Greta G Cummings1, Malcolm Doupe2, Liane Ginsburg3, Margaret J McGregor4, Peter G Norton5, Carole A Estabrooks1. 1. Faculty of Nursing, University of Alberta, Edmonton, Canada. 2. Department of Community Health Sciences, College of Medicine, Manitoba Centre for Health Policy, Winnipeg, Canada. 3. School of Health Policy and Management, York University, Toronto, Canada. 4. Faculty of Medicine, University of British Columbia, Vancouver, Canada. 5. University of Calgary, UCMC North Hill, Alberta, Canada.
Abstract
Purpose of the study: To (a) describe A Scheduled Shifts Staffing measure (ASSiST) to derive care aide worked hours per resident day (HCA WHRD) at facility and unit levels in nursing homes, (b) report reliability through comparisons to administrative staffing data; (c) report validity by examining associations between HCA WHRD, staff outcomes (job satisfaction, emotional exhaustion), and resident quality indicators (QIs) (e.g. falls, delirium, stage 2+ pressure ulcers), and (d) explore intrafacility variation in staffing intensity levels related to unit-level variation in resident and staff outcomes. Design and Methods: We used data from 40 care units in 12 Canadian nursing homes between 2007 and 2012. Descriptive statistics and tests of association and difference described relationships of two measures of staffing with resident and staff outcomes. Results: Annualized rates of HCA WHRD from both data sources compared well at the facility level (Pearson Product Correlation; R = 0.847, p < .001), and were correlated similarly to staff work life and many QIs. Using ASSiST data, we show that staffing levels can vary by up to 40% at the unit-level within nursing homes. Implications: ASSiST is easy to collect, more timely to retrieve than administrative data, has good criterion and construct validity, and reflects intrafacility variation in health care aide staffing levels.
Purpose of the study: To (a) describe A Scheduled Shifts Staffing measure (ASSiST) to derive care aide worked hours per resident day (HCA WHRD) at facility and unit levels in nursing homes, (b) report reliability through comparisons to administrative staffing data; (c) report validity by examining associations between HCA WHRD, staff outcomes (job satisfaction, emotional exhaustion), and resident quality indicators (QIs) (e.g. falls, delirium, stage 2+ pressure ulcers), and (d) explore intrafacility variation in staffing intensity levels related to unit-level variation in resident and staff outcomes. Design and Methods: We used data from 40 care units in 12 Canadian nursing homes between 2007 and 2012. Descriptive statistics and tests of association and difference described relationships of two measures of staffing with resident and staff outcomes. Results: Annualized rates of HCA WHRD from both data sources compared well at the facility level (Pearson Product Correlation; R = 0.847, p < .001), and were correlated similarly to staff work life and many QIs. Using ASSiST data, we show that staffing levels can vary by up to 40% at the unit-level within nursing homes. Implications: ASSiST is easy to collect, more timely to retrieve than administrative data, has good criterion and construct validity, and reflects intrafacility variation in health care aide staffing levels.