Cardiovascular Burden Linked to AECOPD Stay and Readmission

Key Takeaways

• Among Beijing hospital admissions for AECOPD, 71.7% involved at least one of the five assessed cardiovascular conditions.
• Prolonged stays were significantly associated with two or more cardiovascular conditions after full adjustment, but not with one.
• One-year AECOPD-related readmission had a significant adjusted association starting with one cardiovascular condition.

Fu and colleagues retrospectively included 3,684 primary-diagnosis AECOPD admissions from 2,281 patients aged at least 40 years in their cardiovascular comorbidity analysis at Beijing Friendship Hospital, China, from January 2015 through December 2024. Eligible laboratory tests had to be performed within 48 hours of admission. Repeat eligible admissions were retained and linked by patient. Documented hypertension, ischemic heart disease, arrhythmia, heart failure, and peripheral arterial disease were counted for each admission. Cardiovascular disease (CVD) counts were grouped as none, one, two, or at least three conditions, with no CVD as the regression reference.

The primary endpoint was a length of stay (LOS) of at least 14 days; the secondary endpoint was a recorded AECOPD-related readmission within a year of discharge. For readmission, 295 admissions with less than one year of observation were excluded, leaving 3,389 for Cox analysis. Patient-clustered generalized estimating equations (GEE) logistic regression addressed LOS, while Cox proportional hazards modeling with patient-level robust variance addressed readmission. Fully adjusted models included age, sex, diabetes, smoking, drinking, heart and respiratory rates, and hematologic, inflammatory, and coagulation measures. Recorded measures included neutrophil-to-lymphocyte ratio, hemoglobin, red cell distribution width, eosinophil percentage, platelets, and D-dimer.

Prolonged stays were more frequent with at least three CVDs than with none. Compared with no CVD, the fully adjusted odds ratio (OR) was 1.60 (95% confidence interval [CI] 1.18–2.16) for two CVDs and 1.95 (95% CI 1.30–2.93) for at least three. One CVD was not significantly associated with prolonged stay.

Observed one-year AECOPD-related readmissions were likewise more frequent in the highest-count group than in admissions without CVD. The fully adjusted hazard ratio (HR) was 1.38 (95% CI 1.06–1.80) for one CVD and 1.68 (95% CI 1.14–2.49) for at least three, compared with none; two CVDs were also significantly associated with readmission. Hypertension with arrhythmia was associated with both outcomes. Hypertension with ischemic heart disease, with or without arrhythmia or peripheral arterial disease, was associated with prolonged stay, while hypertension alone was associated with readmission. Analyses using stays of at least 13 days or continuous LOS generally supported the prolonged-stay pattern.

Because these data came from a single-center retrospective electronic health record analysis in China, the associations do not establish causal effects, intervention benefits, or a validated risk tool; no estimate of transferability to U.S. hospitals was provided. Baseline chronic obstructive pulmonary disease (COPD) severity was not adjusted for, readmissions elsewhere were not recorded, and COPD identification relied predominantly on codes without systematic spirometric confirmation. Cardiovascular diagnoses or their severity may have been incompletely captured, and estimates for less frequent combinations may be imprecise. Age, sex, smoking, drinking, and diabetes subgroups showed no significant interaction for prolonged LOS. The authors interpreted the differing count and combination patterns as potentially distinguishing inpatient from postdischarge risk.

In this Beijing cohort, cardiovascular profiles were associated with both outcomes, although significant adjusted associations emerged at different disease counts for prolonged stay and readmission. Fu and colleagues called for prospective multicenter validation.

Clinician Questions

Can a patient's cardiovascular disease count change across AECOPD hospital admissions?

Fu and colleagues reassessed documented hypertension, ischemic heart disease, arrhythmia, heart failure, and peripheral arterial disease for every qualifying AECOPD admission at Beijing Friendship Hospital. The same person could therefore enter a different count category on a later admission; eligible repeat admissions were linked by patient identifier, and the analyses accounted for patient clustering. This classification does not establish that cardiovascular disease progressed between admissions.

Which repeat hospitalization counted as one-year AECOPD-related readmission?

For each qualifying AECOPD admission at Beijing Friendship Hospital, the event was the earliest subsequent AECOPD-related admission recorded there within one year after discharge. The interval began at discharge, and additional hospitalizations were not separately counted as events for that index admission; admissions to other hospitals were not captured.