Dr. Janower’s 1997 study appears to suffer from the same problems we noted in other earlier attempts to assess the utility of routine chest radiographs in this patient population. His study was observational rather than a randomized, controlled trial. One quarter of the “abnormalities” detected were repositioning of the endotracheal tube, an intervention that is unlikely to have prevented a significant adverse outcome. Furthermore, in his study, we cannot determine how many of the abnormalities that were detected would have been found if the radiograph had been obtained as the result of a clinical suspicion of a problem. Finally, there is no attempt to determine whether the real outcomes of interest (eg, length of time on the ventilator, length of stay in the ICU, or mortality) are affected by the use of routine radiographs. We acknowledged in our manuscript that the relatively small size of our sample may have led to a β error with respect to some of these outcomes. Nevertheless, we believe that as a result of our study design our data provide more information than those of the observational study he reports. Although it is possible that a negative routine radiograph may have averted unnecessary testing by residents, our anecdotal experience would suggest that there are a greater number of routine films with “false-positive” results in the ICU setting than truly negative radiograph findings. Thus, the performance of routine radiographs might be expected to lead to a greater expenditure of resources rather than to a reduction of costs.