Purdue University Graduate School
MS_Thesis_Iftekhar_Ahmed.pdf (567.58 kB)

Long-Term Opioid Therapy in Older Adults: Incidence and Risk Factors Related to Patient Characteristics and Initial Opioid Dispensed

Download (567.58 kB)
posted on 2022-12-07, 15:50 authored by Iftekhar AhmedIftekhar Ahmed


Background: Older adults have a higher prevalence of pain compared to other age groups and are more likely to become long-term opioid users. The clinical benefits of long-term opioid therapy (LTOT) are not clearly known, however, LTOT has been found to increase the risk of all-cause mortality, opioid overdose, constipation, fractures, and myocardial infarction. 

Objective: The study was conducted to estimate the incidence of LTOT and risk factors associated with LTOT in older adults aged 65 years and older.

Methods: This was a retrospective cohort study based on Medicare claims data obtained from Research Data Assistance Center (ResDAC). Opioid naïve older adults filling an opioid prescription between 2014 and 2016 were included. The outcome was LTOT which was defined as an opioid use episode lasting longer than 90 days and having more than 60 cumulative days of supply. The independent variables (risk factors) were patient characteristics (demographics, comorbidities, substance use disorders), characteristics of initial/index opioid dispensed (opioid type, duration of action of opioid, opioid dose, number of days’ supply, concomitant medications), and pain conditions. Multivariable logistic regression was performed to assess the association between the risk factors and LTOT. To address statistical interactions among variables, secondary analyses were conducted after stratifying the dataset by pain conditions.

Results: Among 162,287 opioid naive patients, 10,296 (6.3%) transitioned to LTOT. Demographic characteristics associated with LTOT were age greater than 85 years (adjusted odds ratios [AOR]: 1.1, 95% confidence interval [CI]:1.03-1.18) and being black (AOR: 1.11, 95% CI: 1.01-1.22). Risk factors related to substance use disorders included drug use disorder (AOR: 1.59, 95% CI: 1.30-1.95), alcohol use disorder (AOR: 1.26, 95% CI: 1.06-1.49), tobacco use disorder (AOR: 1.33, 95% CI: 1.21-1.45), and a history of opioid use disorder (OUD) (AOR: 1.63, 95% CI: 1.34-1.98). Patients with more than 5 comorbidities had 1.56 times higher odds (95% CI: 1.46-1.66) of LTOT compared to patients with 0-2 comorbidities. Characteristics of initial/index opioid associated with LTOT were dispensing long-acting opioids (AOR: 1.73, 95% CI: 1.22-2.46), concomitant use of benzodiazepines (AOR: 1.19, 95% CI: 1.11-1.28), gabapentinoids (AOR: 1.59, 95% CI: 1.49-1.69), and non-steroidal anti-inflammatory drugs (NSAIDs) (AOR: 1.23, 95% CI: 1.16-1.30). Starting therapy with tramadol increased the odds of LTOT compared to hydrocodone in patients with osteoarthritis and joint pain (AOR: 1.22, 95% CI: 1.06-1.41) as well as abdominal and bowel pain (AOR: 1.53, 95% CI: 1.05- 2.22). However, starting therapy with oxycodone decreased the odds of LTOT in patients with osteoarthritis and joint pain (AOR: 0.69, 95% CI: 0.53-0.90). For all pain conditions, initial opioid supply of ≥30 days led to 10-16 times higher odds of LTOT compared to days’ supply of 1-3 days.

Conclusions: Higher age, black race, comorbidities, substance use disorders, and history of OUD are the patient-related risk factors of LTOT in older adults. Moreover, specific patterns of initial/index opioid prescription/dispensing such as greater number of days’ supply, dispensing long-acting opioids, and concomitant use of benzodiazepines, gabapentinoids, and NSAIDs increase the odds of LTOT. Prescribers should take these factors into consideration when prescribing opioids to older adults.


Degree Type

  • Master of Science


  • Pharmacy Practice

Campus location

  • West Lafayette

Advisor/Supervisor/Committee Chair

David R. Foster

Additional Committee Member 2

Alan J. Zillich

Additional Committee Member 3

Kevin M. Sowinski

Additional Committee Member 4

Noll L. Campbell

Usage metrics



    Ref. manager