(196) The prevalence and characteristics of patients with indicators of opioid abuse within an integrated group practice

Epidemiological data indicates a significant increase in the use of opioids and problems of opioid abuse in the US during the past decade. The prevalence of opioid abuse within Group Practice settings has not been well characterized. In this study we examined the prevalence rates and characteristics of patients ≥18 years old with an encounter ICD-9 diagnosis for opioid abuse and/or clinician documented opioid abuse noted in the electronic health record among patients receiving chronic opioid therapy (COT) (defined as ≥70 days supply of a prescription opioid analgesic in at least one calendar quarter between 2006-2012) in the integrated group practice of Group Health Cooperative. Statistical significance was tested across groups using chi-square for categorical variables. We examined a random sample of 12,471 patients who received COT during this period (about half the patients receiving COT). Overall, 9.1% of patients had an ICD-9 diagnosis of opioid abuse. No significant difference in prevalence was observed according to gender, with 9.6% of males on COT having an ICD-9 code compared to 8.9% of females (p=0.24). A significant difference was observed across age groups with patients 18-44 years of age having the highest rates (13.1%), compared to 8.5% for 45-64 years of age, and 4.5% for those over 65 years of age (p<0.0001). These rates will be compared to those identified using natural language processing of the 4,644,224 clinical notes from patient encounters within the Group Health electronic medical record system to determine agreement between these two approaches and the utility of automated clinical note analysis in measuring rates of opioid abuse.This study was sponsored by Pfizer Inc. Epidemiological data indicates a significant increase in the use of opioids and problems of opioid abuse in the US during the past decade. The prevalence of opioid abuse within Group Practice settings has not been well characterized. In this study we examined the prevalence rates and characteristics of patients ≥18 years old with an encounter ICD-9 diagnosis for opioid abuse and/or clinician documented opioid abuse noted in the electronic health record among patients receiving chronic opioid therapy (COT) (defined as ≥70 days supply of a prescription opioid analgesic in at least one calendar quarter between 2006-2012) in the integrated group practice of Group Health Cooperative. Statistical significance was tested across groups using chi-square for categorical variables. We examined a random sample of 12,471 patients who received COT during this period (about half the patients receiving COT). Overall, 9.1% of patients had an ICD-9 diagnosis of opioid abuse. No significant difference in prevalence was observed according to gender, with 9.6% of males on COT having an ICD-9 code compared to 8.9% of females (p=0.24). A significant difference was observed across age groups with patients 18-44 years of age having the highest rates (13.1%), compared to 8.5% for 45-64 years of age, and 4.5% for those over 65 years of age (p<0.0001). These rates will be compared to those identified using natural language processing of the 4,644,224 clinical notes from patient encounters within the Group Health electronic medical record system to determine agreement between these two approaches and the utility of automated clinical note analysis in measuring rates of opioid abuse.This study was sponsored by Pfizer Inc.