In 60 medical centers, 1,500 patients were randomized to either a restrictive or liberal fluid strategy. The 90-day mortality rate was 14% in the fluid-restrictive group and 14.9% in the liberal group, equating to statistical equivalence between the two approaches.