RESUMEN
PURPOSE: To identify the accuracy of urinalysis in the diagnosis of urinary tract infection in pregnant women at high risk. METHODS: a prospective, cross-sectional study was conducted on 164 pregnant women admitted to the high-risk the ward of the Instituto de Medicina Integral Prof. Fernando Figueira (IMIP) during the period from January to June 2011. Patients who had been taking antibiotics in the last ten days were excluded. All patients were subjected to simple urine tests and urine culture at the beginning of their admission. The agreement between the results of the examinations was evaluated by Kappa indices (K), and accuracy, sensitivity, specificity, and positive (PPV) and negative (NPV) predictive values were also determined. RESULTS: When only the presence of pus cells in urinalysis was used as a diagnostic criterion suggesting bacteriuria, there was a poor agreement when compared to uroculture (K=0.16). Accuracy was 61%, sensitivity 62.5%, and specificity 60.6%. PPV was 27.78% and NPV was 87%. CONCLUSION: The presence of alteration of urinalysis does not necessarily indicate an ongoing urinary tract infection, with urine culture being necessary. However, when urinalysis data are normal, uroculture may be avoided.
