Evaluation of the Diagnostic Performance of urine dipstick test for Detection of urinary tract infections in patients treated in Kenyan hospitals.

Abstract Introduction Culture is the gold-standard diagnosis for Urinary Tract Infections (UTIs). However, most hospitals in poor-resource countries lack adequately equipped laboratories and relevant expertise to perform culture and, therefore, rely heavily on dipstick tests for UTI diagnosis. Research gap In many Kenyan hospitals, routine evaluations are rarely done to assess the accuracy of popular screening tests such as the dipstick test. As such, there is a substantial risk of misdiagnosis emanating from inaccuracy in proxy screening tests. This may result in misuse, under-use, or over-use of antimicrobials. Aim The present study aimed to assess the accuracy of the urine dipstick test as a proxy for the diagnosis of urinary tract infections in selected Kenyan hospitals. Methods A hospital-based cross-sectional method was used in the present study. Utility of dipstick  in the diagnosis of UTI was assessed using Mid stream urine against culture as gold standard Results The dipstick test predicted 1,416 positive UTIs, but only 1,027 were confirmed positive by culture, translating to a true prevalence of 54.1%. The sensitivity of the dipstick test was better when leukocytes and nitrite tests were combined (63.1%) than when the two tests were separate (62.6% and 50.7%, respectively). Similarly, the two tests combined had a better positive predictive value (87.0%) than either test alone. The nitrite test had the best specificity (89.8%) and negative predictive value (97.4%) than leukocytes esterase (L.E) or both tests combined. In addition, sensitivity in samples from the inpatients (69.2%) was higher than outpatients (62.7%). Furthermore, the dipstick test had a better sensitivity and positive predictive value among female (66.0% and 88.6%) patients than male patients (44.3% and 73.9%). Among the various patients' age groups, the dipstick test's sensitivity and positive predictive value were exceptionally high in patients ≥75 years  (87.5% and 93.3%). Conclusion The prevalence discrepancies from the urine dipstick test and culture, the gold standard, indicate dipstick test inadequacy for accurate UTI diagnosis. The finding also demonstrates the need for urine culture for accurate UTI diagnosis. However considering it’s not always possible to perform a culture, especially in low resource settings, future studies can combine specific UTI symptoms and dipstick results to assess possible increase in the test’s sensitivity. There is also need to develop readily available and affordable algorithms that can detected UTI where culture is not available.