ICU nurses displayed large variability in their ability to perform central line maintenance tasks. After SBML, there was significant improvement, and all nurses reached a predetermined level of competency.

BACKGROUND The nationally reported metric for Clostridium difficile infection (CDI) relies solely on laboratory testing, which can result in overreporting due to asymptomatic C. difficile colonization. OBJECTIVE To review the clinical scenarios of cases of healthcare facility-onset CDI (HO-CDI) and to determine the appropriateness of C. difficile testing on the basis of presence of symptomatic diarrhea in order to identify areas for improvement. DESIGN Retrospective cohort study. SETTING Northwestern Memorial Hospital, a large, tertiary academic hospital in Chicago, Illinois. PATIENTS The cohort included all patients with a positive C. difficile test result who were reported to the National Healthcare Safety Network as HO-CDI during a 1-year study period. METHODS We reviewed the clinical scenario of each HO-CDI case. On the basis of documentation and predefined criteria, appropriateness of C. difficile testing was determined; cases were deemed appropriate, inappropriate, or indeterminate. Statistical analysis was performed to compare demographic and clinical parameters among the categories of testing appropriateness. RESULTS Our facility reported 168 HO-CDI cases to NHSN during the study period. Of 168 cases, 33 (19.6%) were judged to be appropriate tests, 25 (14.8%) were considered inappropriate, and 110 (65.5%) were indeterminate. Elimination of inappropriate testing would have improved our facility's standardized infection ratio from 0.962 to 0.819. CONCLUSION Approximately 15% of HO-CDI cases were judged to be tested inappropriately. Testing only patients with clinically significant diarrhea would more accurately estimate CDI incidence, reduce unnecessary antibiotic use, and improve facilities' performance of reportable CDI metrics. Improved documentation could facilitate targeted interventions. Infect Control Hosp Epidemiol 2016;1395-1400.

No abstract