Is a short trial of prone positioning sufficient to predict the improvement in oxygenation in patients with acute respiratory distress syndrome?

Objective: To determine whether a 1-h trial of prone positioning is suffici ent to identify responders. Design: Prospective clinical. cohort study in a medico-surgical ICU in a te aching hospital. Patients: 49 patients with acute respiratory distress syndrome. Interventions: A 6-h period of prone positioning. Measurements and results: Baseline measurements (blood gas analysis and res piratory parameters) were evaluated in supine position just prior to turnin g the patients prone. Measurements were then repeated 1 h after the beginni ng of prone positioning (PP1h) and at the end of the 6-h period of prone po sitioning (PP6h). The last measurements were performed 1 h after reposition ing the patients supine. Prone position induced an increase in the PaO2/FIO 2 ratio (p < 0.001). A response (increase in PaO2/FIO2 ratio of at least 20 % at PP1h and/or at PP6h) was observed in 37 of 49 patients (76%). Twenty-s even of these patients (73%) were responders at PP1h while 10 (27%) were re sponders only at PP6h. In all, two-thirds of the patients were considered p ersistent responders. However, whereas the PaO2/FIO2 ratio decreased signif icantly 1 h after repositioning the fast responders supine, the PaO2/ FIO2 ratio remained unchanged after repositioning slow responders. Conclusions: A short-term trial of prone positioning does not appear a suff icient method to identify patients who would benefit from the postural trea tment.