Full and simplified assessment of left ventricular diastolic function in covid‐19 patients admitted to ICU: Feasibility, incidence, and association with mortality

Abstract Purpose Left ventricular diastolic dysfunction (LVDD) is associated with poor outcomes in the intensive care unit (ICU). Nonetheless, precise reporting of LVDD in COVID‐19 patients is currently lacking and assessment could be challenging. Methods We performed an echocardiography study in COVID‐19 patients admitted to ICU with the aim to describe the feasibility of full or simplified LVDD assessment and its incidence. We also evaluated the association of LVDD or of single echocardiographic parameters with hospital mortality. Results Between 06.10.2020 and 18.02.2021, full diastolic assessment was feasible in 74% (n = 26/35) of patients receiving a full echocardiogram study. LVDD incidence was 46% (n = 12/26), while the simplified assessment produced different results (incidence 81%, n = 21/26). Nine patients with normal function on full assessment had LVDD with simplified criteria (grade I = 2; grade II = 3; grade III = 4). Nine patients were hospital‐survivors (39%); the incidence of LVDD (full assessment) was not different between survivors (n = 2/9, 22%) and non‐survivors (n = 10/17, 59%; p = .11). The E/e’ ratio lateral was lower in survivors (7.4 [3.6] vs. non‐survivors 10.5 [6.3], p = .03). We also found that s’ wave was higher in survivors (average, p = .01). Conclusion In a small single‐center study, assessment of LVDD according to the latest guidelines was feasible in three quarters of COVID‐19 patients. Non‐survivors showed a trend toward greater LVDD incidence; moreover, they had significantly worse s’ values (all) and higher E/e’ ratio (lateral).


INTRODUCTION
Coronavirus disease  pandemic has caused over 6 million deaths worldwide and these figures are likely underestimated. 1 COVID-19 infection may span from asymptomatic or mild and selflimiting cases to severe illness requiring hospitalization where COVID-19 may trigger a multi-systemic infection involving different organs. [2][3][4][5] The lungs seem the most affected organ with possible development of interstitial pneumonia requiring hospitalization and intensive care unit (ICU) admission with mechanical ventilation in severe cases. [6][7][8] A substantial cardiovascular impact in patients with COVID-19 has been repeatedly demonstrated 9 ; of note, even patients not requiring hospitalization have shown some degree of myocardial dysfunction with features of myocarditis on magnetic resonance imaging. [10][11][12][13] Severely ill COVID-19 patients admitted to ICU may experience cardio-circulatory failure and a fair amount of them may need support with catecholamine infusions. Different degrees of cardiac injury as evaluated by biomarkers [14][15][16] or echocardiography 9,17 have been reported for COVID-19 patients admitted to ICU. Several patterns of cardiovascular dysfunction have also been described: from signs of myocarditis or myocardial ischemia to significant hypovolemia (due to pyrexia and prolonged fasting), from right ventricular (RV) failure (influenced by mechanical ventilation and/or by micro/macro pulmonary embolism) to septic cardiovascular dysfunction due to super-imposed bacterial or fungal infections. 10,11,13,18,19 Moreover, ventricular diastolic dysfunction (LVDD) has received attention for its association with outcomes, [21][22][23] while the same association has not been shown for left ventricular systolic dysfunction (LVSD). 24 Echocardiography is crucial in diagnosing and grading LVDD and may help distinguish patterns of cardiovascular dysfunction, suggest therapeutic options, and track the changes with sequential monitoring. 25 Our single-center joined the international ECHO-COVID study. 17 With the purpose to fully characterize LVDD, we also collected tissue Doppler Imaging (TDI) and left atrial volume index (LAVI) data.
Hereby, we report the feasibility of full and simplified LVDD assessment in severe COVID-19 patients admitted to ICU, the incidence of LVDD, and its association with mortality.

METHODS
This study was conducted in parallel to the international ECHO-COVID, 17

Outcomes
Our primary outcomes were the feasibility of assessment of LVDD

RESULTS
In total, 102 patients were admitted to our ICU during the study period, while our unit served as a general "clean" ICU during the other periods of the current pandemic. At our center, 35 patients (34% of those admitted) received advanced CCE within the first 3 days of ICU admission or after escalation from non-invasive to invasive respiratory support while already admitted to ICU.

Study population and feasibility of LVDD assessment
Assessment of LVDD according to current ASE/EACVI guidelines 20 was feasible in 26/35 patients (74%), and these patients were included in the study. Table 1 shows characteristics in the study population, both overall and according to hospital mortality (survivors, n = 9; non-survivors, n = 17). In particular, we report baseline characteristics, comorbidities, and outcome data on length of stay and mortality, all together with ventilation support, arterial blood gas analysis and hemodynamic conditions at the time of advanced CCE. Non-survivors had higher incidence of intubation and mechanical ventilation as compared to survivors (p = .03), and a trend toward a worse P/F ratio (p = .052). Table 2 reports the CCE data in study population. These data are grouped according to LV size, systolic, and diastolic function, along with data on RV size and systolic function, inferior vena cava size (IVC) and pericardium. According to the ASE/EACVI 2016 guidelines, 20 12 patients had LVDD (46%). When performing the grading, we found one patient with LVDD grade I, four patients with grade II, and seven with indeterminate grade (could be I or II).
When investigating association between hospital mortality and LVDD diagnosis according to simplified Lanspa criteria, 30 n = 14/17 (82%) of hospital non-survivors and n = 7/9 (78%) of survivors had a diagnosis of LVDD (p = 1.00). We also found no differences in the diagnosis of LVDD according to ICU mortality (p = .50).
Numbers were far too small to analyze subgroups according to LVDD grade, and these analyses were not conducted.

Single echocardiographic parameters and association with hospital mortality
Regarding the secondary outcomes focusing on the association between single echocardiographic parameters and hospital mortality, we found that non-survivors had higher lateral E/e' ratio (p = .03). A trend toward higher average E/e' ratio was also found (p = .08), while septal values were not significantly different (p = .31). Deceleration

Baseline characteristics and comorbidities
Age (years) 71 [15.5] 66 [22] 72 [10] .22 Weight (Kg) 76.5 [21.25] 80 [29] 75 [21] .18 Height (cm) 170 [15] 170 [14] 170 [17] .  [9] .01   With several limitations, this study is probably one of the few available experiences reporting full LVDD assessment according to the current ASE/EACVI 2016 guidelines. 20 Indeed, while several studies reported behavior of one or more echocardiography variables used for the assessment of LVDD, it seems no studies have reported full LVDD assessment according to the latest guidelines, 20 as shown by a systematic review. 9 From an overview of the literature on COVID-19 patients, we also could not find any experiences comparing the full and the simplified assessment of LVDD.

ICU
Unfortunately, our study is severely underpowered for detecting the influence of LVDD on the outcome of severe COVID-19 patients.
This was behind our control as the ICU served as COVID-ICU for the Trust only for a brief period of time (∼4 months, n = 102 COVID-19 admissions); moreover, the workload did not always allow timely assessment with advanced CCE for the purpose of this study, as only one operator had advanced CCE skills and joined the ECHO-COVID study. Therefore, all together with the risk of statistical error, it is likely that an inevitable selection bias took place. Bearing in mind the limitations of the study, we think that our analysis is in line with previous experience reporting the importance of LVDD in the context of critical illness and also with the known difficulties in assessing LVDD in mechanically ventilated patients. 34 Different phenotypes of cardiovascular dysfunction have been described in critically ill patients, 35 and LVDD has received attention for its association both with mortality in septic patients 21,22 and for weaning failure. 23 Conversely, LVSD has not shown the same association when evaluated by means of LVEF 36 or s' wave 24 in critically ill patients. It was somewhat unexpected to find that TDI s' wave was significantly lower in hospital (and ICU) non-survivors, as this parameter has not been found associated with prognosis in critically ill patients (i.e., septic patients 24 ); moreover, the population we studied was mostly free from cardiovascular support (77%), and those on norepinephrine received a very low dose (.04 mcg/kg/min). However, considering that a myocarditislike pattern has been found in cardiac magnetic resonance imaging after COVID-19 also in cohorts of asymptomatic and mildly symptomatic patients, 12,37,38 it is possible that the lower TDI s' values are related to an impaired longitudinal LV systolic function not detected by assessment of LVEF.
We also found that lateral E/e' ratio was significantly higher in non-survivors at hospital discharge, followed by a trend in average E/e' ratio (p = .08). The mean difference between survivors and non- for the effects of mechanical ventilation, rather than as a reflection of an ongoing impaired LV relaxation (post-capillary). Therefore, there are several adjunctive differences and peculiarities that may render the evaluation of LVDD even more complex as compared to the usual ICU patient. Among these, COVID-19 usually has a more gradual evolution of the critical illness as compared to typical septic shock evolving more rapidly.

Limitations
We already mentioned the small sample size and the non-consecutive enrollment as main limitation of this ancillary study. In consideration of the small sample size, we thought that performing sophisticated multivariate and/or regression statistical analyses with the aim to address for confounders would have not been meaningful. Although we reported the items for the study interpretation according to the

CONCLUSIONS
In a small single-center study, the assessment of LVDD according to