2006 HSR&D National Meeting Abstract
3037 — Assessing Inappropriate Prescribing for the Elderly Using a New Quality Measure
Pugh MJ (Veterans Evidence-based Research, Dissemination, and Implementation Center )
Hanlon JT (Center for Health Equity Research and Promotion)
Zeber JE (Veterans Evidence-based Research, Dissemination, and Implementation Center )
Amuan ME (Center for Health Quality, Outcomes and Economic Research)
Cunningham F (Pharmacy Benefits Management)
Berlowitz DR (Center for Health Quality, Outcomes and Economic Research)
Inappropriate drug exposure in older Americans stimulated development of a measure to identify potentially inappropriate prescribing in the elderly (PIPE) by the National Committee on Quality Assurance based on the most commonly used measure of PIPE—the Beers criteria. This measure, included in the 2006 Health Plan Employer Data and Information Set (HEDIS), will assess prescribing quality for older Americans. Estimates of patient exposure and characteristics of patients at risk of PIPE based on this measure are unknown.
This cross-sectional study used national VA administrative and pharmacy databases to identify older veterans receiving HEDIS drugs. Veterans 65 years or older on October 1, 1999 and who had at least two outpatient visit days during FY00 or outpatient visits in FY99 and FY00 were included (N=1,096,361). Multivariable logistic regression analyses stratified by gender identified patient characteristics associated with increased risk of HEDIS drug exposure.
Overall 19.6% of older veterans were exposed to PIPE; 41.8% of females and 19.2% of males. The most commonly used drugs were estrogens (females), and antihistamines, skeletal muscle relaxants, and opioid analgesics (both males and females). For both genders younger patients (males OR 1.25; 95% CI 1.21-1.29; females OR 2.61 95% CI 2.21-3.08) and those prescribed multiple medications were at greater risk (males >9 OR 8.17 61, 95% CI 7.99-8.36; females >9 OR 8.06, 95% CI 7.18-9.04); blacks were at lower risk (OR males .91, 95% CI .90-.93; females OR .74, 95% CI .64-.85). For males, Hispanics (OR 1.26, 95% CI 1.23-1.29) and patients with psychiatric comorbidity (OR 1.77, 95% CI 1.74-1.80) were at higher risk. Results for women excluding estrogen were unchanged.
These criteria include only a limited number of drugs from the updated (2003) Beers criteria, yet PIPE was common. Patients receiving many medications, women, and male psychiatric patients were at highest risk.
Research focused on understanding the complex relationships among gender, race, and psychiatric comorbidity is needed if effective interventions to improve prescribing for older Americans are to be developed and implemented. However, research linking HEDIS drug exposure to outcomes has not been conducted and is sorely needed before VA adopts this measure.