The PASS study: performance assessment system for success In residential aged care

Abstract

Objective: This study sought to (i) assess the impact of Quality Improvement Activities on adverse events in residential care for older people (nursing homes); and (ii) to explore the effectiveness of a Data Envelopment Analysis (DEA) quantity/quality model in providing additional information. Design: A cluster randomised controlled trial, utilising longitudinal resident level data comparing the impact of Quality Improvement Activities including benchmarking with benchmarking alone on adverse event rates. The data for the two groups was applied to quantity and quality DEA models to explore the effectiveness of DEA as a best practice benchmarking application for nursing homes. Participants: Nursing homes (N=32) providing care for older people in six District Health Boards in New Zealand. Outcome measures: Participating nursing homes provided monthly data on quality indicators (catheters, falls, pressure ulcers, physical restraints, and urinary tract infections, registered and enrolled nurses and caregivers hours as well as resident days) monthly for 15-months. Results: Logistic analysis revealed a statistically significant impact from the QIAintervention with a significant reduction in the use of catheters in the QIA-group (p=0.002). For the BM-intervention there was a significant reduction in the use of restraints (p=0.0002) and an increase in the days free of pressure ulcers (p=0.031). Significant associations were identified between increasing registered nurse hours and decreasing restraint use (p= 0.04) and increasing registered and enrolled nurse time and increasing urinary tract infections (p=0.0014). This effect was greater for enrolled nurses (p=0.0012) than registered nurses (p=0.04). A significant association between the Modified Shortells survey sub section for teamwork/leadership scores and pressure ulcers was found but no associations were observed. DEA established the quantity and quality efficiencies for each nursing home. DEA quality efficiency scores indicated that the QIA-intervention increased the number of days free of restraints in the QIA-group whereas adverse event free days increased more for the remaining quality indicators in BM-group. The DEA quality efficiency scores suggested that BM-intervention improved DEA quality efficiency for all quality indicators. Conclusions: The Quality Improvement Activities had limited impact on the observed adverse events. However, the use of benchmarking across the two groups and utilisation of DEA in the analysis provided interesting findings. DEA was found to be more sensitive to clinically significant improvements than traditional statistical approaches. It offers an opportunity for benchmarking excellence in nursing homes. In addition, the results signal the importance of further research into the role of registered nurses in this setting.