An exploratory study of a predictive model for nursing -sensitive patient outcomes derived from patient care unit structure and process variables

Relative to other areas in the hospital, high labor costs have made patient care units the usual targets of hospital restructuring initiatives. Staff nurses, nursing leaders, and patients express concern about the changes brought about by hospital restructuring and the impact on nurse staffing, professional nursing practice, and patient outcomes. This study is a secondary data analysis of a prospective multi-hospital study that examines hospital restructuring\u27s impact on outcomes. The purpose of this exploratory study is to begin the development of a multivariable model that predicts nursing-sensitive patient outcomes from patient care unit structure and process variables. Potential confounding variables that are controlled for in the multivariable model include unit type (medical or surgical) and size (average daily census), study year (1997 versus 1998), and hospital-level case mix index. Nursing sensitive patient outcomes are measured by patient fall rates, nosocomial pressure ulcer rates, nosocomial urinary tract infection rates, patient satisfaction with nursing care, and patient satisfaction with pain management. The structure variable, patient care unit staffing, is measured by RN%, HWPPD, UAPHWPPD, and RNHWPPD. The process variable is professional nursing practice and is measured by a constellation of nurses\u27 perceptions of their autonomy and decision-making, nurse to physician collaboration and nurse-to-nurse collaboration and problem solving. The findings from this secondary data analysis extend those of the multi-hospital study Hospital Restructuring\u27s Impact on Outcomes (HRIO). The final predictive model includes CMI, ADC, RNHWPPD, nurse autonomy and decision-making, and nurse-to-nurse collaboration and problem solving. Using hierarchical multilevel regression analysis, this exploratory study assessed and found evidence of a simultaneous impact of patient care unit staffing levels and professional nursing practice in one out of the five hypothesized relationships specifically patient fall rates. In addition, either the patient care unit staffing or professional practice variables were significant predictors for three other outcomes. These additional findings provide further empirical support for the relationship among RNHWPPD, nurse autonomy and decision-making, and patient outcomes and provide a foundation for future study of multivariable models