The role of cardiac troponin-I as a prognostic tool for mortality in patients hospitalized with COVID-19

Background

Coronavirus disease 2019 (COVID-19) has been associated with significant morbidity and mortality, with cardiovascular involvement being usual. Elevations in cardiac Troponin-I level has proposed as an independent biomarker for mortality among patients with COVID-19.

Aim

To evaluate the role of high sensivity Troponin-I (hs-TnI) level at hospital admission in predicting 30 day in-hospital mortality and 6-month mortality in patients hospitalized with a COVID-19 diagnosis.

Methods

We performed a retrospective single-center cohort study including consecutive patients aged 18 years and older who were admitted for COVID-19, during a 1-year period (n=818). We excluded patients with acute coronary syndrome (n=23), patients with acute heart failure (n=42), and patients in which hs-TnI level was not dosed at admission (n=163). Patients were divided into two groups according to hs-TnI levels: hs-TnI <19.8 vs hs-TnI ≥19.8 pg/mL. Primary outcomes were 30-day in-hospital mortality and 6-months mortality. According to the data distribution, appropriate statistical tests were conducted to compare independent samples. Multivariable logistic regression was used to analyze mortality risk. Receiver operator characteristics (ROC) curve and area under the curve (AUC) were obtained to determine the discriminative power of hs-TnI as a predictor of mortality. (Figure 1).

Results

This cohort included 590 patients. Mean age was 71 ≥±15 years and 52.4% were men. Overall, 209 patients (35.4%) had elevated hs-TnI levels and 381 patients had normal hs-TnI levels. Individuals in the hs-TnI ≥19.8 pg/mL group were older (80±11 vs 66±14 years, p<0.001) and presented higher prevalence of chronic heart failure (24.9% vs 7.1%, p<0.001), hypertension (77.0% vs 57.5%, p<0.001), atrial fibrillation/flutter (19.1% vs 5.5%, p<0.001), prior stroke (12.4% vs 5.2%, p=0.001) and ischemic heart disease (12.4% vs 3.7%, p<0.001). There was no difference in length of hospital stay between the groups (8.0 [IQR 9.6] in hs-TnI 19.8 pg/mL group vs 9.0 [IQR 8.0] normal hs-TnI group, p=0.669). Troponin-I was the only independent predictor of in-hospital mortality (OR 3.80, CI 95%: 2.44–5.93, p<0.001), see Table 1. The troponin levels had the highest area under the receiver operating characteristic curv (AUC) with an AUC of 0.705 (95% CI: 0.667–0.742, p<0.001) for association with the in-hospital mortality (figure 1). There was no difference in 6-months mortality between the two groups.

Conclusion

Acute myocardial injury is common in patients hospitalized with COVID-19. In the present study a TnI level ≥19.8 pg/mL was predictor of 30 days in-hospital mortality, suggesting that raised levels of this biomarker is associated with adverse prognosis. This tool might be useful for COVID-19 patient risk stratification. Further studies are needed to provide robust data and reliable recommendations on this theme.

Funding Acknowledgement

Type of funding sources: None.

Specialty-related asymmetries in treatment and outcomes of younger patients with atrial fibrillation

Funding Acknowledgements

Type of funding sources: None.

Background

Atrial fibrillation (AF) is the most common arrhythmia and one of the most frequent motives for presentation in the emergency department (ED). However, there are still significant inequalities concerning its treatment.

Purpose

We aimed to assess if management and outcomes of patients with AF differed according to the specialty to which they were referred at hospital discharge.

Methods

We conducted a retrospective single-centre analysis including adult patients admitted to the ED in 2016 with the International Classification of Diseases codes correspondent to AF. Patients older than 75 years and those whose electronic medical file was not accessible were excluded. Patient referral at discharge was not based on any pre-specified criteria but rather performed at the discretion of the physician in the ED.

We divided our sample into 3 groups according to whom the patients were referred at hospital discharge: A) patients referred to an electrophysiologist; B) patients referred to a general cardiologist; and C) those referred to a non-cardiologist (general practitioner or internist).

We assessed patient characteristics, delay to first appointment, treatment strategies as well as stroke and all-cause mortality at 3 years of follow-up.

Results

Of the 372 patients initially screened, 184 were excluded due to age over 75 and 85 patients due to inaccessible medical records, resulting in a final sample of 103 patients with a median age of 65 (58-71) years and median CHA2DS2-VASc of 2 (1-3).

At discharge from the index admission, one third of patients were referred to an electrophysiologist (group A), 41% to a general cardiologist (group B) and the remaining 26% to a non-cardiologist (group C). There were no significant differences in baseline characteristics or type of AF between the groups.

The delay from index hospital admission to the first appointment was superior in groups A and B compared with C (3 and 2 vs 0 months, p=0.050).

Patients referred to an electrophysiologist were more frequently started on a rhythm-control strategy compared with those referred to other physicians (71% vs 38% and 26%, p=0.004). Similarly, significantly more patients in group A underwent catheter ablation (47% vs 12% and 4%, p<0.001), around 2 years after the diagnosis (tendentially earlier than those in group B).

At 3 years follow-up, mortality was higher in group C compared with the other groups (15% vs 0 in group A and 2% in group B, p=0.019). Stroke rates did not differ between groups.

Conclusion

In this group of AF patients, referral to an electrophysiologist at discharge was associated with a higher rate of a rhythm-control strategy, including catheter ablation. In the long-term, those patients showed better survival. These results suggest that, in patients under 75 years, an earlier referral to an electrophysiologist might be beneficial.

Patients with reflex syncope have a higher prevalence of atrial fibrillation than general population which is predicted by the type of the tilt table test response

Introduction

Reflex syncope is the result of an imbalance between the sympathetic and parasympathetic nervous system. The autonomic nervous system has been associated to some forms of atrial fibrillation (AF).

Objective

The aim of this study was to assess the prevalence of AF in patients with reflex syncope and a positive tilt table test and to identify the type of tilt test response in patients with AF in comparison to patients without AF.

Methods

We retrospectively studied consecutive patients that underwent a tilt table test at our institution between 2016 and 2019. We selected those patients with a positive test. Patients with an implanted pacemaker at the time of the tilt test and patients followed in a different institution were excluded. Previous diagnoses of AF episodes were assessed. We analysed the clinical characteristics and the tilt table test results according to previous history of AF.

Results

We studied 49 patients with a positive tilt test. Seven (14.3%) patients had previously diagnosed paroxysmal AF at the time of the tilt table test. Patients with AF were older, had more frequently hypertension, and the tilt test response was more frequently a vasodepressor than mixed or cardioinhibitory response (71.4% vs. 28.6%) (Table). In univariate analysis, age and hypertension were associated with AF, respectively, OR 1.08 (95% CI 1.01–1.17), p=0.034 and OR 10.80 (95% CI 1.19–98.36), p=0.035. A vasodepressor response was also associated with AF (OR 6.25, 95% CI 1.06–36.74, p=0.043).

Conclusions

Patients with reflex syncope and a positive tilt table test had a higher prevalence of AF than the general population. A vasodepressor response was associated with AF as were age and hypertension, demonstrating the possible impact of the autonomic nervous system and the multifactorial nature of AF.

Funding Acknowledgement

Type of funding sources: None.

Rapid atrial fibrillation increases cardiac biomarkers: decision to perform coronary angiography based on novel high-sensitivity Troponin I peak

Funding Acknowledgements

Type of funding sources: None.

Background

Since January 2018 the availability of high sensitivity Troponin I (hsTnI) has improved ischemia diagnosis. In patients with rapid atrial fibrillation (AF), the decision to undergo coronary angiography is usually due to elevated cardiac biomarkers. However, evidence to support the rentability of this approach is sparse.

Purpose

Evaluate if hsTnI in patients with rapid AF and elevated cardiac biomarkers has a good discriminative power to predict a positive coronary angiography.

Methods

We retrospectively studied consecutive patients admitted to the emergency department (ED) between January 2018 and December 2019 with rapid AF that underwent coronary angiography and had multiple hsTnI values obtained. We analysed risk factors, initial and peak hsTnI, time from ED admission to peak hsTnI and ST-T segment abnormalities (ST depression and/or T wave inversion). We evaluated the presence of significant coronary artery stenosis with the need of revascularization at coronary angiography. Univariable and multivariable analysis was performed to obtain the Odds Ratio (OR, 95% CI, p-value) for significant coronary artery disease (CAD).

Receiver operator characteristics (ROC) curve and area under the curve (AUC) were obtained to determine the discriminative power of peak hsTnI as predictor of a positive coronary angiography. Optimal cut-point value was obtained (Youden index) and patients were divided according to this value.

Results

From 1407 patients admitted to the ED with rapid AF, 30 patients, 60% male, median age 74 (IQR 61.25-80.75) years, were submitted to coronary angiography. Significant coronary artery stenosis was present in 17 (57%) patients.

Age, ST-T segment abnormalities and peak hsTnI were predictors of significant CAD, respectively 1.203, 1.064-1.361, 0.003; 25.00, 3.522-177.477, 0.001; and 1.000, 1.000-1.001, 0.015.

Optimal cut-point value for predicting the presence of significant coronary artery stenosis at coronary angiography was a peak hsTnI of 359 pg/mL (AUC 0.869, p-value 0.001, 95% CI 0.742-0.995). The two groups with hsTnI < 359 and hsTnI >359 differed in age and ST-T segment abnormalities (see Table).

After adjustment, peak hsTnI >359 pg/mL was the only independent predictor of significant CAD (23.894, 1.310-435.669, 0.032).

Conclusion

In this group of patients with rapid AF, peak hsTnI >359 pg/mL was the only independent predictor of significant coronary artery disease. Therefore, those patients should undergo coronary angiography. Abstract Figure.

Optimal percentage of biventricular pacing to obtain CRT response: how high is high enough

Funding Acknowledgements

Type of funding sources: None.

Background

The greatest benefit with cardiac resynchronization therapy (CRT) is achieved when biventricular pacing (BivP) percentage (%) is close to 100%. However, in some patients that goal can be challenging to obtain.

Purpose

Determine whether a lower BivP% could lead to similar CRT response and events, as compared with patients with BivP% >98%.

Methods

Patients with CRT followed up in a remote-monitoring network were retrospectively analyzed. BivP% was assessed and response to CRT was defined as an absolute increase in left ventricle ejection fraction (LVEF) >5% or a relative increase in LVEF >15%.

Low BivP% was defined as <98%. Clinical, echocardiographic data and all-cause death during follow-up were evaluated. ROC curve and AUC were obtained to determine the discriminative power of BivP% as predictor of CRT response. Optimal cut-point value was obtained and patients were divided according to this value. Kaplan-Meyer survival function was used to compare survival in the different groups and the Log-rank test was used for comparison between the groups.

Results

88 patients, 76% male, median age 73.5 (IQR 65.75-79.25) years were included. A CRT-D was implanted in 69%. Etiology was ischemic in 44%. 93% were under beta-blockers. Median LVEF before CRT was 27% (IQR 20.25-32). 

44 patients (50%) had low BivP% (median 91%, IQR 96-99), 55% due to atrial fibrillation and 52% due to frequent premature ventricular complexes. After optimization of medical therapy, device programming and/or interventional procedures, we obtained a BivP >98% in 26 out of the 44 patients (59%). However, in 18 patients (20%) BivP% was <98% (median 95, IQR 92.25-96). 66% patients were CRT responders. Median follow-up was 36 (IQR 23.75-84) months. During follow-up, all-cause mortality was 27% (24 patients).

Optimal cut-point value for predicting CRT response was 91% BivP% (AUC 0.644, p-value 0.047, 95% CI 0.496-0.792). The characteristics of the two groups didn’t differ significantly (Table).

Survival was significantly higher in patients with BivP% >91% (Log-rank 3.667, p-value 0.050) – Figure.

Conclusion

In this population, BivP% >91% was sufficient to achieve CRT-response and was associated with a better survival. BivP% <91%(n = 4)BivP% >91%(n = 84)p-valueAge in years, median (IQR)72.50 (70.50-73.75)74.00 (65.00-80.00)0.666CRT-D, n (%)3 (75.0)58 (69.0)0.999Ischemic cardiopathy, n (%)3 (75.0)35 (41.7)0.311LVEF before CRT, median (IQR)27 (19-39)27 (20-32)0.795Beta-blockers, n (%)4 (100.0)78 (95.1)0.999Abstract Figure.

Clinical impact of transoesophageal echocardiography in acute brain ischaemia: who should we select?

Funding Acknowledgements

Type of funding sources: None.

Background

Stroke is a prevalent disease and is still the leading cause of death in Portugal. Transoesophageal echocardiography (TOE) is a sensitive test often performed to detect embolic sources. However, since its most common findings such as patent foramen ovale (PFO) and atheroma plaques do not necessarily mandate a change in treatment, there is still debate over its clinical impact in the context of brain ischaemia (BI) and which patients (pts) should be submitted to it. 

Purpose

To assess the clinical impact of TOE following BI and to identify clinical and diagnostic testing results that could help predict which pts benefit from it.

Methods

A retrospective study was conducted including all pts submitted to TOE in our hospital after acute BI in 2018 and 2019. Clinical and testing data (brain, vascular and cardiac imaging and 24h-Holter monitoring) was analysed and compared between 2 groups: the pts who had findings in TOE compatible with a source of embolism which resulted in a change in treatment ("relevant TOE" group) vs all other pts who had no such findings or whose findings did not result in change in treatment ("others"). Predictors of relevant TOE were also analysed.

Results

 Of the 87 pts (mean age of 57 and maximum of 83) included in the study, 51 (59%) had findings compatible with a potential source of embolism in TOE, PFO being the most common (n = 42). In only half of them did these findings result in a change in treatment (the relevant TOE group: n = 25; 29% of the overall population).

Age and other baseline characteristics did not significantly differ between groups. Pts with a relevant TOE presented more often with visual-field defects (32% vs 10%, p = 0.020) and were more likely to have visible acute lesions on brain imaging (96% vs 76%, p = 0.032) compared with the others. There was also a borderline significant association between the presence of infarct in the territory of the superior cerebellar artery and a relevant TOE (p = 0.054). On the contrary, the presence of significant lesions in extracranial arteries was negatively associated with a relevant TOE (p = 0.016).

Considering the whole population, there were no transthoracic echocardiography (TTE) predictors of a relevant TOE but when analysing only younger patients (age < 50), the presence of any abnormality in TTE became associated with a relevant TOE (OR 8.5, CI 1.1-63.9; p = 0.044). We found no predictors of relevant TOE in 24h-Holter results.

Conclusions

TOE commonly identified potential sources of brain embolism, which proved relevant in half the cases. In the impossibility of submitting all BI patients to TOE, this study suggests that brain and vascular imaging rather than age or other baseline characteristics may be useful in predicting a relevant result. Moreover, TTE does not seem to be an adequate screening method to select patients for TOE, except possibly in younger patients. Studies with larger samples are needed to confirm these results.

Acute coronary syndrome in COVID-19 times: could it be business as usual?

Funding Acknowledgements

Type of funding sources: None.

Background

In an attempt to cope with the COVID-19 pandemic, several measures of social distancing and reorganization of health care systems have been adopted. In Portugal, these included suspending elective clinical activity and a temporary national lockdown. Data from multiple centres, including central hospitals in Portugal, has shown a reduction in hospital admissions for acute coronary syndrome (ACS) during the COVID-19 outbreak. Fear of going to the hospital and physical inactivity have been pointed as possible explanations.

Purpose

to assess the impact of this pandemic and the measures taken against it on the pattern of admissions and treatment of patients with ACS in a district hospital. 

Methods

A single-centre retrospective study was conducted in the Cardiology Department of a Portuguese district hospital, capable of performing 24h percutaneous coronary intervention. We analyzed the admissions for ACS during the peak of the first COVID-19 outbreak in Portugal (16th of March to 3rd of May 2020, coinciding with the suspension of elective activity) and compared it with two control periods: one immediately preceding the study period (27th of January to 15th of March - 2020 control) and one exactly one year before (16th of March to 3rd of May 2019 - 2019 control).

Results

During the 7 weeks of the first COVID-19 outbreak, 46 patients were hospitalized for ACS in our department, while 54 had been admitted during the 2020 control period and 40 in the 2019 control period, with a mean of 7 admissions/week (vs 8 and 6, respectively; p > 0.05). There were no significant differences between the study group and both controls regarding basal characteristics. Their presentation did not differ: the majority of patients (57%) was admitted for ST-elevation myocardial infarction and median time from symptom onset to admission to the emergency room was 3 hours, with 4% of patients presenting in Killip class ≥III (p > 0.05 for all). We also found no significant differences in the treatment administered nor in the outcomes (96% submitted to coronary angiography, median length of hospitalization of 4 days and in-hospital mortality rate of 2%, p > 0.05 for all). 

Conclusions

Our study suggests that the reduction in admissions for ACS during the COVID-19 outbreak is not universal. Additionally, severity at presentation, treatment administered and outcomes did not seem to differ. The relatively low prevalence of COVID-19 in our hospital"s region might be the cause, since people might fear and refrain less from going to the hospital. Larger studies with other centres in low-prevalence regions are needed to confirm this hypothesis.

P1506Prolonged and recurrent hospital admissions predict risk of cardiac implantable electronic device infections in hospitalized patients with bacteriaemia

Background

As the use of Cardiac Implantable Electronic Devices (CIED) is expanding, the burden of device-related infections, particularly infectious endocarditis, is also increasing. Furthermore, with the increase in life expectancy, these devices are being implanted in patients who are older, with more serious co-morbidities and who are frequently hospitalized.

Purpose

To assess incidence and predisposing factors for infective endocarditis (IE) in hospitalized patients with CIED and nosocomial bacteraemia.

Methods

We performed a retrospective analysis of all hospitalized patients with CIED and positive blood cultures admitted to the Cardiology department between January 2012 and February 2019. Endocarditis was defined according to modified Duke criteria. We analyzed clinical parameters, device- and procedure-related characteristics, length of hospital stay and hospitalizations in the previous year.

Results

25 patients had positive blood cultures, 4 patients (16.0%) were diagnosed with infective endocarditis (median age 59.25 years, 75.0% male).

Patients characteristics are displayed in the Table.

In univariable analysis, the number and presence of hospitalizations in the previous year and the length of hospital stay was significantly associated with the possibility of endocarditis (respectively, OR 3.411, 95% CI 1.164-9.998, p-value 0.025; OR 18.000, 95% CI 1.375-235.686, p-value 0.028; and OR 1.047, 95% IC 1.001-1.096, p-value 0.046).

Conclusion

In this group of patients with positive blood cultures during hospitalization, the possibility of CIED infection was predicted by the length of hospital stay and the presence and number of hospitalizations in the previous year.

With CIED-IE Without CIED-IE p-value Age in years, median (IQR) 63 (16) 76 (16) 0.002 Heart failure, n (%) 4 (100%) 10 (47.6%) 0.105 LVEF in %, median (IQR) 21 (9) 51 (30) 0.008 LV dilation, n (%) 4 (100%) 8 (38.1%) 0.039 Device revision/upgrade/substitution, n (%) 2 (50.0%) 6 (28.57%) 0.570 LOS in days, median (IQR) 58.5 (83) 20 (15) 0.004 Hospital admissions in the previous year, n (%) 3 (75.0%) 3 (14.3%) 0.031 No. of hospital admissions in the previous year, mean (standard deviation) 1.67 (1.528) 0.29 (0.784) 0.005 Use of central venous access, n (%) 4 (100%) 4 (19.05%) 0.008 LOS – Length of Stay; LV – Left Ventricle; LVEF – Left Ventricle Ejection Fraction