Impact of provider mailings on medication adherence by Medicare Part D members

详细信息    查看全文

• 作者：Bijan J. Borah^a ; ^{borah.bijan@mayo.edu" class="auth_mail" title="E-mail the corresponding author} ; Yang Qiu^b ; Nilay D. Shah^a ; Patrick P. Gleason^b ; ^c
• 关键词：CMS ; Centers for Medicaid and Medicare Services ; DID ; difference-in-difference ; OR ; odds ratio ; PBM ; pharmacy benefits manager ; PDC ; proportion of days covered ; PDP ; Part D plan ; PRG ; pharmacy risk group ; QDID ; quantile difference-in-difference ; RAS ; renin angiotensin system
• 刊名：Healthcare
• 出版年：2016
• 出版时间：September 2016
• 年：2016
• 卷：4
• 期：3
• 页码：207-216
• 全文大小：346 K

文摘

The Medicare 5-Star Rating System measures and provides incentive for improving Medicare Part D plans through a quality-based payment program. Adherence to medications for chronic conditions is key to the Star ratings. Our objective was to assess the impact of direct-to-provider letters on improving medication adherence.

Methods

Members of a large US pharmacy benefits manager (PBM) who did not adhere to prescription of oral diabetes (antidiabetics), cholesterol-reducing (statins), or hypertension (renin angiotensin system [RAS] antagonists) drug therapy were identified from the prescriptions claims data of>600,000 continuously enrolled Medicare members. Nonadherence was defined by the Star ratings definition of proportion of days covered (PDC)<80%. The PBM sent letters to prescribing physicians of nonadherent members, requesting that they discuss adherence barriers and potential solutions with their patients. A historical control cohort was constructed from the PBM satisfying the same eligibility criteria as the intervention cohort. Both binary (≥80%) and continuous PDC measures were assessed as outcomes through multivariate logistic regression and difference-in-difference models, respectively.

Results

Final sample sizes were 21,044; 106,829; and 73,560 patients for antidiabetic, statin, and RAS antagonist use, respectively, with approximately equal number of intervention and control subjects in each drug class. Physician mailing was associated with 11%, 16%, and 7% higher odds of being adherent by members in antidiabetic, statin, and RAS antagonist cohorts, respectively (all P<.001).

Conclusions

Within limitations of historical controls, physician mailing was associated with improved medication adherence.

Implications

Physician mailing can be an impactful tool for improving medication adherence.

Level of evidence

II