3158 — Opioid use in Veterans eligible for VA and Medicare Part D and predictors for overlap within and among VA and Medicare Part D-reimbursed pharmacies
Suda KJ, Edward Hines, Jr. VA Hospital; Bailey L, Edward Hines, Jr. VA Hospital; Smith BM, Edward Hines, Jr. VA Hospital; Gellad WF, VA Pittsburgh Healthcare System; Huo Z, Edward Hines, Jr. VA Hospital; Cunningham F, VAMedSafe; Burk M, VAMedSafe; Stroupe KT, Edward Hines, Jr. VA Hospital;
Pain is the most prevalent problem among Veterans, receiving pain diagnoses five times as frequently as the general population. Opioid analgesics are commonly used for pain management, but have potential for misuse and serious adverse events. The objective of this study was to evaluate opioid prescribing and predictors for overlap in Veterans dually eligible for VA and Medicare Part D.
A sample of male and all women Veterans > = 66 years without cancer who received opioids in 2005-2009 were included. Controls include Veterans eligible, but did not enroll or submit a claim to Part D. Cases include Veterans enrolled and filed a claim to Part D. Overlapping opioid days supply was evaluated within VA, within Part D, and in dual users of VA and Part D-reimbursed pharmacies from 2007-2009 (2006 = transitional year). T-tests, Chi-square tests and GEE were applied.
From 2005 to 2009, opioid prescribing experienced 121.1% growth. Annual mean number of opioid prescriptions/patient increased from 4.7+4.9 (range = 1-54) in 2005 to 4.9+5.9 (range = 1-104) in 2009 (p < 0.0001). In 2006 after Part D implementation, 55.5% of opioids were dispensed at VA, decreasing to 44.7% in 2009 (p < 0.0001). Cases received significantly more opioids compared to the control group (p < 0.0001); 22.2% filled by VA. Opioids dispensed from Part D-reimbursed pharmacies had a higher frequency of overlap as compared to those filled at VA (p < 0.0001). While overlapping opioid and oxycodone prescriptions filled at VA decreased from 2007 to 2009, overlap increased for prescriptions filled at Part D-reimbursed pharmacies (p < 0.0001). There was minimal overlap between systems (1.3%-1.6%). Predictors for any opioid overlap include women (OR = 1.4), Part D enrollment (OR = 1.8), VA priority 1 copay (OR = 1.9) and persons with sleep disorders (OR = 1.5), psychiatric diagnoses (OR = 1.5), and substance/alcohol abuse (OR > 1.4; p < 0.01 for all). Hispanics were less likely to have overlap (OR = 0.7; p < 0.0001).
Opioid use increased after implementation of Part D. A decrease in VA overlap suggests that VA programs/policies encouraging opioid vigilance have been effective.
Veterans with access to non-VA care should be evaluated for opioids dispensed in the private sector. Tools, such as drug monitoring programs, should be employed by VA and non-VA providers to decrease opioid misuse and adverse events.