249 Opioid Prescribing from Emergency Departments and Risk of Long-Term Use in a Large, Managed Care Network

Though EDs only account for 4% of total opioid prescriptions, recent evidence suggests that exposure to prescription opioids in the ED increases the risk of long-term opioid use. It is unclear if this finding generalizes to populations outside of fee-for-services Medicare. We use methods similar to a recent study to determine if quasi-random exposure to high or low intensity opioid prescribers in the ED influences the risk of long-term opioid use in a mixed age population served by a large managed care network. We performed a retrospective analysis of clinical and administrative data for adult Kaiser Permanente Southern California members who presented to 1 of 14 EDs between January 2013 and December 2014. Only patients continuously enrolled in the Kaiser medical plan for at least 6 months prior to their index ED visit and 1 year following the index ED visit. We categorized ED providers according to the proportion of patients to whom they prescribed opioids at each hospital and stratified them into quartiles. Accordingly, “high” intensity prescribers are the quartile of providers who are most likely to prescribe opioids to their ED patients at any given hospital. The primary outcome was long-term opioid use, defined as a patient filling at least 90 days supply of opioids in the 12 months following the index visit. The analysis focused on opioid-naïve patients, defined as the patient having no opioid prescription in the 6 months preceding the ED visit. We test the hypothesis that treatment by high intensity prescribers is associated with higher odds of long term use (compared with treatment by low intensity prescribers) using multivariate logistic regression controlled for observable patient covariates (eg, age, race, diagnosis, sex, triage acuity, pain score) and hospital characteristics. Because ED patients generally cannot choose whether they are treated by a “high” or “low” intensity prescriber in an ED setting, the results are less likely to be confounded by selection bias. We identified 72,838 patients treated by “low intensity” prescribers and 94,192 treated by “high intensity” providers. High intensity opioid providers gave opioid prescriptions to 29.3% of their ED encounters compared to only 8.5% of the low intensity providers. However, there was no significant association between treatment by high versus low intensity prescribers on the key outcome of long-term opioid use among opioid-naïve patients (OR 1.02 95% CI 0.853-1.219). High comorbidity index (OR 3.5, 95% CI 3.2-3.9), high maximal pain score (OR 1.9, 95% CI 1.8-2.0) and advancing age were associated with higher risk of long-term opioid use. We do not observe any association with ED exposure to high versus low intensity prescribers and long-term opioid use. These findings are in contrast to a recent analysis conducted in a fee-for-service Medicare population. Reasons for these disparate results require further study but may include that the managed care environment protects against this exposure, our data is more recent and attention to limiting opioid prescribing has recently intensified (2013-2014 versus 2008-2011) or that the effects of exposure to a high intensity provider are limited to older populations.