(005) Challenges, Lessons Learned and Implications for Improvement in Employing Drug Utilization Studies to Evaluate the Effectiveness of Additional Risk Minimization Measures (aRMMs)

Background: GVP module XVI draft revision 3 recommends drug utilization studies (DUS) leveraging secondary data as one of the main methods to evaluate the effectiveness of aRMMs in the EU. However, data from recent systematic review show that >50% of DUS were unable to render a conclusion. A better understanding of the challenges and limitations of DUS is crucial to successful aRMM evaluation.

Objectives: To describe challenges and provide lessons learned, based on results from a DUS assessing aRMM effectiveness for tofacitinib, a Janus kinase inhibitor, among patients treated for ulcerative colitis (UC).

Methods: This was an interim analysis of secondary data using national registries, electronic health records, and claims in 3 European countries: Sweden, the Netherlands, and Germany. Patients were required to have newly initiated tofacitinib from 04/2017-12/2019 (index date) and have 1 year of available data before and after the index date. Effectiveness of aRMMs was assessed descriptively by proportions of patients adherent to the aRMMs.

Results: A total of 415 UC patients initiated tofacitinib and fulfilled study eligibility criteria. Posology: Across the 3 countries, 31-37% of UC patients appeared to exceed the recommended daily dose in the maintenance phase, however, among which 58-78% had used an alternative therapy before initiating tofacitinib. Contraindicated use: 0-5% of patients initiated tofacitinib with a serious infection when infections were restricted to the inpatient setting versus up to 25% in any setting. The rates of all other contraindicated use were low (0-3%). Noncompatible concomitant medications: Up to 17% of UC patients had concomitant biologic use when defined as ≥30 days overlap and excluding potential medication switching, compared to 25-62% when concomitant usage was defined as ≥1 day overlap.

Conclusions: Outcome indicators for the effectiveness of aRMMs as measured by secondary data in DUS are sensitive to the definitions used. As days’ supply of medications are often not recorded in secondary data, simple approximation based on intervals between 2 dispensings may lead to overestimation in daily dose if early refill is common. High daily dose does not indicate non-adherence if there are permissive situations by the label, which may not be captured in some secondary data. Concomitant usage, commonly estimated by overlapping days’ supply, does not equate to true concomitant consumption especially in the presence of medication switching. When certain outcomes are poorly captured, it may be advisable to supplement DUS with other data collection methods (eg, surveys). Use of sensitivity analyses and validated algorithms, when available, to define outcomes are also highly recommended.