Currently Viewing:
The American Journal of Managed Care July 2019
Changing Demographics Among Populations Prescribed HCV Treatment, 2013-2017
Naoky Tsai, MD; Bruce Bacon, MD; Michael Curry, MD; Steven L. Flamm, MD; Scott Milligan, PhD; Nicole Wick, AS; Zobair Younossi, MD; and Nezam Afdhal, MD
Precision Medicines Need Precision Patient Assistance Programs
A. Mark Fendrick, MD; and Jason D. Buxbaum, MHSA
From the Editorial Board: Robert W. Dubois, MD, PhD
Robert W. Dubois, MD, PhD
Real-Time Video Detection of Falls in Dementia Care Facility and Reduced Emergency Care
Glen L. Xiong, MD; Eleonore Bayen, MD, PhD; Shirley Nickels, BS; Raghav Subramaniam, MS, BS; Pulkit Agrawal, PhD; Julien Jacquemot, MSc, BSc; Alexandre M. Bayen, PhD; Bruce Miller, MD; and George Netscher, MS, BS
Currently Reading
Impact of a Co-pay Accumulator Adjustment Program on Specialty Drug Adherence
Bruce W. Sherman, MD; Andrew J. Epstein, PhD; Brian Meissner, PharmD, PhD; and Manish Mittal, PhD
Medicare’s Bundled Payment Model Did Not Change Skilled Nursing Facility Discharge Patterns
Jane M. Zhu, MD, MPP; Amol Navathe, MD, PhD; Yihao Yuan, MSc; Sarah Dykstra, BA; and Rachel M. Werner, MD, PhD
Number of Manufacturers and Generic Drug Pricing From 2005 to 2017
Inmaculada Hernandez, PharmD, PhD; Chester B. Good, MD, MPH; Walid F. Gellad, MD, MPH; Natasha Parekh, MD, MS; Meiqi He, MS; and William H. Shrank, MD, MSHS
Insurers’ Perspectives on MA Value-Based Insurance Design Model
Dmitry Khodyakov, PhD; Christine Buttorff, PhD; Kathryn Bouskill, PhD; Courtney Armstrong, MPH; Sai Ma, PhD; Erin Audrey Taylor, PhD; and Christine Eibner, PhD
Healthcare Network Analysis of Patients With Diabetes and Their Physicians
James Davis, PhD; Eunjung Lim, PhD; Deborah A. Taira, ScD; and John Chen, PhD
What Are the Potential Savings From Steering Patients to Lower-Priced Providers? A Static Analysis
Sunita M. Desai, PhD; Laura A. Hatfield, PhD; Andrew L. Hicks, MS; Michael E. Chernew, PhD; Ateev Mehrotra, MD, MPH; and Anna D. Sinaiko, PhD, MPP
Physician Satisfaction With Health Plans: Results From a National Survey
Natasha Parekh, MD, MS; Sheryl Savage; Amy Helwig, MD, MS; Patrick Alger, BS; Ilinca D. Metes, BS; Sandra McAnallen, MA, BSN; and William H. Shrank, MD, MSHS
Evaluation of Interdisciplinary Geriatric Transitions of Care on Readmission Rates
Nada M. Farhat, PharmD; Sarah E. Vordenberg, PharmD, MPH; Vincent D. Marshall, MS; Theodore T. Suh, MD, PhD, MHS; and Tami L. Remington, PharmD

Impact of a Co-pay Accumulator Adjustment Program on Specialty Drug Adherence

Bruce W. Sherman, MD; Andrew J. Epstein, PhD; Brian Meissner, PharmD, PhD; and Manish Mittal, PhD
Commercial health plan initiation of a co-pay accumulator adjustment program for specialty medications treating autoimmune diseases was associated with significant reductions in medication adherence and persistence.


The CAAP studied here was started on January 1, 2017, by self-insured employers in a private health insurance exchange. This analysis used enrollment and SpRx claims data for patients enrolled in high-deductible health savings account (HSA) plans or zero pharmacy deductible preferred provider organization (PPO) plans. The CAAP was operationalized by the PBM, separately tracking the dollar amounts from the co-pay assistance card and a patient’s actual OOP expenditure. These amounts were added to the pharmacy claim detail and sent to the health plan, which then updated the combined pharmacy and medical deductible accumulator total to reflect only the patient-paid amount. To assess the CAAP’s impact, this study compared patterns of autoimmune SpRx use between HSA and PPO enrollees before and after CAAP implementation. We hypothesized that autoimmune drug use decreased more rapidly for HSA enrollees than PPO enrollees under the CAAP but that their drug use trends were parallel before CAAP implementation.

Data and Sample

This study examined SpRx use among a convenience sample of more than 150,000 enrollees in employer-sponsored healthcare coverage during 2016 and 2017 through a private health insurance exchange (RightOpt, Conduent HR Services). The RightOpt exchange is administratively managed as a portfolio of integrated health plan offerings. Benefit design options include a no-deductible PPO and HDHPs with HSA or health reimbursement arrangement benefit design options. (This analysis used data only from PPO and HSA enrollees who were subject to the CAAP. All HSAs had combined medical/pharmacy annual deductibles of at least $1400.) To preserve confidentiality, no additional details were available.

Deidentified SpRx claims and enrollment data during January 1, 2016, through October 31, 2017, were collected on 3543 individuals with SpRx claims from 15 employers, each with a nationally distributed workforce, from the PBM (CVS Caremark; Woonsocket, Rhode Island). Besides standard pharmacy claims content, the data from January 1, 2017, onward included information on co-pay assistance card use and the patient net OOP payment to ensure that only the actual patient payment amount was credited toward the deductible.

The analytic sample consisted of patients who had at least 1 claim for any of 23 SpRx medications used to treat any of 8 autoimmune conditions (eAppendix [available at]) listed in the CVS specialty pharmacy drug list12 and had continuous enrollment in 1 HSA or PPO plan (ie, no coverage interruptions or plan switching) during the study period.

Statistical Analyses

There were 3 primary outcomes (monthly fills per person, risk of treatment discontinuation, and proportion of days covered [PDC]) and 1 secondary outcome (risk of absolute treatment discontinuation), as described below. Trends in unadjusted outcomes were plotted graphically. Adjusted analyses controlled for patient age as of January 1, 2017; sex; and zip code–level estimated adjusted gross income (AGI) in 2015.13 Statistical comparisons were made between HSA and PPO enrollees for 2016 (prior to the CAAP start) and 2017 (when enrollees were subject to the CAAP) using regression models. Model standard errors were made robust to heteroskedasticity and clustering where relevant. HSA- and PPO-specific expected outcomes, their differences, and their 95% CIs were calculated at the end of each period’s follow-up (December 31, 2016, and October 31, 2017, respectively) from the model results.

Mean fills per person was calculated by calendar month and plan type for January 2016 through October 2017 as the number of autoimmune SpRx claims divided by the number of individuals with plan coverage in that month. Using data at the person-month level, an adjusted linear regression was estimated on the outcome of number of fills as a function of plan-specific linear time-trend splines with knots at December 2016. Slope coefficients represented monthly changes in mean fills per person and were compared for HSA versus PPO enrollees by calendar year using Wald tests.

Treatment discontinuation was defined as the presence of a gap between the end of a patient’s drug supply and the next refill claim that exceeded 60 days. We identified separate subsamples of patients with autoimmune prescriptions that were not yet discontinued as of index dates of February 1, 2016 (pre-CAAP), and January 1, 2017 (post-CAAP). The time to discontinuation following each index date was characterized with unadjusted Kaplan-Meier survival curves and compared between plan types with adjusted Cox proportional hazards models that produced hazard ratios (HRs). Patients who ended insurance coverage prematurely or reached the end of follow-up (December 31, 2016, or October 31, 2017, respectively) without discontinuation were censored. Besides age, sex, and AGI, the Cox models controlled for the number of fills observed prior to the index date and were stratified by the patient’s drug at the index date (ie, adalimumab, etanercept, or other). Analyses were repeated for the secondary end point of absolute discontinuation, which was defined as treatment discontinuation with no evidence of treatment restart or switching.

Construction of the PDC outcome paralleled treatment discontinuation. Separate subsamples were constructed for patients with positive drug supply from previously filled autoimmune prescriptions as of index dates February 1, 2016, and January 1, 2017. From the index date until the end of follow-up (December 31, 2016, and October 31, 2017, respectively), PDC was calculated daily for each enrollee as days with positive drug supply divided by total days since the index date. Patients’ stockpiling drug supply was calculated from January 1, 2016, forward. Separately for the 2016 and 2017 subsamples, linear trendlines were fit on patient-day–level PDC data for HSA and PPO enrollees starting 30 days after index date through the end of follow-up. The trendlines were estimated using linear regression models that adjusted for patient age, sex, AGI, observed fills before index date, and drug. Slope coefficients were scaled to represent changes in mean PDC per 30 days and were compared between plan types pre- and post CAAP using Wald tests.

Three sensitivity analyses were conducted. First, as a robustness check, the treatment discontinuation and PDC analyses were repeated for the post-CAAP period with an alternate index date of February 1, 2017. Second, all analyses were repeated with stricter sample selection criteria. Enrollees were required to have continuous coverage from January 1, 2016, through October 31, 2017, and to have used a co-pay card during the first quarter of 2017. Third, the main analyses were conducted for HSA and PPO enrollees with MS SpRx, which were on the preventive drug list, not subject to the HSA deductible, and thus not expected to be affected by the CAAP.

Statistical significance was based on 2-sided tests with α = .05. Analyses were conducted using Stata MP 15.1 (StataCorp; College Station, Texas).

Copyright AJMC 2006-2020 Clinical Care Targeted Communications Group, LLC. All Rights Reserved.
Welcome the the new and improved, the premier managed market network. Tell us about yourself so that we can serve you better.
Sign Up