Bayesian re-analysis of the RELAx trial suggests a 75-78% probability that a lower PEEP strategy is superior to higher PEEP in ICU patients without ARDS, with even higher benefits observed in specific clinical subgroups.