Background: Few studies have evaluated dipstick urinalysis for elderly and practically none present confidence intervals. Furthermore, most previous studies combine all bacteria species in a "positive culture". Thus, their evaluation may be inappropriate due to Yule-Simpson's paradox. The aim of this study was to evaluate diagnostic accuracy of dipstick urinalysis for the elderly in nursing homes.
Methods: In this cross-sectional study voided urine specimens were collected from 651 elderly individuals in nursing homes. Dipstick urinalysis for nitrite, leukocyte esterase and urine culture were performed. Sensitivity, specificity, positive and negative predictive values with 95% confidence intervals were calculated. Visual readings were compared to readings with a urine chemistry analyzer.
Results: 207/651 (32%) of urine cultures showed growth of a potentially pathogenic bacterium. Combining the two dipsticks improved test characteristics slightly compared to using only one of the dipsticks. When both dipsticks are negative, presence of potentially pathogenic bacteria can be ruled out with a negative predictive value of 88 (84-92)%. Visual and analyzer readings had acceptable agreement.
Conclusion: When investigating for bacteriuria in elderly people at nursing homes we suggest nitrite and leukocyte esterase dipstick be combined. There are no clinically relevant differences between visual and analyzer dipstick readings. When dipstick urinalysis for nitrite and leukocyte esterase are both negative it is unlikely that the urine culture will show growth of potentially pathogenic bacteria and in a patient with an uncomplicated illness further testing is unnecessary.