Background: In patients with type 2 diabetes (T2D) with poor glycemic control, there is an unmet need for treatment optimization involving the initiation and/or intensification of insulin therapy, which is often delayed because of clinical inertia. Educational initiatives that target patients and physicians might be one way to address this need.
Objective: To evaluate the effectiveness of educational materials mailed to physicians and their patients in affecting initiation of insulin therapy and other health care outcomes.
Methods: This study, named PIVOTs (Personalized care and the role of Insulin as a Vehicle to Optimizing Treatments), used integrated medical and pharmacy claims data from the U.S.-based HealthCore Integrated Research Database between January 1, 2006, and April 4, 2014, to identify patients who were potential candidates for insulin therapy. Eligible patients were aged 18-75 years, currently enrolled in a commercial or Medicare Advantage health plan, with T2D diagnosis codes. Patients selected for insulin treatment education had glycated hemoglobin A1c (A1c) > 10%, irrespective of the number of noninsulin antihyperglycemic drugs used, or A1c > 8.0% and ≤ 10% while receiving ≥ 2 noninsulin antihyperglycemic drugs. For each identified patient, a corresponding treating physician was identified on a hierarchical basis. Physician-level randomization was conducted to assign physicians and their linked patients to the following 4 cohorts: (1) a cross-sectional cohort in which educational materials were sent to patients and physicians on a single outreach date; (2) a longitudinal cohort in which educational materials were sent to patients and physicians on 2 occasions, 3 months apart; (3) an enhanced cohort in which patients and physicians received the same mailings as the longitudinal cohort, plus physicians were invited to attend a 1:1 video conference academic detailing session; and (4) a control cohort in which patients and physicians did not receive any educational materials. Insulin initiation rates, A1c levels, and medical and pharmacy costs were assessed from claims over 6 and 12 months follow-up within each cohort.
Results: Mean insulin initiation rates at 12 months ranged from 9.2%-10.3% (all patients) to 12.3%-14.9% (subset with baseline A1c ≥ 9.0%), with similar rates across the intervention and control cohorts. Reductions in A1c from baseline were also similar across cohorts for all patients (0.1%-0.6%), as well as for those with a baseline A1c ≥ 9.0% (0.9%-1.6%). Approximately 14%-20% of patients achieved A1c < 7.0%, with no differences across cohorts. Changes in mean total all-cause and diabetes-related health care costs were also similar across cohorts.
Conclusions: The findings of this real-world, randomized intervention call into question the value of educational mailings as a means to overcoming clinical inertia and improving health outcomes in patients with T2D, at least in the context of insulin initiation.
Disclosures: This study was funded by Sanofi US. Bieszk and Wei are employees of Sanofi US. Grabner, Barron, and Quimbo are employees of HealthCore, which was under contract with Sanofi US for the conduct of this study. Yan is an employee of PHAR, LLC and was employed by HealthCore at the time this study was conducted and completed. Biel is an employee of Anthem. Chu is a consultant for Sanofi US; a member of the lecture bureaus for AstraZeneca, Eli Lilly, and Sanofi US; and has received research funding from Novo Nordisk. Study concept and design were contributed by Bieszk, Grabner, Wei, Quimbo, and Barron. Yan, Barron, Quimbo, and Grabner collected the data, which were interpreted by Biel, Chu, Bieszk, and Wei, with assistance from the other authors. The manuscript was written by Bieszk, with assistance from the other authors, and revised by Bieszk, Grabner, Biel, and Chu, along with the other authors. Part of this work was presented in poster format at the 76th Scientific Sessions of the American Diabetes Association; June 10-14, 2016; New Orleans, Louisiana.