P313 Steroid and opioid use among Crohn’s patients before and after initiation of ustekinumab

Abstract

BackgroundA goal of pharmaceutical treatment (tx) in CD is resolution of symptoms. Proxies for successful tx from administrative claims databases can provide insight into real-world drug effectiveness. Reductions in use of drugs associated with tx of Crohn-related symptoms may be a proxy for tx effectiveness. To examine the use of opioids and steroids in patients with CD before and after initiation of ustekinumab (UST).MethodsThe analytic cohort was drawn from an immunology subset of Symphony Health Solutions data (1 October 2012–31 August 2017). Adult patients (age ≥18 years) were included if they had ≥2 CD diagnoses (ICD-9:555.xx,ICD-10:K50.xxx) separated by ≥30 days (d); no PsA diagnosis; no UST claims before FDA approval date (26 September 2016) for CD; an initial pharmacy claim for a 90 mg SC dose of UST dispensed after 20 November 2016; an index date ≥60 d prior to end date of the dataset (i.e. minimum observation); and claim activity for ≥360 d prior to and during pre-index period. Observation periods were pre-induction (index-120 d-index-61 d), induction (index-60 d-index-1 d) and post-index (index + 59d). Percentages of patients with opioid or steroid claims were tested using McNemars test, with post-index period as the referent group. Since the data do not contain eligibility information, a sensitivity analysis was conducted using patients with ≥1 prescription claim for a biologic, and ≥1 prescription claims for any medication except biologics. This subset has greater evidence of having complete medication data.ResultsPatients who met analysis criteria (n = 2036) had mean age of 41.7(SD 14.1) years, 58.3% were female, and 74.1% had commercial insurance. The percent of patients with ≥1 steroid claim by observation period was 37.6% pre-induction (p < 0.0001), 34.7% during induction (p < 0.0001) and 22.9% post-index. The percent of patients with ≥1 opioid claim by observation period was 29.6% pre-induction (p = 0.0002), 27.3% during induction (p = 0.1754) and 26.0% post-index (Table 1). In the subset with ≥1 claim for a biologic and ≥1 claim for any medication except biologics (n = 1817; Table 2), the percent of patients with ≥1 steroid claim by observation period was 40.3% pre-induction (p < 0.0001), 37.8% during induction (p < 0.0001), and 25.6% post-index. Percent with ≥1 opioid claim was 32.1% pre-induction (p = 0.0056), 29.9% during induction (p = 0.4446), and 29.1% post-index.ConclusionsThere was a statistically significant decline in the percent of patients with opioid claims, and percent with steroid claims within 60d of initiating the first dose of UST SC for the treatment of CD. Changes in steroids and/or opioid claims may be proxies for measuring real-world effectiveness of UST for treatment of CD.View largeDownload slideView largeDownload slide