The performance of fully automated urine analysis results for predicting the need of urine culture test

Objectives: Urinalysis and urine culture are most common tests for diagnosis of urinary tract infections. The aim of our study is to examine the diagnostic performance of urine analysis and the role of urine analysis to determine the requirements for urine culture. Methods: Urine culture and urine analysis results of 362 patients were retrospectively analyzed. Culture results were taken as a reference for chemical and microscopic examination of urine and diagnostic accuracy of the test parameters, that may be a marker for urinary tract infection, and the performance of urine analysis were calculated for predicting the urine culture requirements. Results: A total of 362 urine culture results of patients were evaluated and 67% of them were negative. The results of leukocyte esterase and nitrite in chemical analysis and leukocytes and bacteria in microscopic analysis were normal in 50.4% of culture negative urines. In diagnostic accuracy calculations, leukocyte esterase (86.1%) and microscopy leukocytes (88.0%) were found with high sensitivity, nitrite (95.4%) and bacteria (86.6%) were found with high specificity. The area under the curve was calculated as 0.852 in ROC analysis for microscopic examination for leukocytes. Conclusion: Full-automatic urine devices can provide sufficient diagnostic accuracy for urine analysis. The evaluation of urine analysis results in an effective way can predict the necessity for urine culture requests and especially may contribute to a reduction in the work load and cost.