As chronic opioid therapy (COT) becomes more common, complexity of pain management in the inpatient setting increases; little is known about medical inpatients on COT.


To determine the prevalence of COT among hospitalized patients and to compare outcomes among these patients relative to those not receiving COT.


Observational study of inpatient and outpatient administrative data.


All veterans with acute medical admissions to 129 Veterans Administration hospitals during fiscal years 2009 to 2011, residing in the community, and with outpatient pharmacy use.


We defined COT as 90 or more days of opioids prescribed in the 6 months prior to hospitalization. Patient characteristics included demographic variables and major comorbidities. Outcomes included 30-day readmission and death during hospitalization or within 30 days, with associations ascertained using multivariable logistic regression.


Of 122,794 hospitalized veterans, 31,802 (25.9%) received COT. These patients differed from comparators in age, sex, race, residence, and presence of chronic noncancer pain, chronic obstructive pulmonary disease, complicated diabetes, cancer, and mental health diagnoses including post-traumatic stress disorder. After adjustment for demographic factors, comorbidities, and admission diagnosis, COT was associated with hospital readmission (odds ratio [OR]: 1.15, 95% confidence interval [CI]: 1.10-1.20) and death (OR: 1.19, 95% CI: 1.10-1.29).


COT is common among medical inpatients. Patients on COT differ from patients without COT beyond dissimilarities in pain and cancer diagnoses. Occasional and chronic opioid use are associated with increased risk of hospital readmission, and COT is associated with increased risk of death. Additional research relating COT to hospitalization outcomes is warranted. Journal of Hospital Medicine 2014;9:82–87. © 2013 Society of Hospital Medicine