Chronic Opioid Use Is Associated With Early Biologic Discontinuation in Inflammatory Bowel Disease

Christian Rhudy; Courtney L. Perry; Michael Singleton; Jeffery Talbert; Terrence A. Barrett


Aliment Pharmacol Ther. 2021;53(6):704-711. 

In This Article


This study addresses the hypothesis that chronic opioid use among IBD patients portends shorter periods of persistence on biologic therapies independent of disease severity. In this analysis, we found that IBD patients with chronic opioid use trialled a greater mean number of biologic therapies and had shorter mean durations of therapy on the first and second agents trialled. These results are strengthened by a large, nationwide sample that allows for greater generalisability of the findings. An additional strength is the use of predictive modelling, which confirmed a significant difference in the risk of early discontinuation of biologic therapy when controlling for factors such as healthcare utilisation, abdominal surgeries, age, sex, steroid use and other medical comorbidities.

Based upon our univariable analysis, the time to discontinuation decreases with increased comorbid conditions, emergency department visits, inpatient admissions and steroid therapies. IBD patients with chronic opioid use in our study population were observed to have higher rates of comorbidities and healthcare utilisation as compared to the reference cohort. Using these variables as surrogate markers for disease acuity, we acknowledge the possibility that chronic opioid users have more severe IBD, which contributes to higher rates of biologic failure and therefore decreased persistence to biologic therapy. We also considered the possibility that the chronic opioid use cohort was simply sicker overall, given this group had a higher Charlson comorbidity index compared to the reference group. However, the chronic opioid use cohort was still 23% more likely to discontinue biologic therapy early even when controlling for the most relevant surrogate factors of disease severity using a multivariable predictive model. This suggests that the difference cannot be solely attributed to greater disease severity or concomitant comorbidities in the chronic opioid use cohort, which lends strength to the study's conclusions.

Results presented show that the risk of biologic discontinuation increased linearly as the number of emergency department visits and outpatient steroid usage increased. IBD patients with chronic opioid use were observed to visit the emergency department more frequently and receive higher doses of steroid therapy for presumed IBD flares. More frequent emergent healthcare utilisation in patients as well as a greater reliance on rescue steroid therapy may lead to earlier determination of biologic therapy failure and subsequently decreased persistence. However, given that IBD patients with chronic opioid use are more likely to have early discontinuation of biologic therapy even when controlling for surrogate markers of disease severity, including measures of healthcare utilisation, we suspect that other factors associated with chronic opioid use decrease the likelihood of persistence to biologic therapies. Symptoms of opioid withdrawal that occur with chronic opioid use resemble IBD flares, and may lead providers to inappropriately switch biologic therapies under a false assumption of poor disease control. Additionally, gaps in biologic therapy due to shorter periods of persistence may diminish treatment effect and increase the risk for the development of antibodies to biologics, necessitating discontinuation or therapeutic switch.[32]

It is interesting to note that patients with chronic opioid use were significantly more likely to receive self-injectable biologic therapies than those without chronic opioid use. This could be an additional reason for decreased persistence to therapy in this patient population, as it places increased onus on patients to obtain and self-administer therapy without further provider contact. Lack of persistence to therapy may be detected in a timelier manner in those receiving infusion therapies and allow for more rapid intervention to ensure patients remain on track with therapy.


This study relies on commercial insurance claims data, which may not be representative of the non-insured and Medicaid/Medicare populations. The retrospective design of this study may introduce increased risk of confounding factors. Due to the large scale of data collection, we were unable to track direct indicators of disease severity, including abdominal CT scan findings, endoscopy findings, faecal calprotectin, C-reactive protein, etc. Therefore, we were unable to directly control for disease severity and were forced to use surrogates of disease severity including abdominal surgeries, fistula and inpatient admissions. We acknowledge that direct indicators of disease severity are a major factor in clinical decision-making with regard to therapeutic alteration.