Background Adherence is under consideration for quality reporting in a number of disease states. Published data on adherence of biologics shows a wide range of calculation methods.
Objectives To compare methods of calculating adherence&persistence for immunology biologics & make recommendations for future research.
Methods Administrative medical & pharmacy claims from 11/1/2008 thru 12/15/2011 were sourced from a leading national health care provider. The first biologic claim between 5/1/2009-6/30/2010 was the pt's index date (Index). Four methods of assigning days' supply were tested. Method I used the pre-Index period to establish induction period status & assign dosing intervals based on product labeling. Methods II & III were adaptive & infliximab (IFX)-specific, using either the first to second infusion interval criteria of ≤21 days (Method II), or alternatively, activity in the pre-Index period (Method III), to establish induction period status, followed by assignment of dosing intervals based on product labeling. Method IV used either days' supply from the pharmacy claim or estimated values for medical benefit transactions.These estimated values were based on a qualitative a priori review of medical benefit transaction reporting frequencies. Adherence was measured using Proportion of Days Covered (PDC) and Medication Possession Ratio (MPR), representing analysis of a “fixed period” (364 days post-index) & “variable period” (index to cessation of treatment or end of observation) respectively. Several new methods of reporting adherence were tested, including:Proportion of Pts with treatment gap ≥20% of expected (PPgap ≥20), Sum of Gap Days ≥20% of expected (SoGD≥20), Sum of any Gap Days (SoGD), & Number of Gaps ≥10% of expected (NoG≥10). Persistence alternatives included Number of days from Index to a gap≥90 days, & Number of days from Index to a gap ≥10% of expected interval.Results are reported descriptively using means, standard deviations (SD),& percentages.
Results 636 pts had IFX claims;while in a separate sample of 523 chart review pts had only subcutaneous claims (SQ).Methods I & IV for estimating days' supply produced similar results for all analyses. Method III produced comparatively low adherence & persistence rates. PDC-type fixed period vs. MPR-type variable period consistently reported a 10-15% lower adherence rate,& a 30-50 day shorter time to event rate. For Method IV, variable period techniques–the IFX cohort had an MPR of 0.94 (SD 0.12) while the SQ cohort had MPR of 0.82 (SD 0.19). For alternative adherence measures, using Method IV, variable period techniques, comparing infusion (IFX cohort) vs. self-administered (SQ cohort) showed the PPgap≥20 was 39.8% vs. 74.2% resp. The SoGD≥20 was more than twice as great for the SQ vs. IFX cohort (56.1 (SD 65.0) vs. 23.2 (SD 46.3) resp).The SQ cohort reported more than 2.5 times the NoG≥10 vs. IFX cohort (2.59 (SD 2.14) vs. 0.97 (SD 1.22) resp).
Conclusions Substantial differences may result from assumptions made regarding missing days' supply & calculation methods for persistence & adherence when using medical claims. Quality reporting should include all details for days' supply assumptions & calculation methods. Alternative methods of reporting adherence may have greater clinical significance than MPR or PDC.
Disclosure of Interest : C. Kozma Consultant for: Janssen Scientific Affairs, LLC, A. Paris Consultant for: Janssen Scientific Affairs, LLC, M. Ingham Employee of: Janssen Scientific Affairs, LLC
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.