Classification Algorithm to Distinguish Between Type 1 and Type 2 Myocardial Infarction in Administrative Claims Data

Area of Science:

• Cardiology
• Health Informatics
• Health Services Research

Background:

• Type 2 myocardial infarction (T2MI) and type 1 myocardial infarction (T1MI) have distinct patient demographics, comorbidities, treatments, and outcomes.
• Accurate quality and outcomes assessment necessitates reliable differentiation between T1MI and T2MI within administrative claims data.

Purpose of the Study:

• To develop and validate a classification algorithm for distinguishing T1MI from T2MI in administrative claims data.
• To enable more precise quality and outcomes assessments for acute myocardial infarction.

Main Methods:

• Utilized Medicare accountable care organization data (2018-2021) from a New England healthcare system.
• Developed a classification algorithm using International Classification of Diseases, Tenth Revision codes, patient demographics, and comorbidities.
• Assessed algorithm performance using a C-statistic and compared classified T2MI with observed T2MI.

Main Results:

• The study included 7759 hospitalizations for myocardial infarction (46.5% T1MI, 53.5% T2MI).
• Factors associated with higher odds of T2MI included female gender, Black race, COVID-19 diagnosis, and hypertensive emergency.
• The classification algorithm achieved a C-statistic of 0.83, indicating good discriminatory power.

Conclusions:

• A straightforward classification algorithm can effectively differentiate between T1MI and T2MI hospitalizations.
• This algorithm can be applied to claims data both before and after the introduction of the T2MI code.
• Improved classification facilitates more accurate longitudinal assessments of acute myocardial infarction quality and outcomes.