key: cord-0930140-ejt8tw87
authors: Giabicani, Mikhael; Le Menestrel, Alice; Roux, Olivier; Rautou, Pierre-Emmanuel; Weiss, Emmanuel
title: Focus on the decisions to forego life-sustaining therapies during ICU stay of patients with cirrhosis and COVID-19: a case control study from the prospective COVID-ICU database
date: 2021-09-30
journal: J Hepatol
DOI: 10.1016/j.jhep.2021.09.018
sha: 8de234a18f5bfdfb1abe5e13137e9cba98ab855b
doc_id: 930140
cord_uid: ejt8tw87

nan

with chronic liver disease [1] . Based on the French National Hospital Discharge database, outcome of 34 259,110 COVID-19 adults including 15,746 patients with chronic liver disease was analyzed. Results 35 suggest that chronic liver disease per se is not a risk factor for COVID-19 mortality, but rather that 36 limitations of the therapeutic efforts, including a reduced access to mechanical ventilation, may have 37 accounted for the excess mortality of patients with cirrhosis. However, to date, no study has specifically 38 analyzed prognosis of patients with cirrhosis hospitalized in intensive care for COVID-19 related acute 39 respiratory distress syndrome (ARDS) and, in particular, decisions to forego life-sustaining therapies 40 (DFLST). Therefore, we took advantage of the COVID-ICU study [2] , a multi-center prospective cohort tended to be less often introduced in patients with cirrhosis; their ICU and hospital mortality rates were 54 slightly, but non significantly higher (Fig. 1B) To explore the reasons for DFLST, we performed univariate analysis among patients with cirrhosis and 63

identified SOFA score at admission, SAPS 2 and MELD score as associated with DFLST (all p<0.05). 64

Patients with cirrhosis had higher SOFA score during the first days of ICU hospitalization than patients scores both strongly related to cirrhosis. "Hemodynamic", "respiratory", "renal" and "neurologic" 68 components of the SOFA score at admission (Fig. 1A) and during the first two weeks of hospitalization 69 (data not shown) did not differ between the two groups. Interestingly, DFLST also tended to be more 70 frequent in patients with alcohol use disorders (50% vs. 18%, p=0.06) , confirming data already 71 suggested by the study from Mallet et al. 72 As intensity of inflammatory response has been associated with both severity of COVID-19 and outcome 73 of critically ill cirrhotic patients [3,4], we investigated the pattern of inflammatory response in our 74 cohort. Apart from a lower C-reactive protein concentration on admission in patients with cirrhosis, 75 daily measurements of leukocytes, neutrophils, lymphocytes, neutrophil-to-lymphocyte ratio, C-reactive 76 protein and procalcitonin from day 1 to day 28 did not differ between the two groups (data not shown). 77

Secondary bacterial infections and thromboembolic complications were not different either (Fig. 1B) . 78

In conclusion, our data show that intensity of treatment of patients with cirrhosis admitted to ICU was 79 more commonly limited compared to non-cirrhotic patients and suggest a significant impact of hepatic 80 failure on these DFLST. Yet, patients with cirrhosis did not exhibit more pronounced inflammatory 

Prognosis of French 114 COVID-19 patients with chronic liver disease: a national retrospective cohort study for 2020

Clinical characteristics and day-90 outcomes of 4244 critically ill adults with COVID-19: a 117 prospective cohort study

An 120 inflammatory cytokine signature predicts COVID-19 severity and survival

Effect of COVID-19 on patients with 123 compensated chronic liver diseases

The COVID-ICU study was funded by the Fondation APHP and its donators through the program 105