Published in:
01-10-2016 | ORIGINAL ARTICLE
Algorithms to Identify Statin Intolerance in Medicare Administrative Claim Data
Published in: Cardiovascular Drugs and Therapy | Issue 5/2016
Login to get accessAbstract
Purpose
To compare characteristics of patients with possible statin intolerance identified using different claims-based algorithms versus patients with high adherence to statins.
Methods
We analyzed 134,863 Medicare beneficiaries initiating statins between 2007 and 2011. Statin intolerance and discontinuation, and high adherence to statins, defined by proportion of days covered ≥80 %, were assessed during the 365 days following statin initiation. Definition 1 of statin intolerance included statin down-titration or discontinuation with ezetimibe initiation, having a claim for a rhabdomyolysis or antihyperlipidemic event followed by statin down-titration or discontinuation, or switching between ≥3 types of statins. Definition 2 included beneficiaries who met Definition 1 and those who down-titrated statin intensity. We also analyzed beneficiaries who met Definition 2 of statin intolerance or discontinued statins.
Results
The prevalence of statin intolerance was 1.0 % (n = 1320) and 5.2 % (n = 6985) using Definitions 1 and 2, respectively. Overall, 45,266 (33.6 %) beneficiaries had statin intolerance by Definition 2 or discontinued statins and 55,990 (41.5 %) beneficiaries had high adherence to statins. Compared with beneficiaries with high adherence to statins, those with statin intolerance and who had statin intolerance or discontinued statins were more likely to be female versus male, and black, Hispanic or Asian versus white. The multivariable adjusted odds ratio for statin intolerance by Definitions 1 and 2 comparing patients initiating high versus low/moderate intensity statins were 2.82 (95%CI: 2.42–3.29), and 8.58 (8.07–9.12), respectively, and for statin intolerance or statin discontinuation was 2.35 (2.25–2.45).
Conclusions
Definitions of statin intolerance presented herein can be applied to analyses using administrative claims data.