Frequent Cardiovascular Manifestations Associated With SARS Cov-2 Infection: Experience at a Tertiary Hospital In Cali, Colombia

,


Introduction
Coronavirus disease 2019 (COVID- 19) is caused by the novel severe acute respiratory syndrome coronavirus 2 (SARS CoV-2).To date, more than 700 million confirmed cases and more than 6 million deaths due to secondary complications have been reported.Moreover, it was declared a pandemic in March 2020, 1 thus constituting one of the greatest threats to humanity in recent centuries.
Interaction of the virus with the host occurs through the S protein (spike) of the virus and its entry receptor angiotensinconverting enzyme 2 (ACE 2); however, invasion requires priming of a cellular serine protease called Transmembrane Serine Protease 2 (TMPRSS2). 2This receptor is present in various cells of the body, mainly in type II pneumocytes, and is also present at the systemic level, including in the heart and blood vessels, which are organs frequently involved in the complications presented during the infection. 3

Procedure
The study included subjects with a diagnosis of COVID-19 infection confirmed via reverse transcription-polymerase chain reaction (RT-PCR), presence of IgG and IgM type antibodies, or presence of antigens for SARS CoV-2, who also had an electrocardiogram upon admission to the institution.Patients with a known history of cardiac arrhythmia, The mechanisms of cardiovascular complications have been explained by a series of pathophysiological events, which initially occur at the respiratory level but finally occur at the multisystemic level, thereby altering cellular processes, causing modifications in the cardiac excitation and contraction coupling processes, and consequently increasing the potential arrhythmogenic effect. 4,5This is an event that occurs much more frequently in individuals with cardiovascular risk factors, such as high blood pressure (HBP) and diabetes, and pre-existing arrhythmias and among those aged >55 years.These patients also show an increased need for treatment in the intensive care unit (ICU). 6,7rrently, in Colombia and Latin America, knowledge about cardiovascular involvement related to these infections is gaining visibility, in part due to improved understanding of the disease.Early electrocardiographic findings are increasingly described as markers of poor prognosis, such as correlations of QT interval prolongation, higher incidence of recent sustained arrhythmias, and the presence of myocardial injury, all of which are associated with worse outcomes. 8,9Highlighting these findings, this study aimed to identify the frequency of electrocardiographic changes and cardiac arrhythmias, sociodemographic and clinical characteristics, and factors associated with mortality in patients hospitalized with COVID-19 infection in a tertiary care hospital.hospital stay of <24 h, or evidence of cardiorespiratory arrest during their initial care were excluded from the study.
Poor prognosis, comorbidities, and superinfection variables for the disease were defined according to the recommendations of the Colombian consensus on the care, diagnosis, and management of SARS CoV-2/ COVID-19 infection in healthcare institutions. 10The following were used as cutoff points: lymphocyte count < 1000 × 10 3 /µL; ferritin > 1000 ng/ml, lactate dehydrogenase (LDH) > 350 IU/L, D-dimer > 1000 mcg/ml, and procalcitonin > 0.5 ng/mL; while a PaO 2 / FiO 2 ratio <100 indicates severe hypoxemia, 100-200 indicates moderate hypoxemia, 200-300 indicates mild hypoxemia, and a ratio ≥300 indicates no hypoxemia.Presence of hyperkalemia or hypokalemia was defined as values >5.5 mEq/l or <3.5 mEq /l, respectively; transaminitis with increased levels of aminotransferases was defined as levels >3 times the normal upper limit, and a cutoff point of 47.34 ng/l was used for the 99 th percentile (99p) of troponin (ADVIA Centaur TNIH) -values above this cutoff are considered indicative of myocardial injury.
During the course of SARS CoV2 infection, any history of illness prior to diagnosis was considered as comorbidity, while any clinical picture suggestive of an infectious process more than 48 hours beyond hospital admission, confirmed by cultures, was defined as superinfection.
Only the first in-hospital electrocardiographic study performed in subjects admitted with a suspected diagnosis of COVID-19 was analyzed, in order to avoid possible confusion with patients admitted for reasons other than COVID-19 infection or cases in which COVID-19 infection was acquired within the hospital.
The electrocardiogram readings were obtained by two medical specialists in clinical cardiology with more than 10 years of experience and who were unaware of the patients' clinical diagnoses.In cases of disagreement regarding identification of the main finding, diagnosis was made by a third physician specialized in clinical cardiology.
Normal tracing was defined as no significant findings after a reading based on the recommendations for standardized electrocardiogram interpretation, including specific considerations such as significant ST-segment elevation when >1 mm in at least two contiguous leads, except v2-v3. 11

Data analysis
A descriptive analysis of the variables was conducted for the overall population as well as by survivors and non-survivors.For categorical data, this analysis used frequency distributions and relative and absolute frequencies.For quantitative data, a numerical analysis of measures of central tendency (mean) and dispersion (standard deviation) was conducted, followed by their respective categorizations.The Kolmogorov-Smirnov normality test was used to assess the normality of the data.To evaluate the differences in means, the unpaired Student t test and the Chi-square test for difference in proportions were applied, with their respective p values.Statistical significance was established as p values lower than 0.05.
Finally, multivariate analyses were conducted to determine the risk factors associated with mortality in the study population.Explanatory or independent variables were converted to dichotomous categories if necessary.According to statistical criteria, independent variables with p ≤ 0.25 in the univariate models were introduced into the initial logistic model with crude OR, followed by stepwise-backward elimination with an entry probability of <0.10 and a withdrawal probability of >0.25, which generated the final explanatory model with adjusted OR.All estimates were established with a 95% confidence level (CI) and a 5% significance level.In addition, the Hosmer-Lemeshow goodness-of-fit and verisimilitude tests were performed to validate the model's assumptions.The data analysis was performed using STATA 16® statistical program.

Results
The selection process and demographic data, clinical characteristics, and severity markers at admission as well as diagnostic aids, therapies, and complications of the subjects who met the inclusion criteria are described in Figure 1 and Table 1, respectively.Overall, the mortality rate was 32.8%, reaching 78.3% in those who required ICU management (53%); the most frequent rhythm of arrest was asystole, which was reported in 96.3% of the non-survivors.Several characteristics were not statistically significant between survivors and non-survivors.Among nonsurvivors, age, particularly age >75 years, history of cancer, and cardiovascular disease showed statistically significant relevance.The central figure shows the main results of this study.

Complications and severity markers
Complications occurred more frequently among nonsurvivors than among survivors, with a greater need for mechanical ventilation, along with higher incidence of acute kidney injury and myocardial injury.Diagnosis of superinfection was also more frequent in this group of patients and was associated with administration of antibiotic regimens lasting more than 5 days.The most frequently observed severity biomarkers in deceased subjects were lymphocyte count <1000 × 10 3 /µL, D-dimer level >1000 µg/ml, and elevated troponin levels >99p.Elevated troponin levels were significantly more frequent in subjects admitted to the ICU compared to those not admitted (31.8% vs 10.5%, p < 0.001)

Cardiovascular compromise and myocardial injury
Electrocardiographic findings did not differ between the two groups (p > 0.05); 15% of patients presented sinus bradycardia; changes compatible with significant ST-segment elevation occurred in 6.25% of cases; and 30% had troponin values above the 99p, which suggests possible acute myocardial infarction (AMI) with STsegment elevation.The most frequent rhythm disorder identified was atrial fibrillation.in 8.8%, considered to be the patient's first diagnosis, given the study's exclusion criteria.There was no evidence of QT interval prolongation in this population (Table 2).These electrocardiographic findings were more frequent and significantly evident when serum potassium and ferritin levels were <3.5 mEq/L and >1000 ng/mL, in 3.5% (p = 0.037) and 46.1% (p = 0.018) of the subjects, respectively.
The frequency of myocardial injury was significantly higher in patients with abnormal electrocardiographic findings compared to those without (26.5% vs. 12.1%, p = 0.032) F i g u r e 2 i l l u s t r a t e s t h e b e h a v i o r o f t h e main biomarkers in subjects with abnormal electrocardiographic findings.
Significant ST-segment changes and troponin elevation were present in 7% of the subjects.The multidisciplinary management team considered presence of myocardial injury associated with infection as the first possibility, ruling out AMI or pericarditis and taking into account the absence of symptoms; 33% of these subjects had an echocardiographic study, half of them with LVEF < 40%.However, since they had a history of previous ischemic heart disease, none required specific treatment.
Additional studies such as echocardiography were performed in about a quarter of the patients admitted, while computed tomography (CT) angiography was much less frequent, being positive for pulmonary embolism in 21% of all studies performed.Analysis of the results of the multiple logistic regression model enabled identification of individuals aged >75 years, those with a history of comorbidities (at least one), complications such as superinfection, and need for mechanical ventilation in days (> 2 days) as risk factors with statistical significance for a mortality event.The variability in hospital mortality events can be explained by these variables included in the final model in 33.4% of cases (PseudoR 2 ) (Table 3).Assessment of the final logistic regression model revealed a good fit to the data (Hosmer-Lemeshow) with a 5.5 (p = 0.86) Chi 2 (Degrees of freedom = 5).Differences between the expected and observed frequencies within the risk quantiles could not be confirmed; in other words, the observed and expected values were similar.Classification of the model indicated that it has a high ability to rule out patients who have no probability of mortality, with sensitivity and specificity of 74% and 83.6%, respectively.The probability of identifying a non-survivor among those predicted by the model was 67.27%.The probability of identifying a survivor among the patients who would not die according to the model was 87.6%.

Discussion
The impact on the cardiovascular system generated by COVID-19 infection has been evident since the beginning of the pandemic and is currently considered one of the main extrapulmonary manifestations, increasing the risk of arrhythmias and myocardial injury associated with worse clinical outcomes. 12ong these outcomes, hospital mortality has varied over time according to the epidemiological behavior of the pandemic, which is associated with the scant knowledge of a new disease and the changes in the incidence of cases that are still observed.In addition, mortality is associated with confirmed myocardial injury, which has been reported to be up to 10 times higher. 13,14n this regard, our study found a mortality rate that is 3.4 times higher than the general mortality of 32.8%, slightly exceeding the frequency of deaths reported in the literature (16%-30%). 15,16On the other hand, the post-vaccination era has generated significant changes in the behavior of the disease, mentioning complications such as myocarditis, for which a sevenfold higher risk is described in unvaccinated subjects. 17e need for ICU management was strongly associated with worse outcomes; however, these can vary if risk factors, such as age, comorbidities, and the presence of complications are considered. 18In this study, the proportion of patients who required an ICU stay was 53%.Among these, mortality related to myocardial injury was two times higher, which is statistically significant (p = 0.01).These results were demonstrated by the logistic regression model, in which variables such as age and the need for mechanical ventilation are significantly associated with a higher risk of mortality (OR > 1).N o t a b l y , b o t h m y o c a r d i a l i n j u r y a n d electrocardiographic changes have gained significant clinical importance, owing to their increasing association with hospital mortality, with a reported prevalence of up to 22% in this type of population. 19le sex is one of the factors reported as having the highest risk of mortality among subjects with COVID-19 infection. 20,21In our study, the number of men was 2 times higher than that of women, with no significant difference found in this regard.Furthermore, age has been considered an incremental risk factor for mortality.][24] This finding is consistent with that reported in our study, where a higher frequency of arterial hypertension was found among the deceased, followed by other cardiovascular antecedents and cancer.
Severe COVID-19 has been reported in up to 23% of cases 25 and is determined by multiorgan involvement.Kidney injury also represents an important factor; kidney involvement may be associated with greater cardiovascular dysfunction and the need for additional therapies.Acute kidney injury can vary and has been reported in 7%-30% 26,27 of cases.Our study found a high frequency, at 25%, with the need for renal replacement therapy in 17% of individuals.
Among the electrolyte changes, those related to potassium represent a prevalent factor during COVID-19 and are associated with greater severity; these are mediated by multiple pathophysiological mechanisms, with known repercussions for the cardiac conduction system. 28,29Our study evidenced significant electrocardiographic changes with serum potassium levels of <3.5 mEq/L in 3.5% of cases (p = 0.037).Knowledge of cardiovascular involvement is increasing every day, including information recently described in Latin America and Colombia on the identification of arrhythmias or electrocardiographic findings as risk predictors, 30 such as identification of prolonged QT interval in up to 10% of patients at hospital admission, associated with almost double the risk of mortality. 8n this sense, a 12-lead electrocardiogram enables early detection of findings, which are representative of cardiovascular system dysfunction.Besides, it constitutes an accessible, rapid, and cost-effective test for the care of patients with COVID-19. 31Although these findings have not been fully established and may be variable, their presence represents the possibility of various complications, whether transient or permanent, including lethal arrhythmias. 32thin the literature, reports of these manifestations have been changing from a 94% frequency of normal electrocardiographic findings in cases at the beginning of the pandemic, 33 to abnormal findings in 26%-37% of cases over the course of the pandemic, given the greater knowledge regarding cardiovascular system compromise. 34,35These were found mainly in critically ill subjects, among whom rates reached 90%. 36Our study found the frequency of these abnormalities to be 64%, with sinus tachycardia and bradycardia being the disorders with the highest rates.
The incidence of arrhythmias detected only using electrocardiography has been reported as approximately 16%-20% of cases and the vast majority of them were of supraventricular origin, 37,38 with atrial fibrillation present in 3%-18% of the cases.Our study identified arrhythmias in 8.8% of the cases, with a higher risk of incidence in the presence of concurrent multiple comorbidities, thus increasing the risk of mortality; 39,40 findings that are very similar to contemporary Latin American studies. 9rrent studies confirm the presence of these findings using telemetry devices, identifying at least one episode of arrhythmia in 73% of cases analyzed, with a possible increase during the evolution of the disease. 41her findings, such as changes in the ST segment, have also been reported during COVID-19, either mimicking an acute coronary syndrome or as a manifestation of other cardiovascular complications, including myocarditis or pericarditis. 42other important aspect is use of pharmacological therapies with effects on the cardiovascular system, such as macrolides and hydroxychloroquine, which have been reported in our study in 46.5% and 11.5% of cases, respectively, compared with the results of other studies that reported frequency of use of up to 80%. 43,44hese therapies are currently not recommended, as their benefit on the course of the disease 45 has not been proven; however, they do increase the prolongation of the QT interval produced by the infection itself, which has been reported in approximately 20% of the cases, 46,47 a finding that has not been evidenced in our study.Regarding use of antibiotics, which has been reported in approximately 50% of cases, 48 mainly associated with respiratory system compromise due to the COVID-19 infection, this investigation observed a higher frequency (65%) and the most common indication was suspected bacterial superinfection, significantly affecting survival.Despite this, the coinfection rate reported in the literature is variable, ranging from 5% to 10%, 49 which could increase the risk in the local bacterial epidemiological profile.

Limitations
This study has some limitations: First, it was conducted retrospectively, which is the reason why performing electrocardiogram at admission was not parameterized; in addition, the high risk of infection at the beginning of the pandemic should be taken into account and therefore the number of patients was lower than the total number of subjects with COVID-19 infections.Second, the electrocardiographic analysis presented corresponds to hospital admission, so it is not possible to confirm persistence of the initial findings reported, which could even be exacerbated in the context of other factors including pharmacological therapies used.However, we describe the acute electrocardiographic changes, including presence of arrhythmias such as atrial fibrillation which had not been previously diagnosed, although it is true that these could constitute existing asymptomatic arrhythmia.Third, exclusion of patients with known arrhythmia limits the description of the behavior of the event in this population.

Conclusions
Patients with SARS CoV2 infection may present a heterogeneous group of complications, with cardiovascular complications being a frequent finding.This study describes presence of initial electrocardiographic changes in this population, including newly diagnosed arrhythmias such as atrial fibrillation and the presence of myocardial injury.Moreover, need for invasive mechanical ventilation for more than 48 hours, presence of comorbidities, superinfection, and age greater than 75 years were associated with mortality.Further studies are required to evaluate the behavior of these events over the medium and long terms, but they justify the use of costeffective tools such as electrocardiograms as part of the initial approach and follow-up of patients with COVID 19.

Figure 1 -
Figure 1 -Flowchart illustrating patient recruitment and analysis process.

Figure 2 -
Figure 2 -Behavior of the main biomakers in relation with electrocardiographic findings.*p value < 0.05.Source: Prepared by the autors.

Table 3 -Adjusted OR of the logistic regression model for mortality
CI: confidence interval; OR: odds ratio.Source: Prepared by the authors.