Predictors of Mortality and Cardiovascular Outcome at 6 Months after Hospitalization for COVID-19

Clinical outcome data of patients discharged after Coronavirus disease 2019 (COVID-19) are limited and no study has evaluated predictors of cardiovascular prognosis in this setting. Our aim was to assess short-term mortality and cardiovascular outcome after hospitalization for COVID-19. A prospective cohort of 296 consecutive patients discharged after COVID-19 from two Italian institutions during the first wave of the pandemic and followed up to 6 months was included. The primary endpoint was all-cause mortality. The co-primary endpoint was the incidence of the composite outcome of major adverse cardiac and cerebrovascular events (MACCE: cardiovascular death, myocardial infarction, stroke, pulmonary embolism, acute heart failure, or hospitalization for cardiovascular causes). The mean follow-up duration was 6 ± 2 months. The incidence of all-cause death was 4.7%. At multivariate analysis, age was the only independent predictor of mortality (aHR 1.08, 95% CI 1.01–1.16). MACCE occurred in 7.2% of patients. After adjustment, female sex (aHR 2.6, 95% CI 1.05–6.52), in-hospital acute heart failure during index hospitalization (aHR 3.45, 95% CI 1.19–10), and prevalent atrial fibrillation (aHR 3.05, 95% CI 1.13–8.24) significantly predicted the incident risk of MACCE. These findings may help to identify patients for whom a closer and more accurate surveillance after discharge for COVID-19 should be considered.


Introduction
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is a highly transmissible and pathogenic beta-coronavirus responsible for the pandemic 'coronavirus disease 2019' (COVID- 19) [1]. Here, most of the available evidence is focused on patients' characteristics, risk factors, clinical course, and outcome in the acute phase of the infection, particularly among hospitalized cohorts [2][3][4][5][6][7][8][9]. Patients with COVID-19 usually present a respiratory syndrome, including interstitial pneumonia and acute respiratory distress syndrome. However, common complications are a prothrombotic coagulopathy, resulting in venous and arterial thromboembolic events, as well as acute liver or kidney injury and heart involvement characterized by myocarditis, acute coronary events, heart failure, and/or dysrhythmias [8]. To date, follow-up data of patients discharged after COVID-19 are limited [10][11][12][13][14][15] and, in particular, no study has specifically evaluated independent predictors of cardiovascular prognosis in this setting. Hence, the aim of this study was to prospectively assess 6-month mortality and cardiovascular outcome in a multicenter cohort of patients discharged after COVID-19 during the first wave of the pandemic in Italy.

Methods
Out of 549 patients admitted for COVID-19 in two Italian institutions-Maggiore della Carità Hospital, Novara and Santissima Annunziata Hospital, Chieti-from 20 February through 12 May 2020, we investigated clinical outcome during follow-up among 296 consecutive discharged patients (aged ≥18 years), representing 80% of those discharged alive. SARS-CoV-2 infection was confirmed by reverse-transcriptase-polymerase-chain-reaction assay in all patients. Individual in-hospital data, including demographics, previous medical history, co-morbidities, laboratory results, drug treatments, and clinical outcome, were collected. Patients were enrolled regardless of the type of COVID-19 clinical presentation and in-hospital therapies for the SARS-CoV-2 infection. After discharge, patients were prospectively followed up to 6 months. Follow-up assessment was performed by telephone interviews or ambulatory visits/in-hospital evaluation in the case of clinical recurrence.
The primary endpoint was all-cause mortality at 6 months. The co-primary endpoint was the incidence of the composite outcome measure including major adverse cardiac and cerebrovascular events (MACCE: cardiovascular death, myocardial infarction, stroke, pulmonary embolism, acute heart failure, or hospitalization for cardiovascular causes) at 6 months. The study protocol was approved by the institutional ethical committee (IRB code CE 97/20) and was conducted in strict accordance with the principles of the Declaration of Helsinki.

Statistical Analysis
The normality of distribution of the parameters was assessed by Kolmogorov-Smirnov test. Since all continuous variables had a normal distribution, they were described as mean ± standard deviation. Categorical variables were expressed as frequencies and percentages. One-way ANOVA test was used for group differences in continuous variables and Fisher exact test for group differences in categorical variables. The follow-up time was estimated as the time between hospital discharge and date of event or end of follow-up through 31 December 2020. Kaplan-Meier analysis for all-cause mortality and MACCE was performed. The Schoenfeld residuals test was used to check the proportional hazards assumption. Cox regression model was applied to estimate hazard ratios with a 95% confidence interval (CI). The Cox regression multivariable model was adjusted for age, sex, and those variables showing an association with p < 0.10 at the univariate model. There were no missing values in any of the outcomes. All calculations were performed using the Wizard 2 statistical software version 2.0.4 for Mac and Prism 9 (1995-2022 GraphPad Software, LLC, La Jolla, CA, USA). All tests were two-sided and a p value < 0.05 was considered statistically significant.

Results
The main characteristics of the study population at baseline (n = 296) are reported in Table 1. The mean age was 64 ± 16 years, and the prevalence of male sex was 58%. Women were more frequently smokers (p = 0.021), more frequently affected by chronic kidney disease (p = 0.032), atrial fibrillation (p = 0.01), and cognitive impairment (p = 0.003) compared with men.
As compared with those without events, patients with MACCE were significantly older; had a higher prevalence of peripheral artery disease, atrial fibrillation, chronic obstructive pulmonary disease, and chronic kidney disease; during the index hospitalization, suffered more frequently from deep venous thrombosis and showed higher neutrophilto-lymphocyte ratio and PaCO 2 ; presented a greater use of beta-blockers, diuretic agents, and anticoagulant therapy (Table 1). At multivariate analysis, female sex (aHR 2.6, 95% CI 1.05-6.52), prevalent atrial fibrillation (aHR 3.05, 95% CI 1.13-8.24), and in-hospital acute heart failure (aHR 3.45, 95% CI 1.19-10) were independent predictors of MACCE (Table 3). MACCE-free survival curves at 6 months according to tertiles of age, sex, prevalent atrial fibrillation, and in-hospital acute heart failure are depicted in Figure 2. MACCE after discharge occurred in 21 patients (7.2%). Crude rates of individual adverse events included in the composite cardiovascular outcome are reported in Table 2. As compared with those without events, patients with MACCE were significantly older; had a higher prevalence of peripheral artery disease, atrial fibrillation, chronic obstructive pulmonary disease, and chronic kidney disease; during the index hospitalization, suffered more frequently from deep venous thrombosis and showed higher neutrophil-tolymphocyte ratio and PaCO2; presented a greater use of beta-blockers, diuretic agents, and anticoagulant therapy (Table 1). At multivariate analysis, female sex (aHR 2.6, 95%CI 1.05-6.52), prevalent atrial fibrillation (aHR 3.05, 95%CI 1. 13-8.24), and in-hospital acute heart failure (aHR 3.45, 95%CI 1.  were independent predictors of MACCE (Table  3). MACCE-free survival curves at 6 months according to tertiles of age, sex, prevalent atrial fibrillation, and in-hospital acute heart failure are depicted in Figure 2.

Discussion
In this prospective, multicenter investigation we first provided 6-month follow-up data on mortality and cardiovascular morbidity among patients discharged after COVID-

Discussion
In this prospective, multicenter investigation we first provided 6-month follow-up data on mortality and cardiovascular morbidity among patients discharged after COVID-19 during the first wave of the current pandemic in Italy. We observed a mortality rate of 4.7% and a crude MACCE incidence of 7.2%. Age resulted as the sole independent predictor of all-cause death, whereas female sex, in-hospital acute heart failure, and prevalent atrial fibrillation were independent predictors of MACCE.
Evidence on clinical outcomes during follow-up of patients discharged for COVID-19 is scant. Two studies explored the persistence of symptoms at 2 months, showing at least one symptom, particularly fatigue and dyspnea, in 87% of patients with more severe COVID-19, and 68% of those with a non-critical disease, mainly anosmia/ageusia, dyspnea, or asthenia [12,13]. Other investigations reported a high incidence of residual impairment of pulmonary function and lung injury by computed tomography performed at 3 months after discharge in survivors of critical COVID-19 [14,15]. Furthermore, data on residual physical and functional impairment at 3 to 6-month follow-up [10], as well as on the persistence of psychological sequelae at 4 months [11], have been recently published. Readmission and death rate at 60 days was evaluated in the nationwide Veterans Affairs health care system, without analyzing organ-specific endpoints [16]. The largest study evaluating organ-specific dysfunction in individuals with COVID-19 after discharge included 47,780 English patients over a follow-up of 140 days, and observed an increased risk of mortality, readmission, and multiorgan dysfunction compared with similar individuals in the general population [17]. More recently, in a German cohort of patients hospitalized for COVID-19, 6-month all-cause mortality and readmission rates were related to coagulopathy, congestive heart failure, neurological diseases, and acute renal failure, while the female sex resulted in a protective factor [18].
The present study represents the first report specifically focused on independent predictors of mortality and cardiovascular outcome in survivors after COVID-19 hospitalization.
In 2019 the probability of annual death for individuals aged 64 years (i.e., the mean age in our study population) in Italy was 0.7% [19]. In patients discharged after COVID-19, we observed a crude mortality > 6 times higher over 6 months, with a cardiovascular event being the cause of death in 43% of patients. Importantly, in our investigation overall mortality after the hospitalization was unrelated to severity of COVID-19-related respiratory impairment at presentation, length of stay, or occurrence of in-hospital complications, also including the need for intensive care unit admission. Older age has been shown an independent predictor of lower in-hospital survival in patients with COVID-19 [9]. In particular, described in-hospital mortality overall ranges between 15% and 20%, but varies across decades of age and exceeds 60% in octogenarians [6]. This reflects frailty, prevalent co-morbidities, and higher rates of complications with aging. The present study indicates over 6 months an 8% age-related overall relative increase in all-cause death and a 10% absolute increase in mortality in the subgroup of patients with age in the highest tertile (>77 years).
Prevalent cardiovascular diseases are frequent in patients hospitalized for COVID-19 [6], but little is known about their incidence and prognostic significance after discharge. We demonstrated a not negligible overall incidence of cardiovascular events at 6-month follow-up. Atrial fibrillation is a common feature in patients hospitalized for COVID-19, partly because it shares with such disease a high prevalence of older age, cardiovascular risk factors, and co-morbidities, and partly because it represents a frequent new-onset complication. In these patients, atrial fibrillation has been reported in approximately 20% of cases (either historical or new-onset) [20], and such arrhythmia, especially newonset, resulted in an independent predictor of in-hospital all-cause death, cardiovascular death, and more severe clinical pattern [21]. It has been hypothesized that SARS-CoV-2 infection-related inflammation, edema, and fibrosis of atrial tissue, besides immune response, hypoxia, and electrolyte abnormalities, can contribute to the occurrence of atrial arrhythmias, in particular atrial fibrillation [20,22]. Notably, at multivariate analysis, we found that atrial fibrillation was associated with a three-fold increase of cardiovascular events at 6 months after discharge. Atrial fibrillation as a marker of increased cardiovascular risk, as well as a more severe cardiac impairment in patients with atrial fibrillation, may explain the excess in mortality related to this arrhythmia, either during the in-hospital stay or afterward during follow-up.
We also observed that an acute heart failure event during index hospitalization was independently associated with a 3.5-fold higher risk of MACCE after discharge. This may reflect an underlying cardiac impairment persisting over time and predisposing to further adverse events during follow-up. Unfortunately, we had no data on the specific causes of acute heart decompensation during the in-hospital stay and we cannot discern whether it occurred in patients with pre-existing cardiac diseases or was precipitated by new cardiovascular events, either spontaneous or related to SARS-CoV-2 infection, such as acute coronary syndromes, myocarditis, arrhythmias, respiratory failure, renal insufficiency, sepsis. A possible explanation for cardiovascular events occurring during the months after discharge is that inflammation and immune reaction persist for a longer period relative to hospitalization and continue to affect the cardiovascular system. On the other hand, clinical features and co-morbidities of COVID-19 patients may account for the increased cardiovascular risk.
Moreover, in our study female sex was an independent predictor of MACCE. This appears to be in contrast with a reported higher incidence of complications and mortality among male patients during the acute phase of SARS-CoV-2 infection [5][6][7][8][9]. Sex differences in both innate and adaptative immune systems, related to hormones and cytokines production, have been hypothesized to explain such survival advantage in women [23]. Indeed, a previous investigation found that the female sex was associated with a higher risk of respiratory sequelae at 4 months after discharge for COVID-19 [11]. To date, only one study showed a lower rate of all-cause death at 6-month follow-up in women compared to men [18] and no data are available on possible sex-related differences in terms of cardiovascular prognosis during follow-up in patients with COVID-19. We observed a 2.6-fold increased risk of MACCE at 6 months in female vs male patients that might likely be explained by an unbalanced distribution of frailty-related conditions, including chronic kidney disease, atrial fibrillation, and cognitive impairment, more frequently observed in women. Furthermore, the largely reported excess in-hospital mortality in men [5][6][7][8][9] could justify a relatively greater number of women at risk of suffering adverse events after COVID-19 hospitalization.
Our study has strengths and limitations. Strengths include the robustness of data obtained from a multicenter, real-life population with a wide spectrum of COVID-19-related clinical features, also including a severe pulmonary disease; the reliability of prospectively collected data with a comprehensive assessment of individual medical history, medical treatments, in-hospital outcome, and follow-up evaluation. Limitations include the risk of inclusion bias, despite the study aiming to enroll consecutive patients; residual confounding, due to the lack of adjustment for all potential confounders; the absence of information on B-type natriuretic peptides, d-dimer levels, and echocardiographic features at the time of discharge, as well as on specific causes of non-cardiovascular death at 6 months; and the follow-up assessment being performed by telephone interviews in a large proportion of patients. However, the latter was indispensable due to rigorous access restrictions in the hospital for all patients requiring elective cardiological visits during the COVID-19 pandemic in Italy.
In conclusion, this prospective, multicenter investigation first addresses the issue of cardiovascular outcome at 6 months in patients hospitalized for COVID-19. Our findings may help to detect patients at higher risk of adverse events after discharge for whom a closer and more accurate clinical and imaging surveillance should be considered.

Data Availability Statement:
The authors agree to make data and materials supporting the results or analyses presented in their paper available upon reasonable request.