Aims To determine the accuracy of phase contrast microscopy in predicting urine culture results.
Design and methods A prospective study comparing the results of phase contrast microscopy interpretation of urine samples with that of urine culture. Samples were microscoped at the time culture was being performed.
The sample size though ongoing was based on the availability of the clinician to get to the laboratory at the time of urine culture. Uncentrifuged, unstained urine samples were examined with an Olympus BH2 microscope enabled with phase contrast at 400x magnification mounted on a Hawksley Nebauler counting chamber. Samples were interpreted as being either “Positive,” “Negative” or “Indeterminate” based on their level of bacteriurea. “Indeterminate” specimens would be repeated in the clinical setting and were not further analysed in our laboratory based study.
Results 65 samples were microscoped. Immediate determination was made for 62 samples (95.4%). 3 samples were deemed as “Indeterminate.”
Of the 62 microscopy interpretations 57 (91.9%) showed concordance with the microbiology culture results. Sensitivity of 15/17 (88.2%), specificity 42/45 (93.3%), positive predictive value 15/18 (83.3%) and negative predictive values of 42/44 (95.5%) were obtained. Video evidence is available showing bacteria in 2 of the samples deemed “Microscopy Positive” but “Culture Negative.” If these samples, as they ought to, returned as “Culture Positive” then microscopy would have attained concordance 59 (95.2%), sensitivity 17/19 (89.5%), specificity 42/43 (97.7%), positive predictive value 17/18 (94.4%), negative predictive value 42/44 (95.5%).
Conclusions Phase contrast microscopy afforded immediate interpretation in 62/65 (95.4%) of samples studied. Microscopy interpretation showed high concordance rate, sensitivity, specificity, positive and negative predictive values when compared to the accepted gold standard i.e., microbiological culture even with the apparent error in the culture process.