Abstract
Despite known side effects, oral steroid use (Prednisolone) remains prevalent in the manage- ment of inflammatory bowel disease (IBD). We hypothesize that steroid use is multifactorial - historical practices, easy access, and quick response, are some factors that may explain the "overuse" of these medications. We sought to quantify and explore patterns and predictors of steroid use in real world IBD care. Crohn’s Colitis Care (CCCare) is a cloud-based IBD- specific electronic medical record (EMR) used in Australia and New Zealand. Deidentified data prospectively entered by clinicians and consumers were retrospectively analysed. People with IBD with an assessment during the last 14 months were included. Data were extracted on 15th August 2024. Steroid exposure was defined as short (1-28 days), moderate (29-56 days), or prolonged (>56 days). Recency was defined as recent (within 365 days), previous (365-1085 days), or no exposure. Stepwise logistic regression was used to explore relation- ships amongst steroid use and demographic and/or disease characteristics. In 5441 eligible people, 3136 (57.6%) had Crohn’s Disease (CD); 50.2% were female; median age was 42 years (IQR 56-32) and 66% were non-smokers. Only 994 people had steroid exposure (18.3%). In those exposed, median oral steroid daily dose was 0.22 mg (IQR 3.2-6.8); median oral steroid total dose was 246.1mg (IQR 351.2-69.2), and median annualized total steroid dose was 18.6 mg (IQR 31.6-13.3). Up to 10 courses of steroids were prescribed per person. The following covariates - IBD diagnosis, gender, age, previous IBD surgery, stoma surgery, IBD therapy and age at disease onset reached statistical significance. Compared to people with UC, those with CD had less likelihood of both recent steroid exposure and prolonged exposure (AOR 0.78, P=0.014 & AOR 0.72, P=0.018 respectively). Adults between 20-29 years had a much higher likelihood of both recent and prolonged exposure compared to people > 70 years (AOR 6.67, P=0.000 & AOR 6.58, P=0.000 respectively). Females had a higher likelihood of both prolonged and recent exposure to steroids compared to males (AOR 1.22, P=0.048 & AOR 1.23, P=0.041 respectively). People on advanced therapies had a greater chance of both recent and prolonged exposure compared to people on 5-ASA alone (AOR 1.65, P=0.05 & OR 1.71, P=0.000). Age at diagnosis reached statistical significance for steroid use parameters (AOR 1.03, P=0.000), but with only modest effect size. In this large real-world cohort, the majority had no recent steroid exposure. In those exposed, younger people, those with UC, females and those with more active disease had higher likelihood of prolonged exposure. This suggests that steroid use may be further reduced by proactive tailoring of effective therapies for those with UC and focusing on disease control in younger people with IBD.