To the Editor:
We thank Doctor Giabicani and colleagues for their comments 1 on our article published in the Journal of Hepatology.2 Their investigations from a prospective, multicenter, international, cohort of more than 4,000 intensive care unit (ICU) patients with COVID-19 3 confirm our results at the national level: despite similar COVID-19 severity, including initial non-hepatic organ failure, bacterial infections and thromboembolic complications, a decision to forego life-sustaining therapies was more frequently made for patients admitted to the ICU with cirrhosis. As both studies suggest the presence of selection (prescription of invasive mechanical ventilation or decision to forego life-sustaining therapies), we further investigated the profile of patients who received invasive mechanical ventilation vs. those who died without ventilation from the latest release of the French National Hospital Discharge (PMSI) database.
Between February 1, 2020 and August 31, 2021, 567,491 patients were discharged with a primary or an associate discharge diagnosis code for COVID-19 (ICD-10 codes: U0710, U0711, U0712, U0714, U0715), COVID-19 sequelae (ICD-10 codes: U089, U099), or COVID-19 inflammatory heart diseases (ICD-10 code: U109). Of them, 41,078 patients recorded with COVID-19 sequelae without prior admission for COVID-19 and 11,656 patients younger than 18 years were removed from the analysis. The final sample comprised 514,602 patients who were first admitted at hospital during the first (26.8%), second (45.8%), and third (27.4%) waves of the SARS-CoV-2 pandemic in France (before October 1, 2020, from October 1, 2020 to March 15, 2021, and after May 31, 2021, respectively). Their characteristics are detailed in Table 1 and were similar to those reported previously.2 The median (interquartile range) age at first hospital admission was 69 (54, 83) and 51.8% of the cohort were male. Hypertension was the most prevalent recorded risk factor (51.6%), followed by obesity (26.4%), diabetes mellitus (26.3%), smoking (10.3%), and alcohol use disorders (6.1%). More than half (62.4%) of patients had at least 1 major comorbidity. Similar associations were found, including less invasive mechanical ventilation for patients with cerebrovascular disease, dementia, and connective tissue disorders; and lower 30-day mortality for female patients, aged ≤70 years, with obesity, and with AIDS.2
Table 1.
Characteristics of 2020-2021 French COVID-19 patients by invasive mechanical ventilation and 30-day mortality.
| Characteristic | Overall, N = 514,6021 | Invasive mechanical ventilation |
p value2 | 30-day mortality |
p value2 | ||
|---|---|---|---|---|---|---|---|
| No, n = 475,908 (92%)1 | Yes, n = 38,694 (7.5%)1 | No, n= 432,259 (84%)1 | Yes, n = 82,343 (16%)1 | ||||
| Age | 69 (54, 83) | 70 (53, 83) | 67 (58, 74) | <0.001 | 66 (51, 79) | 83 (74, 89) | <0.001 |
| Age category | <0.001 | <0.001 | |||||
| [18,30) | 24,153 (4.7%) | 23,703 (5.0%) | 450 (1.2%) | 24,066 (5.6%) | 87 (0.1%) | ||
| [30,40) | 33,613 (6.5%) | 32,490 (6.8%) | 1,123 (2.9%) | 33,334 (7.7%) | 279 (0.3%) | ||
| [40,50) | 45,223 (8.8%) | 42,499 (8.9%) | 2,724 (7.0%) | 44,399 (10.3%) | 824 (1.0%) | ||
| [50,60) | 69,508 (13.5%) | 62,859 (13.2%) | 6,649 (17.2%) | 66,644 (15.4%) | 2,864 (3.5%) | ||
| [60,70) | 86,183 (16.7%) | 74,263 (15.6%) | 11,920 (30.8%) | 77,441 (17.9%) | 8,742 (10.6%) | ||
| [70,80) | 97,244 (18.9%) | 84,891 (17.8%) | 12,353 (31.9%) | 78,995 (18.3%) | 18,249 (22.2%) | ||
| [80,Inf) | 158,678 (30.8%) | 155,203 (32.6%) | 3,475 (9%) | 107,380 (24.8%) | 51,298 (62.3%) | ||
| Sex | <0.001 | <0.001 | |||||
| Female | 247,990 (48.2%) | 235,730 (49.5%) | 12,260 (31.7%) | 212,990 (49.3%) | 35,000 (42.5%) | ||
| Male | 266,611 (51.8%) | 240,178 (50.5%) | 26,433 (68.3%) | 219,268 (50.7%) | 47,343 (57.5%) | ||
| Hypertension | 265,713 (51.6%) | 242,243 (50.9%) | 23,470 (60.7%) | <0.001 | 205,731 (47.6%) | 59,982 (72.8%) | <0.001 |
| Obesity | 135,765 (26.4%) | 118,934 (25%) | 16,831 (43.5%) | <0.001 | 114,209 (26.4%) | 21,556 (26.2%) | 0.15 |
| Diabetes mellitus | 135,551 (26.3%) | 121,371 (25.5%) | 14,180 (36.6%) | <0.001 | 107,124 (24.8%) | 28,427 (34.5%) | <0.001 |
| Past or current smoking | 53,242 (10.3%) | 47,514 (10%) | 5,728 (14.8%) | <0.001 | 43,884 (10.2%) | 9,358 (11.4%) | <0.001 |
| Alcohol use disorders | 31,379 (6.1%) | 28,342 (6%) | 3,037 (7.8%) | <0.001 | 25,001 (5.8%) | 6,378 (7.7%) | <0.001 |
| Congestive heart failure | 98,851 (19.2%) | 91,574 (19.2%) | 7,277 (18.8%) | 0.037 | 67,979 (15.7%) | 30,872 (37.5%) | <0.001 |
| Chronic obstructive pulmonary disease | 83,608 (16.2%) | 75,767 (15.9%) | 7,841 (20.3%) | <0.001 | 65,574 (15.2%) | 18,034 (21.9%) | <0.001 |
| Solid or liquid cancer | 83,199 (16.2%) | 76,930 (16.2%) | 6,269 (16.2%) | 0.9 | 61,347 (14.2%) | 21,852 (26.5%) | <0.001 |
| Moderate to severe chronic kidney disease | 60,861 (11.8%) | 55,952 (11.8%) | 4,909 (12.7%) | <0.001 | 42,683 (9.9%) | 18,178 (22.1%) | <0.001 |
| Dementia | 59,488 (11.6%) | 58,714 (12.3%) | 774 (2%) | <0.001 | 41,225 (9.5%) | 18,263 (22.2%) | <0.001 |
| Peripheral vascular disease | 58,498 (11.4%) | 53,503 (11.2%) | 4,995 (12.9%) | <0.001 | 42,326 (9.8%) | 16,172 (19.6%) | <0.001 |
| Diabetes mellitus with complications | 51,406 (10%) | 46,094 (9.7%) | 5,312 (13.7%) | <0.001 | 38,556 (8.9%) | 12,850 (15.6%) | <0.001 |
| Neurovascular disease | 40,713 (7.9%) | 37,596 (7.9%) | 3,117 (8.1%) | 0.3 | 30,138 (7%) | 10,575 (12.8%) | <0.001 |
| Cirrhosis | 34,039 (6.6%) | 29,769 (6.3%) | 4,270 (11%) | <0.001 | 25,254 (5.8%) | 8,785 (10.7%) | <0.001 |
| Myocardial infarction | 32,039 (6.2%) | 29,257 (6.1%) | 2,782 (7.2%) | <0.001 | 23,019 (5.3%) | 9,020 (11%) | <0.001 |
| Peptic ulcer disease | 13,403 (2.6%) | 12,149 (2.6%) | 1,254 (3.2%) | <0.001 | 10,128 (2.3%) | 3,275 (4%) | <0.001 |
| Connective tissue disorder | 12,077 (2.3%) | 11,143 (2.3%) | 934 (2.4%) | 0.4 | 9,398 (2.2%) | 2,679 (3.3%) | <0.001 |
| Transplant recipient | 6,156 (1.2%) | 5,176 (1.1%) | 980 (2.5%) | <0.001 | 5,076 (1.2%) | 1,080 (1.3%) | <0.001 |
| AIDS | 2,959 (0.6%) | 2,662 (0.6%) | 297 (0.8%) | <0.001 | 2,685 (0.6%) | 274 (0.3%) | <0.001 |
| Acute respiratory distress syndrome | 147,293 (28.6%) | 112,352 (23.6%) | 34,941 (90.3%) | <0.001 | 95,767 (22.2%) | 51,526 (62.6%) | <0.001 |
| Acute kidney injury | 28,585 (5.6%) | 17,428 (3.7%) | 11,157 (28.8%) | <0.001 | 15,525 (3.6%) | 13,060 (15.9%) | <0.001 |
| Acute liver failure | 2,813 (0.5%) | 1,661 (0.3%) | 1,152 (3%) | <0.001 | 1,552 (0.4%) | 1,261 (1.5%) | <0.001 |
| Cholangitis | 4,829 (0.9%) | 4,107 (0.9%) | 722 (1.9%) | <0.001 | 3,809 (0.9%) | 1,020 (1.2%) | <0.001 |
| Portal vein thrombosis | 359 (0.1%) | 290 (0.1%) | 69 (0.2%) | <0.001 | 226 (0.1%) | 133 (0.2%) | <0.001 |
| Pulmonary embolism | 18,813 (3.7%) | 15,412 (3.2%) | 3,401 (8.8%) | <0.001 | 15,305 (3.5%) | 3,508 (4.3%) | <0.001 |
| Pandemic wave | <0.001 | <0.001 | |||||
| First | 137,961 (26.8%) | 126,478 (26.6%) | 11,483 (29.7%) | 116,501 (27%) | 21,460 (26.1%) | ||
| Second | 235,585 (45.8%) | 219,214 (46.1%) | 16,371 (42.3%) | 191,955 (44.4%) | 43,630 (53%) | ||
| Third | 141,056 (27.4%) | 130,216 (27.4%) | 10,840 (28%) | 123,803 (28.6%) | 17,253 (21%) | ||
Data are for COVID-19 patients who were discharged between February 1, 2020 and August 31, 2021 in France. The 2011-2020 French National Hospital Discharge (PMSI) database was used to identify underlying conditions before COVID-19. Dates to define first, second, and third pandemic waves were October 1, 2020, March 15, 2021, and May 31, 2021. AIDS, acquired immune deficiency syndrome.
Median (IQR); n (%).
Wilcoxon rank sum test; Pearson's Chi-squared test.
Table 2 presents the factors associated with invasive mechanical ventilation (7.5%) and with 30-day mortality without invasive mechanical ventilation (16.0%) from a multivariate multinomial regression. Assumingly, prognostic factors positively associated with 30-day mortality without invasive mechanical ventilation indicate higher need for advanced ventilation and should thus be positively associated with invasive mechanical ventilation. For instance, the adjusted odds ratios for 30-day mortality without invasive mechanical ventilation almost doubled by age decade indicating higher need for invasive mechanical ventilation with age. However, age was associated with invasive mechanical ventilation in an inverted U-shape, suggesting that invasive mechanical ventilation resources were preferably allocated to COVID-19 patients younger than 80 years old. Similarly, COVID-19 patients first admitted during the second pandemic wave (vs. third wave), patients with alcohol use disorders or patients with a modified (without age) Charlson comorbidity index ≥3 were at higher need for invasive mechanical ventilation (adjusted odds-ratio of 30-day mortality without invasive mechanical ventilation significantly above 1), although they had less chance of receiving invasive mechanical ventilation (adjusted odds-ratio of 30-day mortality without ventilation significantly below 1), all other things being equal. On the contrary, patients with hypertension, obesity, or diabetes mellitus were at lower risk of dying without invasive mechanical ventilation (adjusted odds-ratio of 30-day mortality without ventilation significantly below 1), but had a higher chance of receiving invasive mechanical ventilation (adjusted odds-ratio of invasive mechanical ventilation significantly above 1), all other things being equal. Our results support the idea that, in France, COVID-19 resources have been preferentially allocated to younger patients with obesity, hypertension, or diabetes mellitus, and not to patients with severe comorbidities and those with alcohol use disorders.
Table 2.
Multivariate risks for invasive mechanical ventilation and 30-day mortality without ventilation.
| Characteristics1 | Invasive mechanical ventilation | p value3 | 30-day mortality without invasive mechanical ventilation | p value3 |
|---|---|---|---|---|
| Age category2 | ||||
| [30,40) | 1.62 (1.45–1.81) | <0.001 | 2.16 (1.54–3.04) | <0.001 |
| [40,50) | 2.63 (2.38–2.92) | <0.001 | 4.7 (3.44–6.42) | <0.001 |
| [50,60) | 4.06 (3.68–4.48) | <0.001 | 10.91 (8.08–14.73) | <0.001 |
| [60,70) | 5.97 (5.42–6.58) | <0.001 | 27.87 (20.70–37.53) | <0.001 |
| [70,80) | 5.74 (5.21–6.32) | <0.001 | 65.67 (48.80–88.37) | <0.001 |
| [80,Inf) | 1.28 (1.16–1.42) | <0.001 | 183.03 (136.06–246.22) | <0.001 |
| Sex: Male | 1.9 (1.86–1.95) | <0.001 | 1.41 (1.38–1.43) | <0.001 |
| Hypertension | 1.36 (1.32–1.39) | <0.001 | 0.97 (0.95–0.99) | 0.004 |
| Obesity | 1.86 (1.82–1.91) | <0.001 | 0.97 (0.95–0.99) | 0.002 |
| Diabetes mellitus | 1.12 (1.09–1.15) | <0.001 | 0.94 (0.92–0.96) | <0.001 |
| Smoking | 1.03 (1.00–1.07) | 0.052 | 1.1 (1.07–1.13) | <0.001 |
| Alcohol use disorders | 0.83 (0.80–0.87) | <0.001 | 1.29 (1.24–1.33) | <0.001 |
| Charlson comorbidity index ≥3 | 0.99 (0.97–1.02) | 0.644 | 1.98 (1.94–2.02) | <0.001 |
| Pandemic waves4 | ||||
| First | 1.19 (1.16–1.23) | <0.001 | 1.04 (1.02–1.07) | 0.001 |
| Second | 0.97 (0.94–0.99) | 0.016 | 1.09 (1.06–1.11) | <0.001 |
Data are for patients who were discharged after COVID-19 between February 1, 2020 and August 31, 2021 in France. The 2011-2020 French National Hospital Discharge (PMSI) database was used to identify underlying conditions before COVID-19. Dates to define first, second, and third pandemic waves were October 1, 2020, March 15, 2021, and May 31, 2021.
Risks were computed with multinomial logistic regression models.
Reference was the [18, 30) age category.
Reference was the third pandemic wave (May 31 to August 31, 2021).
Financial support
The study did not receive any external funding.
Authors’ contributions
VM, FD, SB, MS: conception of the study, analysis and interpretation of the data, draft of the manuscript. All other members of the Demosthenes group facilitated the study or took care of the reported patients.
Conflicts of interest
The authors declare no conflicts of interest that pertain to this work.
Please refer to the accompanying ICMJE disclosure forms for further details.
Footnotes
Supplementary data to this article can be found online at https://doi.org/10.1016/j.jhep.2021.11.020.
Contributor Information
the Demosthenes research group:
François Depret, Samir Bouam, Michaël Schwarzinger, Hélène Fontaine, Marion Corouge, Anaïs Vallet Pichard, Clémence Hollande, Philippe Sogni, Stanislas Pol, and Vincent Mallet
Demosthenes research group
François Depret,1 Samir Bouam,2 Michaël Schwarzinger,3 Hélène Fontaine,4 Marion Corouge,4 Anaïs Vallet Pichard,4 Clémence Hollande,4 Philippe Sogni,4 Stanislas Pol,4 Vincent Mallet4
1AP-HP.Nord Université de Paris, Groupe Hospitalier St-Louis-Lariboisière, DMU Parabol, Département d’Anesthésie Réanimation et Centre de Traitement des Brûlés, Paris, France; FHU Promice Paris France; INI-CRCT, Nancy, France ; 2AP-HP.Centre Université de Paris, Groupe Hospitalier Cochin Port Royal, DMU PRIME, Unité d’Information Médicale, Paris, France ; 3Service de soutien méthodologique et d’innovation en prévention (SSMIP), CHU de Bordeaux; 33000 Bordeaux, France; University of Bordeaux, Inserm UMR 1219-Bordeaux Population Health, 33000 Bordeaux, France ; 4AP-HP.Centre Université de Paris, Groupe Hospitalier Cochin Port Royal, DMU Cancérologie et spécialités médico-chirurgicales, Service d’Hépatologie, Paris, France
Supplementary data
The following are the supplementary data to this article:
References
- 1.Giabicani M., Le Menestrel A., Roux O., Rautou P.E., Weiss E., COVID-ICU study group 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. J Hepatol. 2022;76(3):742–744. doi: 10.1016/j.jhep.2021.09.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Mallet V., Beeker N., Bouam S., Sogni P., Pol S., Demosthenes research group Prognosis of French COVID-19 patients with chronic liver disease: a national retrospective cohort study for 2020. J Hepatol. 2021;75(4):848–855. doi: 10.1016/j.jhep.2021.04.052. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.COVID-ICU Group on behalf of the REVA Network and the COVID-ICU Investigators Clinical characteristics and day-90 outcomes of 4244 critically ill adults with COVID-19: a prospective cohort study. Intensive Care Med. 2021;47(1):60–73. doi: 10.1007/s00134-020-06294-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.