Any time I see the output of a statistical model like "estimated length of stay" that gives a number like "16.6 days" without also saying something like "with a 95% confidence interval of -5.1 / +20.9 days", I get pretty confident the goal is not actually to help decision-making.