Inability to Predict Antimicrobial Resistance of UTI Pathogens in Emergency Department Patients

• Published in: Academic emergency medicine Vol. 10; no. 5; p. 436
• Main Author: Norton, R.
• Format: Journal Article
• Language: English
• Published: Des Plaines Wiley Subscription Services, Inc 01.05.2003
• ISSN: 1069-6563; 1553-2712
• DOI: 10.1197/aemj.10.5.436-a

OBJECTIVE: To develop a clinical decision rule to predict trimethoprim/sulfamethoxasole (TMP/SMX) resistance for urinary tract infection (UTI) pathogens in patients presenting to the emergency department (ED). METHODS: Resistance patterns were retrospectively reviewed for all adult female patients discharged from a university hospital ED with UTI during 26 months. 18 predictor variables were collected: age, vital signs, nursing home residence, recent hospitalization, previous UTI, recent and current antibiotic use, diabetes, recent travel, indwelling catheter and pregnancy. Classification and regression tree analysis (CART) was used to derive a clinical decision rule for predicting TMP/SMX resistance among patients with true UTIs (defined by colony counts and organisms). Trends for resistance over time were assessed by adjusted and unadjusted analyses. We used logistic regression to assess factors associated with TMP/SMX resistance. RESULTS: Of 512 cultures, 104 (20%) were TMP/SMX resistant and 10 (2.0%) were resistant to ciprofloxacin (CIPRO). There was no combination of variables in the CART analysis that reliably predicted resistance to TMP/SMX. Time trend analysis demonstrated a stable resistance-rate. Logistic regression analysis showed only pulse >100 to be significantly associated with TMP/SMX resistance (OR 1.8, 95% CI 1.1-3.1). CIPRO was prescribed most often (57.8%). TMP/SMX was prescribed for 12 (2%) and CIPRO for 3 (0.5%) patients with resistant organisms before culture results were known. CONCLUSION: TMP/SMX resistance is common among adult females discharged from the ED with UTI. It is difficult to accurately predict which patients will have a resistant organism. Unless a reliable rule can be developed, empiric treatment with CIPRO is appropriate in settings with high TMP/SMX resistance.