OR, 0.82; 95% CI, 0.68-0.98). During the study period, the REA group Y strain accounted for most toxin-negative encounters (32.5%; = .05), whereas REA group BI strain accounted for most toxin-positive encounters (24.3%; = .02).

CONCLUSIONS

A testing strategy of PCR plus toxin EIA helped predict recurrent CDI.