Increased Risk for Potentially Inappropriate Prescribing among Rural Elder Veterans

Lund BC, Kaboli PJ. Increased Risk for Potentially Inappropriate Prescribing among Rural Elder Veterans. Poster session presented at: VA HSR&D National Meeting; 2011 Feb 16; National Harbor, MD.

Related HSR&D Project(s)

• RRP 11-001 – Development of Strategies to Care for PTSD: Decreasing Benzodiazepine Treatment

Objectives: Access barriers may place elder rural veterans at risk for potentially inappropriate prescribing (PIP). Thus, we compared PIP frequency between rural and urban elder veterans. Methods: Using FY2007 national VA administrative data, we selected veterans age 65 years and older with regular outpatient use of VA primary care and medications. PIP was determined by four established measures: 1) drugs to avoid by Zhan et al.; 2) additional drugs to avoid by Fick et al.; 3) therapeutic duplication; and 4) drug-drug interactions. PIP risk was estimated using odds-ratios (OR), adjusted for age and sex. Results: The study sample included 1,549,824 elderly veterans with an average age of 75.8 years and 98.2% were men. Nationally, rural veterans were at significantly increased risk for PIP according to all four measures: (1) Zhan OR = 1.10 (1.09, 1.11); (2) Fick OR = 1.04 (1.03, 1.05); (3) duplication OR = 1.05 (1.04, 1.06); and (4) drug-drug interaction 1.02 (1.00, 1.04). However, regional analyses using Zhan criteria revealed that rural veterans were actually at significantly decreased risk compared to urban veterans in the western (17.8% vs. 19.6%, OR = 0.89) and midwest regions (15.7% vs. 16.1%, OR = 0.97). Conversely, rural veterans experienced the greatest risk elevation in the south (23.0% vs. 19.6%, OR = 1.23), and a modest elevation in the northeast (14.2% vs. 12.8%, OR = 1.12) compared to urban veterans. This regional pattern was also observed for the other three PIP measures. Implications: Nationally, rural veterans were at increased risk for PIP across four separate measures. However, regional analyses revealed a more complex picture, with decreased risk for rural veterans in some areas. The extent of regional variation for all veterans was prominent, with Zhan criteria frequencies ranging from 13.2% in the northeast to 21.2% in the south. The precise clinical impact of these differences in inappropriate prescribing is unknown, but the unexpectedly high level of variation suggests an opportunity for improved quality. Impacts: The association of rurality on prescribing quality is heterogeneous and relying solely on national estimates may yield misleading conclusions. Additional investigation is warranted to understand why rural veterans experienced lower PIP risk in some regions, but higher rates in others, and to use this information to design appropriate interventions.