Nursing home residents with suspected urinary tract infections: a diagnostic accuracy study

Background Urinary tract infections (UTIs) are one of the most common infections in nursing homes (NHs). A high error rate of a UTI diagnosis based solely on clinical criteria is to be expected in older persons as they often present infections in an atypical way. A study was set up to assess the diagnostic value of signs/symptoms and urine dipstick testing in identifying UTIs in NH residents and to explore whether C-reactive protein (CRP) measured by point-of-care testing (POCT) can help in the diagnosis. Methods During a three month prospective multicentre study, urine sampling for culture, POCT CRP and urinary dipstick testing were performed in each NH resident with a suspected UTI. UTIs were defined according to Stone et al., i.e. criteria based upon the presence of a set of signs/symptoms and a positive urine culture. Results Eleven NHs and 1 263 residents participated. Sixteen out of 137 recorded UTI suspicions were confirmed. Acute dysuria (positive likelihood ratio (LR +): 7.56, 95% confidence interval (CI): 3.94-14.5) and acute suprapubic pain (LR + : 11.4, 95% CI: 3.58-35.9) were found to be significant predictors. The combined nitrite and leucocyte esterase urine dipstick test (one or both positive) had a 96.0% negative predictive value (95% CI: 80.5-99.3%). The sensitivity of a positive CRP test (>= 5 mg/L) was 60.0% (95% CI: 32.3-83.7%). Antimicrobials were prescribed in 60.2% of suspected but unconfirmed UTIs and in 92.3% of confirmed UTIs. Conclusions Using a stringent definition, only 11.7% of our suspicions were confirmed. Besides acute dysuria and suprapubic pain, we were not able to prove that any other clinical sign/symptom or POCT CPR adds useful information to the UTI diagnosis. We confirmed the findings of earlier research that urine dipstick tests are useful in ruling out UTIs and identified a potential overuse of antimicrobials in our NH population.