Risk adjustment of quality indicators in French long term care facilities for elderly people. A preliminary study

In France, the funding reform for long term care facilities mandates quality improvement efforts and the measurement of quality of care indicators. In the United States, the Center for Health Systems Research and Analysis (CHSRA) has proposed 24 indicators, which measure practices and outcomes ("events") in long term care facilities. The objective of this preliminary study was to propose risk adjustments on the CHSRA indicators for future use in France.

The unit of analysis was residents assessments (558) in eight French facilities wich voluntarily participated to the experimentation of the Resident Assessment Instrument (RAI) from January 1996 to June 1999. This preliminary study focused on nine of the CHSRA's indicators: "prevalence of falls", "use of nine or more different medications", "prevalence of bladder or bowel incontinence", "prevalence of occasional or frequent bladder or bowel incontinence without a toileting plan", "prevalence of urinary tract infections", "prevalence of dehydration", "prevalence of bedfast residents", "prevalence of little or no activity" and "prevalence of stage 1-4 pressure ulcers". A working group selected residents'characteristics for risk adjustment based on the literature, risk adjustments selected by CHSRA, and their own propositions. For each event, candidate independent characteristics were screened using univariate analyses and, when significant, were entered into a stepwise selection analysis to obtain the final model. The proposed indicator was the ratio between observed and expected numbers of events.

Six indicators were left unadjusted. For the event "prevalence of occasional or frequent bladder or bowel incontinence without a toileting plan", we proposed a stratification on "dementia - neurological diseases" (p<0.001). We proposed two models of prediction: for the event "prevalence of bladder or bowel incontinence" we adjusted on "severe cognitive impairment" (OR=4.00), "dementia- neurological diseases" (OR=1.75) and "total dependence in mobility" (OR=9.30); for the event "prevalence of stage 1-4 pressure ulcers" we adjusted on "dependence for bed mobility" (OR=4.97), "leaves 25% or more food uneaten at most meals" (OR=1.82) and "diabetes mellitus" (OR=3.45).

This preliminary study underlines the importance of considering risk-adjustment for quality indicators in elderly care facilities.