https://orcid.org/0009-0005-8049-6871
Abstract
While prognosis of acute myocarditis with uncomplicated presentation is perceived as benign, data on long-term outcomes are scarce. We evaluated rates of myocarditis-associated cardiovascular events after a first-time hospitalization with uncomplicated acute myocarditis in patients without known heart disease.
In this retrospective nationwide population-based cohort study from 2013 to 2020, hospitalized patients with uncomplicated acute myocarditis but without known heart disease were 1:1 propensity score-matched with surgical controls hospitalized for laparoscopic appendectomy. As assessed in time-to-event analyses, the primary outcome was a composite of rehospitalization for myocarditis, pericardial disease, heart failure and its complications, arrhythmias, implantation of cardiac devices, and heart transplant. After matching, we identified 1439 patients with uncomplicated acute myocarditis (median age of 35 years, 74.0% male) and 1439 surgical controls (median age of 36 years, 74.4% male). Over a median follow-up of 39 months, compared with surgical controls, the hazard ratio for the primary composite outcome was 42.3 [95% confidence interval (CI) 17.4–102.8], corresponding to an incidence rate of 43.7 vs. 0.9 per 1000 patient-years (py) and an incidence rate difference of 42.7 (95% CI 36.7–48.8) per 1000 py.
Patients hospitalized with uncomplicated acute myocarditis and no known prior heart disease were associated with substantial risk for cardiovascular events over a follow-up of up to 8 years. This calls for a more efficient therapeutic management of this population of patients.
Introduction
Acute myocarditis represents a challenge for many clinicians due to the large spectrum of aetiologies and broad range of clinical courses.1 Recently, acute myocarditis has been brought into focus due to reports of COVID-19-associated myocarditis and COVID-19 mRNA vaccine-associated myocarditis, although both conditions are reported to be rare.2,3
While clinical presentation of acute myocarditis ranges from asymptomatic to cardiogenic shock,4 recent evidence shows that it predicts outcomes in patients with acute myocarditis; mainly in those presenting with reduced left ventricular ejection fraction (LVEF), heart failure, cardiogenic shock, advanced atrioventricular block, or sustained ventricular arrhythmias being at higher risk for heart transplantation or death.5–8 Thus, a recent expert consensus suggested the distinction of acute myocarditis into complicated and uncomplicated cases, with the goal of identifying high-risk patients based on initial clinical presentation.1 According to this definition, complicated acute myocarditis is characterized by presentation with one or more of the following features: left ventricular dysfunction (i.e. LVEF <50% on the first echocardiogram), sustained ventricular arrhythmias, advanced heart block, heart failure, low cardiac output syndrome, and cardiogenic shock.1 In uncomplicated acute myocarditis, however, these specified characteristics are missing.1 While prognosis of uncomplicated acute myocarditis is perceived as benign, data on long-term outcomes is scarce, even though this subpopulation is reported to account for the majority of acute myocarditis patients with 73.4%.8 The Multicenter Lombardy Registry study8 reported a higher risk of cardiac events in short- and long-term follow-up in patients with complicated acute myocarditis, while patients with uncomplicated acute myocarditis had mostly benign outcomes.8 However, since focus is frequently set on hard endpoints such as cardiac death and heart transplant,8 the overall cardiovascular burden of disease of uncomplicated acute myocarditis may be underestimated. We therefore sought to assess rates of cardiovascular events after first-time hospitalization with uncomplicated acute myocarditis in patients without known heart disease in a nationwide population-based cohort study.
Methods
Study design and data source
This retrospective population-based analysis was conducted using a nationwide cohort of hospitalizations in Switzerland between 2012 and 2020. The Swiss Federal Statistical Office (Neuchâtel, Switzerland) provided the hospitalization data based on nationwide compulsory full census of Swiss hospitals. This dataset comprises all Swiss inpatient discharge records from acute care-, general-, and special clinics for both paediatric and adult patients. Individual-level data on patient demographics, healthcare utilization, hospital typology, medical diagnoses, diagnostic tests, clinical procedures, and in-hospital patient outcomes were provided. Patient confidentiality is ensured by a multi-step anonymization procedure and the usage of an unique patient identifier to ascertain rehospitalizations. Medical diagnoses were coded using the International Classification of Diseases version 10, German Modification (ICD-10-GM) codes. Diagnostic tests and clinical procedures were coded using the Swiss Operations Classification (Schweizerische Operationsklassifikation, CHOP), versions 2012–20. Since this study exclusively used anonymized data, the institutional review board of Northwestern and Central Switzerland (EKNZ) waived the need for an ethical authorization (EKNZ Project-ID: Req-2021-01397). This study adheres to the ‘Strengthening The Reporting of OBservational Studies in Epidemiology (STROBE)’ statement.9
Patient population and eligibility criteria
As patients without know heart disease and an event of acute myocarditis are usually younger, we defined a control group with similar demographic patient characteristics consisting of patients with a first-time hospitalization for laparoscopic appendectomy.
Eligible patients had a first-time (index) acute care hospitalization for acute myocarditis by usage of the following ICD-10 codes at primary or secondary position in the discharge report: I40 (acute myocarditis), I41 (myocarditis in diseases classified elsewhere), I51.4 (myocarditis, unspecified), B33.2 (viral carditis), I01.2 (acute rheumatic myocarditis), and I09.0 (rheumatic myocarditis). The selection of ICD-10 codes was in line with previous studies.3,8,10–12 Hospitalizations with laparoscopic appendectomy were identified using the CHOP code 47.01. Patients who had both a first-time acute myocarditis hospitalization and a first-time laparoscopic appendectomy hospitalization were allocated to the acute myocarditis group.
To reduce the potential for unmeasured confounding due higher risk of recurrence, we excluded patients with known prior heart disease. Patients with any cardiovascular event 1 year prior to the index hospitalization (washout period) were excluded (see Supplementary material online, eTable 1, Supplementary material online, eFigure 1). To comply with an uncomplicated presentation of an acute myocarditis, patients with any cardiac event during index hospitalization were also excluded. However, for diagnostic and therapeutic purposes, the following interventions during index acute myocarditis hospitalization were allowed: diagnostic coronary angiography and interventional endomyocardial biopsy (see Supplementary material online, eTable 2). Moreover, patients who died during index hospitalization were excluded. There was no age restriction in terms of eligibility.
Outcomes
The primary outcome was a composite of rehospitalization for myocarditis or myocarditis-associated cardiovascular events, which included: myocarditis (primary position), hospitalization for heart failure (primary position), hospitalization for dilated cardiomyopathy (primary position), hospitalization with ventricular tachycardia (any position), hospitalization with ventricular fibrillation/flutter (any position), hospitalization with cardiac arrest (any position), hospitalization with cardiogenic shock (any position), hospitalization for supraventricular tachycardia (primary position), hospitalization for atrial fibrillation or flutter (primary position), hospitalization for second or third degree atrioventricular (AV) block (primary position), hospitalization for sick sinus syndrome (primary position), hospitalization for a pericardial disease (e.g. acute pericarditis) (primary position), hospitalization with a cardiac device implantation (any position), and hospitalization with heart transplant (any position) (see Supplementary material online, eTable 3).
Secondary outcomes included all individual components of the primary outcome. Individual components of the primary outcome that were included at primary position only, were also assessed at any position as further secondary outcome (see Supplementary material online, eTable 4). Additional secondary outcomes included hospitalization with atherosclerotic cardiovascular disease (ASCVD, any position) and all-cause rehospitalization (see Supplementary material online, eTable 3).
Statistical analysis
Descriptive statistics were calculated for patient demographics and comorbidities. To account for the heterogeneity of this patient cohort, we performed a propensity score matched analysis including all relevant and documented demographics and comorbidities, and, therefore, achieved highly balanced patient characteristics when compared with pairwise matched controls. Eligible patients with an index hospitalization of uncomplicated acute myocarditis were 1:1 propensity score matched to eligible patients who underwent laparoscopic appendectomy. The estimated propensity score was used to match patients with uncomplicated acute myocarditis to nearest neighbour surgical controls within a calliper width of 0.01 on the propensity scale. We assessed covariate balance before and after propensity score matching by computing absolute standardized differences. An absolute standardized difference of <0.10 was considered as an adequate balance between groups.13 For each outcome, incidence rates (IR) and incidence rate differences (IRD) per 1000 person-years (py), as well as hazard ratios (HR)—using Cox proportional hazard models—were calculated, with corresponding 95% confidence intervals (CI), respectively. Kaplan–Meier failure estimates were used to illustrate differences in time to outcome, log-rank tests to assess potential differences between the curves.
To account for potential more previous hospitalizations with a diagnosis of heart disease and to further increase internal validity of our approach, we performed a sensitivity analysis by extending the washout period up to 2 years (see Supplementary material online, eFigure 2). In a first subgroup analysis, we stratified our results by those patients who had immediate diagnostic work-up (endomyocardial biopsy, diagnostic coronary angiography, cardiac magnetic resonance imaging, cardiac computed tomography, myocardial perfusion scintigraphy, cardiac positron emission tomography, transthoracic echocardiography, and transoesophageal echocardiography) during the index hospitalization and those without. The propensity score matching was done separately within each subgroup. We evaluated for heterogeneity across the subgroups using the Wald test for homogeneity. Employing the same method, we additionally stratified patients according to age in a second subgroup analysis, comparing outcomes of patients with and without myocarditis aged 40 years and below to those above 40 years of age. A separate propensity-score matching was conducted within each age subgroup.
In the unmatched cohort, we performed a risk factor analysis for the primary composite outcome to assess predictors for the occurrence of the events. We used a multivariate logistic regression analysis and calculated the area under the receiver operating characteristic curve (ROC-AUC) as measure of discrimination.
All tests were two-sided, have not been adjusted for multiple testing, and P < 0.05 was considered statistically significant. All statistical analyses were performed using STATA, version 15.1 (StataCorp LLC).
Results
Study population
The full dataset included 5 783 557 identifiable individual patients hospitalized in Switzerland between 1 January 2012 and 31 December 2020 (see Supplementary material online, eFigure 3). After application of eligibility criteria and pairwise propensity score matching of patients with uncomplicated acute myocarditis with surgical controls, the study population included 1439 patients who had a first-time hospitalization with uncomplicated acute myocarditis (median age of 35 years, 74.0% male) and 1439 surgical controls (median age of 36 years, 74.4% male). Baseline demographic variables and comorbidities included in the propensity score were well balanced after matching (Table 1). Hypertension and dyslipidaemia were among the most common comorbidities after matching, with a generally low overall comorbidity burden, mirroring the low-risk setting.
Baseline characteristics before and after propensity score matching
| Before matching | After matching | |||||
|---|---|---|---|---|---|---|
| Patients with myocarditis (n = 1443) | Patients without myocarditis (n = 86 589) | Std Diff | Patients with myocarditis (n = 1439) | Patients without myocarditis (n = 1439) | Std Diff | |
| Demographics | ||||||
| Female sex, n (%)a | 375 (26.0) | 42 409 (49.0) | 0.49 | 374 (26.0) | 368 (25.6) | 0.01 |
| Male sex, n (%)a | 1068 (74.0) | 44 180 (51.0) | 0.49 | 1065 (74.0) | 1071 (74.4) | 0.01 |
| Swiss nationality, n (%)a | 1014 (70.3) | 64 259 (74.2) | 0.09 | 1011 (70.3) | 1036 (72.0) | 0.04 |
| Non-Swiss nationality, n (%)a | 429 (29.7) | 22 330 (25.8) | 0.09 | 428 (29.7) | 403 (28.0) | 0.04 |
| Age (years), median (IQR)a | 35 (24–51) | 30 (19–47) | 0.30 | 35 (24–51) | 36 (24–53) | 0.03 |
| Age 0–17 yearsa | 87 (6.0) | 18 936 (21.9) | 0.47 | 87 (6.0) | 83 (5.8) | 0.02 |
| Age 18–40 yearsa | 779 (54.0) | 39 170 (45.2) | 779 (54.1) | 770 (53.5) | ||
| Age >40 yearsa | 577 (40.0) | 28 483 (32.9) | 573 (39.8) | 586 (40.7) | ||
| Year of index hospitalizationa | ||||||
| 2013 | 145 (10.0) | 9995 (11.5) | 0.10 | 145 (10.1) | 145 (10.1) | 0.02 |
| 2014 | 165 (11.4) | 10 454 (12.1) | 165 (11.5) | 165 (11.5) | ||
| 2015 | 161 (11.2) | 10 746 (12.4) | 161 (11.2) | 158 (11.0) | ||
| 2016 | 199 (13.8) | 11 010 (12.7) | 198 (13.8) | 203 (14.1) | ||
| 2017 | 189 (13.1) | 10 947 (12.6) | 188 (13.1) | 189 (13.1) | ||
| 2018 | 182 (12.6) | 11 151 (12.9) | 182 (12.6) | 177 (12.3) | ||
| 2019 | 230 (15.9) | 11 441 (13.2) | 230 (16.0) | 238 (16.5) | ||
| 2020 | 172 (11.9) | 10 845 (12.5) | 170 (11.8) | 164 (11.4) | ||
| Comorbidities | ||||||
| Elixhauser comorbidity score, mean (SD)a | 1 (1) | 0 (1) | 0.44 | 1 (1) | 1 (1) | 0.02 |
| Diabetes mellitus type 1, n (%)a | 6 (0.4) | 180 (0.2) | 0.04 | 6 (0.4) | 4 (0.3) | 0.02 |
| Diabetes mellitus type 2, n (%)a | 36 (2.5) | 915 (1.1) | 0.11 | 35 (2.4) | 27 (1.9) | 0.04 |
| Obesity, n (%)a | 18 (1.2) | 716 (0.8) | 0.04 | 18 (1.3) | 16 (1.1) | 0.01 |
| Hypertension, n (%)a | 153 (10.6) | 4828 (5.6) | 0.19 | 153 (10.6) | 163 (11.3) | 0.02 |
| Cerebrovascular disease, n (%)a | 15 (1.0) | 82 (0.1) | 0.13 | 12 (0.8) | 10 (0.7) | 0.02 |
| Peripheral arterial disease, n (%)a | 4 (0.3) | 68 (0.1) | 0.05 | 4 (0.3) | 2 (0.1) | 0.03 |
| CKD, n (%)a | 33 (2.3) | 415 (0.5) | 0.16 | 33 (2.3) | 23 (1.6) | 0.05 |
| Hepatopathy, n (%)a | 32 (2.2) | 284 (0.3) | 0.17 | 31 (2.2) | 16 (1.1) | 0.08 |
| COPD, n (%)a | 14 (1.0) | 213 (0.2) | 0.09 | 14 (1.0) | 12 (0.8) | 0.01 |
| Sleep apnoea, n (%)a | 13 (0.9) | 286 (0.3) | 0.07 | 13 (0.9) | 9 (0.6) | 0.03 |
| Dyslipidaemia, n (%)a | 112 (7.8) | 1093 (1.3) | 0.32 | 110 (7.6) | 122 (8.5) | 0.03 |
| Malignant neoplasm, n (%)a | 39 (2.7) | 1207 (1.4) | 0.09 | 39 (2.7) | 24 (1.7) | 0.07 |
| Length of index hospitalization | ||||||
| Length of hospital stay (days), median (IQR) | 3 (2–5) | 2 (2–4) | 0.24 | 3 (2–5) | 2 (2–4) | 0.15 |
| In-hospital diagnostics | ||||||
| Endomyocardial biopsy, n (%) | 2 (0.1) | 0 (0.0) | 0.05 | 2 (0.1) | 0 (0.0) | 0.05 |
| Diagnostic coronary angiography, n (%) | 443 (30.7) | 0 (0.0) | 0.94 | 443 (30.8) | 0 (0.0) | 0.94 |
| Cardiac MRI, n (%) | 360 (24.9) | 0 (0.0) | 0.82 | 360 (25.0) | 0 (0.0) | 0.82 |
| Cardiac CT, n (%) | 45 (3.1) | 2 (0.002) | 0.25 | 45 (3.1) | 0 (0.0) | 0.25 |
| Myocardial perfusion scintigraphy, n (%) | 4 (0.3) | 0 (0.0) | 0.07 | 4 (0.3) | 0 (0.0) | 0.07 |
| Cardiac PET, n (%) | 6 (0.4) | 0 (0.0) | 0.09 | 6 (0.4) | 0 (0.0) | 0.09 |
| TTE, n (%) | 300 (20.8) | 34 (0.04) | 0.72 | 299 (20.8) | 0 (0.0) | 0.72 |
| TEE, n (%) | 12 (0.8) | 8 (0.009) | 0.13 | 11 (0.8) | 0 (0.0) | 0.12 |
| Only TTE, n (%)b | 136 (9.4) | 0 (0.0) | n/a | 135 (9.4) | 0 (0.0) | n/a |
| Only TTE or TEE, n (%)b | 142 (9.8) | 0 (0.0) | n/a | 140 (9.7) | 0 (0.0) | n/a |
| None of aforementioned cardiac diagnostics, n (%)b | 591 (41.0) | 0 (0.0) | n/a | 589 (40.9) | 0 (0.0) | n/a |
| Any of aforementioned cardiac diagnostics, n (%)b | 852 (59.0) | 0 (0.0) | n/a | 850 (59.1) | 0 (0.0) | n/a |
| Before matching | After matching | |||||
|---|---|---|---|---|---|---|
| Patients with myocarditis (n = 1443) | Patients without myocarditis (n = 86 589) | Std Diff | Patients with myocarditis (n = 1439) | Patients without myocarditis (n = 1439) | Std Diff | |
| Demographics | ||||||
| Female sex, n (%)a | 375 (26.0) | 42 409 (49.0) | 0.49 | 374 (26.0) | 368 (25.6) | 0.01 |
| Male sex, n (%)a | 1068 (74.0) | 44 180 (51.0) | 0.49 | 1065 (74.0) | 1071 (74.4) | 0.01 |
| Swiss nationality, n (%)a | 1014 (70.3) | 64 259 (74.2) | 0.09 | 1011 (70.3) | 1036 (72.0) | 0.04 |
| Non-Swiss nationality, n (%)a | 429 (29.7) | 22 330 (25.8) | 0.09 | 428 (29.7) | 403 (28.0) | 0.04 |
| Age (years), median (IQR)a | 35 (24–51) | 30 (19–47) | 0.30 | 35 (24–51) | 36 (24–53) | 0.03 |
| Age 0–17 yearsa | 87 (6.0) | 18 936 (21.9) | 0.47 | 87 (6.0) | 83 (5.8) | 0.02 |
| Age 18–40 yearsa | 779 (54.0) | 39 170 (45.2) | 779 (54.1) | 770 (53.5) | ||
| Age >40 yearsa | 577 (40.0) | 28 483 (32.9) | 573 (39.8) | 586 (40.7) | ||
| Year of index hospitalizationa | ||||||
| 2013 | 145 (10.0) | 9995 (11.5) | 0.10 | 145 (10.1) | 145 (10.1) | 0.02 |
| 2014 | 165 (11.4) | 10 454 (12.1) | 165 (11.5) | 165 (11.5) | ||
| 2015 | 161 (11.2) | 10 746 (12.4) | 161 (11.2) | 158 (11.0) | ||
| 2016 | 199 (13.8) | 11 010 (12.7) | 198 (13.8) | 203 (14.1) | ||
| 2017 | 189 (13.1) | 10 947 (12.6) | 188 (13.1) | 189 (13.1) | ||
| 2018 | 182 (12.6) | 11 151 (12.9) | 182 (12.6) | 177 (12.3) | ||
| 2019 | 230 (15.9) | 11 441 (13.2) | 230 (16.0) | 238 (16.5) | ||
| 2020 | 172 (11.9) | 10 845 (12.5) | 170 (11.8) | 164 (11.4) | ||
| Comorbidities | ||||||
| Elixhauser comorbidity score, mean (SD)a | 1 (1) | 0 (1) | 0.44 | 1 (1) | 1 (1) | 0.02 |
| Diabetes mellitus type 1, n (%)a | 6 (0.4) | 180 (0.2) | 0.04 | 6 (0.4) | 4 (0.3) | 0.02 |
| Diabetes mellitus type 2, n (%)a | 36 (2.5) | 915 (1.1) | 0.11 | 35 (2.4) | 27 (1.9) | 0.04 |
| Obesity, n (%)a | 18 (1.2) | 716 (0.8) | 0.04 | 18 (1.3) | 16 (1.1) | 0.01 |
| Hypertension, n (%)a | 153 (10.6) | 4828 (5.6) | 0.19 | 153 (10.6) | 163 (11.3) | 0.02 |
| Cerebrovascular disease, n (%)a | 15 (1.0) | 82 (0.1) | 0.13 | 12 (0.8) | 10 (0.7) | 0.02 |
| Peripheral arterial disease, n (%)a | 4 (0.3) | 68 (0.1) | 0.05 | 4 (0.3) | 2 (0.1) | 0.03 |
| CKD, n (%)a | 33 (2.3) | 415 (0.5) | 0.16 | 33 (2.3) | 23 (1.6) | 0.05 |
| Hepatopathy, n (%)a | 32 (2.2) | 284 (0.3) | 0.17 | 31 (2.2) | 16 (1.1) | 0.08 |
| COPD, n (%)a | 14 (1.0) | 213 (0.2) | 0.09 | 14 (1.0) | 12 (0.8) | 0.01 |
| Sleep apnoea, n (%)a | 13 (0.9) | 286 (0.3) | 0.07 | 13 (0.9) | 9 (0.6) | 0.03 |
| Dyslipidaemia, n (%)a | 112 (7.8) | 1093 (1.3) | 0.32 | 110 (7.6) | 122 (8.5) | 0.03 |
| Malignant neoplasm, n (%)a | 39 (2.7) | 1207 (1.4) | 0.09 | 39 (2.7) | 24 (1.7) | 0.07 |
| Length of index hospitalization | ||||||
| Length of hospital stay (days), median (IQR) | 3 (2–5) | 2 (2–4) | 0.24 | 3 (2–5) | 2 (2–4) | 0.15 |
| In-hospital diagnostics | ||||||
| Endomyocardial biopsy, n (%) | 2 (0.1) | 0 (0.0) | 0.05 | 2 (0.1) | 0 (0.0) | 0.05 |
| Diagnostic coronary angiography, n (%) | 443 (30.7) | 0 (0.0) | 0.94 | 443 (30.8) | 0 (0.0) | 0.94 |
| Cardiac MRI, n (%) | 360 (24.9) | 0 (0.0) | 0.82 | 360 (25.0) | 0 (0.0) | 0.82 |
| Cardiac CT, n (%) | 45 (3.1) | 2 (0.002) | 0.25 | 45 (3.1) | 0 (0.0) | 0.25 |
| Myocardial perfusion scintigraphy, n (%) | 4 (0.3) | 0 (0.0) | 0.07 | 4 (0.3) | 0 (0.0) | 0.07 |
| Cardiac PET, n (%) | 6 (0.4) | 0 (0.0) | 0.09 | 6 (0.4) | 0 (0.0) | 0.09 |
| TTE, n (%) | 300 (20.8) | 34 (0.04) | 0.72 | 299 (20.8) | 0 (0.0) | 0.72 |
| TEE, n (%) | 12 (0.8) | 8 (0.009) | 0.13 | 11 (0.8) | 0 (0.0) | 0.12 |
| Only TTE, n (%)b | 136 (9.4) | 0 (0.0) | n/a | 135 (9.4) | 0 (0.0) | n/a |
| Only TTE or TEE, n (%)b | 142 (9.8) | 0 (0.0) | n/a | 140 (9.7) | 0 (0.0) | n/a |
| None of aforementioned cardiac diagnostics, n (%)b | 591 (41.0) | 0 (0.0) | n/a | 589 (40.9) | 0 (0.0) | n/a |
| Any of aforementioned cardiac diagnostics, n (%)b | 852 (59.0) | 0 (0.0) | n/a | 850 (59.1) | 0 (0.0) | n/a |
Standardized differences are reported as absolute values. An absolute standardized difference of <0.10 indicates a relatively small imbalance.
CKD, chronic kidney disease; COPD, chronic obstructive pulmonary disease; MRI, magnetic resonance imaging; CT, computed tomography; PET, positron emission tomography; TTE, transthoracic echocardiography; TEE, transoesophageal echocardiography; n, number; n/a, not available; SD, standard deviation; IQR, interquartile range; Std Diff, standardized difference.
aFactors used for matching.
bMyocarditis patients only.
Baseline characteristics before and after propensity score matching
| Before matching | After matching | |||||
|---|---|---|---|---|---|---|
| Patients with myocarditis (n = 1443) | Patients without myocarditis (n = 86 589) | Std Diff | Patients with myocarditis (n = 1439) | Patients without myocarditis (n = 1439) | Std Diff | |
| Demographics | ||||||
| Female sex, n (%)a | 375 (26.0) | 42 409 (49.0) | 0.49 | 374 (26.0) | 368 (25.6) | 0.01 |
| Male sex, n (%)a | 1068 (74.0) | 44 180 (51.0) | 0.49 | 1065 (74.0) | 1071 (74.4) | 0.01 |
| Swiss nationality, n (%)a | 1014 (70.3) | 64 259 (74.2) | 0.09 | 1011 (70.3) | 1036 (72.0) | 0.04 |
| Non-Swiss nationality, n (%)a | 429 (29.7) | 22 330 (25.8) | 0.09 | 428 (29.7) | 403 (28.0) | 0.04 |
| Age (years), median (IQR)a | 35 (24–51) | 30 (19–47) | 0.30 | 35 (24–51) | 36 (24–53) | 0.03 |
| Age 0–17 yearsa | 87 (6.0) | 18 936 (21.9) | 0.47 | 87 (6.0) | 83 (5.8) | 0.02 |
| Age 18–40 yearsa | 779 (54.0) | 39 170 (45.2) | 779 (54.1) | 770 (53.5) | ||
| Age >40 yearsa | 577 (40.0) | 28 483 (32.9) | 573 (39.8) | 586 (40.7) | ||
| Year of index hospitalizationa | ||||||
| 2013 | 145 (10.0) | 9995 (11.5) | 0.10 | 145 (10.1) | 145 (10.1) | 0.02 |
| 2014 | 165 (11.4) | 10 454 (12.1) | 165 (11.5) | 165 (11.5) | ||
| 2015 | 161 (11.2) | 10 746 (12.4) | 161 (11.2) | 158 (11.0) | ||
| 2016 | 199 (13.8) | 11 010 (12.7) | 198 (13.8) | 203 (14.1) | ||
| 2017 | 189 (13.1) | 10 947 (12.6) | 188 (13.1) | 189 (13.1) | ||
| 2018 | 182 (12.6) | 11 151 (12.9) | 182 (12.6) | 177 (12.3) | ||
| 2019 | 230 (15.9) | 11 441 (13.2) | 230 (16.0) | 238 (16.5) | ||
| 2020 | 172 (11.9) | 10 845 (12.5) | 170 (11.8) | 164 (11.4) | ||
| Comorbidities | ||||||
| Elixhauser comorbidity score, mean (SD)a | 1 (1) | 0 (1) | 0.44 | 1 (1) | 1 (1) | 0.02 |
| Diabetes mellitus type 1, n (%)a | 6 (0.4) | 180 (0.2) | 0.04 | 6 (0.4) | 4 (0.3) | 0.02 |
| Diabetes mellitus type 2, n (%)a | 36 (2.5) | 915 (1.1) | 0.11 | 35 (2.4) | 27 (1.9) | 0.04 |
| Obesity, n (%)a | 18 (1.2) | 716 (0.8) | 0.04 | 18 (1.3) | 16 (1.1) | 0.01 |
| Hypertension, n (%)a | 153 (10.6) | 4828 (5.6) | 0.19 | 153 (10.6) | 163 (11.3) | 0.02 |
| Cerebrovascular disease, n (%)a | 15 (1.0) | 82 (0.1) | 0.13 | 12 (0.8) | 10 (0.7) | 0.02 |
| Peripheral arterial disease, n (%)a | 4 (0.3) | 68 (0.1) | 0.05 | 4 (0.3) | 2 (0.1) | 0.03 |
| CKD, n (%)a | 33 (2.3) | 415 (0.5) | 0.16 | 33 (2.3) | 23 (1.6) | 0.05 |
| Hepatopathy, n (%)a | 32 (2.2) | 284 (0.3) | 0.17 | 31 (2.2) | 16 (1.1) | 0.08 |
| COPD, n (%)a | 14 (1.0) | 213 (0.2) | 0.09 | 14 (1.0) | 12 (0.8) | 0.01 |
| Sleep apnoea, n (%)a | 13 (0.9) | 286 (0.3) | 0.07 | 13 (0.9) | 9 (0.6) | 0.03 |
| Dyslipidaemia, n (%)a | 112 (7.8) | 1093 (1.3) | 0.32 | 110 (7.6) | 122 (8.5) | 0.03 |
| Malignant neoplasm, n (%)a | 39 (2.7) | 1207 (1.4) | 0.09 | 39 (2.7) | 24 (1.7) | 0.07 |
| Length of index hospitalization | ||||||
| Length of hospital stay (days), median (IQR) | 3 (2–5) | 2 (2–4) | 0.24 | 3 (2–5) | 2 (2–4) | 0.15 |
| In-hospital diagnostics | ||||||
| Endomyocardial biopsy, n (%) | 2 (0.1) | 0 (0.0) | 0.05 | 2 (0.1) | 0 (0.0) | 0.05 |
| Diagnostic coronary angiography, n (%) | 443 (30.7) | 0 (0.0) | 0.94 | 443 (30.8) | 0 (0.0) | 0.94 |
| Cardiac MRI, n (%) | 360 (24.9) | 0 (0.0) | 0.82 | 360 (25.0) | 0 (0.0) | 0.82 |
| Cardiac CT, n (%) | 45 (3.1) | 2 (0.002) | 0.25 | 45 (3.1) | 0 (0.0) | 0.25 |
| Myocardial perfusion scintigraphy, n (%) | 4 (0.3) | 0 (0.0) | 0.07 | 4 (0.3) | 0 (0.0) | 0.07 |
| Cardiac PET, n (%) | 6 (0.4) | 0 (0.0) | 0.09 | 6 (0.4) | 0 (0.0) | 0.09 |
| TTE, n (%) | 300 (20.8) | 34 (0.04) | 0.72 | 299 (20.8) | 0 (0.0) | 0.72 |
| TEE, n (%) | 12 (0.8) | 8 (0.009) | 0.13 | 11 (0.8) | 0 (0.0) | 0.12 |
| Only TTE, n (%)b | 136 (9.4) | 0 (0.0) | n/a | 135 (9.4) | 0 (0.0) | n/a |
| Only TTE or TEE, n (%)b | 142 (9.8) | 0 (0.0) | n/a | 140 (9.7) | 0 (0.0) | n/a |
| None of aforementioned cardiac diagnostics, n (%)b | 591 (41.0) | 0 (0.0) | n/a | 589 (40.9) | 0 (0.0) | n/a |
| Any of aforementioned cardiac diagnostics, n (%)b | 852 (59.0) | 0 (0.0) | n/a | 850 (59.1) | 0 (0.0) | n/a |
| Before matching | After matching | |||||
|---|---|---|---|---|---|---|
| Patients with myocarditis (n = 1443) | Patients without myocarditis (n = 86 589) | Std Diff | Patients with myocarditis (n = 1439) | Patients without myocarditis (n = 1439) | Std Diff | |
| Demographics | ||||||
| Female sex, n (%)a | 375 (26.0) | 42 409 (49.0) | 0.49 | 374 (26.0) | 368 (25.6) | 0.01 |
| Male sex, n (%)a | 1068 (74.0) | 44 180 (51.0) | 0.49 | 1065 (74.0) | 1071 (74.4) | 0.01 |
| Swiss nationality, n (%)a | 1014 (70.3) | 64 259 (74.2) | 0.09 | 1011 (70.3) | 1036 (72.0) | 0.04 |
| Non-Swiss nationality, n (%)a | 429 (29.7) | 22 330 (25.8) | 0.09 | 428 (29.7) | 403 (28.0) | 0.04 |
| Age (years), median (IQR)a | 35 (24–51) | 30 (19–47) | 0.30 | 35 (24–51) | 36 (24–53) | 0.03 |
| Age 0–17 yearsa | 87 (6.0) | 18 936 (21.9) | 0.47 | 87 (6.0) | 83 (5.8) | 0.02 |
| Age 18–40 yearsa | 779 (54.0) | 39 170 (45.2) | 779 (54.1) | 770 (53.5) | ||
| Age >40 yearsa | 577 (40.0) | 28 483 (32.9) | 573 (39.8) | 586 (40.7) | ||
| Year of index hospitalizationa | ||||||
| 2013 | 145 (10.0) | 9995 (11.5) | 0.10 | 145 (10.1) | 145 (10.1) | 0.02 |
| 2014 | 165 (11.4) | 10 454 (12.1) | 165 (11.5) | 165 (11.5) | ||
| 2015 | 161 (11.2) | 10 746 (12.4) | 161 (11.2) | 158 (11.0) | ||
| 2016 | 199 (13.8) | 11 010 (12.7) | 198 (13.8) | 203 (14.1) | ||
| 2017 | 189 (13.1) | 10 947 (12.6) | 188 (13.1) | 189 (13.1) | ||
| 2018 | 182 (12.6) | 11 151 (12.9) | 182 (12.6) | 177 (12.3) | ||
| 2019 | 230 (15.9) | 11 441 (13.2) | 230 (16.0) | 238 (16.5) | ||
| 2020 | 172 (11.9) | 10 845 (12.5) | 170 (11.8) | 164 (11.4) | ||
| Comorbidities | ||||||
| Elixhauser comorbidity score, mean (SD)a | 1 (1) | 0 (1) | 0.44 | 1 (1) | 1 (1) | 0.02 |
| Diabetes mellitus type 1, n (%)a | 6 (0.4) | 180 (0.2) | 0.04 | 6 (0.4) | 4 (0.3) | 0.02 |
| Diabetes mellitus type 2, n (%)a | 36 (2.5) | 915 (1.1) | 0.11 | 35 (2.4) | 27 (1.9) | 0.04 |
| Obesity, n (%)a | 18 (1.2) | 716 (0.8) | 0.04 | 18 (1.3) | 16 (1.1) | 0.01 |
| Hypertension, n (%)a | 153 (10.6) | 4828 (5.6) | 0.19 | 153 (10.6) | 163 (11.3) | 0.02 |
| Cerebrovascular disease, n (%)a | 15 (1.0) | 82 (0.1) | 0.13 | 12 (0.8) | 10 (0.7) | 0.02 |
| Peripheral arterial disease, n (%)a | 4 (0.3) | 68 (0.1) | 0.05 | 4 (0.3) | 2 (0.1) | 0.03 |
| CKD, n (%)a | 33 (2.3) | 415 (0.5) | 0.16 | 33 (2.3) | 23 (1.6) | 0.05 |
| Hepatopathy, n (%)a | 32 (2.2) | 284 (0.3) | 0.17 | 31 (2.2) | 16 (1.1) | 0.08 |
| COPD, n (%)a | 14 (1.0) | 213 (0.2) | 0.09 | 14 (1.0) | 12 (0.8) | 0.01 |
| Sleep apnoea, n (%)a | 13 (0.9) | 286 (0.3) | 0.07 | 13 (0.9) | 9 (0.6) | 0.03 |
| Dyslipidaemia, n (%)a | 112 (7.8) | 1093 (1.3) | 0.32 | 110 (7.6) | 122 (8.5) | 0.03 |
| Malignant neoplasm, n (%)a | 39 (2.7) | 1207 (1.4) | 0.09 | 39 (2.7) | 24 (1.7) | 0.07 |
| Length of index hospitalization | ||||||
| Length of hospital stay (days), median (IQR) | 3 (2–5) | 2 (2–4) | 0.24 | 3 (2–5) | 2 (2–4) | 0.15 |
| In-hospital diagnostics | ||||||
| Endomyocardial biopsy, n (%) | 2 (0.1) | 0 (0.0) | 0.05 | 2 (0.1) | 0 (0.0) | 0.05 |
| Diagnostic coronary angiography, n (%) | 443 (30.7) | 0 (0.0) | 0.94 | 443 (30.8) | 0 (0.0) | 0.94 |
| Cardiac MRI, n (%) | 360 (24.9) | 0 (0.0) | 0.82 | 360 (25.0) | 0 (0.0) | 0.82 |
| Cardiac CT, n (%) | 45 (3.1) | 2 (0.002) | 0.25 | 45 (3.1) | 0 (0.0) | 0.25 |
| Myocardial perfusion scintigraphy, n (%) | 4 (0.3) | 0 (0.0) | 0.07 | 4 (0.3) | 0 (0.0) | 0.07 |
| Cardiac PET, n (%) | 6 (0.4) | 0 (0.0) | 0.09 | 6 (0.4) | 0 (0.0) | 0.09 |
| TTE, n (%) | 300 (20.8) | 34 (0.04) | 0.72 | 299 (20.8) | 0 (0.0) | 0.72 |
| TEE, n (%) | 12 (0.8) | 8 (0.009) | 0.13 | 11 (0.8) | 0 (0.0) | 0.12 |
| Only TTE, n (%)b | 136 (9.4) | 0 (0.0) | n/a | 135 (9.4) | 0 (0.0) | n/a |
| Only TTE or TEE, n (%)b | 142 (9.8) | 0 (0.0) | n/a | 140 (9.7) | 0 (0.0) | n/a |
| None of aforementioned cardiac diagnostics, n (%)b | 591 (41.0) | 0 (0.0) | n/a | 589 (40.9) | 0 (0.0) | n/a |
| Any of aforementioned cardiac diagnostics, n (%)b | 852 (59.0) | 0 (0.0) | n/a | 850 (59.1) | 0 (0.0) | n/a |
Standardized differences are reported as absolute values. An absolute standardized difference of <0.10 indicates a relatively small imbalance.
CKD, chronic kidney disease; COPD, chronic obstructive pulmonary disease; MRI, magnetic resonance imaging; CT, computed tomography; PET, positron emission tomography; TTE, transthoracic echocardiography; TEE, transoesophageal echocardiography; n, number; n/a, not available; SD, standard deviation; IQR, interquartile range; Std Diff, standardized difference.
aFactors used for matching.
bMyocarditis patients only.
Incident cardiovascular events
After matching, over a median follow-up of 39 months (min. 1 month, max. 95 months), we observed 43.7 events of the primary outcome per 1000 py among patients hospitalized with an uncomplicated acute myocarditis and 0.9 events per 1000 py among surgical controls, IRD of 42.7 (95% CI, 36.7–48.8) per 1000 py; HR of 42.3 (95% CI, 17.4–102.8). These findings were confirmed by the Kaplan–Meier curve with a steep increase of cumulative incidence after the first month of follow-up among the myocarditis group (Figure 1A).
Kaplan–Meier failure estimates of the cardiovascular composite outcome (primary endpoint) in the matched population. (A) Patients who had a first time (index) hospitalization with uncomplicated acute myocarditis had higher rates of the cardiovascular composite outcome, a composite of rehospitalization for myocarditis (primary position), hospitalization for heart failure (primary position), hospitalization for dilated cardiomyopathy (primary position), hospitalization with ventricular tachycardia (any position), hospitalization with ventricular fibrillation/flutter (any position), hospitalization with cardiac arrest (any position), hospitalization with cardiogenic shock (any position), hospitalization for supraventricular tachycardia (primary position), hospitalization for atrial fibrillation or flutter (primary position), hospitalization for second or third degree atrioventricular block (primary position), hospitalization for sick sinus syndrome (primary position), hospitalization for a pericardial disease (e.g. acute pericarditis) (primary position), hospitalization with a cardiac device implantation (any position) and hospitalization with heart transplant (any position), compared to patients without myocarditis (i.e. surgical controls). (B) Kaplan–Meier failure estimate of the cardiovascular composite outcome with a blanking period of 1 month, i.e. patients entered time-to-event analysis after 1 month of follow-up. The risk table shows the number at risk starting after 1 month of follow-up. After blanking of the first month of follow-up, patients who had an index hospitalization with uncomplicated acute myocarditis still had higher rates of the cardiovascular composite outcome compared to surgical controls.
Even after blanking the first month of follow-up (i.e. only events after 1 month of follow-up were entered in the Cox proportional hazards regression model), the excess hazard in patients with acute myocarditis compared to surgical controls remained statistically significant (HR 12.4, 95% CI 5.0–31.1) (Figure 1B).
Among secondary outcomes, rehospitalization for myocarditis, hospitalization for pericardial disease, hospitalization with ASCVD, and all-cause rehospitalization were the most frequently reached endpoints, with higher rates in patients with myocarditis compared to surgical controls (Table 2, Supplementary material online, eTable 4). Further individual secondary outcomes could not be assessed in regression analyses due to the small number of events.
Primary and secondary outcomes after propensity score matching
| Patients with myocarditis (n = 1439) | Patients without myocarditis (n = 1439) | IRD per 1000 py (95% CI) | HR (95% CI) | |||
|---|---|---|---|---|---|---|
| n (%) | IR per 1000 py | n (%) | IR per 1000 py | |||
| Primary outcome | ||||||
| Cardiovascular composite outcomea | 203 (14.1) | 43.65 | 5 (0.3) | 0.93 | 42.72 (36.66 to 48.78) | 42.34 (17.43 to 102.82) |
| Secondary outcomes | ||||||
| Rehospitalization for myocarditis | 87 (6.0) | 17.19 | 0 (0.0) | 0 | 17.19 (13.58 to 20.81) | n/a |
| Hospitalization for heart failure | 11 (0.8) | 2.06 | 1 (0.1) | 0.19 | 1.87 (0.60 to 3.14) | 11.10 (1.43 to 85.96) |
| Hospitalization for dilated cardiomyopathy | 1 (0.1) | 0.19 | 0 (0.0) | 0 | 0.19 (−0.18 to 0.55) | n/a |
| Hospitalization with ventricular tachycardia | 8 (0.6) | 1.49 | 1 (0.1) | 0.19 | 1.31 (0.21 to 2.41) | 8.06 (1.01 to 64.42) |
| Hospitalization with ventricular fibrillation and/or flutter | 2 (0.1) | 0.37 | 1 (0.1) | 0.19 | 0.19 (−0.45 to 0.82) | 2.01 (0.18 to 22.20) |
| Hospitalization with cardiac arrest | 6 (0.4) | 1.12 | 1 (0.1) | 0.19 | 0.93 (−0.03 to 1.90) | 6.00 (0.72 to 49.87) |
| Hospitalization with cardiogenic shock | 4 (0.3) | 0.75 | 1 (0.1) | 0.19 | 0.56 (−0.26 to 1.38) | 4.02 (0.45 to 35.94) |
| Hospitalization for supraventricular tachycardia | 4 (0.3) | 0.75 | 1 (0.1) | 0.19 | 0.56 (−0.26 to 1.38) | 4.11 (0.46 to 36.78) |
| Hospitalization for atrial fibrillation and/or flutter | 6 (0.4) | 1.12 | 0 (0.0) | 0 | 1.12 (0.22 to 2.02) | n/a |
| Hospitalization for 2nd or 3rd degree AVB | 0 (0.0) | 0 | 1 (0.1) | 0.19 | −0.19 (−0.55 to 0.18) | n/a |
| Hospitalization for sick sinus syndrome | 1 (0.1) | 0.19 | 0 (0.0) | 0 | 0.19 (−0.18 to 0.55) | n/a |
| Hospitalization for pericardial disease | 124 (8.6) | 20.67 | 0 (0.0) | 0 | 20.67 (16.68 to 24.66) | n/a |
| Hospitalization for acute pericarditis | 43 (3.0) | 8.25 | 0 (0.0) | 0 | 8.25 (5.79 to 10.72) | n/a |
| Hospitalization with cardiac device insertion (ICD, CRT, pacemaker) | 4 (0.3) | 0.75 | 1 (0.1) | 0.19 | 0.56 (−0.26 to 1.38) | 4.02 (0.45 to 35.99) |
| Hospitalization with heart transplant | 1 (0.1) | 0.19 | 0 (0.0) | 0 | 0.19 (−0.18 to 0.55) | n/a |
| Hospitalization with ASCVD | 126 (8.8) | 25.54 | 16 (1.1) | 2.99 | 22.55 (17.86 to 27.25) | 8.11 (4.82 to 13.64) |
| All-cause rehospitalization | 537 (37.3) | 144.14 | 359 (24.9) | 81.56 | 62.59 (47.76 to 77.41) | 1.68 (1.47 to 1.92) |
| Patients with myocarditis (n = 1439) | Patients without myocarditis (n = 1439) | IRD per 1000 py (95% CI) | HR (95% CI) | |||
|---|---|---|---|---|---|---|
| n (%) | IR per 1000 py | n (%) | IR per 1000 py | |||
| Primary outcome | ||||||
| Cardiovascular composite outcomea | 203 (14.1) | 43.65 | 5 (0.3) | 0.93 | 42.72 (36.66 to 48.78) | 42.34 (17.43 to 102.82) |
| Secondary outcomes | ||||||
| Rehospitalization for myocarditis | 87 (6.0) | 17.19 | 0 (0.0) | 0 | 17.19 (13.58 to 20.81) | n/a |
| Hospitalization for heart failure | 11 (0.8) | 2.06 | 1 (0.1) | 0.19 | 1.87 (0.60 to 3.14) | 11.10 (1.43 to 85.96) |
| Hospitalization for dilated cardiomyopathy | 1 (0.1) | 0.19 | 0 (0.0) | 0 | 0.19 (−0.18 to 0.55) | n/a |
| Hospitalization with ventricular tachycardia | 8 (0.6) | 1.49 | 1 (0.1) | 0.19 | 1.31 (0.21 to 2.41) | 8.06 (1.01 to 64.42) |
| Hospitalization with ventricular fibrillation and/or flutter | 2 (0.1) | 0.37 | 1 (0.1) | 0.19 | 0.19 (−0.45 to 0.82) | 2.01 (0.18 to 22.20) |
| Hospitalization with cardiac arrest | 6 (0.4) | 1.12 | 1 (0.1) | 0.19 | 0.93 (−0.03 to 1.90) | 6.00 (0.72 to 49.87) |
| Hospitalization with cardiogenic shock | 4 (0.3) | 0.75 | 1 (0.1) | 0.19 | 0.56 (−0.26 to 1.38) | 4.02 (0.45 to 35.94) |
| Hospitalization for supraventricular tachycardia | 4 (0.3) | 0.75 | 1 (0.1) | 0.19 | 0.56 (−0.26 to 1.38) | 4.11 (0.46 to 36.78) |
| Hospitalization for atrial fibrillation and/or flutter | 6 (0.4) | 1.12 | 0 (0.0) | 0 | 1.12 (0.22 to 2.02) | n/a |
| Hospitalization for 2nd or 3rd degree AVB | 0 (0.0) | 0 | 1 (0.1) | 0.19 | −0.19 (−0.55 to 0.18) | n/a |
| Hospitalization for sick sinus syndrome | 1 (0.1) | 0.19 | 0 (0.0) | 0 | 0.19 (−0.18 to 0.55) | n/a |
| Hospitalization for pericardial disease | 124 (8.6) | 20.67 | 0 (0.0) | 0 | 20.67 (16.68 to 24.66) | n/a |
| Hospitalization for acute pericarditis | 43 (3.0) | 8.25 | 0 (0.0) | 0 | 8.25 (5.79 to 10.72) | n/a |
| Hospitalization with cardiac device insertion (ICD, CRT, pacemaker) | 4 (0.3) | 0.75 | 1 (0.1) | 0.19 | 0.56 (−0.26 to 1.38) | 4.02 (0.45 to 35.99) |
| Hospitalization with heart transplant | 1 (0.1) | 0.19 | 0 (0.0) | 0 | 0.19 (−0.18 to 0.55) | n/a |
| Hospitalization with ASCVD | 126 (8.8) | 25.54 | 16 (1.1) | 2.99 | 22.55 (17.86 to 27.25) | 8.11 (4.82 to 13.64) |
| All-cause rehospitalization | 537 (37.3) | 144.14 | 359 (24.9) | 81.56 | 62.59 (47.76 to 77.41) | 1.68 (1.47 to 1.92) |
AVB, atrioventricular block; ICD, implantable cardioverter defibrillator; CRT, cardiac resynchronization therapy; ASCVD, atherosclerotic cardiovascular disease; n, number; CI, confidence interval; IR, incidence rate; IRD, incidence rate difference; HR, hazard ratio; py, person-years; n/a, not applicable.
aComposite of rehospitalization for myocarditis (primary position), hospitalization for heart failure (primary position), hospitalization for dilated cardiomyopathy (primary position), hospitalization with ventricular tachycardia (any position), hospitalization with ventricular fibrillation/flutter (any position), hospitalization with cardiac arrest (any position), hospitalization with cardiogenic shock (any position), hospitalization for supraventricular tachycardia (primary position), hospitalization for atrial fibrillation or flutter (primary position), hospitalization for second or third degree atrioventricular block (primary position), hospitalization for sick sinus syndrome (primary position), hospitalization for a pericardial disease (e.g. acute pericarditis) (primary position), hospitalization with a cardiac device implantation (any position), and hospitalization with heart transplant (any position).
Primary and secondary outcomes after propensity score matching
| Patients with myocarditis (n = 1439) | Patients without myocarditis (n = 1439) | IRD per 1000 py (95% CI) | HR (95% CI) | |||
|---|---|---|---|---|---|---|
| n (%) | IR per 1000 py | n (%) | IR per 1000 py | |||
| Primary outcome | ||||||
| Cardiovascular composite outcomea | 203 (14.1) | 43.65 | 5 (0.3) | 0.93 | 42.72 (36.66 to 48.78) | 42.34 (17.43 to 102.82) |
| Secondary outcomes | ||||||
| Rehospitalization for myocarditis | 87 (6.0) | 17.19 | 0 (0.0) | 0 | 17.19 (13.58 to 20.81) | n/a |
| Hospitalization for heart failure | 11 (0.8) | 2.06 | 1 (0.1) | 0.19 | 1.87 (0.60 to 3.14) | 11.10 (1.43 to 85.96) |
| Hospitalization for dilated cardiomyopathy | 1 (0.1) | 0.19 | 0 (0.0) | 0 | 0.19 (−0.18 to 0.55) | n/a |
| Hospitalization with ventricular tachycardia | 8 (0.6) | 1.49 | 1 (0.1) | 0.19 | 1.31 (0.21 to 2.41) | 8.06 (1.01 to 64.42) |
| Hospitalization with ventricular fibrillation and/or flutter | 2 (0.1) | 0.37 | 1 (0.1) | 0.19 | 0.19 (−0.45 to 0.82) | 2.01 (0.18 to 22.20) |
| Hospitalization with cardiac arrest | 6 (0.4) | 1.12 | 1 (0.1) | 0.19 | 0.93 (−0.03 to 1.90) | 6.00 (0.72 to 49.87) |
| Hospitalization with cardiogenic shock | 4 (0.3) | 0.75 | 1 (0.1) | 0.19 | 0.56 (−0.26 to 1.38) | 4.02 (0.45 to 35.94) |
| Hospitalization for supraventricular tachycardia | 4 (0.3) | 0.75 | 1 (0.1) | 0.19 | 0.56 (−0.26 to 1.38) | 4.11 (0.46 to 36.78) |
| Hospitalization for atrial fibrillation and/or flutter | 6 (0.4) | 1.12 | 0 (0.0) | 0 | 1.12 (0.22 to 2.02) | n/a |
| Hospitalization for 2nd or 3rd degree AVB | 0 (0.0) | 0 | 1 (0.1) | 0.19 | −0.19 (−0.55 to 0.18) | n/a |
| Hospitalization for sick sinus syndrome | 1 (0.1) | 0.19 | 0 (0.0) | 0 | 0.19 (−0.18 to 0.55) | n/a |
| Hospitalization for pericardial disease | 124 (8.6) | 20.67 | 0 (0.0) | 0 | 20.67 (16.68 to 24.66) | n/a |
| Hospitalization for acute pericarditis | 43 (3.0) | 8.25 | 0 (0.0) | 0 | 8.25 (5.79 to 10.72) | n/a |
| Hospitalization with cardiac device insertion (ICD, CRT, pacemaker) | 4 (0.3) | 0.75 | 1 (0.1) | 0.19 | 0.56 (−0.26 to 1.38) | 4.02 (0.45 to 35.99) |
| Hospitalization with heart transplant | 1 (0.1) | 0.19 | 0 (0.0) | 0 | 0.19 (−0.18 to 0.55) | n/a |
| Hospitalization with ASCVD | 126 (8.8) | 25.54 | 16 (1.1) | 2.99 | 22.55 (17.86 to 27.25) | 8.11 (4.82 to 13.64) |
| All-cause rehospitalization | 537 (37.3) | 144.14 | 359 (24.9) | 81.56 | 62.59 (47.76 to 77.41) | 1.68 (1.47 to 1.92) |
| Patients with myocarditis (n = 1439) | Patients without myocarditis (n = 1439) | IRD per 1000 py (95% CI) | HR (95% CI) | |||
|---|---|---|---|---|---|---|
| n (%) | IR per 1000 py | n (%) | IR per 1000 py | |||
| Primary outcome | ||||||
| Cardiovascular composite outcomea | 203 (14.1) | 43.65 | 5 (0.3) | 0.93 | 42.72 (36.66 to 48.78) | 42.34 (17.43 to 102.82) |
| Secondary outcomes | ||||||
| Rehospitalization for myocarditis | 87 (6.0) | 17.19 | 0 (0.0) | 0 | 17.19 (13.58 to 20.81) | n/a |
| Hospitalization for heart failure | 11 (0.8) | 2.06 | 1 (0.1) | 0.19 | 1.87 (0.60 to 3.14) | 11.10 (1.43 to 85.96) |
| Hospitalization for dilated cardiomyopathy | 1 (0.1) | 0.19 | 0 (0.0) | 0 | 0.19 (−0.18 to 0.55) | n/a |
| Hospitalization with ventricular tachycardia | 8 (0.6) | 1.49 | 1 (0.1) | 0.19 | 1.31 (0.21 to 2.41) | 8.06 (1.01 to 64.42) |
| Hospitalization with ventricular fibrillation and/or flutter | 2 (0.1) | 0.37 | 1 (0.1) | 0.19 | 0.19 (−0.45 to 0.82) | 2.01 (0.18 to 22.20) |
| Hospitalization with cardiac arrest | 6 (0.4) | 1.12 | 1 (0.1) | 0.19 | 0.93 (−0.03 to 1.90) | 6.00 (0.72 to 49.87) |
| Hospitalization with cardiogenic shock | 4 (0.3) | 0.75 | 1 (0.1) | 0.19 | 0.56 (−0.26 to 1.38) | 4.02 (0.45 to 35.94) |
| Hospitalization for supraventricular tachycardia | 4 (0.3) | 0.75 | 1 (0.1) | 0.19 | 0.56 (−0.26 to 1.38) | 4.11 (0.46 to 36.78) |
| Hospitalization for atrial fibrillation and/or flutter | 6 (0.4) | 1.12 | 0 (0.0) | 0 | 1.12 (0.22 to 2.02) | n/a |
| Hospitalization for 2nd or 3rd degree AVB | 0 (0.0) | 0 | 1 (0.1) | 0.19 | −0.19 (−0.55 to 0.18) | n/a |
| Hospitalization for sick sinus syndrome | 1 (0.1) | 0.19 | 0 (0.0) | 0 | 0.19 (−0.18 to 0.55) | n/a |
| Hospitalization for pericardial disease | 124 (8.6) | 20.67 | 0 (0.0) | 0 | 20.67 (16.68 to 24.66) | n/a |
| Hospitalization for acute pericarditis | 43 (3.0) | 8.25 | 0 (0.0) | 0 | 8.25 (5.79 to 10.72) | n/a |
| Hospitalization with cardiac device insertion (ICD, CRT, pacemaker) | 4 (0.3) | 0.75 | 1 (0.1) | 0.19 | 0.56 (−0.26 to 1.38) | 4.02 (0.45 to 35.99) |
| Hospitalization with heart transplant | 1 (0.1) | 0.19 | 0 (0.0) | 0 | 0.19 (−0.18 to 0.55) | n/a |
| Hospitalization with ASCVD | 126 (8.8) | 25.54 | 16 (1.1) | 2.99 | 22.55 (17.86 to 27.25) | 8.11 (4.82 to 13.64) |
| All-cause rehospitalization | 537 (37.3) | 144.14 | 359 (24.9) | 81.56 | 62.59 (47.76 to 77.41) | 1.68 (1.47 to 1.92) |
AVB, atrioventricular block; ICD, implantable cardioverter defibrillator; CRT, cardiac resynchronization therapy; ASCVD, atherosclerotic cardiovascular disease; n, number; CI, confidence interval; IR, incidence rate; IRD, incidence rate difference; HR, hazard ratio; py, person-years; n/a, not applicable.
aComposite of rehospitalization for myocarditis (primary position), hospitalization for heart failure (primary position), hospitalization for dilated cardiomyopathy (primary position), hospitalization with ventricular tachycardia (any position), hospitalization with ventricular fibrillation/flutter (any position), hospitalization with cardiac arrest (any position), hospitalization with cardiogenic shock (any position), hospitalization for supraventricular tachycardia (primary position), hospitalization for atrial fibrillation or flutter (primary position), hospitalization for second or third degree atrioventricular block (primary position), hospitalization for sick sinus syndrome (primary position), hospitalization for a pericardial disease (e.g. acute pericarditis) (primary position), hospitalization with a cardiac device implantation (any position), and hospitalization with heart transplant (any position).
Sensitivity and subgroup analyses
To assess the robustness of our results and to address potentially missed patients with a known heart disease, the sensitivity analysis acknowledged a 2-year washout period before the index stay. Baseline characteristics of this modified cohort are separately shown in the Supplementary material online (see Supplementary material online, eTable 5), but did not relevantly differ from the cohort of the primary analysis. Overall, the sensitivity analysis showed consistent results with a higher cardiovascular event rate among patients with acute myocarditis, even though the HR tended a bit more towards 1 [IR 48.2 vs. 1.4, IRD 46.8 (95% CI 39.6–53.9) per 1000 py, HR 31.1 (95% CI 13.8–70.1), Supplementary material online, eTable 6, Supplementary material online, eFigure 4].
In the first subgroup analysis (see Supplementary material online, eTable 7), we found that in patients who received immediate cardiac diagnostic work-up the cardiovascular event rate was lower when compared to those without immediate work-up (32.0 per 1000 py vs. 63.5 per 1000 py), resulting in an IRD of 31.1 per 1000 py and 59.7 per 1000 py when compared with the matched surgical controls (P for homogeneity <0.0001), respectively. The corresponding HRs were 32.4 (95% CI 10.3–102.1) and 14.6 (95% CI 7.1–29.9) (P for homogeneity 0.28), respectively (see Supplementary material online, eTable 8, Figure 2A and B).
Subgroup analysis: Kaplan–Meier failure estimates of the cardiovascular composite outcome among patients with and without cardiac diagnostics during the index hospitalization after propensity score matching. Both subgroups were individually 1:1 propensity-score matched to surgical controls. (A) In this subgroup, the myocarditis population included only patients who did not undergo any of the following diagnostics during index hospitalization: Endomyocardial biopsy, diagnostic coronary angiography, cardiac magnetic resonance imaging, cardiac computed tomography, myocardial perfusion scintigraphy, cardiac positron emission tomography, transthoracic echocardiography, and transoesophageal echocardiography. After an index hospitalization with uncomplicated acute myocarditis and none of the aforementioned cardiac diagnostics, patients had higher rates of the cardiovascular composite outcome, a composite of rehospitalization for myocarditis (primary position), hospitalization for heart failure (primary position), hospitalization for dilated cardiomyopathy (primary position), hospitalization with ventricular tachycardia (any position), hospitalization with ventricular fibrillation/flutter (any position), hospitalization with cardiac arrest (any position), hospitalization with cardiogenic shock (any position), hospitalization for supraventricular tachycardia (primary position), hospitalization for atrial fibrillation or flutter (primary position), hospitalization for second or third degree atrioventricular block (primary position), hospitalization for sick sinus syndrome (primary position), hospitalization for a pericardial disease (e.g. acute pericarditis) (primary position), hospitalization with a cardiac device implantation (any position) and hospitalization with heart transplant (any position), compared to surgical controls. (B) In this subgroup, the myocarditis population included only patients with uncomplicated acute myocarditis who underwent at least one of the above-mentioned cardiac diagnostics during index hospitalization. After an index hospitalization with uncomplicated acute myocarditis and at least one of the aforementioned cardiac diagnostics, patients had higher rates of the cardiovascular composite outcome compared to surgical controls.
In the second subgroup analysis (see Supplementary material online, eTable 9), where patients were stratified by age (40 years and below vs. above 40 years of age), we observed that patients aged 40 years and below exhibited a higher incidence rate of the cardiovascular composite endpoint in comparison to those aged above 40 years (54.6 per 1000 py vs. 28.8 per 1000 py), resulting in an IRD of 54.3 per 1000 py and 25.4 per 1000 py when compared with the matched surgical controls (P for homogeneity <0.0001). The associated HRs were found to be 155.9 (95% CI 21.8–1113.9) and 8.1 (95% CI 3.7–17.8) (P for homogeneity 0.0004), respectively (see Supplementary material online, eTable 10, Supplementary material online, eFigure 5A and B).
Risk factor analysis
For the risk factor analysis, the unmatched myocarditis population was used to identify predictors of the primary outcome (see Supplementary material online, eTable 11). Female sex [receiver operating characteristic (ROC) area 0.56 (95% CI 0.54–0.59), adjusted odds ratio (OR) 0.56, (95% CI 0.37–0.85), P-value 0.01] and older age [ROC area 0.61 (95% CI 0.57–0.66), OR 0.98, (95% CI 0.97–0.99), P-value 0.003] seem to be weak protective predictors of the primary outcome.
Discussion
In this population-based cohort study of 1439 patients without known heart disease who were hospitalized with uncomplicated acute myocarditis, over a follow-up up to 8 years, we observed higher rates of cardiovascular events as compared with surgical controls who underwent laparoscopic appendectomy. Those patients with acute myocarditis who received immediate cardiac diagnostic work-up during index hospitalization had a lower event rate than those who did not get immediate cardiac diagnostics. Patients with acute myocarditis aged 40 years and below exhibited a higher rate of cardiovascular events during follow-up compared to those aged above 40 years. Older age and female sex were protective factors regarding cardiovascular events after a hospitalization for uncomplicated acute myocarditis.
Even though uncomplicated acute myocarditis is reported to represent the majority of acute myocarditis cases,8 the number of studies investigating uncomplicated acute myocarditis is still limited. In line with previous epidemiologic data on acute myocarditis, patients’ demographics in this study were comparable, with a typical age of onset between 30 and 45 years and with men making up between 60% and 80% of cases.1,8,14–20 To capture the full spectrum of uncomplicated acute myocarditis, however, we included both paediatric and adult patients in our analysis.8,10,21
The majority of cardiovascular events during the follow-up occurred within the first month after index hospitalization and diagnosis of acute myocarditis. In line with this finding, a recent Swedish retrospective study observed the highest risk for development of heart failure, dilated cardiomyopathy and death in the immediate post-discharge period after a first-time index hospitalization for myocarditis.22 While it is unlikely that most of our observed cardiovascular events are related to emergencies, an important proportion of these cases might be driven by a higher detection rate (based on additional diagnostic work-up) due to the recent new diagnosis of acute myocarditis. Nonetheless, when looking at individual components included in the cardiovascular composite outcome, rehospitalization for acute myocarditis, and rehospitalization for pericardial disease were also main drivers of this early steep increase in events.
Recurrent symptomatic episodes of myocarditis are a known complication of acute myocarditis.1 While in our study 6.0% of patients were rehospitalized for recurrent myocarditis during the follow-up, none were hospitalized for acute myocarditis among surgical controls. This finding is congruent with previous studies: not only among data from the Multicenter Lombardy Registry,8 a retrospective study that compared acute myocarditis with complicated and uncomplicated presentation, but also in a retrospective Finnish and Polish study,10,23 the proportion of recurrent acute myocarditis after initial hospitalization for acute myocarditis ranged between 3% and 10% over a comparable period of follow-up.
Apart from recurrent episodes of acute myocarditis, we report that 8.6% of patients with an index hospitalization for uncomplicated acute myocarditis were rehospitalized for pericardial disease during follow-up, compared to zero hospitalizations for pericardial disease in the surgical control population. Patients with acute pericarditis and acute myocarditis can suffer from similar symptoms; however a differentiation can be made by measuring cardiac troponins.1 Since patients with pericardial disease prior or during index hospitalization were excluded in this study, our findings suggest that a part of patients with uncomplicated acute myocarditis may have developed pericardial disease as a complication of myocarditis (e.g. pericardial chest pain, pericardial effusion, or radiological signs).
Other components of the composite outcome worth mentioning are hospitalization for heart failure and cardiac device implantation with lower rates among our study population as compared to the literature. While, during follow-up, we observed a proportion of only 0.8% of patients with acute myocarditis requiring a hospitalization for heart failure and 0.3% requiring a cardiac device implantation, other population-based studies reported higher percentages: in two recent Danish studies of patients with acute myocarditis, between 6.5–7.5% of patients had a rehospitalization for heart failure and 1.6–3.7% a cardiac device implantation over follow-ups of 90 days to 8.5 years.12,21 However, these studies also included complicated cases of acute myocarditis which consequently resulted in sicker study populations with myocarditis. In turn, the Multicenter Lombardy Registry study reported no implantable cardioverter defibrillator (ICD) or cardiac resynchronization therapy device (CRT) implantations in patients with uncomplicated acute myocarditis during a median follow-up of 35 months,8 in line with our study findings.
To exclude potential bias through consecutive diagnostic work-up after a first hospitalization for acute myocarditis, we performed a sensitivity analysis by blanking the first month of follow-up. Similar to the finding of the primary analysis, we found a cumulative incidence of the primary composite outcome remaining higher in patients with acute myocarditis as compared to surgical controls (Figure 1B), supporting the robustness of our results.
In the first subgroup analysis of this study, patients with uncomplicated acute myocarditis who did not undergo myocarditis-related diagnostic work-up during index hospitalization had higher incidence rates of the cardiovascular composite outcome as compared with those who received immediate cardiac diagnostics. While endomyocardial biopsy (EMB) is still considered to be the diagnostic gold standard, according to the European Society of Cardiology (ESC) taskforce, the diagnosis of ‘clinically suspected myocarditis’ can be set by fulfilling several criteria such as clinical presentation, electrocardiogram (ECG) findings, cardiac biomarkers and cardiac imaging, and excluding relevant ASCVD.24 Due to its invasive nature, EMB is rarely performed (in 0.1% of cases in our study), even though it is primarily recommended in patients with fulminant myocarditis and should also be considered in haemodynamically stable patients with clinically suspected myocarditis.1,5,25 In our study, during the index hospitalization, the most commonly performed diagnostics were diagnostic coronary angiography, cardiac magnetic resonance imaging (MRI), and transthoracic echocardiography. Since cardiac MRI has the highest sensitivity only after 2–3 weeks after the initial clinical presentation with myocarditis,5 it is only recommended before discharge in high risk cases,1 and this may explain the large proportion of 40.9% of patients in our study who did not undergo imaging or invasive cardiac diagnostics during index hospitalization. An alternative explanation may be that the Swiss health care system financially incentivizes hospitals to perform imaging and invasive diagnostics in an ambulatory setting, whenever feasible. Although immediate cardiac diagnostics are rather recommended in more high-risk patients, our estimates in these patients suggest a lower incidence of adverse cardiac events than in those with presumably postponed diagnostic work-up. Because no causal explanation can be done given our study design, further exploration of this association in a clinical trial may be worthwhile.
Patients with acute myocarditis aged 40 or below had higher rates of cardiovascular events in comparison with those aged above 40 years. Concordantly, a risk factor analysis of the unmatched myocarditis population suggested that older age was associated with a lower risk of developing the primary outcome in patients with uncomplicated acute myocarditis. The incidence of acute myocarditis varies with age, with higher rates in infants and young adults.10,26,27 Fulminant presentations of acute myocarditis seem to be more common in hospitalized children compared to adults.27 Nonetheless, several studies have reported an association between older age and increased mortality in individuals diagnosed with acute myocarditis.10,28 However, the comparability to these previous studies is limited due to our study’s relatively young population, the absence of prior heart disease, minimal burden of comorbidities, as well as the inclusion of paediatric patients. In contrast, a cardiovascular magnetic resonance study reported that individuals below 40 years of age with acute myocarditis had a higher occurrence of regional oedema and myocardial fibrosis compared to older patients.29 The myocardial injury sustained in these younger patients seemed to be more severe with a higher incidence of irreversible injury compared to older patients.29 A more severe and more lasting myocardial injury pattern after myocarditis in younger individuals could, at least partly, explain the age association found in our study.
Moreover, the risk factor analysis conducted on the unmatched myocarditis population suggested that female sex was a protective predictor of the primary outcome in patients with uncomplicated acute myocarditis. In line with this finding, female sex is known to be a protective factor for myocarditis and inflammatory cardiomyopathy.30–32 Indeed, murine models of myocarditis as well as clinical studies focusing on recent onset cardiomyopathy, including myocarditis, have shown better myocardial recovery in females, translating into a better transplant-free survival in the clinical setting compared to males.30–32 While the mechanism remains uncertain, sex-related differences in immune function are believed to play an important role.30,32 Further, since males are reported to have a higher incidence of coronary heart disease compared to women (with a decreasing sex gap with progressing age),33 this protective association between female sex and the cardiovascular composite outcome may be influenced, at least in part, by the general incidence of coronary heart disease.
This study has limitations. First, although we used propensity score matching to adjust for baseline characteristics, residual confounding by unmeasured or not fully measured characteristics in claims (e.g. laboratory values, vital signs, or severity of disease) cannot be entirely excluded. Second, the assessment of a cardiovascular composite outcome to provide an overall measure for severe adverse events could dilute the interpretation of single cardiovascular outcomes. However, to avoid a misleading message, we addressed each cardiovascular outcome individually in secondary analyses. Third, we were unable to evaluate all-cause death since the dataset provided by the Swiss Federal Statistical Office did not include information on out of hospital death and a merger of this dataset with Swiss national death records was not feasible due to logistical reasons at the time of study conduction. Thus, we could not consider overall mortality as a competing risk. Nevertheless, in this young, very low-risk population, we do not think that considering mortality as a competing risk factor may have altered our findings relevantly. Fourth, a washout period of 1 year to exclude any history of cardiac disease may be short, however, given the findings from the sensitivity analysis, we are confident that an even longer washout period (e.g. 5 years) would not change the interpretation of our results but would compromise external validity due to the more strict patient selection. Fifth, although patients with cardiac disease were excluded during the washout period and the index hospitalization, we acknowledge the possibility of being retrospectively diagnosed with ASCVD in the outpatient setting. This could have led to the reclassification of a condition originally diagnosed as myocarditis to a myocardial infarction. Sixth, using administrative data is prone to information bias as hospitalizations were selected according to the ICD-10 codes with the risk of misclassification and underreporting of diagnoses.
Conclusion
In this nationwide cohort study, patients hospitalized with uncomplicated acute myocarditis without known prior heart disease had substantially higher rates of cardiovascular events as compared with surgical controls over a follow-up of up to 8 years. Our data suggest that the burden of uncomplicated acute myocarditis, representing the majority of cases with acute myocarditis, may be currently underestimated. This calls for a more efficient therapeutic management of this population of patients. Prospective studies focusing on longer-term outcomes after initial hospitalization for uncomplicated acute myocarditis are required to further assess the risk of life-time sequelae in this younger population.
Supplementary material
Supplementary material is available at European Heart Journal: Acute Cardiovascular Care online.
Acknowledgements
We thank the Swiss Federal Statistical Office (Bundesamt für Statistik, Neuchâtel, Switzerland) for the acquisition and provision of data.
Author contributions
Data access: A.K. had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Concept and design: A.S., S.H., T.A.F., and A.K. Acquisition, analysis, or interpretation of data: A.S., C.G., and A.K. Drafting of the manuscript: A.S., C.G., and A.K. Critical revision of the manuscript for important intellectual content: all authors. Statistical analysis: A.S., C.G., S.H., and A.K.
Funding
No financial support and no potential conflict of interest relevant to this article was reported.
Data availability
The data underlying this article were provided by the Swiss Federal Statistical Office and are available upon request from the Swiss Federal Statistical Office (Bundesamt für Statistik, Neuchâtel, Switzerland). Restrictions apply to the availability of these data, which were used under license for this study. Data are available as part of the data on “Medizinische Statistik der Krankenhäuser” with the permission of the Swiss Federal Statistical Office, Section Health Services and Population Health.
References
Author notes
Conflict of interest: none declared.
Comments