Algorithms to Identify Statin Intolerance in Medicare Administrative Claim Data

Area of Science:

• Cardiovascular medicine
• Pharmacoeconomics
• Health services research

Background:

• Statin intolerance is a significant barrier to lipid-lowering therapy.
• Accurate identification of statin intolerance using administrative claims data is crucial for patient management and research.
• Existing claims-based algorithms for identifying statin intolerance vary, necessitating comparative analysis.

Purpose of the Study:

• To compare the characteristics of patients identified as possibly statin intolerant using different claims-based algorithms.
• To contrast these patients with those demonstrating high adherence to statin therapy.
• To evaluate the utility of various claims-based definitions for assessing statin intolerance.

Main Methods:

• Analysis of 134,863 Medicare beneficiaries initiating statins (2007-2011).
• Assessment of statin intolerance, discontinuation, and high adherence (proportion of days covered ≥80%) within 365 days of initiation.
• Utilized two distinct definitions for statin intolerance based on down-titration, discontinuation, ezetimibe initiation, specific events, or statin switching.

Main Results:

• Prevalence of statin intolerance was 1.0% (Definition 1) and 5.2% (Definition 2).
• 33.6% experienced intolerance or discontinuation, while 41.5% showed high adherence.
• Patients with statin intolerance were more likely to be female and belong to Black, Hispanic, or Asian racial/ethnic groups compared to high adherence patients.
• Higher intensity statin initiation was strongly associated with increased odds of statin intolerance (ORs ranging from 2.35 to 8.58).

Conclusions:

• The developed claims-based definitions for statin intolerance are applicable for analyses using administrative claims data.
• These algorithms can aid in identifying patient subgroups experiencing statin intolerance.
• Further research can leverage these definitions to understand the impact of statin intolerance on patient outcomes.