Post-Myocardial Infarction Guideline-Recommended Therapy Utilizing Sankey Diagrams Among Medicare Beneficiaries

To describe 2-year post-myocardial infarction (MI) longitudinal patterns of guideline- directed medical therapy (GDMT) and cardiac rehabilitation (CR) participation with Sankey diagrams.

Eligible Medicare beneficiaries were aged 66 to 95 years with an acute MI (International Classification of Diseases-9-CM discharge codes of 410.xx excluding 410.x2) hospital admission between January 1, 2014 and September 30, 2015 and ≥1 follow-up CR sessions. We defined GDMT (angiotensin converting enzyme-inhibitor or angiotensin receptor blocker, statin, and β-blocker) use as having at least a 21-day supply available during a 30-day window. We stratified CR participation by days with claims (1-11, 12-23, ≥ 24). Population level trends of 6 GDMT combinations, CR participation, and death were depicted with Sankey diagrams.

Study population consisted of 5793 beneficiaries, 72% of whom had ≥1 GDMT pre-MI, 93% had ≥1 GDMT at baseline, and 45% initiated CR by 30 days post-MI. A median 23% of CR participants did not flow from low to moderate CR participation each month. At 1-year post-MI, 37% of beneficiaries without pre-MI GDMT and 33% of beneficiaries with pre-MI GDMT concluded CR early. Between 9% and 16% of beneficiaries without pre-MI GDMT and 2% to 6% beneficiaries with pre-MI GDMT did not have a GDMT fill post-MI. On average, 4% to 5% of beneficiaries switched from β-blocker + statin to another GDMT group post-MI each month.

Describing patterns of secondary prevention method utilization with Sankey diagrams can identify intervention populations, such as groups with inconsistent CR participation, primary nonadherence to new medications, and volatile medication persistence.