Trends in Short-, Intermediate-, and Long-Term Mortality Following Hospitalization for Myocardial Infarction Among Medicare Beneficiaries, 2008 to 2018

Abstract

Background Advances in technology and care quality have transformed the care of acute myocardial infarction (AMI), but little is known about trends in mortality rates across separate time periods after hospitalization. Methods and Results We identified all Medicare fee-for-service beneficiaries hospitalized with incident AMI from 2008 to 2018. We calculated unadjusted mortality rates by dividing the number of all-cause deaths by the number of patients with incident AMI for the following time periods: acute (in hospital), post acute (0-30 days after hospital discharge), short term (31 days to 1 year after discharge), intermediate term (1-2 years after discharge), and long term (2-3 years after discharge). Each period was considered separately (ie, patients who died during one period were not counted in subsequent periods). Using logistic regression to account for differences in patient characteristics, we calculated annual risk standardized mortality ratios defined as observed over expected mortality based on 2008 rates. Among 768 084 patients with incident AMI (mean age 81 years, 48% male, 87% White), declines in observed-to-expected mortality ratios were observed for each time period: acute (0.68 [95% CI, 0.66-0.71]), postacute (0.72 [95% CI, 0.71-0.75]), short term (0.77 [95% CI, 0.75-0.78]), intermediate term (0.79 [95% CI, 0.77-0.81]), and long term (0.77 [95% CI, 0.75-0.79]). Declines were observed both for patients with and without ST-segment-elevation AMI. Conclusions For patients with incident AMI, there have been improvements in mortality rates across periods spanning the hospital stay through 3 years after discharge, reflecting advances in AMI care from hospitalization through long-term outpatient follow-up.

Keywords: acute myocardial infarction; mortality; quality of care.

Figure 1. Risk standardized mortality ratios (relative to 2008) for patients with acute myocardial infarction (AMI) for each time period.

Sample includes all Medicare fee‐for‐service patients with an incident AMI hospitalization from 2008 through 2018. We used 2008 as a base year and used the estimated parameters from this year to estimate expected mortality for in‐hospital and postdischarge time increments from 2009 to 2018. Figure shows ratio of actual/expected mortality for each year and each time increment, with bootstrapped 95% CIs. Data points are jittered horizontally so that data from different time periods within a calendar year are visible.

Figure 2. Risk standardized mortality ratios (relative to 2008) for patients with and without ST‐segment–elevation acute myocardial infarction (AMI).

Overall sample, sample period, and methodology are the same as Figure 1, but overall sample is separated into ST‐segment–elevation AMI (A) and non–ST‐segment–elevation AMI (B).