OBJECTIVES
To examine the rates and predictors of long‐term opioid therapy in older cancer survivors.
DESIGN
Retrospective cohort study.
SETTING
Texas, United States.
PARTICIPANTS
Cancer survivors (5 years or more postcancer diagnosis) diagnosed from 1995 to 2008 and who were also Medicare Parts A, B, and D beneficiaries.
MEASUREMENTS
We used Medicare Part D event data to calculate the proportion of cancer survivors with a prolonged opioid prescription (90‐day or more supply of opioids/year). Adjusted odds ratios were calculated to identify predictors of prolonged opioid prescribing. All analyses were repeated with a subcohort of opioid‐naïve cancer survivors.
RESULTS
The rate of prolonged opioid therapy for cancer patients diagnosed in 2008 was 7.1% prior to cancer diagnosis; it rose to 9.8% within a year of cancer treatments, and to 13.3% at 5 years postdiagnosis. The rate at the sixth year varied by cancer sites: 19.4% in lung cancer and 9.6% in prostate cancer. Among opioid‐naïve survivors, the rate increased from 1.4% to 7.1%, from 5 to 18 years postcancer diagnosis. Cancer survivors diagnosed in 2004 to 2008 had higher rates of opioid prescribing compared to those diagnosed in 1995 to 1998 and 1999 to 2003. Years since diagnosis, a later year of diagnosis, female sex, urban location, lung cancer diagnosis, disability as reason for Medicare entitlement, Medicaid eligibility, one or more comorbidity, and history of depression or drug abuse were predictors of prolonged opioid therapy. Among opioid‐naïve cancer survivors, diagnosis in 2004 to 2008 was the strongest predictor, while a history of drug abuse was the strongest predictor for all the survivors.
CONCLUSION
The rates of prolonged opioid prescribing for older cancer survivors remained high at 5 or more years after cancer diagnosis. Our findings have potential to inform the development of clinical guidelines and public policy to ensure safer and more effective pain treatment in older cancer survivors. J Am Geriatr Soc 67:945–952, 2019.