Abstract
Background:
Corticosteroids induce remission in ulcerative colitis (UC) but are associated with adverse effects, including prednisone ⩾20 mg/day associated with serious COVID-19 outcomes.
Objectives:
To evaluate the risk of rescue prednisone use following budesonide multimatrix system (MMX) initiation in newly diagnosed individuals with UC before and during the COVID-19 pandemic.
Design:
Population-based cohort study.
Methods:
We conducted a population-based study in Alberta, Canada, identifying adults with incident UC (2018–2022). Courses of budesonide MMX were examined for treatment failure, defined as prednisone rescue or inflammatory bowel disease-specific hospitalization within 90 days. Outcomes were compared between prepandemic (before April 2020) and pandemic periods using Cox proportional hazards models.
Results:
Among 269 dispensings of budesonide MMX (38 with multiple courses), 30.5% required rescue prednisone and 6.7% resulted in hospitalization. Budesonide MMX dispensing declined during the pandemic wherein, the hazard of prednisone rescue increased (HR 1.64; 95% CI: 1.01–2.67).
Conclusion:
Budesonide MMX was prescribed less frequently during the COVID-19 pandemic, with increased treatment failure, underscoring pandemic-related disruptions in UC management.
Plain language summary
Ulcerative colitis (UC) is a chronic disease that causes inflammation in the large intestine. When symptoms flare, steroids like prednisone can be used to control them. Prednisone can cause serious side effects, which were made worse with COVID-19. A different steroid, called budesonide multi matrix system (MMX), has fewer side effects.
We studied health records from over 200 people who were diagnosed with UC between 2018 and 2022 and prescribed budesonide MMX. We looked to see how many of them also received prednisone or were hospitalized within 90 days. Roughly 30% received prednisone after starting budesonide MMX, and about 7% were hospitalized. We also looked to see if budesonide MMX was being used more often before and during the pandemic. Budesonide MMX use was on the rise before the pandemic, but not during. Using budesonide MMX less during the pandemic may be due to changes to healthcare at that time. During the pandemic, budesonide may have been given to people with severe UC more often to avoid serious side effects. But it may not have been as effective for these patients. More research is needed to guide when budesonide MMX should be used and how best to support people with UC when healthcare is disrupted.
Background
The prevalence of ulcerative colitis (UC) is increasing, thereby expanding the population at risk for experiencing disease flares. 1 Systemic corticosteroids, particularly oral prednisone, are effective for inducing remission in moderate-to-severe UC flares but are associated with a wide range of adverse effects.2,3 While these risks may be acceptable in cases of severe disease, they are less justified for people with milder UC who are otherwise maintained on 5-ASA therapy. 4
During the COVID-19 pandemic, evidence showed that prednisone doses ⩾20 mg/day were associated with a significantly increased risk of severe COVID-19 outcomes, including hospitalization, ICU admission, and death.5,6 As a result, international guidelines recommended minimizing systemic corticosteroid use during UC flares wherever possible.7,8
Budesonide multimatrix system (MMX) is a delayed, extended-release oral corticosteroid designed to deliver drug throughout the colon while undergoing extensive first-pass hepatic metabolism (80%–90%), resulting in lower systemic exposure and minimal systemic immunosuppression compared with oral prednisone. 9 Owing to its mechanism of action and pharmacokinetic profile, budesonide MMX may be associated with a lower risk of infection, making it an attractive alternative to prednisone, particularly during the COVID-19 pandemic, when clinicians were more judicious in prescribing systemic corticosteroids. 5 Therefore, we conducted a population-based study to assess the risk of prednisone rescue following budesonide MMX prescription in newly diagnosed individuals with UC before and during the COVID-19 pandemic.
Methods
Data sources
Since 2002, the Discharge Abstract Database has captured administrative, clinical, and demographic information on all hospital discharges in Alberta, including deaths and interfacility transfers. Diagnoses are coded using the International Classification of Diseases, 10th Revision, Canadian Enhancement (ICD-10-CA). The Pharmaceutical Information Network Database (PIN) database is a centralized repository of prescription medications dispensed from community pharmacies across Alberta. It includes all prescription drugs regardless of supplementary drug plan coverage. Over 95% of community pharmacies submit data to PIN. However, medications dispensed during hospital stays are not captured. These data sources have been previously validated10,11 and widely used to study drug utilization patterns and health outcomes in individuals with UC residing in Alberta.12,13
Study population
Using a validated case-finding algorithm, we identified a prevalent population of individuals with inflammatory bowel disease (IBD) in a population-based administrative database from Alberta, Canada, between January 1, 2018, and December 31, 2022, with follow-up through March 31, 2023. 10 From this cohort, we restricted the population to adults (⩾18 years at diagnosis) with newly diagnosed UC during the study period. Incident cases of UC were defined by applying an eight-year washout period (January 1, 2010–December 31, 2017), to exclude individuals with any prior IBD diagnosis codes. We then used the PIN database (Supplemental Table 1) to further restrict the inception cohort to individuals who had a minimum of 90 days of follow-up after a first dose was dispensed from a pharmacy. Individuals concurrently dispensed prednisone at the time of budesonide MMX initiation were excluded. Individuals could contribute multiple entries to the dataset if they had more than one distinct budesonide MMX dispensing; however, repeat dispensings were only included if dispensing dates did not overlap and occurred at least 6 months apart.
Outcomes
A typical course of budesonide MMX is prescribed at 9 mg daily for 8 weeks. The primary outcome was failure of budesonide MMX to adequately control UC within 90 days of its dispensing, defined as either a dispensing of rescue prednisone within 90 days of budesonide MMX or an IBD-specific hospitalization within the same 90-day period (Supplemental Table 2). The primary outcome was a comparison of budesonide MMX failure rates before versus during the COVID-19 pandemic (defined as April 2020 onward), assessed within 90 days of budesonide MMX dispensing. A 90-day window was selected a priori to capture treatment failure during the expected therapeutic course while minimizing unrelated downstream events.
Statistical analyses
Descriptive statistics were used to summarize characteristics of the cohort at the time of budesonide MMX dispensing including age, sex, and days supplied. These summaries were stratified by overall dispensings and by treatment failure. The average monthly percent change (AMPC), along with 95% confidence intervals (CI), of the counts of dispensing of budesonide MMX was estimated using Poisson or negative binomial regression models, as appropriate. AMPCs of failure were calculated using proportions of treatment failure based on the number of dispensings in the same month (i.e., failures per month per total dispensings). Changes in AMPC before and during the COVID-19 pandemic were assessed to evaluate shifts in dispensing trends and failure over time. Cox proportional hazards models were used to estimate the hazard of treatment failure pre- versus during the pandemic, with results reported as hazard ratios (HR) and corresponding 95% CIs. The proportional hazards assumption was assessed using Schoenfeld residuals and was not violated. Robust standard errors clustered at the individual level were applied to account for within-individual correlation arising from the possibility that individuals could contribute more than one treatment episode. The study design and analyses are provided as a diagram (Supplemental Figure 1). The reporting of this study conforms to the Reporting of studies conducted using Observational Routinely collected health Data guidelines (Supplemental File). 14
Results
Between March 2, 2018, and December 31, 2022, 269 dispensings of budesonide MMX were recorded in 229 individuals; among them, 38 had two or more courses. Approximately 36% of all dispensings occurred before the onset of the COVID-19 pandemic. Median duration of treatment was 56 days (interquartile range (IQR): 30–60) (Table 1). The AMPC in budesonide MMX dispensing remained stable at 0.54% (95% CI: −0.19–1.27) across the entire study period. Comparing the prepandemic and pandemic periods showed a significant difference in AMPC: 3.46% (95% CI: 0.55–6.46) before versus −1.21% (95% CI: −2.75–0.36) during the pandemic (p = 0.005).
Characteristics of inception cohort of ulcerative colitis patients receiving budesonide MXX, proportion of rescue prednisone use or hospitalizations, and average monthly percentage change in budesonide MMX dispensing.
All patients hospitalized received rescue prednisone.
MMX, multimatrix system; CI, confidence interval; SD, standard deviation.
Within 90 days of dispensing, 82 budesonide MMX courses (30.5%; 95% CI: 25.0–36.4) were followed by rescue prednisone use, and 18 (6.7%; 95% CI: 4.0–10.4) by an IBD-specific hospitalization (Table 1); all those hospitalized received a rescue dose of prednisone. AMPC values for monthly proportions of treatment failure remained stable: 1.02% (95% CI: −0.46–2.53) for rescue prednisone and −0.99% (95% CI: −4.03–2.14) for IBD-specific hospitalizations (Table 1). No significant differences in AMPC emerged when comparing prepandemic and pandemic periods.
Cox proportional hazards models comparing failure before versus during the pandemic indicated an increased hazard of receiving rescue prednisone within 90 days of budesonide MMX dispensing during the pandemic period (HR: 1.64; 95% CI: 1.01–2.67) (Figure 1). No significant difference emerged in the hazard of IBD-specific hospitalization between the two time periods.
Kaplan–Meier survival curves showing the probability of remaining free from rescue prednisone use within 90 days following a budesonide MMX dispensing before (solid) and during (dashed) the COVID-19 pandemic.
Discussion
Budesonide MMX undergoes extensive first-pass hepatic metabolism (80%–90%), resulting in lower systemic exposure and reduced immunosuppressive effects compared to oral prednisone. 9 During the COVID-19 pandemic, it may have been favored over prednisone given concerns about corticosteroid-related complications. 5 However, in individuals with more severe disease, budesonide MMX may be insufficient, leading to treatment failure. 9 In this study, one-third of patients required rescue prednisone and nearly 7% were hospitalized within 90 days, with the likelihood of prednisone rescue higher during the pandemic. 15
Prior to the COVID-19 pandemic, prescriptions for budesonide MMX increased steadily by nearly 3.5% per month. From 2020 to 2022, this trend reversed, reflecting the widely reported disruptions in IBD care. Access to healthcare professionals and services such as endoscopy was restricted, and some individuals may have deferred management due to concerns about contracting COVID-19 while immunosuppressed. 16
The higher rate of budesonide MMX failure requiring rescue prednisone during the pandemic suggests a shift in prescribing behavior, with budesonide MMX potentially being used in more severely active patients to minimize prednisone exposure. The increased frequency of rescue prednisone use during the pandemic may also reflect lack of access to care or more severe presentation of disease activity. 17 Consistent with our findings, a UK study reported that rescue prednisone increased from 10% in 2019 to 31% during March–June 2020. 15 Together, these findings underscore the importance of reserving budesonide MMX for individuals with mild to moderate UC flares, where its efficacy is most appropriate and has been studied. 18
Limitations of this study include the potential for misclassification inherent in administrative data. 19 Although we used a validated case-finding algorithm with high sensitivity and specificity, some misclassification of UC cases cannot be excluded. 10 Moreover, administrative data lack clinical information such as Mayo scores or endoscopic findings, preventing stratification of patients by disease severity or controlling for confounding by indication at the time of budesonide MMX prescription. This limitation is particularly important, as treatment response and the likelihood of requiring rescue prednisone are closely tied to baseline disease activity. We defined rescue prednisone as occurring within 90 days of budesonide MMX dispensing to capture treatment failure during the typical 60-day course and the subsequent 30-day period after discontinuation, recognizing that some participants may have received extended courses beyond 60 days.
Additionally, residual confounding related to socioeconomic status and insurance coverage may have influenced prescribing patterns of budesonide MMX, particularly during the COVID-19 pandemic, when access to care was more limited. Temporal trend analyses using AMPC provide a population-level summary of dispensing patterns and enable evaluation of disruptions in medication use at the health-system level. Accordingly, trends observed before and during the pandemic should be interpreted as indicators of system-level change rather than direct measures of individual outcomes. Finally, the relatively small number of prednisone rescue events and hospitalizations limited statistical power and constrained the strength of inferences that could be drawn from these analyses. Extending the study beyond 2022 will be important to increase sample size and to determine whether prescribing patterns changed following the pandemic. Given these limitations, the interpretation that budesonide MMX may have been used in patients with more severe UC flares during the pandemic should be considered hypothesis-generating.
Conclusion
Overall, the pandemic was associated with altered prescribing patterns of budesonide MMX, with fewer dispensings and a higher likelihood of treatment failure requiring rescue prednisone among individuals treated for UC flares. These findings highlight the disruption in care experienced by people with UC during the pandemic.
Supplemental Material
sj-docx-1-tag-10.1177_17562848261434442 – Supplemental material for Increased failure of budesonide MMX for ulcerative colitis during the COVID-19 pandemic: a population-based study
Supplemental material, sj-docx-1-tag-10.1177_17562848261434442 for Increased failure of budesonide MMX for ulcerative colitis during the COVID-19 pandemic: a population-based study by Stephanie Coward, Karen J. B. Martins, Scott Klarenbach, Lawrence P. Richer, Christopher Ma, Remo Panaccione, Cynthia H. Seow, Karen Kroeker, Frank Hoentjen, Laura E. Targownik and Gilaad G. Kaplan in Therapeutic Advances in Gastroenterology
Supplemental Material
sj-docx-2-tag-10.1177_17562848261434442 – Supplemental material for Increased failure of budesonide MMX for ulcerative colitis during the COVID-19 pandemic: a population-based study
Supplemental material, sj-docx-2-tag-10.1177_17562848261434442 for Increased failure of budesonide MMX for ulcerative colitis during the COVID-19 pandemic: a population-based study by Stephanie Coward, Karen J. B. Martins, Scott Klarenbach, Lawrence P. Richer, Christopher Ma, Remo Panaccione, Cynthia H. Seow, Karen Kroeker, Frank Hoentjen, Laura E. Targownik and Gilaad G. Kaplan in Therapeutic Advances in Gastroenterology
