Background: Studies have found that 20% to 25% of older patients receive drugs identified as inappropriate by the 1997 Beers criteria. After the addition of 22 new drugs to the 2003 Beers criteria, the National Committee on Quality Assurance convened an expert consensus panel to identify which drugs from the 2003 Beers criteria should always be avoided in the elderly. The resulting list of drugs to avoid was added to the 2006 Health Plan Employer Data and Information Set (HEDIS) to measure the quality of prescribing for the elderly.
Objective: To use HEDIS 2006 criteria to determine the rate of potentially inappropriate prescribing in the elderly (PIPE) and to determine if patient risk factors are similar to those found using Beers criteria.
Methods: This cross-sectional database study identified older patients receiving drugs included in the HEDIS 2006 criteria using national data from the Veterans Health Administration. Patients aged 65 years or older on October 1, 1999, with at least 2 outpatient visit days during fiscal year 2000, ending September 30, or outpatient visits in fiscal years 1999 and 2000 were included (N=1,096,361). Multivariable logistic regression analyses stratified by gender identified patient characteristics associated with increased risk of HEDIS 2006 drug exposure. Since oral estrogens were considered appropriate at the time of this study, they were excluded from the list of HEDIS 2006 drugs.
Results: Overall, 19.6% of older veterans were exposed to HEDIS 2006 drugs. 23.3% of older veteran women and 19.2% of older veteran men. The most commonly prescribed HEDIS 2006 drugs were antihistamines (received by 9.0% of men and 10.7% of women), opioid analgesics (received by 4.6% of men and 5.8% of women), and skeletal muscle relaxants (received by 4.3% of men and 5.3% of women). Propoxyphene was the most commonly used HEDIS 2006 drug, received by 4.5% of men and 5.7% of women, followed by diphenhydramine, received by 3.5% of men and 4.7% of women, and hydroxyzine, received by 3.2% of both men and women. Patients receiving 10 or more medications of any type were at greatest risk of exposure. Men were 8.2 times more likely to receive at least 1 HEDIS 2006 drug than those taking 1 to 3 drugs of any type (95% confidence interval [CI], 8.0-8.4), while women were 9.6 times more likely (95% CI, 8.2-11.2).
Conclusions: Even though we included a slightly different list of drugs to avoid, results for the HEDIS 2006 measure were similar to those of the 1997 Beers criteria. The HEDIS 2006 drugs are commonly prescribed, and there is a distinct need for direct evidence linking HEDIS 2006 PIPE exposure to adverse patient outcomes. To reduce PIPE, it seems necessary to provide additional evidence for clinicians through the conducting of a well-designed study to assess patient outcomes associated with PIPE exposure as defined by the HEDIS criteria.