Demand, Selection and Patient Outcomes in German Acute Care Hospitals

In times of peak demand hospitals may fail to deliver the high standard of treatment quality that they are able to offer their patients at regular times. To assess the magnitude of these effects, this study analyzes the effects of low staff-to-patients ratios on patient outcomes empirically. We use the variation of patient admissions over time as a proxy for varying staff level. Further, we control for within diagnosis unobservable variation in severity across days with as opposed to days without excess demand. We find that when this variation is ignored in the regression framework, the effect of demand on outcomes is biased upwards. The reason is that when demand is high more patients with a higher unobservable frailty are admitted to the hospitals. After having controlled for this selection of patients, excess demand does not negatively affect patient outcomes.