The authors reply.

We would like to thank Jha and Jha (1) for their many insightful comments about our article (2). First, although a dissociation between tricuspid annular plane systolic excursion (TAPSE) and right ventricular (RV) fractional area change (RVFAC) appears surprising, this has been reported by multiple studies assessing RV systolic function in COVID-19 acute respiratory distress syndrome (ARDS), including our own. TAPSE, lateral tricusipid annulus peak systolic velocity (RV S’), and RV free-wall longitudinal strain were all relatively preserved despite low RVFAC and were considered to be poorly sensitive of RV systolic impairment in COVID-19 ARDS (3). The authors proposed an RV phenotype in COVID-19 ARDS characterized by radial impairment with sparing of longitudinal function, a pattern of myocardial remodeling that resembles the response to pulmonary hypertension (3), and our results support these findings.

We also disagree with the suggestion that all RV phenotypes were associated with normal left ventricular (LV) function. Thirty-seven percent of our cohort had hyperdynamic LV ejection fraction (HDLVEF), and this should not be considered “normal.” The prevalence of HDLVEF was higher in nonsurvivors than that in survivors (46% vs 32%). Although HDLVEF is associated with poor prognosis in critically ill patients with sepsis (4), likely due to unmitigated vasoplegia precipitating a high cardiac output state, in ARDS, its development may be due to low LV preload, caused by RV systolic impairment and/or obstruction of LV filling from RV dilation (RVD). HDLVEF may, therefore, reflect a low cardiac output state that precipitates multiple organ dysfunction and death. We were unable to estimate cardiac output via transthoracic echocardiography in these patients to corroborate this hypothesis.

Jha and Jha (1) highlight the relatively low prevalence of RV volume and pressure overload as measured through the LV eccentricity index and septal dyskinesia, a surprising finding. This may be because we were only able to measure this parameter in 72% of patients due to poor views. However, we disagree with their suggestion that these findings are out of keeping with the measured maximum tricuspid regurgitation (TR) velocity (TRVmax), which at 2.7 (interquartile range, 2.3–3.1) is also relatively low: only 29% of patients had intermediate high risk of pulmonary hypertension. These findings are also limited as we were only able to measure TRVmax in 42% of patients, due to incomplete or absent TR Doppler signal in 58% of patients.

Finally, we agree with the notion that the definition of RV dysfunction in ARDS should include multiple parameters that holistically assess RV size, function, and coupling to the pulmonary circulation but also include downstream consequences on organ congestion/LV function and cardiac output. However, these parameters should be chosen carefully: septal dyskinesia was previously included in definitions of RV dysfunction in ARDS but has recently been disregarded (5), perhaps due to its subjective nature. Current guideline definitions of RVD include RV dilation with evidence of systemic congestion (5) (most recently defined as central venous pressure greater than 8 [6]) or RV systolic impairment (defined as low indices of TAPSE, RV S’, and RVFAC [7]). This study proposes using RVD with systolic impairment, which has been independently associated with mortality in COVID-19 ARDS (2) and sepsis (8).