Clinician practice and the National Healthcare Safety Network definition for the diagnosis of catheter-associated urinary tract infection

Abstract

The National Healthcare Safety Network (NHSN) definition for catheter-associated urinary tract infection (CAUTI) is used to evaluate improvements in CAUTI prevention efforts. We assessed whether clinician practice was reflective of the NHSN definition. We evaluated all adult inpatients hospitalized between July 2010 and June 2011, with a first positive urine culture > 48 hours of admission obtained while catheterized or within 48 hours of catheter discontinuation. Data comprised patients' signs, symptoms, and diagnostic tests; clinician's diagnosis; and the impression of the infectious diseases (ID) consultant. The clinician's practice was compared with the NHSN definition and the ID consultant's impression. Antibiotics were initiated by clinicians to treat CAUTI in 216 of 387 (55.8%) cases, with 119 of 387 (30.7%) fitting the NHSN CAUTI definition, and 63 of 211 (29.9%) considered by ID to have a CAUTI. The sensitivity, specificity, and positive and negative predictive values of a clinician diagnosis of CAUTI were 62.2%, 47%, 34.3%, and 73.7% when compared with NHSN CAUTI definition (n = 387) and 100%, 57.4%, 50%, and 100% when compared with the ID consultant evaluation (n = 211), respectively. The positive predictive value of the NHSN CAUTI definition was 35.1% when compared with the ID consultant's impression (n = 211). NHSN CAUTI definition did not reflect clinician or ID consultant practices. Our findings reflect the differences between surveillance definitions and clinical practice.