National trends and variation in nurse staffing on inpatient psychiatric units

The purpose of this study was to examine national trends and variation in nurse staffing on inpatient psychiatric units in US general hospitals from 2005-2017. The National Database of Nursing Quality Indicators^® provided data on nurse staffing from 1,143 psychiatric units in 610 US hospitals. A weighted linear mixed model was fitted for each of two staffing measures: Registered nurse (RN) hours per patient day (HPPD) and non-RN HPPD. Monthly staffing levels were modeled as a function of study year, unit type, and hospital bed size, teaching status, government ownership, for-profit status, metropolitan location, and US census division. Very gradual upward trends in staffing were observed. Compared with adult units, child/adolescent units had lower RN staffing and higher non-RN staffing. Levels of both types of staffing were lower in for-profit facilities. The Pacific census division had higher RN staffing than every other census division by an estimated margin of 0.52-1.54 HPPD, and census divisions with the lowest levels of RN staffing had the highest levels of non-RN staffing. Despite concerns expressed over the past 15 years about patient violence, staffing levels, and use of seclusion and restraint on psychiatric units, average staffing levels have apparently increased only modestly since 2005, and increases in RN staffing on psychiatric units have not kept pace with increases in general care units. Marked regional differences in staffing merit further investigation.