COVID-19 COPD Exacerbation Risks

Evidence

This nationwide cohort study emulated a target trial using Swedish nationwide register data from the Swedish COVID-19 Investigation for Future Insights—a Population Epidemiology Approach using Register Linkage project. It examined individuals with chronic obstructive pulmonary disease (COPD) who had moderate or severe exacerbations between March 2020 and June 2023. The investigators compared 28-day risks after COVID-19-related versus non-COVID-19-related exacerbations for stroke, pulmonary embolism, acute myocardial infarction, major adverse cardiovascular events, and all-cause mortality. Exacerbations were considered distinct when they occurred at least 28 days apart. [pmid:41950377]

Moderate exacerbations were defined through treatment with oral corticosteroids with or without antibiotics, whereas severe exacerbations required hospitalization. COVID-19-related exacerbations were defined by COVID-19 infection, hospitalization, or intensive care admission. The study linked individual-level information across Swedish national registers, including demographic data from Statistics Sweden registers, prescriptions from the National Prescribed Drugs Register, and hospital information from the National Patient Register. Analyses estimated adjusted hazard ratios with 95% confidence intervals, stratified by exacerbation severity and COVID-19 variant periods. [pmid:41950377]

The cohort contained 266,273 exacerbations: 87.2% were moderate and 12.8% were severe. Of all exacerbations, 5,425 (2%) were COVID-19 related. This exposure distribution matters for interpreting precision, because the COVID-19-related group was much smaller than the comparison group. The reported cohort therefore reflects exacerbation episodes in Swedish register data rather than a randomized assignment of infection-related cause. [pmid:41950377]

Among moderate exacerbations, COVID-19-related episodes were associated with a higher 28-day risk of pulmonary embolism than non-COVID-19-related episodes, with an adjusted hazard ratio of 2.26 (95% confidence interval, 1.49-3.42). The same comparison was associated with higher risks of overall cardiovascular events and major adverse cardiovascular events. For all-cause mortality after moderate exacerbations, the adjusted hazard ratio was 4.58 (95% confidence interval, 4.06-5.17). COVID-19-related moderate exacerbations were not significantly associated with acute myocardial infarction or stroke in the reported analyses. [pmid:41950377]

For severe exacerbations, the reported pattern differed. COVID-19-related severe exacerbations were associated with higher all-cause mortality, with an adjusted hazard ratio of 1.46 (95% confidence interval, 1.28-1.66), while the reported cardiovascular risk associations were mainly observed in the moderate-exacerbation stratum. The investigators also reported that cardiovascular risks for moderate exacerbations were highest during pre-Alpha and Delta periods and attenuated during Alpha and Omicron. Mortality remained elevated for moderate and severe exacerbations, particularly during pre-Alpha and Delta. [pmid:41950377]

Analysis — Severity-stratified short-term risk pattern

The central result is a severity-stratified observational pattern rather than a uniform excess across every cardiovascular outcome. In the moderate-exacerbation comparison, COVID-19-related episodes were associated with pulmonary embolism, overall cardiovascular events, major adverse cardiovascular events, and all-cause mortality; acute myocardial infarction and stroke were not significantly associated in the reported analysis. In contrast, severe COVID-19-related exacerbations had an observed association with all-cause mortality, while reported cardiovascular associations were mainly concentrated among moderate episodes. [pmid:41950377]

This contrast should be read in the context of how the groups were assembled and measured. Moderate episodes were identified from outpatient treatment records, while severe episodes were identified through hospitalization, so the strata represent different clinical settings and ascertainment pathways. The authors noted that hospitalized exacerbations have higher baseline cardiovascular risk, which can narrow relative differences, and that moderate episodes may show larger relative effects from modest absolute increases. They also noted that differences in clinical management may have contributed to the larger cardiovascular risk difference in moderate episodes. These are interpretations consistent with the study’s register-based comparison, not evidence that COVID-19 caused the observed differences. [pmid:41950377]

The reported variation across pandemic periods adds temporal context but does not isolate why estimates changed. Variant periods can coincide with changes in prior immunity, testing, clinical management, health-care use, and the composition of people experiencing exacerbations. Accordingly, the observed attenuation of cardiovascular associations during Alpha and Omicron is a description of this cohort’s estimates, not a causal estimate of variant-specific biological effect. [pmid:41950377]

Limitations

The retrospective design cannot establish causation. Although the study emulated a target trial and estimated adjusted hazard ratios, COVID-19-related and non-COVID-19-related exacerbations were not randomly assigned. Selection processes could differ across the exposure groups because COVID-19-related status depended on documented infection, hospitalization, or intensive care admission, and testing or documentation may vary by setting and illness severity. The study also compared outpatient-treated moderate episodes with hospitalization-defined severe episodes only within their respective strata; these severity categories should not be treated as interchangeable populations. [pmid:41950377]

The authors reported that most covariates had acceptable balance between groups, but differences were slightly more pronounced for moderate exacerbations in some measures, with standardized mean differences just above 0.1. They also reported that people with COVID-19-related exacerbations were less frequently vaccinated against COVID-19 before the exacerbation. Such measured differences underscore that adjusted comparisons may retain residual confounding from factors imperfectly represented in registers. [pmid:41950377]

COVID-19-related exacerbations were rare in the cohort, which may have limited precision of exposure-effect estimates. Etiologic differentiation of non-COVID-19 exacerbations was also limited by the absence of microbiological data, so the comparator group was not microbiologically characterized by alternative causes. Finally, the study emphasized effect estimates and uncertainty rather than statistical significance. The reported confidence intervals quantify uncertainty around the presented associations, but they do not resolve residual confounding, exposure classification limitations, or differences in management between groups. [pmid:41950377]