Bringing doctors together

or

A free, private network for verified physicians

Leading voices

Dr. Connor O'Brien

Cardiac Critical Care · UCSF Medical Center

View

Dr. Rosy Thachil, MD, MBA

Cardiac Critical Care · Mount Sinai

View

Dr. Dorairaj Prabhakaran

General Cardiology

View

Dr. Pugazhendhi Vijayaraman

EP · Geisinger

View

Dr. Robert Schaller

EP · HUP

View

Dr. Trey Woods

General Cardiology · Emory Healthcare

View

Dr. Brahmajee Nallamothu

Interventional Cardiology · UMHS

View

Dr. Felix Yang

EP · NYU Langone

View

Dr. Jennifer Rymer

Interventional & Peripheral Cards · Duke University

View

Dr. Ami Bhatt

Adult Congenital Heart Disease · MGH

View
  • Discover your community

  • Stay current

  • A profile that represents you

  • Find your people

About RoonPrivacyTerms

    Overview

    Myocarditis-Associated Cardiogenic Shock: A Nationwide Multicenter Study.

    Nov 11, 2025DOI: 10.1161/jaha.125.042850Journal ArticleAHA

    Written by: Talal El ZarifAlexander P AmbrosiniAndi ShahuMiles ShenMaria G GastanaduiIsrael SafiriyuAvirup GuhaMark JacobsTariq AliJennifer M KwanSarah C HullLauren BaldassarreP Elliott Miller

    Find these authors on Roon

    AS
    Dr. Andi Shahu
    Cardiac Critical Care · CUIMC

    Methods

    Requests for data access may be sent to Vizient, Inc, at cdpinfo@vizientinc.com. We queried the Vizient Clinical Data Base, which includes patient‐level data from 97% of academic medical centers and their affiliates in the United States, including >300 community hospitals.17 It includes administrative, financial, laboratory, and pharmacy‐related inpatient information. We identified all adults aged ≥18 years admitted between October 1, 2015, and June 30, 2023, with a diagnosis of cardiogenic shock. For this analysis, we selected patients diagnosed with myocarditis using the following International Classification of Diseases, Tenth Revision, Clinical Modification (ICD‐10‐CM) codes: I401, I408, I409, and I514. All data obtained from Vizient were deidentified and exempt from the Yale University Institutional Review Board review. Data from the Vizient Clinical Data Base were used with the permission of Vizient (all rights reserved).Demographic characteristics included age, sex, race and ethnicity, primary payer, and estimated median income quartile by patient home zip code.18 The severity of illness was assessed by calculating the Society for Cardiovascular Angiography and Interventions shock staging and Critical Care Cardiology Trials Network scores.19, 20 Comorbidities were assessed using the Charlson comorbidity index.21 Notable comorbidities present on admission included smoking, dyslipidemia, coronary artery disease, prior acute myocardial infarction (AMI), prior percutaneous coronary intervention, prior coronary artery bypass grafting, congestive heart failure, history of heart transplant, history of left ventricular assist device (LVAD) placement, peripheral vascular disease, chronic kidney disease, end‐stage renal disease, cerebrovascular disease, chronic obstructive pulmonary disease, cancer, dementia, diabetes, chronic liver disease, AIDS, rheumatic disease, and peptic ulcer disease. Hospital characteristics included Association of American Medical Colleges teaching hospital status (yes/no), bed range (<125, 125–300, 301–500, >500 beds), and US census regions (Midwest, Northeast, South, West).The database includes relative time stamps, reported as the day since admission, for procedures, laboratory data, and medications. We extracted occurrences of left heart catheterization, pulmonary artery catheterization, percutaneous coronary intervention, coronary artery bypass grafting, intra‐aortic balloon pump (IABP), Impella, extracorporeal membrane oxygenation (ECMO), heart transplantation, durable LVAD, invasive and noninvasive ventilation, renal replacement therapy, and tracheostomy. A comprehensive list of ICD‐10, Procedure Coding System (ICD‐10‐PCS) codes of the above procedures is shown in Table S1. For the ≈70% of patients with available laboratory data, we identified values on the first day of admission except for N‐terminal pro‐B‐type natriuretic peptide, where the first available laboratory data collected were used.The primary outcome was in‐hospital death. Secondary outcomes included the length of the hospital and ICU stays and the total hospital costs to produce care.Baseline characteristics were compared between patients with cardiogenic shock, stratified by myocarditis status. Continuous variables were presented as mean±SD, while categorical variables were presented as frequencies and percentages. Paired t tests were used to compare continuous variables, and the χ2 test was used for categorical variables. We assessed the association between myocarditis and in‐hospital death, adjusting for confounders using augmented inverse probability weighting with weight stabilization.22 SEs were calculated using bootstrapping methodology across 1000 iterations. Covariates in our model were selected on the basis of clinical relevance and prior literature on predictors of cardiogenic shock outcomes.23, 24, 25, 26 They included demographics (age, sex, race, ethnicity, primary payer, income quartile), hospital characteristics (region, hospital size in beds, Association of American Medical Colleges teaching status), comorbidities (smoking status, Charlson comorbidity index, coronary artery disease, dyslipidemia, prior AMI, prior percutaneous coronary intervention, prior coronary artery bypass grafting, diabetes, congestive heart failure, peripheral artery disease, cerebrovascular disease, chronic kidney disease, end‐stage renal disease, chronic obstructive pulmonary disease, chronic liver disease, cancer, dementia, rheumatologic disease, AIDS, out‐of‐hospital cardiac arrest, and acute renal failure on admission). To ensure covariate balance, we assessed weighted standardized differences (Table S2). Poisson regression, expressed as an incidence rate ratio, was used to evaluate the length of stay and ventilator days. A γ regression model was used to evaluate adjusted total hospital costs.In addition, we performed several sensitivity analyses. First, we assessed for the impact of center capabilities on outcomes. Due to the potential impact of transfer status on outcomes, we assessed in‐hospital death stratified by whether the patient was transferred or presented directly to the index hospitalization. Next, we evaluated differences in outcomes stratified by whether the treating center had heart transplantation or LVAD implantation capabilities. Then, we performed a mixed‐effects logistic regression model with hospital center as a random effect among just patients with myocarditis to evaluate whether outcomes were different, stratified by hospital center. We used the same approach to assess whether treatment strategies (eg, MCS, pulmonary artery catheterization, or early corticosteroid use) were different across centers. Early corticosteroid use was defined as within the first 2 days of admission.To better understand the impact of MCS use on outcomes, we assessed for the association between early MCS (day 1 versus day 2 of admission) and whether patients with or without myocarditis were more likely to undergo MCS device escalation (IABP to ECMO, IABP to Impella, or Impella to ECMO) using the entire cohort. Interaction testing was completed between early MCS, each individual device, and myocarditis. Finally, we compared outcomes for patients with a primary diagnosis of AMI or heart failure with patients with myocarditis. All analyses were performed using STATA 16.0 (StataCorp, College Station, TX) with statistical significance considered at a 2‐tailed P<0.05.

    Results

    Including a total of 621 695 patients with cardiogenic shock, 2925 (0.5%) patients had a diagnosis of myocarditis (Figure 1). The top principal admission diagnoses associated with cardiogenic shock are presented in Table S3. Patients with myocarditis‐associated cardiogenic shock were more likely to be younger (49.9 versus 65.2 years) and women (45.8% versus 36.7%), have Society for Cardiovascular Angiography and Interventions shock stage E (15.9% versus 13.1%; P<0.001), treated at large academic centers (P<0.001), and to be transferred to the treating institution (45.0% versus 30.4%) (Table 1). They also had a lower Charlson comorbidity index (3.6 versus 4.4; P<0.001) and fewer chronic cardiac and noncardiac comorbidities, albeit more cancer (13.5% versus 7.5%) and rheumatic diseases (8.3% versus 3.4%) (all P<0.001) (Table 2).CABG indicates coronary artery bypass grafting; CAD, coronary artery disease; CKD, chronic kidney disease; COPD, chronic obstructive pulmonary disease; ECMO, extracorporeal membrane oxygenation; MCS, mechanical circulatory support; and PCI, percutaneous coronary intervention. Data from the Vizient Clinical Data Base were used with the permission of Vizient (all rights reserved). The figure was created with BioRender.com.Characteristics of Patients With Cardiogenic Shock With Versus Without MyocarditisData are presented as mean±SD for continuous measures and n (%) for categorical measures. P values were calculated using Pearson’s χ2 test for categorical variables and t tests (2 group comparisons) or ANOVA (>2 group comparisons) for continuous variables. CCCTN indicates Critical Care Cardiology Trials Network; and SCAI, Society for Cardiovascular Angiography and Interventions. Data from the Vizient Clinical Data Base were used with the permission of Vizient (all rights reserved).Comorbidities for Patients With Cardiogenic Shock With Versus Without MyocarditisData are presented as mean±SD for continuous measures and n (%) for categorical measures. P values were calculated using Pearson’s χ2 test for categorical variables and t tests (2 group comparisons) or ANOVA (>2 group comparisons) for continuous variables. AMI indicates acute myocardial infarction; and LVAD, left ventricular assist device. Data from the Vizient Clinical Data Base were used with the permission of Vizient (all rights reserved).Including 420 067 (67.6%) patients with available laboratory data, the initial and maximum lactate on the day of admission and initial pro‐B‐type natriuretic peptide were similar between groups (all P>0.05). In contrast, the first creatinine on admission (1.7 versus 2.0 mg/dL) and maximum admission creatinine (1.8 versus 2.1 mg/dL) were lower in patients with myocarditis‐associated cardiogenic shock, whereas the initial aspartate aminotransferase was significantly higher in the myocarditis‐associated group (578 versus 262 U/L) (all P<0.001). The remainder of the evaluated laboratory values showed statistically significant differences but were numerically similar (Table 3).Admission Laboratory Results for Patients With Cardiogenic Shock Presenting With Versus Without MyocarditisData are presented as mean±SD for continuous measures. P values were calculated using Pearson’s χ2 test for categorical variables and t tests (2 group comparisons) or ANOVA (>2 group comparisons) for continuous variables. NT‐proBNP indicates N‐terminal pro‐B‐type natriuretic peptide. Data from the Vizient Clinical Data Base were used with the permission of Vizient (all rights reserved).Throughout their hospitalization, patients with myocarditis were more likely to receive any vasoactive medication (82.6% versus 80.6%; P=0.007) with more inotropic support (56.9% versus 41.3%; P<0.001), but a similar proportion of any vasopressor use (72.0% versus 72.6%; P=0.43), (Table 4). Specific vasoactive medication use is shown in Figure 2. Vasoactive medication use was similar in both groups during the first day of admission (46.6% versus 45.5%; P=0.21), but it was higher in the myocarditis group during the subsequent days of the hospitalization (50.5% versus 43.9%; P<0.001).Early Hemodynamic Support Use in Patients With Cardiogenic Shock Presenting With Versus Without MyocarditisP values were calculated using Pearson’s χ2 test for categorical variables and t‐tests (two‐group comparisons) or ANOVA (>2 group comparisons) for continuous variables. ECMO indicates extracorporeal membrane oxygenation; IABP, intra‐aortic balloon pump; LVAD, left ventricular assist device; and MCS, mechanical circulatory support.The sum of MCS is higher than that of IABP and Impella combined because certain patients require >1 MCS. Data from the Vizient Clinical Data Base were used with the permission of Vizient (all rights reserved).CABG indicates coronary artery bypass grafting; CV, cardiovascular; CS, cardiogenic shock; ECMO, extracorporeal membrane oxygenation; IABP, intra‐aortic balloon pump; LHC, left heart catheterization; LVAD, left ventricular assist device; MCS, mechanical circulatory support; PAC, pulmonary artery catheterization; and PCI, percutaneous coronary intervention. Data from the Vizient Clinical Data Base were used with the permission of Vizient (all rights reserved).In addition to differences in vasoactive medication use, patients with myocarditis‐associated cardiogenic shock were more likely to receive any MCS device during their hospitalization (35.2% versus 19.3%) and MCS early on day 1 of admission (16.2% versus 9.2%) with a higher proportion of each device, including IABP, Impella, and ECMO with higher escalation from 1 MCS device type to another (all P<0.001). They were also more likely to undergo durable therapies for advanced heart failure, including LVAD (4.5% versus 2.3%) and heart transplantation (5.1% versus 1.8%) (both P<0.001; Figure 2 and Table 4).We found significant differences in other critical care therapies between groups. Patients with myocarditis‐associated cardiogenic shock were more likely to undergo left heart catheterization (33.1% versus 26.0%) but less likely to undergo percutaneous coronary intervention (2.0% versus 12.1%) or coronary artery bypass grafting (0.8% versus 7.5%) (all P<0.001). They also underwent a considerably higher amount of pulmonary artery catheterization (49.5% versus 26.3%; P<0.001). In terms of noncardiac critical care therapies, the myocarditis group also required invasive mechanical ventilation (52.3% versus 46.2%) and renal replacement therapy (17.4% versus 14.6%) more frequently (both P<0.001; Figure 2). Patients with myocarditis were also more likely to receive corticosteroids during their hospitalization (P<0.001) (Table 4).Among patients not transferred to another facility (95.1%), we found that the unadjusted in‐hospital death of patients with cardiogenic shock who had myocarditis was 29.4%, compared with 34.6% among those without myocarditis (P<0.001). After augmented inverse probability weighting adjustment (SMD found in Table S2), myocarditis was associated with a higher mortality rate, with a +9.8% (95% CI, 4.0–15.5; P=0.001) increased adjusted mean mortality rate. Patients with myocarditis had a more extended hospital stay, with a mean of 19.4 (±25.1) versus 14.2 (±19.3) days (adjusted incidence rate ratio, 1.12 [95% CI, 1.11–1.13]; P<0.001), and a longer ICU stay with a mean of 11.2 (±17.8) versus 7.5 (±18.6) days (adjusted incidence rate ratio, 1.16; 95% CI, 1.15–1.18, P<0.001), respectively. Myocarditis was associated with a higher total hospital cost to produce care with a mean of $117 958 (±171 446) versus $74 493 (±119 303) in patients with versus without myocarditis, which persisted after multivariable adjustment (mean cost difference, $15 826 [95% CI, 11 913–19 738]; P<0.001) (Table 5).Adjusted Primary and Secondary OutcomesICU, intensive care unit; and IRR, incidence rate ratio. Data from the Vizient Clinical Data Base were used with the permission of Vizient (all rights reserved).In stratified analysis of in‐hospital death by admission source, among nontransferred patients, the adjusted mean in‐hospital death was higher in the myocarditis‐associated cardiogenic shock group by +12.3% (95% CI, 6.7–18.0; P<0.001). Among transferred patients, the adjusted in‐hospital mortality rate was not statistically significant (mean difference, −3.4% [95% CI, −11.9 to 5.1]; P=0.43).To explore the difference in outcomes at centers with heart transplant or LVAD placement capabilities, we identified 290 567 patients (49.2%) who were treated at a center performing at least 1 transplant and 365 782 patients (61.9%) at a center performing at least 1 transplant or LVAD implantation within our cohort, respectively. The adjusted mean in‐hospital mortality rate among patients with myocarditis at centers where patients did not undergo heart transplants or LVAD implantations was +9.3% (95% CI, 0.8–17.8; P=0.03), +8.6% (95% CI, 3.0–14.1, P=0.003) at centers performing both transplants and LVAD placements, and +7.8% (95% CI, 1.5–14.0; P=0.02) at transplant‐only centers (Figure S1). When examining in‐hospital death across centers, we observed minimal variation in adjusted in‐hospital mortality rates with a median odds ratio (OR) of 1.02 (P=0.50), indicating highly consistent mortality rates. In contrast, heterogeneity in cardiovascular interventions was observed across sites, especially with the use of any MCS (median OR, 1.39; P<0.001) and pulmonary artery catheterization (median OR, 1.56; P<0.001). In addition, we found moderate variability in use of early corticosteroids by center (median OR, 1.43; P=0.001) and that early corticosteroid use was associated with an increased mortality rate (adjusted OR, 1.90 [95% CI, 1.54–2.34]; P<0.001) in patients with myocarditis‐associated cardiogenic shock.After multivariable adjustment, day 1 versus day 2 MCS was not associated with a lower odds of mortality (OR, 0.87 [95% CI, 0.57–1.33]; P=0.52) in patients with myocarditis. Interaction testing between any MCS (or each individual device) and myocarditis diagnosis was not statistically significant (all P>0.05) except for ECMO (P<0.001), suggesting that despite a higher mortality rate in patients requiring ECMO, ECMO use was associated with a lower mortality rate in patients with myocarditis compared with those without myocarditis. However, patients with myocarditis were more likely to undergo device escalation (adjusted OR, 1.46 [95% CI, 1.16–1.84]; P=0.001).Finally, we individually compared outcomes for patients with a primary diagnosis of AMI and then a primary diagnosis of heart failure, each compared with patients with myocarditis. Patients with myocarditis had a statistically significant in‐hospital mortality rate compared with patients presenting with a primary diagnosis of AMI (adjusted mean mortality rate, +13.6% [95% CI, 2.9–24.4]; P=0.01) or heart failure (adjusted mean mortality rate, +19.2% [95% CI, 9.9–28.5]; P<0.001]).

    Discussion

    In this nationwide multicenter study, we evaluated the characteristics and clinical outcomes of a large cohort of patients presenting with cardiogenic shock and found that 0.5% of patients had concomitant myocarditis. Despite being younger with fewer comorbidities, patients with myocarditis‐associated cardiogenic shock had a significantly higher adjusted in‐hospital mortality rate compared with those without myocarditis. They also experienced longer hospitalizations, longer ICU stays, and substantially higher total hospital costs. Furthermore, we found that patients with myocarditis were more likely to receive any vasoactive medication, MCS, or durable congestive heart failure therapies such as LVAD and heart transplants. Overall, our findings highlight the need for further data to guide management of this relatively younger patient population, which requires increased use of vasoactive and mechanical hemodynamic support.The younger age and lower burden of comorbidities in the myocarditis cohort were expected and consistent with what has been reported in the literature. In prior reports, the mean age of patients presenting with acute myocarditis was typically <50 years, similar to our cohort.12, 13, 14, 27, 28, 29, 30 Risk factors precipitating myocarditis include viral infections and certain anticancer medications (ie, immune checkpoint inhibitors).4, 5, 7, 8 Patients with myocarditis usually have lower rates of age‐related comorbidities, such as coronary artery disease.13, 28 Notably, our myocarditis cohort exhibited a higher prevalence of cancer and rheumatic disease. Immune cell activation against cardiomyocytes is a hallmark of myocarditis, which is usually more prevalent in patients with autoimmune rheumatic diseases6 or among those receiving anticancer medications,31 especially immune checkpoint inhibitors, which are being increasingly used.8, 9 Previous reports showed that the prevalence of concurrent malignancy in myocarditis increases with age, affecting ≈1.2% of patients aged 20 to 39 years, 6.2% of those aged 40 to 59 years, and 11.1% of those aged 60 to 79 years.13 Despite patients with myocarditis being younger compared with those without myocarditis, they exhibited a higher prevalence of cancer. This may be related to the higher frequency of anticancer medication use that may have increased the risk for myocarditis, thereby increasing the representation of patients with cancer in the myocarditis cohort. For instance, the number of patients eligible for immune checkpoint inhibitors is rising; therefore, patients are receiving more immune checkpoint inhibitors.32 Overall, we found that myocarditis‐associated cardiogenic shock occurred in a distinct patient population from those without myocarditis, notably with a lower comorbidity burden and higher prevalence of cancer and rheumatologic diseases.We observed that myocarditis‐associated cardiogenic shock was associated with a higher adjusted in‐hospital mortality rate. There are limited data on mortality rates in myocarditis‐associated cardiogenic shock. In a study of 1145 patients with myocarditis requiring ICU care, the in‐hospital mortality rate was close to 15%, with higher rates reaching 20% and 39% in patients requiring IABP or ECMO, respectively.14 Although patients required ICU care, it is unclear whether all patients had cardiogenic shock, given the much lower rates of mechanical or medical hemodynamic support.14 In contrast, in a study evaluating patients with fulminant myocarditis, ≈28% of patients died or received a heart transplant within 60 days of their presentation,29 which is closer to the mortality rate observed in our study. Unlike our study, neither included a comparison cohort of patients without myocarditis. In addition, it is important to note that the unadjusted mortality rate was lower in the myocarditis group, and that this association reversed after multivariable adjustment, driven predominantly by differences in age and comorbidities. This was further reflected by the significantly longer hospitalizations and ICU stays that required higher costs of care. Furthermore, patients with myocarditis also exhibited higher adjusted mortality rates than those with a primary diagnosis of AMI and heart failure in our cohort. Collectively, our findings highlight the clinical and economic burden of myocarditis complicated by cardiogenic shock and the importance of considering treatment algorithms directed to this high‐risk patient subgroup.We found significant differences in critical care therapies, such as vasoactive medication use, between those presenting with and without myocarditis. The choice of critical therapies in cardiogenic shock is not clearly defined.33, 34 In patients presenting with cardiogenic shock, norepinephrine remains among the first vasopressor choices.35 In our cohort, a similar number of patients received norepinephrine in both subgroups, which was higher than reported in a prior study.14 In contrast, patients with myocarditis were more likely to receive inotropes such as dobutamine or milrinone. We also found higher use of MCS in the myocarditis group in our cohort in up to 35% of patients compared with previous reports at 26%.14 This underscores the severity and unique challenges that arise. Although MCS devices were more frequently used in patients with myocarditis, the benefit of their use or the timing of these interventions in patients with myocarditis remains to be determined during the management of myocarditis‐associated cardiogenic shock. Notably, we found that patients with myocarditis were more likely to undergo device escalation but did not find evidence that the effect of MCS differed by myocarditis status, except for ECMO use, which was associated with an overall higher mortality rate in the general cohort, but this was attenuated in patients with myocarditis.We found that outcomes were relatively uniform across centers. However, there was a substantial difference in treatment approaches. For instance, corticosteroid use was variable across centers, and earlier use was associated with worse outcomes despite adjusting for clinical covariates and hospital‐level clustering. However, this finding must be interpreted with caution. Corticosteroids are a cornerstone of therapy in giant‐cell, eosinophilic, and immune checkpoint inhibitor myocarditis, yet their role in viral myocarditis remains controversial. Prior studies have suggested both potential harm (eg, increased viral load, prolonged inflammation, reduced left ventricular ejection fraction) and safety without adverse impact on prognosis.36, 37 Importantly, the observed association may reflect confounding by severity, as corticosteroids are often initiated in relatively sicker patients. This underscores the heterogeneity of myocarditis and the need for prospective data stratifying patients by myocarditis subtype to guide immunomodulatory therapy in this population. These findings warrant further standardization of therapeutic strategies as more evidence‐based guidelines emerge for managing myocarditis‐associated cardiogenic shock.There are several limitations to our work that need to be considered. First, the retrospective, observational design of our study should be evaluated primarily as an epidemiologic study including a majority of academic centers, and any clinical interventions should be viewed as hypothesis‐generating only. Second, given the reliance on administrative data and the inability to capture medication data before the hospital admission, we could not further characterize the type of myocarditis (ie, viral versus medication‐induced). In addition, reliance on ICD‐10‐CM codes is a limitation due to their variable positive predictive values for identifying myocarditis.38, 39, 40 Furthermore, distinguishing myocarditis from other acute overlapping cardiac conditions such as AMI can be challenging and depends on integrating multiple diagnostic tools (ECG, echocardiography, cardiac magnetic resonance imaging, or biopsy), which is not fully captured in our data set. Our database includes only inpatient data; therefore, we cannot identify outpatient medications. Additionally, biomarkers such as troponin and creatine kinase were not consistently available across hospitals in our database, limiting our ability to compare myocardial injury between groups. Finally, the inability to account for unmeasured confounders is likely despite adjustment with augmented inverse probability weighting. In particular, not all covariates (eg, age) reached the ideal SMD cutoff of <0.10.In conclusion, to our knowledge, this is the largest and most extensive study of patients with myocarditis‐associated cardiogenic shock. Although myocarditis occurs in a relatively younger population with fewer comorbidities, these patients require more critical care and hemodynamic support therapies, in addition to immunomodulatory therapies like corticosteroids. Our findings shed light on a largely unexplored area that warrants further investigation and highlight the need for clinical risk stratification and the development of specific evidence‐based approaches to guide treatment choices for this unique patient population.

    Sources of Funding

    None.

    Disclosures

    None.

    Supporting information

    Tables S1–S3Figure S1
    ASDr. Andi Shahu

    Dr. Andi Shahu

    Cardiology · Cardiac Critical Care
    Columbia University Irving Medical Center
    andi