Urinalysis: diagnostic performance of urine dipstick compared to an automated microscopic method

Abstract

Introduction: Urinalysis is one of the most important clinical laboratory tests because numerous pathologies can manifest or be suspected through this test. Although the previous reports mention that urinary microscopy is a fundamental part of urinalysis for diagnostic support of various conditions, there is a debate about the utility of this test section in a certain patient population. The aim of this study was to determine the diagnostic performance of the urinary dipstick analysis and the potential risks of false-negative (FN) results. Material and methods: This is a retrospective and observational study, and urinalysis information was obtained from non-hospitalized patients. The dipstick and microscopic analyses were performed using the Clinitek-ATLAS (index test) and iQ200-SPRINT (reference standard) devices. Dipstick or microscopy analyses were positive if ≥ 1 parameters were abnormal. A Bayesian hierarchal beta-binomial model was carried out for each performance parameter. Risk analysis was performed as proposed in the literature. Results: Five hundred and fifty-two patients were included in the study. The posterior median at group level was 94% (credible interval 95% [CrI 95%] 89.9-97%) for sensitivity (Se), 57.1% (CrI 95%, 50.1-64.1%) for specificity, and 5.8% (CrI 95%, 2.59-9.64%) for FN rate (FNR). The posterior probability Se > 90% was 95.9% at a group level. The risk analysis found only low-risk false-negative events. Conclusions: The performance of the dipstick analysis was appropriate, with a good certainty of Se > 90% and a FNR < 10% at the operator level. Omission of microscopic analysis can be a safe action in a patient with a negative dipstick since FNs with a clinical impact are not expected.