Reply : FAIM3:PLAC8 ratio compared with existing biomarkers for diagnosis of severe community-acquired pneumonia: comparing apples to oranges?

Abstract

We are grateful to Dr. Narasimhan and colleagues for their interest in our study on the derivation and validation of the FAIM3:PLAC8 gene expression candidate biomarker for the rapid diagnosis of communityacquired pneumonia (CAP) (1). However, Dr. Narasimhan and colleagues clearly overlooked our stated premise for defining thresholds that favored high sensitivities at the expense of specificity, which reads as follows, “By favoring a high sensitivity we sought to address the potentially serious consequences of false negative predictions (CAP patient classified as no-CAP).” (1). The same rationale was adopted for the evaluation of plasma IL-8, IL-6, and procalcitonin measurements. Thus, the procalcitonin threshold of 0.159 ng/ml reflected our high-sensitivity criterion. We acknowledge that the procalcitonin threshold resides on the lower end of the ranges previously reported by others (2–4), which, together with Dr. Narasimhan and colleagues’ comments, motivated us to evaluate the performance of procalcitonin across different thresholds (Table 1). Considering the 0.159-, 0.25-, 0.5-, 1-, and 2-ng/ml thresholds, the performance of plasma procalcitonin measurements in discriminating CAP and no-CAP patients remained poor (Table 1). [Extract from the article]