Methods: We performed a single-center retrospective cohort study of patients who had clinical specimens sent for universal PCR from August 2013 to April 2016. Clinical data were extracted from medical records. Odds ratios were calculated and patients testing positive/negative were compared with univariate logistic regression. Sensitivity, specificity, positive predictive value and negative predictive values were calculated by comparing the test result with a gold standard composite final clinical diagnosis determined by 3 independent reviewers based on all available clinical information.
Results: 71 tissue samples were included, of which 21(29.6%) were positive. 12 bacteria, 3 mycobacteria and 7 fungi were identified. The number of leukocytes in the gram stain (odds ratio, OR 1.57, p=0.04) and presence of inflammation on histopathological examination (OR 5.69, p=0.02) were found to be significantly associated with a positive result. The sensitivity, specificity, positive predictive value, and negative predictive values were 56%, 95%, 91% & 70% respectively. Management was altered in 22 patients, 9 of whom had a positive and 13 had a negative result.
Conclusion: These findings suggest that the universal PCR assay has significant clinical utility, but the yield of this test can be optimized by careful patient/specimen selection. Utility was highest in patients with microscopic evidence of inflammation by gram stain or histopathlogical examination. Specificity was high. Use of this complex, difficult to interpret, and expensive test should be limited to infectious disease physicians incorporating all available clinical information to optimize performance.
G. M. Andujar-Vazquez, None
A. S. Joel Chandranesan, None
A. Rabson, None
S. Doron, None
D. R. Snydman, None