A comparison of diagnostic algorithms and clinical parameters to diagnose ventilator-associated pneumonia: A prospective observational study

Abstract

BACKGROUND: Suspicion and clinical criteria continue to serve as the foundation for ventilator-associated pneumonia (VAP) diagnosis, however the criteria used to diagnose VAP vary widely. Data from head-to-head comparisons of clinical diagnostic algorithms is lacking, thus a prospective observational study was performed to determine the performance characteristics of the Johanson criteria, Clinical Pulmonary Infection Score (CPIS), and Centers for Disease Control and Prevention’s National Healthcare Safety Network (CDC/NHSN) criteria as compared to Hospital in Europe Link for Infection Control through Surveillance (HELICS) reference standard. METHODS: A prospective observational cohort study was performed in three mixed medical-surgical ICUs from one academic medical center from 1 October 2016 to 30 April 2018. VAP diagnostic criteria were applied to each patient including CDC/NHSN, CPIS, HELICS and Johanson criteria. Tracheal aspirate cultures (TAC) and serum procalcitonin values were obtained for each patient. RESULTS: Eighty-five patients were enrolled (VAP 44, controls 41). Using HELICS as the reference standard, the respective true positive rate (TPR; sensitivity), true negative rate (TNR; specificity), false positive rate (FPR; fall-out) and false negative rate (FNR; miss-rate) for each of the assessed diagnostic algorithms were: CDC/NHSN (Sensitivity 54.2%; Specificity 100%; FPR 95%; FNR 0%), CPIS (Sensitivity 68.75%; Specificity 95.23%; FPR 50%, FNR 2%), Johanson (Sensitivity 67.69%; Specificity 95%; FPR 52.5%, FNR 2%). The positive TAC rate was 81.2%. The sensitivity for positive TAC with the serum procalcitonin level >0.5 ng/ml was 51.8%. CONCLUSION: VAP remains a considerable source of morbidity and mortality in modern ICUs. The optimal diagnostic method remains unclear. Using HELICS criteria as the reference standard, CPIS had the greatest comparative diagnostic accuracy, whereas the sensitivity of the CDC/NHSN was only marginally better than a positive TAC plus serum procalcitonin >0.5 ng/ml. Algorithm accuracy was improved by adding serum procalcitonin >0.5 ng/ml, but not positive quantitative TAC.