Risk of Atherosclerotic Cardiovascular Disease Hospitalizations after COPD Hospitalization among Older Adults

Abstract

Background: Meta-analyses have suggested the risk of atherosclerotic cardiovascular disease (ASCVD) events is significantly higher after a chronic obstructive pulmonary disease (COPD) exacerbation. However, these studies have been limited to highly selected patient populations potentially not generalizable to the broader population of COPD.

Methods: We assessed the risk of ASCVD hospitalizations after COPD hospitalization compared to before COPD hospitalization and identified patient factors associated with ASCVD hospitalizations after COPD hospitalization. This retrospective cohort study used claims data from 920,550 Medicare beneficiaries hospitalized for COPD from 2016-2019 in the US. The primary outcome was risk of a ASCVD hospitalization composite outcome (myocardial infarction, percutaneous coronary intervention, coronary artery by-pass graft surgery, stroke, or transient ischemic attack) in the 1 year after-COPD hospitalization relative to the 1 year before-COPD hospitalization. Time from discharge to a composite ASCVD hospitalization outcome was modeled using an extension of the Cox Proportional-Hazards model, the Anderson-Gill model with adjustment for patient characteristics. Additional analyses evaluated for interactions in subgroups and risk factors associated with the composite ASCVD hospitalization outcome.

Results: Among 920,550 patients (mean age, 73 years) the hazard ratio estimate (HR; 95% CI) for the composite ASCVD hospitalization outcome after-COPD hospitalization vs before-COPD hospitalization was 0.99 (0.97, 1.02; p = 0.53) following adjustment. We observed 3 subgroups that were significantly associated with higher risk for ASCVD hospitalizations after COPD hospitalization: 76+ years old, women, COPD hospitalization severity. Among the 19 characteristics evaluated, 10 were significantly associated with higher risk of CVD events 1 year after COPD hospitalization with hyperlipidemia (2.78; 2.67, 2.90) and history of cardiovascular disease (1.77; 1.72 1.83) associated with the greatest risk.

Conclusion: Among Medicare beneficiaries hospitalized for COPD, the risk of ASCVD hospitalizations was not significantly increased after COPD-hospitalization relative to before-COPD hospitalization. Although, we identified age 76+ years old, female sex, and COPD hospitalization severity as high risk subgroups and 10 risk factors associated with increased risk of ASCVD events after-COPD hospitalization. Further research is needed to characterize the COPD exacerbation populations at highest ASCVD hospitalization risk.

Figure 1:

Consort study diagram Cohort selection consort diagram *death date does not come before index date

Figure 2:

study schema

Figure 3:

Forest plot showing subgroup specific adjusted hazard ratios for association of COPD time period (after vs before) and composite CVD outcome using a recurrent Cox regression model for subgroups that had significant interactions with COPD time period. Also showing adjusted hazard ratios for association of COPD hospitalization burden (i.e., hospt in plot; combination of COPD period and hospitalization severity) and composite CVD outcome using a recurrent Cox regression model. Note that there was no hospitalization burden in the before period, while in the after period, hospitalization burden could either be Severe or Very Severe. Each stratified model was adjusted as applicable for age, sex, race, Medicaid dual eligibility, comorbidities (hypertension, diabetes, CHF, history of cardiovascular disease, cardiac arrhythmia, PVD, renal disease, VHD, depression, dementia, metastatic cancer, and hyperlipidemia), index year, and frailty.

Figure 4:

Forest plot showing adjusted hazard ratio estimates for association between CVD composite outcome and predictors of interest over 30, 90, 180, and 365 days from date of discharge; estimates were obtained from separate Cox proportional hazard models for each time frame (0–30 days, 0–90 days, 0–180 days, and 0–365 days; indicated in plot with color key).