Estimating Efficiency in Ontario's Long-Term Care Facilities: An Assessment Using Stochastic Frontier Analysis (SFA)

Presenter: Peter Coyte, University of Toronto

Abstract

INTRODUCTION: With uncertainty surrounding the true impact of the ageing population, financing the future of high quality long-term care (LTC) depends on two pivotal factors: cost control and more efficient use of resources. However, the relationships between inputs (and their costs), quality of care and operational efficiency - and what factors determine efficiency - are relatively understudied in the LTC sector.

DATA SOURCE: This study used data from The Residential Care Facilities Survey (RCFS), for the province of Ontario, collected between 2004 and 2007. The RCFS is a longitudinal census survey administered annually by Statistics Canada, to all LTC facilities operating in Canada that receive public funding. The RCFS contains information about the: 1) facility’s type, location, chain ownership and profit status; 2) quantity and cost of inputs employed (i.e. nursing staff, other medical staff, administrative staff, equipment, number of beds, etc.); 3) age, sex and morbidity distribution of residents; 4) number of patient days; and 5) number of deaths and discharges. Quality measures (e.g. use of restraints, loss of activities of daily living, and incidences of pressure ulcers and infections) were derived from annual compliance reviews conducted by the Ministry of Health and Long-Term Care (MOHLTC) for 619 LTC facilities in Ontario.

METHODS: Using the stochastic frontier analysis (SFA) of panel data, this study assesses the determinants of operational efficiency in LTC facilities in Ontario, while controlling for quality of care using a wide array of indicators. The analysis accounts for unobserved heterogeneity and the potential endogeneity of quality. In keeping with the literature, output was measured in terms of patient days, and labour inputs were measured in terms of accumulated FTE paid hours (RNs, physiotherapists/occupational therapists, other direct care staff, administration, dietary services staff, other general services staff, recreational staff, and etc.) per resident-day in each LTC facility.

CONCLUSIONS: Our analysis revealed significant differences in performance by profit status, despite the fact that facilities are compensated at the same per diem rates, with larger LTC facilities operating more efficiently than smaller LTC facilities. The results also underline the importance of considering an array of quality indicators and controlling for endogeneity of quality. Generally, we found that differences in operational efficiency between for-profit, not-for-profit, and public facilities were not significant after controlling for differences in service quality.