Cardiovascular implications of coronavirus disease 2019: a systematic review

Protocol and registration

We followed the recommendations established by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement^7,8. The systematic review protocol was not registered due to the urgency of the matter and very limited available evidence on the topic.

Assessment of methodological quality and risk of bias

For quality assessment, we used the Newcastle-Ottawa scale (NOS). The total score was found to be 9 and all the studies are of high quality⁹.

Eligibility criteria

Published peer-reviewed articles from January 1, 2020 until June 15, 2020 which aimed to assess cardiovascular implications in patients with COVID-19 were included. Article language restrictions were not imposed. In addition to original articles, case reports, and case series were also included. Editorials, letters to editor, and correspondences were excluded.

Search strategy

We searched standard relevant publications indexed in PubMed, PubMed Central (PMC), EMBASE, and Google Scholar. Databases were searched using the search terms under the two search themes below, combined using the Boolean operator “AND”. For COVID-19, we used “Novel coronavirus”, “Novel coronavirus 2019”, “2019 nCoV”, “Coronavirus Disease 2019”, “COVID-19”, “Wuhan coronavirus,” “Wuhan pneumonia,” and “SARS-CoV-2.” For cardiovascular implications, we used “Cardiovascular implications”, “Acute myocardial injury”, “Acute cardiac injury”, “Arrhythmias”, “Heart failure”, and “Myocardial infarction”. A thorough review of the references revealed further relevant articles.

Study selection and data abstraction

First of all, we screened articles by title and abstract. Then, we examined full texts of relevant articles for inclusion and exclusion criteria. A data abstraction spreadsheet using Microsoft Excel version 2013 was developed, which included the following information: author, year of publication, journal, and country where the study was done, study design, sample size, baseline characteristics and laboratory parameters of the patients and cardiovascular implications of COVID-19. A third researcher checked the article list and data abstraction spreadsheet to ensure there were no duplicate articles. Any duplicated article was counted as a single article.

Study selection

A total of 315 articles were retrieved using the search strategy. After screening by abstract and title, 52 articles were selected for full-text assessment. Of these, 46 articles were excluded as they were not relevant to research aims and objectives (Figure 1).

Figure 1. PRISMA diagram detailing the study identification and selection process.

Study characteristics

All the studies were conducted in China in 2020 and all of them were original articles except for one case report, by Cui et al.¹⁰. Among the original articles, all were retrospective studies except study by Huang et al.¹¹, which was a prospective, observational study. All the studies have categorized patients into two groups: non-ICU/non-critical/without myocardial injury/with normal cardiac Troponin group and ICU/critical and severe/with myocardial injury/with elevated cardiac Troponin group. Critical and severe COVID-19 was defined as presence of one of the following conditions: respiratory failure requiring mechanical ventilation; shock; failure of other organs that requires treatment in ICU. Wang et al.¹² and Huang et al.¹¹ defined acute cardiac injury as blood levels of hypersensitive troponin I above the 99th percentile upper reference limit (>28 pg/mL) or new abnormalities shown on electrocardiography and echocardiography. Xingwei et al.¹³ defined myocardial injury as blood levels of myocardial troponin ≥3 times the upper reference limit (34.2 ng/L). However, Chen et al.¹⁴ and Guo et al.¹⁵ defined myocardial injury as serum levels of troponin above the 99th percentile upper reference limit (Table 1).

Baseline characteristics and co-morbidities of the patients

The median age of the patients in all the studies was above 50 years, except in the studies by Huang et al.¹¹ (median age, 49 years) and Cui et al.¹⁰ (patient’s age; 55 days). In all studies, patients requiring ICU/critical care or patients with myocardial injury/elevated cardiac troponin were older comparatively, except in study conducted by Huang et al.¹¹ (median ages of patients in non-ICU and ICU groups were same). Male predominance was seen in all studies except those by Cui et al.¹⁰ and Guo et al.¹⁵. All the studies have included the underlying co-morbidities of the patients (Table 2).

Cardiovascular complications associated with COVID-19

Wang et al.¹² established that among the 138 patients, acute cardiac injury was seen in n=10 (7.2%), shock in n=12 (8.7%), and arrhythmia in n=23 (16.7%) patients. ICU patients were more likely to have one of these complications than non-ICU patients (p < 0.001) [Table 3]. ICU patients demonstrated significantly high levels of creatine phosphokinase myocardial band (CPK-MB), cardiac troponin I (cTnI) and D-dimer. Similarly, the number of patients who had procalcitonin levels >0.05 ng/mL was higher in the ICU group compared with the non-ICU group (n=27 [75%] vs n=22 [21.6%]) (Table 4).

The study conducted by Huang et al.¹¹ found that among the 41 patients, acute cardiac injury was seen in n=5(12%) and shock in n=3 (7%) patients. Acute cardiac injury n=4 ([31%] vs n=1 [4%]) and shock (n=3 [23%] vs n=0 [0%]) were seen more commonly in ICU patients compared with non-ICU patients (Table 3). Besides, ICU patients demonstrated significantly high levels of creatine phosphokinase (CPK), and D-dimer. Number of patients who had procalcitonin levels >0.05 ng/mL and cTnI >28 pg/mL (99th percentile) were more in ICU group compared with non-ICU group (Table 4).

Similarly, Chen et al.¹⁴ observed that among the n=150 patients, n=22 (7.1%) had acute cardiac injury. In the same study, out of n=24 patients in the critical group, n=15 (62.5%) had acute cardiac injury and n=7 (5.6%) out of 126 patients in non-critical group had acute cardiac injury and the difference was statistically significant (p-value <0.001) (Table 3). Hypersensitive C-reactive protein (hs-CRP), N-terminal-pro brain natriuretic peptide (NT-proBNP) and cTnI levels of the patients were significantly higher in critical care cases than in mild cases (p-value <0.001) (Table 4). On univariate logistic regression analysis, critical disease status had a significant correlation with age, male gender, elevated NT-proBNP, elevated cTnI, elevated hs-CRP, hypertension, and coronary artery disease (all p<0.05). Multivariate logistic regression analysis revealed that elevated cTnI (OR=26.909, 95%CI 4.086–177.226, p=0.001) and coronary artery disease (OR=16.609, 95%CI 2.288–120.577, p=0.005) were the independent risk factors of critical disease status.

In study performed by Xingwei et al.¹³, among the n=54 severe/critically severe patients, acute myocardial injury was discovered in 24 (44.4%) patients while 26 (48.1%) patients died during hospital stay. The in-hospital mortality rate was significantly higher in the myocardial injury group compared to the group without myocardial injury [75.0% (18/24) vs. 26.7% (8/30), p-value = 0.001]. In myocardial injury group, C-reactive protein (CRP) and NT-proBNP levels were also found to be significantly higher than those without myocardial injury (all p-value<0.01)

Cui et al.¹⁰ described a case report of acute cardiac injury in an infant infected with COVID-19 as suggested by elevated plasma levels of CPK-MB, and cTnI.

Guo et al.¹⁵ observed that among the n=187 patients, n=52 (27.8%) exhibited acute cardiac injury as indicated by elevated cardiac troponin T (cTnT) levels. During hospitalization, arrhythmias were seen in n=11 (5.9%) patients. Patient with elevated cTnT levels had more frequent malignant arrhythmias compared with normal cTnT (n=9 [17.3%] vs n=2 [1.5%]). The overall mortality was 23% (43 patients died) (Table 4). The mortality rates were 7.62% (8 of 105), 13.33% (4 of 30), 37.50% (6 of 16), 69.44% (25 of 36) for patients without underlying cardiovascular disease (CVD) and normal cTnT levels, with underlying CVD and normal cTnT levels, without underlying CVD but elevated cTnT levels, and with underlying CVD and elevated cTnT levels, respectively. Elevated cTnT levels was seen more commonly in patients with underlying CVD compared with patients without CVD (36 [54.5%] vs 16 [13.2%]). A significant positive linear correlation of plasma cTnT levels with plasma hsCRP levels (β = 0.530, p< 0.001) and NT-proBNP levels (β = 0.613, p < 0.001) was also observed.

Limitations of the review

Our review has included limited number of available studies with small sample size. Hence, the result of our systematic review should be considered with caution. All studies included in our review were conducted in China and most of them were retrospective. Thus, our review lacks heterogeneous and diverse population. We recommend multicenter, multi-national, prospective studies be carried on cardiovascular implications of COVID-19. It is difficult to postulate myocardial injury as the sole cause of death as it may have occurred due to multi-organ dysfunction.

Underlying data

All data underlying the results are available as part of the article and no additional source data are required.

Reporting guidelines

Figshare: PRISMA checklist for ‘Cardiovascular implications of coronavirus disease 2019: a systematic review’. https://doi.org/10.6084/m9.figshare.12658409.v1⁸.

The completed PRISMA checklist is available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).