URINE SPECIMEN COLLECTION FROM INCONTINENT FEMALE NURSING-HOME RESIDENTS

OBJECTIVE: To determine if a clean catch technique can accurately diag nose bacteriuria among incontinent female nursing home residents. DESI GN: Cultures and dipstick screening test results from paired urine spe cimens, one collected by a clean catch technique and the other collect ed by sterile in-and-out catheterization, were compared. PARTICIPANTS: A total of 101 incontinent female nursing home residents who were bei ng assessed for participation in a larger clinical intervention trial for incontinence. MEASUREMENTS: Each urine was tested by a dipstick me thod for leukocyte esterase and nitrite and sent to a bio-clinical lab oratory for quantitative culture. RESULTS: Positive and negative cultu re results matched in 92 of the 101 paired specimens. Using the cathet er specimen as a gold standard, the clean catch had a sensitivity of 9 0%, specificity of 92%, positive predictive value of 81%, and a negati ve predictive value of 95%. In a population with symptoms suggestive o f infection, among whom the prevalence of bacteriuria would likely be higher than in the asymptomatic residents we studied (e.g., 60% vs 30% ), the positive predictive value would increase to 95%, but the negati ve predictive value would decrease to 86%. The concordance of the resu lts of the urine screening tests was not as good, except for the detec tion of a positive leukocyte esterase test and a negative nitrite test . CONCLUSION: Incontinent female nursing home residents do not necessa rily have to be catheterized in order to obtain an accurate quantitati ve urine culture. Our results using a careful clean catch technique ar e comparable to those previously reported using urine obtained from a urine-soiled diaper as well as those using a condom catheter technique in men.