The relationship between nurse staffing levels and nursing-sensitive outcomes in hospitals: Assessing heterogeneity among unit and outcome types.

Objective: To identify which unit types are most sensitive to nurse staffing levels.

Data Sources/study Setting: Collection of secondary data took place from March to July 2016. For our study, we analyzed administrative hospital claims data and self-reported structural data from hospitals in Germany. We used 26,502,579 admissions nested in 13,089 units in 3,680 hospitals from 2012 to 2014.

Study Design: We used regression analysis to examine the relationship between 11 established nursing-sensitive outcomes (NSOs) and nurse-to-patient ratios on a unit level. Nurse-to-patient ratios were our key explanatory variable. We conducted separate OLS regressions for each NSO in each unit type using linear and non-linear terms.

Data Collection/extraction Methods: We linked hospital claims data with self-reported structural data from hospitals from 2012 to 2014.

Principal Findings: We identified 15 unit types with at least one significant NSO. The effect of potential understaffing on NSOs depends on the unit type.

Conclusions: Our study indicates that the relationship between nurse staffing levels and NSOs varies greatly depending on the unit type concerning both significance and magnitude. Future research might consider performing analyses on unit level instead of hospital level.