Abstract
<h3>In Reply.</h3> —Evaluation of predictors of death in ICU patients is not only a question of statistics, but may serve as a paradigm of the difficulties in plausibly evaluating biological and clinical phenomena as markers of prognosis. Multiple factors affect the outcome of ICU patients, including age, severity of underlying medical conditions, acute physiologic disturbances, primary diagnosis, and response to therapy.<sup>1</sup>Consequently, interactions between such factors are complex and innumerable.<sup>2</sup>Dr Nishioka's suggestion is, nonetheless, relevant, particularly concerning the relationship between nosocomial pneumonia and nosocomial bacteremia in inducing death. Unfortunately, our database did not permit the exploration of all possible interactions and the determination of the respective effects of pneumonia and bacteremia on mortality; in this setting, an accurate analysis requires simultaneous consideration of the timedependent relationship between these 2 ICU-acquired complications (simultaneous or with order effect), the responsible pathogen(s) (identical or different), the severity of disease at the time
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
