Copyright 2004 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2004
We appreciate the comments by Drs Shah and Metlay and would like to address them. Typically, a clinical prediction model is developed by comparing patients with a disease with those who do not have the disease.1 In this study, we compared patients with influenza with those who were specifically culture negative for influenza. Though culture negative for influenza, these patients likely had an undetectable viral process. Thus, excluding the 10 patients with known viral pathogens would have no effect on the clinical applicability of the model. Furthermore, adding the 10 patients with positive cultures for viruses other than influenza made little or no difference in the predictive abilities of the published model, with a slight trade-off for higher sensitivity and lower specificity. For those 2 reasons, the concern that these 10 subjects were not included in the analysis is of no consequence if the model is used.
Friedman MJ, Attia MW. Clinical Prediction Model for Influenza—Reply. Arch Pediatr Adolesc Med. 2004;158(10):1017. doi:10.1001/archpedi.158.10.1018-b