Why intensivists want chest radiographs

We would like to contribute to the ongoing discussion regarding different chest radiograph (CXR) strategies in the intensive care unit (ICU). In their review and metaanalysis in a recent issue of Critical Care, Ganapathy and colleagues [1] concluded that they found no harm associated with a restricted CXR strategy for ICU patients. On the other hand, they stated that the safety of abandoning routine CXRs for ICU patients was still uncertain. Several investigators, including one large multicenter randomized trial [2], confirmed that performing an ondemand CXR strategy instead of a daily routine CXR strategy decreased the total number of CXRs significantly and was accompanied by an increase in their diagnostic efficacy but without any increase in adverse events or the use of other imaging studies. This leads to the question of what the exact indications for ondemand CXRs in critically ill patients are. And might there be certain patient groups that may still benefit from routine CXRs? Another interesting point of view is the impact of an on-demand CXR strategy on workflow and efficiency [3], where a number of issues still need to be addressed. For example, can certain ICU patients safely be transferred to the ward without a CXR before? What is the impact of ‘negative’ CXR findings on this workflow and on our personal clinical decision-making? And is it possibly more (cost) efficient for a radiology department to perform multiple routine CXRs during a morning round instead of performing several single CXRs during the day and at night? Our recent web study among Dutch intensivists showed that, nowadays, in line with the current evidence, a daily routine CXR strategy is used significantly less frequently than one decade ago [4]. However, surrogate routine strategies like performing a routine CXR on certain fixed days a week or only on the first days of admission have become