ABSTRACT
Background
Exacerbations of chronic obstructive pulmonary disease (COPD) often require treatment with oral corticosteroids (OCS). While OCS can improve forced expiratory volume in first second (FEV1) in the short term, there are still concerns about their potential cardiovascular adverse effects. Evidence on whether short-term OCS use increases the risk of major adverse cardiovascular events (MACE) remains lacking.
Hypothesis
We hypothesized that short-term OCS use in patients with COPD would be associated with an increased one-year risk of MACE compared with antibiotic treatment alone.
Methods
We conducted a registry-based, nationwide cohort study in outpatients diagnosed with COPD in the Danish Register of COPD (DrCOPD) between 2010 and 2022. Exposure was defined as prescription of OCS and antibiotics for respiratory infections in the year before baseline (defined as DrCOPD entry). Exposed patients were propensity score matched on clinical parameters to patients prescribed antibiotics for respiratory infections but no OCS. The primary outcome was all MACE, as a composite outcome, within 365 days following baseline.
Results
A total of 4,666 patients were included. Short-term OCS exposure was not associated with increased risk of MACE during one-year follow-up compared to use of antibiotics for respiratory infections (hazard ratio = 0.84, 95% CI 0.66 – 1.07, p = 0.17). The findings were consistent in a sensitivity analysis using an adjusted Cox model in the unmatched population. Secondary analyses of individual MACE components showed no significant differences in risk. An additional sensitivity analysis stratified by cumulative OCS dose revealed no dose-response relationship for the primary outcome (all MACE).
Conclusion
Short-term OCS use was not associated with an increased risk of MACE within 365 days among outpatients with COPD.
KEYWORDS: Chronic obstructive pulmonary disease, oral corticosteroids, antibiotics for respiratory infections, major adverse cardiovascular events, cohort study, propensity score matching, hazard ratios
Introduction
Exacerbations of chronic obstructive pulmonary disease (COPD) are often treated with oral corticosteroids (OCS) and/or antibiotics, prescribed across various healthcare settings including general practice, emergency departments, and outpatient clinics. Current guidelines recommend OCS at a dose of up to 40 mg of prednisolone equivalent daily for 5 days [1]. Patients may also use rescue home-treatment of OCS to quickly address exacerbations and prevent further decline in lung function before consulting a physician [2]. OCS are recommended for acute exacerbations of chronic obstructive pulmonary disease (AECOPD) due to their effectiveness in improving forced expiratory volume in first second (FEV1), reducing relapse risk, and accelerating recovery [1,3,4].
However, concerns have been raised regarding both short-term and long-term adverse effects of OCS, including increased risk of sepsis, venous thromboembolism, fractures, type 2 diabetes, obesity, osteoporosis, gastrointestinal bleeding, cataracts and hypertension [5]. Studies indicate that adverse effects exhibit a dose-response relationship with OCS exposure in the treatment of patients with asthma [6,7]. Prolonged OCS exposure is an independent predictor of all-cause mortality in patients with COPD. Reducing the recommended treatment duration from 14 to 5 days has proven non-inferior in terms of re-exacerbation within 6 months as well as survival rates [8–10]. COPD itself is associated with an increased risk of cardiac events, partly due to shared risk factors such as age, smoking, and physical inactivity. OCS treatment may further contribute to this risk through development of hypertension and hyperlipidemia, which are known risk factors for acute myocardial infarction [11]. While observational studies suggest a link between OCS use and cardiac events in COPD patients, large-scale systematic investigations are lacking [11–14].
Cardiovascular risks have also been reported for several antibiotics used for respiratory infections. For example, fluoroquinolones have been associated with increased rates of cardiovascular events, and the CLARICOR trial found higher cardiovascular mortality among patients with stable coronary heart disease treated with clarithromycin [15,16]. However, studies comparing the cardiovascular impact of short-term OCS in combination with antibiotics for respiratory infections versus antibiotics alone in large, well-characterized COPD populations are lacking. We aimed to determine whether short-term OCS use in patients with COPD is associated with an increased risk of MACE compared to treatment with antibiotics for respiratory infections, without OCS. We hypothesize that short-term OCS use is associated with a higher risk of MACE in patients with COPD compared to antibiotics for respiratory infections alone.
Methods
Study design
A nationwide retrospective cohort study was conducted by combining information from the following registries:
The Danish Register of Chronic Obstructive Pulmonary Disease (DrCOPD). Established in 2008, it is a nationwide database containing information on the quality of treatment of all patients with COPD who are treated by a respiratory medicine specialist at a Danish hospital. The validity and completeness of COPD diagnoses in DrCOPD have previously been documented [17]. Covariates included in the study were age, lung function assessed as forced expiratory volume in first second as percent of predicted (FEV1%), body mass index (BMI) assessed as kilograms per square meter, dyspnea assessed using the Medical Research Council (MRC) Dyspnea Scale, and smoking status.
The Danish Civil Registration System. All citizens in Denmark acquire a unique personal identification number at birth or upon immigration. In this register, the unique personal identification number links individual information for each resident to data on name, sex, date of birth, and vital status, that is, whether the person is alive, has emigrated, or has died (with dates of death or emigration) [18].
The Danish National Health Service Prescription Database holds information on all prescriptions dispensed by Danish pharmacies since 2004 (coded according to ATC classification), including date of dispensation, quantity dispensed, strength, and formulation. All pharmacies are required by Danish legislation to provide information ensuring complete and accurate registration [19].
The Danish National Patient Registry holds information on all admissions to Danish hospitals since 1977, and hospital outpatient clinic visits since 1995. Each visit is coded by physicians with one primary diagnosis and one or more secondary diagnoses, according to the International Classification of Diseases, eighth revision (ICD-8) codes until 1994 and ICD-10 thereafter [20].
The Danish Register of Causes of Death dates to 1970 and holds information on causes of death on all deaths in Denmark. All deaths in Denmark are registered by a doctor and both the manner and cause of death are registered in this register [21].
Study population
All Danish residents with a specialist verified diagnosis of COPD registered in DrCOPD, who attended an outpatient clinic between 1 January 2010 and 1 January 2022, were eligible for inclusion in the study. Further, individuals were required to be at least 30 years of age at the time of study entry. Participants were excluded if they had received maintenance oral corticosteroid (OCS) therapy, defined as having filled at least one prescription for low-dose OCS (<25 mg of prednisolone equivalent) within the 365 days prior to baseline, or if they had a cancer diagnosis within 5 years before study entry, excluding non-melanoma skin cancer (ICD-10 code DC44).
Study entry and exposure
Study entry was defined as the date of a patient’s enrollment in the DrCOPD registry.
Exposure was defined as at least one prescription of OCS and antibiotics for respiratory infections or antibiotics for respiratory infections without OCS, in the year before enrollment in the DrCOPD registry.
We compared the risk of MACE for patients with COPD treated with short-term OCS (a prescription of OCS for 25–37.5 mg was assumed to indicate a 5–7-day treatment for AECOPD (≤250 mg OCS)) in combination with antibiotics for respiratory infections versus treatment with antibiotics for respiratory infections alone. Antibiotic exposure was defined as prescriptions of commonly used agents for the clinical management of respiratory infections in patients with COPD, including amoxicillin (J01CA04), amoxicillin with clavulanic acid (J01CR02), ciprofloxacin (J01MA02), doxycycline (J01AA02), azithromycin (J01FA10), roxithromycin (J01FA06), clarithromycin (J01FA09) and moxifloxacin (J01MA14).
Follow-up
All patients were followed for 365 days from study entry. This was the observation study period during which patients were monitored for the occurrence of events.
For the primary and secondary analyses, censoring criteria included mortality from causes other than lethal MACE, except for the secondary analysis of all-cause mortality, which had no censoring criteria.
Outcomes
Patients were followed for events of the primary composite outcome; All MACE involving lethal cardiovascular events, cardiovascular events requiring revascularization, cardiovascular events requiring admission or cardiovascular events requiring prescriptions of ADP receptor inhibitors or nitrates. We used the common definition of MACE, including medical treatment for acute coronary syndrome [22].
They were also followed for the secondary outcomes: i) severe MACE, including lethal cardiovascular events, cardiovascular events requiring revascularization and cardiovascular events requiring admission, ii) lethal cardiovascular events, iii) cardiovascular events requiring revascularization, iv) cardiovascular events requiring admission and v) cardiovascular events requiring prescriptions of ADP receptor inhibitors or nitrates. ICD-10 codes for events can be found in Table S1 in the Supplementary Material.
The outcomes were assessed from the Danish National Patient Registry and the Danish National Health Service Prescription Database.
Statistics
Patients treated with OCS and antibiotics for respiratory infections were propensity score matched with patients treated with antibiotics for respiratory infections alone on known and potential confounders:
- Age (as a continuous variable)
- Sex (male or female)
- Tobacco exposure (divided into the categories ‘never or passive smoking’, ‘previous smoking’, ‘active smoking’ and ‘unknown tobacco exposure’)
- Exacerbation of COPD requiring hospitalization in the year before study entry (admission vs. no admission)
- Prescription of inhaled corticosteroids (ICS) in the year before study entry (prescription vs. no prescription)
- Prescription of long-acting muscarinic antagonists (LAMA) in the year before study entry (prescription vs. no prescription)
- MRC (with the options 1, 2, 3, 4 and 5)
- BMI (as a continuous variable)
- FEV1% (as a continuous variable)
- Charlson Comorbidity Index (CCI) score (as a continuous variable)
- Entry year (as a continuous variable)
Propensity score matching was performed using the greedy matching algorithm from the Mayo Clinic [23].
To test the similarity of the baseline characteristics between the propensity score matched groups, we applied unpaired t-tests for continuous variables with a normal distribution and Mann-Whitney U tests for non-normally distributed variables. Categorical variables were compared using chi-square tests. To evaluate the strength of any associations found in the baseline characteristics, we calculated Cramér’s V which was interpreted as follows: values between 0.00–0.10 indicated negligible association, 0.11–0.30 weak, 0.31–0.50 moderate, and >0.51 indicated a strong association. These thresholds were applied according to established conventions for interpreting Cramér’s V [24]. Cox regression models and cumulative incidence curves based on Fine-Gray models were used to assess the risk of events between the groups. In the Fine-Gray models, deaths from causes other than MACE were considered competing events, whereas in the all-cause mortality analysis, death was the primary event without consideration of competing risks. For the primary analysis, an unadjusted Cox regression model was employed.
For sensitivity analysis, we first conducted an adjusted Cox proportional hazards regression model of the primary outcome in the unmatched population, comprising all 7,435 patients: 3,761 patients receiving OCS (≤250 mg OCS) and antibiotics for respiratory infections, and 3,674 patients receiving antibiotics for respiratory infections only in the year before baseline. This analysis was adjusted for OCS exposure and the variables included in the propensity score matching procedure (age, sex, tobacco exposure, exacerbations requiring hospitalization in the year before study entry, prescription of ICS and LAMA in the year before study entry, MRC, BMI, FEV1%, CCI score, and entry year).
In a second sensitivity analysis, we stratified patients in the OCS group by their cumulative OCS exposure during the 365 days preceding baseline: <500 mg, 500–1000 mg and >1000 mg. This categorization reflects clinically meaningful treatment durations with 37.5 mg OCS per day, corresponding to short (≤13 days), intermediate (14–26 days) and prolonged (≥27 days) courses.
All statistical analyses were performed using SAS 9.4 (Cary, NC, USA). A two-sided 95% confidence interval was considered statistically significant. Cumulative incidence plots were customized using the NewSurv macro [25].
Results were presented as hazard ratios (HR) with 95% confidence intervals (CI).
Risks were visualized by cumulative incidence plots with Fine-Gray analyses. HR profiles of variables in the sensitivity analysis were visualized by a forest plot.
Ethics
In Denmark, retrospective use of register data does not require ethical approval or patient consent.
Results
After applying eligibility criteria, a total of 7,435 patients were included. After propensity score matching, 2333 patients remained in each group. Please refer to Figure 1 for more details.
Figure 1.
STROBE flowchart of the study cohort flowchart of included patients, cohorts treated with antibiotics for respiratory infections and OCS.
Abbreviations: COPD: Chronic Obstructive Pulmonary Disease; DrCOPD: Danish Register of Chronic Obstructive Pulmonary Disease; OCS: Oral corticosteroid.
Baseline characteristics
Baseline characteristics were comparable between the two groups across most variables, including age, sex distribution, dyspnoea, FEV1% predicted, and BMI. A statistically significant difference was observed for tobacco exposure (chi-square test, p = 0.0005, Cramér’s V = 0.05). Significant differences were also noted in medical treatments: SAMA (p = 0.0005, Cramér’s V = 0.05), SABA (p<0.0001, Cramér’s V = 0.16), and insulin analogues (p = 0.008, Cramér’s V = 0.04). Additionally, atrial fibrillation was more prevalent in the OCS group (p = 0.01, Cramér’s V = 0.04), whereas hypertension (p = 0.03, Cramér’s V = 0.03) and osteoporosis (p = 0.0001, Cramér’s V = 0.05) were more common in the antibiotic-only group. Although these differences were statistically significant, most effect sizes, assessed via Cramér’s V, were small and therefore unlikely to be clinically relevant. A notable exception was SABA use (Cramér’s V = 0.16), indicating a weak association. After propensity score matching, the baseline characteristics of the two groups were highly similar, indicating good comparability. A detailed overview of baseline characteristics is presented in Table 1.
Table 1.
Baseline characteristics of 4666 propensity matched outpatients with COPD.
| Patients treated with OCS (N = 2,333) |
Patients treated with antibiotics for respiratory infections alone (N = 2,333) |
|
|---|---|---|
| Age years - mean ± SD | 70.6 ± 10.4 | 70.6 ± 10.5 |
| Female - n (%) | 1,260 (54.0) | 1,244 (53.3) |
| Tobacco exposure - n (%) | ||
| Never smoking and passive smoking | 68 (2.9) | 78 (3.3) |
| Previous smoking | 1,232 (52.8) | 1,284 (55.0) |
| Active smoking | 825 (35.4) | 722 (30.9) |
| Unknown tobacco exposure | 208 (8.9) | 249 (10.7) |
| MRC dyspnoea score - median (IQR) | 3 (2–4) | 3 (2–3) |
| BMI kg×m−2 - mean ± SD | 25.8 ± 5.9 | 25.7 ± 5.8 |
| FEV1 % pred - mean ± SD | 51.5 ± 16.5 | 51.4 ± 16.6 |
| CCI - median (IQR) | 3 (2–4) | 4 (2–4) |
| Comorbidities - n (%) | ||
| Hypertension | 593 (25.4) | 659 (28.2) |
| Hypercholesterolemia | 200 (8.6) | 204 (8.7) |
| Atrial fibrillation | 336 (14.4) | 400 (17.1) |
| Diabetes | 251 (10.8) | 260 (11.1) |
| Osteoporosis or osteopenia | 462 (19.8) | 377 (16.2) |
| Renal insufficiency | 73 (3.1) | 86 (3.7) |
| Liver insufficiency | 34 (1.5) | 40 (1.7) |
| Atopy or allergy | 55 (2.4) | 48 (2.1) |
| Depression | 67 (2.9) | 75 (3.2) |
| Exacerbations requiring admission within the year prior inclusion - n (%) | 814 (34.9) | 793 (34.0) |
| MACE requiring admission within five-year prior inclusion - n (%) | 238 (10.2) | 234 (10.0) |
| Medical treatment for respiratory disease within five years prior to inclusion - n (%) | ||
| Inhaled corticosteroids | 1,687 (72.3) | 1,686 (72.3) |
| Long-acting β2-agonist | 2,042 (87.5) | 2,002(85.8) |
| Long-acting muscarinic antagonists | 1,896 (81.3) | 1,907 (81.7) |
| Short-acting β2-agonists | 1,951 (83.6) | 1,634 (70.0) |
| Short-acting muscarinic antagonists | 219 (9.4) | 154 (6.6) |
| Medical treatment for cardiovascular disease within five years prior to inclusion - n (%) | ||
| Antihypertensives | 1,562 (67.0) | 1,602 (68.7) |
| Cholesterol-lowering medication | 880 (37.7) | 857 (36.7) |
| Antiarrhythmics | 154 (6.6) | 168 (7.2) |
| ADP-receptor inhibitors | 211 (9.0) | 246 (10.5) |
| Acetyl sialic acid | 575 (24.6) | 633 (27.1) |
| Nitrates | 122 (5.2) | 140 (6.0) |
| Medical treatment for diabetes within five years prior inclusion - n (%) | ||
| Insulin | 74 (3.2) | 109 (4.7) |
| Metformin | 245 (10.5) | 224 (9.6) |
| GLP1-analogs | 26 (1.1) | 39 (1.7) |
| SGLT2-inhibitors | 10 (0.4) | 13 (0.6) |
Abbreviations: IQR: Interquartile range; MRC: Medical Research Council; BMI: Body Mass Index; CCI: Charlson Comorbidity Index; MACE: Major adverse cardiovascular events as defined in this study.
Primary outcome
Treatment with OCS and antibiotics for respiratory infections did not show a higher risk of all MACE (123 patients), (HR 0.84, confidence interval (CI) 0.66 – 1.07, p = 0.17) within the 365 days following baseline, compared to treatment with antibiotics for respiratory infections alone (145 patients). Please see, Table 2 (unadjusted Cox analysis) and Figure 2 (cumulative incidence plot) among propensity score matched patients.
Table 2.
Primary and secondary outcome.
| Outcome | Propensity score matched patients |
|
|---|---|---|
| Cases exposed to OCS and antibiotics for respiratory infections (N = 2,333) |
Controls exposed to antibiotics for respiratory infections (N = 2,333) |
|
| Primary outcome | ||
| **All MACE within 365 days from entry date N (%) *HR p-value |
123 (5.27) 0.84 (0.66 – 1.07) 0.17 |
145 (6.22) Reference |
| Secondary outcomes | ||
| All-cause mortality N (%) *HR p-value |
224 (9.60) 0.99 (0.83 – 1.19) 0.93 |
226 (9.69) Reference |
| ***Severe MACE N (%) *HR p-value |
77 (3.30) 0.74 (0.55 – 0.99) 0.05 |
103 (4.42) Reference |
| Lethal cardiac events N (%) *HR p-value |
13 (0.56) 0.93 (0.44 – 1.98) 0.85 |
14 (0.60) Reference |
| Cardiac events requiring revascularization N (%) *HR p-value |
13 (0.56) 0.56 (0.29 – 1.11) 0.10 |
23 (1.00) Reference |
| Cardiac events requiring admission N (%) *HR p-value |
57 (2.44) 0.76 (0.54– 1.07) 0.11 |
75 (3.21) Reference |
| Cardiac events requiring prescriptions of ADP receptor inhibitors N (%) *HR p-value |
41 (1.76) 0.72 (0.48 – 1.07) 0.11 |
57 (2.44) Reference |
| Cardiac events requiring prescriptions of nitrates N (%) *HR p-value |
36 (1.5) 1.39 (0.84 – 2.30) 0.20 |
26 (1.11) Reference |
*Hazard ratios analyzed by unadjusted Cox regression analyses of propensity score matched patients with collection of OCS and antibiotics for respiratory infections or antibiotics only.
**All MACE: Lethal cardiovascular events (DG45, DI20, DI21, DI22, DI23 or DI24), cardiovascular events requiring revascularization (surgery diagnosed as KFNA00. KFNA20, KFNC10-30, KFNE00, KFNG02A, KFNG05 or KFNG05A) cardiovascular events requiring admission (hospitalization diagnosed as DG45, DI20, DI21, DI22, DI23 or DI24) or cardiovascular events requiring prescriptions of ADP receptor inhibitors or nitrates.
***Severe MACE: Lethal cardiovascular events (DG45, DI20, DI21, DI22, DI23 or DI24), cardiovascular events requiring revascularization (surgery diagnosed as KFNA00. KFNA20, KFNC10-30, KFNE00, KFNG02A, KFNG05 or KFNG05A) cardiovascular events requiring admission (hospitalization diagnosed as DG45, DI20, DI21, DI22, DI23 or DI24).
Figure 2.
Cumulative incidence plot of primary outcome: MACE within 365 days following baseline.
This plot illustrates the cumulative incidence rates of MACE between the two treatment groups over the 365 days period.
Secondary outcome analysis
Treatment with oral corticosteroids (OCS) was not associated with a significant difference in all-cause mortality compared to antibiotics for respiratory infections alone (HR 0.99, CI 0.83–1.19, p = 0.93), as shown in Table 2 and Figure S1A in the Supplementary Material.
There were no significant differences in the risk of severe MACE (HR 0.74, CI 0.44–1.98, p = 0.05) (Figure S1B), lethal cardiac events (HR 0.93, CI 0.44–1.98, p = 0.85) (Figure S1C), cardiac events requiring revascularization (HR 0.56, CI 0.29–1.11, p = 0.10) (Figure S1D), cardiac events requiring hospital admission (HR 0.76, CI 0.54–1.07, p = 0.11) (Figure S1E), cardiac events requiring prescriptions of nitrates (HR 1.39, CI 0.84–2.30, p = 0.20) (Figure S1F), or ADP receptor inhibitors (HR 0.72, CI 0.48–1.07, p = 0.11) (Figure S1G) between the OCS and antibiotic-only groups.
Sensitivity analysis: multivariable Cox regression and forest plot in all patients before propensity score matching
Furthermore, an adjusted Cox proportional hazard model performed in the unmatched population, containing 7435 patients, showed the same tendency (HR of 0.83, CI 0.65–1.05 p = 0.13) as observed in the primary outcome analysis.
In this model, increasing age was not significantly associated with MACE (HR = 0.99, CI 0.97–1.00, p = 0.09). Female sex was associated with a lower risk of MACE compared with male sex (HR = 0.68, CI 0.53–0.87, p < 0.01). Decreasing BMI showed no significant effect (HR = 0.98, CI 0.96–1.01, p = 0.13). Increasing severity of dyspnoea (by MRC score) was associated with a modestly higher risk of MACE (HR = 1.15, CI 1.01–1.31, p = 0.03), whereas increasing level of FEV1% GOLD stage was not (HR = 1.01, CI 1.00–1.02, p = 0.11). Smoking history (HR = 1.09, CI 0.93–1.28, p = 0.30) and calendar year of study entry (HR = 1.04, CI 0.90–1.21, p = 0.60) were not associated with MACE risk. In contrast, a higher CCI score (HR = 1.44, CI 1.31–1.58, p < 0.0001), was strongly associated with risk of MACE, and the same was true for a history of COPD exacerbation requiring hospitalization (HR = 1.50, CI 1.17–1.91, p < 0.01). Prescription of ICS or LAMA in the year preceding baseline was not significantly associated with a risk of MACE (ICS: HR = 0.97, CI 0.74–1.25, p = 0.79; LAMA: HR = 0.96, CI 0.71–1.28, p = 0.77).
The model was adjusted for OCS treatment and all the variables included in the propensity score matching. Among these patients, 3761 received OCS and 3674 did not. After adjustment, OCS treatment was not significantly associated with the risk of all MACE within 365 days of baseline (HR = 0.83, CI 0.65–1.05, p = 0.13).
Detailed results are presented in Table 3 and Figure S2 (forest plot) in the Supplementary Material.
Table 3.
Adjusted Cox analysis of the primary outcome in 7435 unmatched patients with COPD exposed to OCS compared to antibiotics for respiratory infections.
| Variable | Risk of primary outcome: Any MACE HR (CI) p-value |
|---|---|
| OCS treatment | 0.83 (0.65–1.05) p = 0.13 |
| Age group | 0.99 (0.97–1.00) p = 0.09 |
| Sex female vs. male | 0.68 (0.53–0.87) p <0.01 |
| Entry year | 1.04 (0.90–1.21) p = 0.60 |
| Tobacco exposure ‘smoking history’ vs. ‘never smoking’ |
1.09 (0.93–1.28) p = 0.30 |
| Exacerbation for COPD requiring hospitalization in the year before study entry | 1.50 (1.17–1.91) p <0.01 |
| Prescription of ICS in the year before study entry | 0. 97 (0.74–1.25) p = 0.79 |
| Prescription of LAMA in the year before study entry | 0.96 (0.71–1.28) p = 0.77 |
| Medical Research Council (MRC) Dyspnoea Scale | 1.15 (1.01–1.31) p = 0.03 |
| Decreasing body mass index (BMI) | 0.98 (0.96–1.01) p = 0.13 |
| FEV1% GOLD stage | 1.01 (0.10–1.02) p = 0.11 |
| Charlson Comorbidity Index (CCI) score | 1.44 (1.31–1.58) p = <0.0001 |
Hazard ratios were calculated by adjusted Cox analysis for collection of OCS 25 or 37 mg as well as the variables later included in the propensity score matching.
Age group (≤70 years, >70 and ≤80 years, >80 and ≤90 or >90 years), sex (female vs. male), entry year by calendar year, tobacco exposure (active smoking history vs. never smoker), exacerbations requiring hospitalization in the year before study entry (admission vs. no admission), prescription of ICS and LAMA in the year before study entry (prescription vs. no prescription), medical Research Council (MRC) Dyspnoea Scale (1, 2, 3, 4 or 5), decreasing body mass index (BMI) (≥20 kg/m2, <20 kg/m2 and ≥15 kg/m2 or <15 kg/m2), FEV1% GOLD stage 1–4(≥ 80%, <80% and ≥50%, <50% and ≥30% or <30%), Charlson Comorbidity Index (CCI) score (as a continuous variable).
Sensitivity analysis: cumulative OCS exposure in the OCS group
In a sensitivity analysis stratified by cumulative OCS exposure during the year preceding baseline (<500 mg, 500–1000 mg and >1000 mg), no significant difference in the risk of all MACE was observed across the three groups (Fine-Gray test p = 0.33). Compared with the lowest cumulative exposure group, the hazard ratios for all MACE were 1.09 (CI 0.72–1.63, p = 0.69) for the intermediate exposure group and 1.41 (CI 0.89–2.25, p = 0.14) for the highest exposure group, please see Figure S3 in the Supplementary Material.
Discussion
This retrospective cohort study included 4,666 propensity score matched patients with COPD either treated with OCS and antibiotics for respiratory infections or antibiotics for respiratory infections alone in the 365 days prior to baseline. The study evaluated the risk of MACE within 365 days from baseline.
In the primary analysis, no significant difference in the risk of all MACE was observed between patients treated with OCS and antibiotics for respiratory infections compared to those treated with antibiotics for respiratory infections alone.
Similarly, the secondary outcomes showed no statistically significant association between OCS treatment and all-cause mortality, severe MACE, lethal cardiac events, cardiac events necessitating revascularization, or hospital admission. Additionally, there was no observed association between OCS treatment and the administration of ADP-receptor inhibitors or nitrates.
This study specifically examined the short-term cardiovascular risks of OCS use in the context of AECOPD. While the findings suggest that short-term OCS treatment does not increase the risk of MACE, the potential cardiovascular effects of repeated or cumulative OCS exposure warranted further investigation. To address this concern, we performed an additional sensitivity analysis stratifying patients by cumulative OCS dose during the year prior to baseline. This analysis did not reveal a significant dose-response relationship, suggesting that even repeated short-term OCS treatments within one year were not associated with an increased short-term risk of MACE. However, this dose-stratified analysis should be interpreted with caution, as the exposure groups were relatively small, limiting the statistical power to detect dose-response differences. Furthermore, we acknowledge that longer follow-up periods would be required to fully access potential long-term effects.
Previous studies have raised concerns regarding the cardiac safety of OCS treatment for AECOPD [11–14,26]. However, our findings stand in contrast to these concerns. Notably, two of the earlier studies identified increased cardiovascular risk primarily in relation to long-term cumulative exposure to OCS [11,13], while three others were nested case-control studies that indicated a higher risk with short-term treatment, though not all were limited to patients with COPD [12,14,26]. According to our results, short-term OCS treatment in patients with COPD appears safe and does not increase the risk of MACE. These results provide reassurance for the current clinical practice, indicating that cardiac monitoring may not be necessary. Sensitivity analyses further supported the primary finding that OCS is not associated with an increased risk of MACE.
Beyond potential cardiovascular adverse effects to OCS, it has been associated with a broader spectrum of complications, including hypertension, increased risk of infections, osteoporosis, fractures, obesity and type 2 diabetes [4,5]. These risks are important considerations when examining overall corticosteroid safety. Emerging evidence suggests that blood eosinophil counts can help identify patients with COPD that are more likely to benefit from corticosteroid therapy [27,28]. Our findings support the safe use of short-term OCS, but more individualized treatment based on eosinophil levels could further improve therapeutic effect and reduce unnecessary exposure to corticosteroids.
Strengths and limitations
This study design has several strengths. It is based on a comprehensive national register, including all patients with COPD seen in outpatient clinics across Denmark. This register includes over 100,000 patients, providing a large sample size, that enhances the generalizability of the results. Furthermore, the cohorts were propensity score matched on eleven variables that could potentially influence the results. This allowed us to adjust for differences in baseline characteristics, thereby providing a more robust basis for comparing the two groups. Another strength of this study is the selection of comparison groups, ensuring that the case group also received antibiotics for respiratory infections, helped to equalize the risk of infection-related MACE between the groups and thus reduce the risk of indication bias [29,30]. Despite these strengths, the study is not without limitations. First, although we performed an extensive propensity score matching, some significant differences in the baseline characteristics remained between the two groups, particularly in terms of medication prescriptions. The OCS group had a higher overall use of certain respiratory medications: SABA and SAMA, that were not included in the matching process. Additionally, the OCS group had a higher prevalence of atrial fibrillation, and there were differences in tobacco use as well as higher prevalence of hypertension and osteoporosis, suggesting the possibility of residual confounding. While these imbalances may complicate the comparisons between the groups, further statistical analyses indicated that the differences were small and unlikely to be clinically relevant. Secondly, some outcomes had relatively few events, resulting in wide confidence intervals and limited precision. This raises the possibility of type II error, meaning that small to moderate associations between OCS and cardiovascular outcomes may have gone undetected. Larger studies are needed to confirm these findings. Lastly, there is a potential bias in that patients who received both OCS and antibiotics for respiratory infections may have been more severely ill than those treated with antibiotics alone, which could have influenced the outcomes. Although this concern cannot be entirely ruled out, the results for all-cause mortality support that the groups were generally comparable in terms of illness severity.
Conclusion
In conclusion, the study found no increased risk of major adverse cardiovascular events (MACE) within 365 days following short-term treatment with oral corticosteroid (OCS) in patients with chronic obstructive pulmonary disease (COPD) during the year prior to baseline. A sensitivity analysis showed no cumulative dose-response relationship with the one-year risk of MACE. While these findings provide some reassurance regarding short-term OCS use for acute exacerbations, the observational design and non-significant trends warrant caution, and larger prospective studies are needed to confirm these results.
Supplementary Material
Funding Statement
The work was supported by The Novo Nordisk Foundation (grant number: NNF20OC0060657).
Disclosure statement
No potential conflict of interest was reported by the author(s).
Supplementary material
Supplemental data for this article can be accessed online at https://doi.org/10.1080/20018525.2026.2616846
References
- [1].Global Initiative for Chronic Obstructive Lung Disease . [homepage on the Internet]. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease [updated], 2024.
- [2].The appropriate use of rescue packs | Primary Care Respiratory Society [Internet]. [cited 2024 Feb 17]. Available from:https://www.pcrs-uk.org/resource/current/appropriate-use-rescue-packs
- [3].Aaron SD, Vandemheen KL, Hebert P, et al. Outpatient oral prednisone after emergency treatment of chronic obstructive pulmonary disease. N Engl J Med. 2003. Jun 26;348(26):2618–12. doi: 10.1056/NEJMoa023161 [DOI] [PubMed] [Google Scholar]
- [4].Walters JAE, Tan DJ, White CJ, et al. Systemic corticosteroids for acute exacerbations of chronic obstructive pulmonary disease. Cochrane Database Syst Rev. 2014. Sep;1(9):CD001288 doi: 10.1002/14651858.CD001288.pub4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [5].Waljee AK, Rogers MAM, Lin P, et al. Short term use of oral corticosteroids and related harms among adults in the United States: population based cohort study. BMJ. 2017. Apr 12;357:j1415. doi: 10.1136/BMJ.j1415 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [6].Sullivan PW, Ghushchyan VH, Globe G, et al. Oral corticosteroid exposure and adverse effects in asthmatic patients. J Allergy Clin Immunol. 2018. Jan;141(1):110–116.e7. doi: 10.1016/j.jaci.2017.04.009 [DOI] [PubMed] [Google Scholar]
- [7].Bloechliger M, Reinau D, Spoendlin J, et al. Adverse events profile of oral corticosteroids among asthma patients in the UK: cohort study with a nested case-control analysis. Respir Res. 2018. Apr 27;19(1):75. doi: 10.1186/s12931-018-0742-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- [8].Sivapalan P, Ingebrigtsen TS, Rasmussen DB, et al. Copd exacerbations: the impact of long versus short courses of oral corticosteroids on mortality and pneumonia: nationwide data on 67 000 patients with copd followed for 12 months. BMJ Open Respir Res. 2019. Mar 30;6(1):e000407. doi: 10.1136/bmjresp-2019-000407 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].Leuppi JD, Schuetz P, Bingisser R, et al. Short-term vs conventional glucocorticoid therapy in acute exacerbations of chronic obstructive pulmonary disease: the REDUCE randomized clinical trial. JAMA. 2013. Jun 5;309(21):2223–2231. doi: 10.1001/jama.2013.5023 [DOI] [PubMed] [Google Scholar]
- [10].Sivapalan P, Rutishauser J, Ulrik CS, et al. Effect of different corticosteroid regimes for hospitalised patients with exacerbated COPD: pooled analysis of individual participant data from the REDUCE and CORTICO-COP trials. Respir Res. 2021;22(1):155. doi: 10.1186/s12931-021-01745-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [11].Bazell C, Pollack M, Comellas AP, et al. A 4-year retrospective claims analysis of oral corticosteroid use and health conditions in newly diagnosed Medicare FFS patients with COPD. Int J Chron Obstruct Pulmon Dis. 2022. Oct;17:2635–2652. doi: 10.2147/COPD.S373590 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [12].Varas-Lorenzo C, Rodriguez LAG, Maguire A, et al. Use of oral corticosteroids and the risk of acute myocardial infarction. Atherosclerosis. 2007. Jun 1;192(2):376–383. doi: 10.1016/j.atherosclerosis.2006.05.019 [DOI] [PubMed] [Google Scholar]
- [13].Tse G, Emmanuel B, Ariti C, et al. A long-term study of adverse outcomes associated with oral corticosteroid use in COPD. Int J Chron Obstruct Pulmon Dis. 2023. Nov 15;18:2565–2580. doi: 10.2147/COPD.S433326 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Huiart L, Ernst P, Ranouil X, et al. Oral corticosteroid use and the risk of acute myocardial infarction in chronic obstructive pulmonary disease. Can Respir J. 2006. Apr;13(3):134–138. doi: 10.1155/2006/935718 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [15].Jain H, Marsool MDM, Abbasi HQ, et al. Association between fluoroquinolones and major adverse cardiovascular events: a systematic review. Cardiol Rev. 2024 Apr. doi: 10.1097/CRD.0000000000000710 [DOI] [PubMed] [Google Scholar]
- [16].Jespersen CM, Als-Nielsen B, Damgaard M, et al. Randomised placebo controlled multicentre trial to assess short term clarithromycin for patients with stable coronary heart disease: cLARICOR trial. BMJ. 2006. Jan 7;332(7532):22–27. doi: 10.1136/BMJ.38666.653600.55 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [17].Lange P, Tøttenborg S, Sorknæs A, et al. Danish register of chronic obstructive pulmonary disease. Clin Epidemiol. 2016. Oct;8:673–678. doi: 10.2147/CLEP.S99489 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [18].Schmidt M, Pedersen L, Sørensen HT.. The Danish civil registration system as a tool in epidemiology. Eur J Epidemiol. 2014. Aug;29(8):541–549. doi: 10.1007/s10654-014-9930-3 [DOI] [PubMed] [Google Scholar]
- [19].Pottegård A, Schmidt SAJ, Wallach-Kildemoes H, et al. Data resource profile: the Danish National Prescription Registry. Int J Epidemiol. 2017. Jun;46(3):798–798f. doi: 10.1093/ije/dyw213. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [20].Schmidt M, Schmidt SAJ, Sandegaard JL, et al. The Danish National Patient Registry: a review of content, data quality, and research potential. Clin Epidemiol. 2015. Nov 17;7:449–490. doi: 10.2147/CLEP.S91125 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [21].Dødsårsagsregisteret (DAR) - Sundhedsdatastyrelsen [Internet]. [cited 2024 Jun 13]. Available from: https://sundhedsdatastyrelsen.dk/dar
- [22].Bosco E, Hsueh L, McConeghy KW, et al. Major adverse cardiovascular event definitions used in observational analysis of administrative databases: a systematic review. BMC Med Res Methodol. 2021. Nov 6;21(1):241. doi: 10.1186/s12874-021-01440-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [23].Mayo . Department of quantitative health sciences Mayo Clinic research. 2022. [cited 2022 January 21]. Available from: http://bioinformaticstools.mayo.edu/research/gmatch/
- [24].Kim HY. Statistical notes for clinical researchers: chi-squared test and Fisher’s exact test. Restor Dent Endod. 2017. Mar;42(2):152–155. doi: 10.5395/rde.2017.42.2.152 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [25].Kaplan-meier survival plotting macro %NEWSURV [Internet]. 2018. [cited 2024 Jun 8]. Available from: https://communities.sas.com/t5/SAS-Communities-Library/Kaplan-Meier-Survival-Plotting-Macro-NEWSURV/ta-p/479747
- [26].Souverein PC, Berard A, Van Staa TP, et al. Use of oral glucocorticoids and risk of cardiovascular and cerebrovascular disease in a population-based case-control study. Heart Br Card Soc. 2004. Aug;90(8):859–865. doi: 10.1136/hrt.2003.020180 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [27].Sivapalan P, Lapperre TS, Janner J, et al. Eosinophil-guided corticosteroid therapy in patients admitted to hospital with COPD exacerbation (CORTICO-COP): a multicentre, randomised, controlled, open-label, non-inferiority trial. Lancet Respir Med. 2019;7(8):699–709. doi: 10.1016/S2213-2600(19)30176-6 [DOI] [PubMed] [Google Scholar]
- [28].Sivapalan P, Bikov A, Jensen J-U. Using blood eosinophil count as a biomarker to guide corticosteroid treatment for chronic obstructive pulmonary disease. Diagn (Basel). 2021;11(2):236. doi: 10.3390/diagnostics11020236 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [29].Smeeth L, Thomas Sara L, Hall Andrew J, et al. Risk of myocardial infarction and stroke after acute infection or vaccination. N Engl J Med. 2004;351(25):2611–2618. doi: 10.1056/NEJMoa041747 [DOI] [PubMed] [Google Scholar]
- [30].Meier CR, Jick SS, Derby LE, et al. Acute respiratory-tract infections and risk of first-time acute myocardial infarction. Lancet. 1998. May 16;351(9114):1467–1471. doi: 10.1016/S0140-6736(97)11084-4 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.