Four-Month Clinical Status of a Cohort of Patients After Hospitalization for COVID-19

Key Points

Question  What are the clinical outcomes after hospitalization for COVID-19?

Findings  Four months after hospitalization, in an uncontrolled cohort study of 478 survivors of COVID-19, at least 1 new-onset symptom was reported by telephone interview by 244 patients (51%), including fatigue in 134 of 431 (31%), cognitive symptoms in 86 of 416 (21%), and dyspnea in 78 of 478 (16%). Computed tomographic lung scan abnormalities were reported in 63% of 171 patients assessed at an ambulatory visit, mainly subtle ground-glass opacities. Fibrotic lesions were observed in 19% of these 171 patients.

Meaning  This study provides clinical status of a cohort of patients 4 months after hospitalization for COVID-19, but further research is needed to understand longer-term outcomes.

Importance  Little is known about long-term sequelae of COVID-19.

Objective  To describe the consequences at 4 months in patients hospitalized for COVID-19.

Design, Setting, and Participants  In a prospective uncontrolled cohort study, survivors of COVID-19 who had been hospitalized in a university hospital in France between March 1 and May 29, 2020, underwent a telephone assessment 4 months after discharge, between July 15 and September 18, 2020. Patients with relevant symptoms and all patients hospitalized in an intensive care unit (ICU) were invited for further assessment at an ambulatory care visit.

Exposures  Survival of hospitalization for COVID-19.

Main Outcomes and Measures  Respiratory, cognitive, and functional symptoms were assessed by telephone with the Q3PC cognitive screening questionnaire and a checklist of symptoms. At the ambulatory care visit, patients underwent pulmonary function tests, lung computed tomographic scan, psychometric and cognitive tests (including the 36-Item Short-Form Health Survey and 20-item Multidimensional Fatigue Inventory), and, for patients who had been hospitalized in the ICU or reported ongoing symptoms, echocardiography.

Results  Among 834 eligible patients, 478 were evaluated by telephone (mean age, 61 years [SD, 16 years]; 201 men, 277 women). During the telephone interview, 244 patients (51%) declared at least 1 symptom that did not exist before COVID-19: fatigue in 31%, cognitive symptoms in 21%, and new-onset dyspnea in 16%. There was further evaluation in 177 patients (37%), including 97 of 142 former ICU patients. The median 20-item Multidimensional Fatigue Inventory score (n = 130) was 4.5 (interquartile range, 3.0-5.0) for reduced motivation and 3.7 (interquartile range, 3.0-4.5) for mental fatigue (possible range, 1 [best] to 5 [worst]). The median 36-Item Short-Form Health Survey score (n = 145) was 25 (interquartile range, 25.0-75.0) for the subscale “role limited owing to physical problems” (possible range, 0 [best] to 100 [worst]). Computed tomographic lung-scan abnormalities were found in 108 of 171 patients (63%), mainly subtle ground-glass opacities. Fibrotic lesions were observed in 33 of 171 patients (19%), involving less than 25% of parenchyma in all but 1 patient. Fibrotic lesions were observed in 19 of 49 survivors (39%) with acute respiratory distress syndrome. Among 94 former ICU patients, anxiety, depression, and posttraumatic symptoms were observed in 23%, 18%, and 7%, respectively. The left ventricular ejection fraction was less than 50% in 8 of 83 ICU patients (10%). New-onset chronic kidney disease was observed in 2 ICU patients. Serology was positive in 172 of 177 outpatients (97%).

Conclusions and Relevance  Four months after hospitalization for COVID-19, a cohort of patients frequently reported symptoms not previously present, and lung-scan abnormalities were common among those who were tested. These findings are limited by the absence of a control group and of pre-COVID assessments in this cohort. Further research is needed to understand longer-term outcomes and whether these findings reflect associations with the disease.

The effects of COVID-19 are highly variable, ranging from individuals who are asymptomatic to patients who develop severe acute respiratory distress syndrome, with potential involvement of almost all organs and systems.¹ These acute symptoms have been well described since the first cohort studies that were published at the beginning of the pandemic.²

However, the possible long-term sequelae of COVID-19 have become an increasing concern. The inflammatory storm that characterizes severe forms of the disease suggests that serious tissue sequelae may affect various organ systems. Other coronaviruses have been shown to induce long-term effects, especially in the lungs.³

Although the long-term sequelae of individual organ damage have been reported,^3-9 there have been few comprehensive evaluations of long-term consequences of COVID-19.^10-12 In addition, most of the studies have included patients who actively decided to participate in follow-up.^10,12 The aim of the present study was to systematically assess, 4 months after discharge, the clinical status of survivors of COVID-19 disease requiring hospitalization.

Patients provided written informed consent. The Ethics Committee of the French Intensive Care Society (CE20-56) approved the study. The COMEBAC (Consultation Multi-Expertise de Bicêtre Après COVID-19) uncontrolled cohort study included adult patients admitted to the Bicêtre Hospital (Paris-Saclay University hospitals) in France for COVID-19 from March 1 to May 29, 2020. Inclusion criteria were survival 4 months after hospital discharge or after intensive care unit (ICU) discharge for patients who had been admitted to an ICU (referred to as “ICU patients” hereafter), who were older than 18 years, who had been hospitalized for greater than 24 hours primarily because of COVID-19, and who had received a diagnosis of SARS-CoV-2 infection by reverse transcriptase–polymerase chain reaction (RT-PCR) , by typical computed tomographic (CT) lung scan associated with clinical features, or both.

Exclusion criteria were death within 4 months after discharge, persistent hospitalization, end-stage cancer, dementia, nosocomial COVID-19 infection, and incidental positive SARS-CoV-2 RT-PCR result during a hospital stay for a different medical indication. Medical charts of patients were reviewed to establish that patients met the eligibility criteria.

The Bicêtre Hospital is one of the university hospitals in the Paris region (Assistance Publique-Hôpitaux de Paris). It has 907 beds, including 53 in adult ICUs, 30 in adult intermediate severity units, 8 in a pediatric ICU, and 12 in a pediatric intermediate-severity unit. In 2019, 18 678 adult patients were admitted to the hospital. At the peak of the first wave of COVID-19, from March to May 2020, the total number of ICU beds with mechanical ventilation capacity was increased to 101 from 53, with adult patients undergoing invasive ventilation in the adult and pediatric ICUs and intermediate-severity units, the operating rooms, and the recovery room. High-flow nasal canula and noninvasive and invasive mechanical ventilation were used only in these areas.

Patient medical history, illness evolution, and treatment during hospitalization were retrospectively collected from medical records. Kidney failure was diagnosed according to the KDIGO criteria.¹³

Three to 4 months after hospital or ICU discharge, patients were contacted by telephone by a medical officer and administered a questionnaire that included general condition and respiratory, cognitive, and neurologic symptoms (with the Q3PC cognitive screening questionnaire).¹⁴ The questionnaire and the list of symptoms in question are shown in the Supplement. Patients were asked whether symptoms existed before they developed COVID-19. All symptoms were listed, without any interpretation. No psychological evaluation was performed.

In addition, patients with no history of chronic kidney disease and with high plasma creatinine levels (>1.47 mg/dL [130 μmol/L]) or estimated glomerular filtration rate less than 60 mL/min/1.73 m² at hospital discharge were requested to have their serum creatinine levels reassessed. Patients were asked whether a lung CT scan had been performed after hospitalization, and if so, the lung CT scan was reviewed.

All ICU patients and those who were symptomatic were invited for further evaluation in the ambulatory setting. Symptomatic patients were defined as those reporting symptoms at the telephone interview (except for anosmia), all patients who had persistent creatinine-level elevation, and all those who had persistent abnormalities on a lung CT scan conducted after hospitalization (including any residual ground-glass opacities, bronchial or bronchioloalveolar abnormalities, lung condensations, or interstitial thickening).

The following assessments were performed in the outpatient facility between July 15 and September 18, 2020.

In addition to receiving a general clinical examination and a having a blood sample taken, patients were evaluated for their quality of life (36-Item Short-Form Health Survey questionnaire)¹⁵ and fatigue (Multidimensional Fatigue Inventory scale)¹⁶ (eTable 1 in the Supplement). The 36-Item Short-Form Health Survey questionnaire evaluates 8 domains: physical functioning, role limitations owing to physical health, role limitations owing to emotional problems, energy/fatigue, emotional well-being, social functioning, pain, and general health. Each subscale ranges from 0 to 100, corresponding to the worst and best health state, respectively. The 20-item Multidimensional Fatigue Inventory questionnaire evaluates 4 domains: general fatigue, reduced activity, motivation, and mental fatigue. Each subscale ranges from 1 to 5, corresponding to the best and worst feeling owing to fatigue for each domain, respectively.¹⁶

In addition to its evaluation at the telephone interview, dyspnea was assessed by the modified Medical Research Council scale. A 6-minute walk test was performed according to current recommendations.¹⁷ Patients completed pulmonary function tests (see the Supplement). Dysfunctional breathing was assessed with the Nijmegen questionnaire¹⁸ (eTable 1 in the Supplement) and a hyperventilation provocation test. Medical records of patients reporting new-onset or increased dyspnea compared with before COVID-19 infection were analyzed by 2 pulmonologists to determine its cause, who reached consensus in case of disagreement.

A high-resolution lung CT scan was performed for all patients (Supplement). Two readers, blinded to clinical evaluation, reviewed the CT scans, reaching consensus regarding any disagreements.

All ICU patients, those with pulmonary embolism during hospitalization, and those with cardiac symptoms on examination at the outpatient clinic were evaluated with transthoracic echocardiography.

All patients underwent psychometric tests, an interview with a neuropsychologist, and an interview with a psychologist. Global cognitive assessment was performed with the Montreal Cognitive Assessment¹⁹ adapted to age and education level²⁰ (eTable 1 in the Supplement). Attention was assessed through the d2-R test.²¹ Memory complaint was assessed through the McNair self-questionnaire.²¹ “Cognitive complaint” was defined by an impaired McNair score, reported cognitive symptoms, or both. “Cognitive impairment” was defined by an impaired Montreal Cognitive Assessment or d2-R score.

Anxiety symptoms were evaluated with the anxiety subscale of the Hospital Anxiety and Depression Scale,²² depression symptoms with the 13-item Beck Depression Inventory score,²³ insomnia with the Insomnia Severity Index, and posttraumatic symptoms with the Posttraumatic Stress Disorder Checklist (PCL-5 scale)²⁴ (eTable 1 in the Supplement).

In all outpatients, anti–SARS-CoV-2 total immunoglobulin level (Elecsys; Roche Diagnostic) and IgM and IgG levels (NG-Biotech) were assessed.²⁵

Study data were collected and managed with Research Electronic Data Capture tools hosted at Assistance Publique-Hôpitaux de Paris.^26,27 Data are presented as counts and percentages, means (SD), and medians (interquartile range [IQR]). All participants for whom the variables of interest were available were included in the final analysis and no assumptions were made for missing data. Because of the importance of understanding the effect of intubation on patient outcomes, the outcomes have been categorized as nonintubated vs intubated. Analysis was performed with the R statistical package version 4.0.0 (R Foundation for Statistical Computing).

Among the 1151 adult patients hospitalized in the Bicêtre Hospital because of COVID-19 from March 1 to May 29, 2020, 834 were eligible for telephone assessment and 478 (57%) consented, including 142 of 172 ICU patients and 336 of 662 non-ICU patients (Figure 1). The characteristics of the 673 patients who did not participate in the study are shown in eTable 2 in the Supplement. The patients who did not consent were similar to those who did.

COVID-19 had been diagnosed with RT-PCR in 415 patients (86.8%) and by an association of typical clinical signs and CT-scan lung images in 63 patients (13.2%). Of 294 patients who were eligible for an ambulatory assessment (n = 135 ICU patients and 159 non-ICU patients), 177 consented (n = 97 ICU patients and 80 non-ICU patients). The median time to telephone assessment was 113 days after hospital discharge (IQR, 94-128 days) and the median time of the ambulatory assessment was 125 days after discharge (IQR, 107-144 days). The mean age of the 478 patients was 60.9 years (SD, 16 years) and 57.9% were men. The patients who attended the ambulatory visit were younger (mean age, 56.9 years) and generally had more comorbid conditions. The 2 most common treatments for all patients were azithromycin and tocilizumab (Table 1). Of the 142 patients in the ICU, 73 had been intubated (Table 1). Comparisons of the characteristics of patients who attended vs did not attend the ambulatory facility, of those who visited the outpatient facility for neurocognitive vs respiratory symptoms, and by younger and older than 75 years are shown in eTables 3, 4, and 5 in the Supplement, respectively.

Of the 478 patients, 244 (51%) reported at least 1 symptom that did not exist before COVID-19 infection (eFigure 1 in the Supplement; Table 2), including fatigue, 31.1% (134 of 431); memory difficulties, 17.5% (73 of 416); dyspnea, 16.3% (78 of 478); and persistent paresthesia, 12.1% (51 of 421).

Eighty-six of 416 patients (20.7%) reported at least 1 cognitive symptom. Memory difficulties, mental slowness, and concentration problems more than once a week were reported by 73 of 416 patients (17.5%), 42 of 415 (10.1%), and 41 of 412 (10%), respectively (Table 2; eFigure 1 in the Supplement).

The results of the telephone assessment according to age and to the reported symptoms are shown in eTables 5 and 6 in the Supplement, respectively.

The results of the ambulatory assessment according to the symptoms that triggered the visit to this facility are provided in eTable 7 in the Supplement. The median 36-Item Short-Form Health Survey score, evaluated in 130 patients, was 25 (IQR, 25.0-75.0) for the subscale “role limited owing to physical problems,” 46.9 (IQR, 31.2-68.8) for “vitality,” and 57.5 (IQR, 40.0-75.0) for “general health” (potential best score, 100; worst score, 0). The median score for the 20-item Multidimensional Fatigue Inventory questionnaire, evaluated in 145 patients, was 4.5 (IQR, 3.0-5.0) for reduced motivation and 3.7 (IQR, 3.0-4.5) for mental fatigue (potential best score, 1; worst score, 5) (eFigure 2 in the Supplement).

Cognitive impairment was confirmed in 38.4% of patients (61/159) (Table 3, Figure 2), more commonly in patients aged 75 years or older (eTable 5 in the Supplement). In ICU patients, anxiety, depression, and significant posttraumatic symptoms were observed in 22 of 94 (23.4%), 17 of 94 (18.1%), and 7 of 94 (7.4%), respectively.

Muscular weakness compatible with ICU-related neuromyopathy was identified in 14 of 51 patients (27.5%) who had been intubated.

Lung CT scan was performed for 171 patients and showed abnormalities in 37 of 49 intubated individuals (75.5%) and 71 of 122 nonintubated ones (58.2%). Abnormalities were most commonly ground-glass opacities (Table 3). Fibrotic lesions were observed in 19.3% of patients (33/171), with subpleural predominance in 30 of 33 individuals (90.9%). Fibrotic lesions affected 4 lobes (IQR, 3-5) and involved less than 25% of lung parenchyma in all but 1 case. In the 49 patients who had received a diagnosis of acute respiratory distress syndrome, fibrotic lesions and reticulations were observed in 38.8% (19/49) and 69.4% (34/49), respectively.

The Nijmegen score was greater than 22 in 37 of 177 cases (20.9%), indicating dysfunctional breathing. This was confirmed by a positive hyperventilation provocation test result in 21 of 177 patients (11.9%) (Table 3), of whom 4 had a history of asthma, with normal pulmonary function test results and satisfactory control of asthma at the assessment. No fibrotic lesions were reported in these patients with confirmed dysfunctional breathing.

Among the 78 patients attending the ambulatory facility who reported new-onset dyspnea, a cause was identified in 61 cases (78.2%). Dyspnea was attributed to abnormalities on lung CT scan in 44 of 78 patients (56.4%), including fibrotic lesions in 18 of 78 cases (23.1%), and to hyperventilation provocation test-confirmed dysfunctional breathing in 14 of 78 patients (17.9%). Dyspnea was potentially explained by underlying chronic lung diseases in 7 patients (4 with uncontrolled asthma; 2, chronic obstructive pulmonary disease; and 1, obstructive sleep apnea), by left ventricular dysfunction in 3 patients, and by physical deconditioning in 2 patients.

The mean diffusing capacity of the lungs for carbon monoxide was 87% of the predicted value (SD, 23%) in 152 patients. It was 77% (SD, 17%) in the 49 patients with previous acute respiratory distress syndrome. Twenty-three of 33 patients (69.7%) with diffusing capacity of the lungs for carbon monoxide less than 70% had persistent lung CT-scan abnormalities, including 12 of 33 patients (36.4%) with fibrotic lesions.

Echocardiography identified a left ventricular ejection fraction less than 50% in 8 of 83 patients (9.6%), with no value less than 40% (Table 3). All patients with left ventricular ejection fraction less than 50% had been ICU patients. Left ventricular dilatation was identified in only 1 patient, in whom it preexisted.

Among the 95 of 478 patients (19.9%) who had experienced acute kidney injury during hospitalization, 2 displayed a persistent alteration of kidney function at 4 months. Both had required ICU care.

A positive serology result was obtained for 172 of 177 patients (97.2%). Three of these 172 patients never had a positive RT-PCR result for SARS-CoV-2 and had received a diagnosis of COVID-19 according to clinical and lung CT-scan findings. The serologic index (ie, serologic titer of total Ig) was similar between ICU and non-ICU patients and lower in immunocompromised ones (eFigure 3 in the Supplement).

In a cohort study of 478 patients who were evaluated 4 months after hospitalization for COVID-19, 51% reported at least 1 symptom that did not exist before the disease. The most common symptoms were fatigue, cognitive problems, and new-onset dyspnea. Among patients who returned for further evaluation, CT scan frequently revealed lung persistent abnormalities, as well as lung fibrotic lesions in a minority of patients. Persistent cardiac dysfunction and kidney failure were uncommon. Almost all patients had positive anti–SARS-CoV-2 serology.

There have been substantial concerns about the respiratory sequelae of COVID-19.¹ However, severe pulmonary sequelae were infrequent in the patients in the present study, although all had experienced a severe or very severe form of COVID-19. Results for lung CT scan, performed for all symptomatic and ICU patients, were abnormal in most patients, but the most frequent lesions were subtle ground-glass opacities, confirming previous findings.¹¹ Fibrotic lesions, suspected because of the severe initial inflammation,¹ were present in only 19% of patients with CT scans, which is consistent with 2 smaller studies reporting CT scan at 3 months after COVID-19.^6,12 In the present cohort, the fibrotic changes affected less than 25% of the lung parenchyma in all but 1 patient and occurred almost exclusively in ICU patients. Although it is possible that patients with mild fibrotic lesions did not report dyspnea at telephone consultation and subsequently did not undergo a CT scan, it appears that this represents a small proportion of the 478 evaluated patients.

Although severe lung sequelae were uncommon, new-onset dyspnea was reported in 16% of patients. If confirmed in other studies and found to be persistent, this could be clinically important, given the large number of seriously ill patients with COVID-19 worldwide. Furthermore, although parenchymal sequelae were the most common finding, dysfunctional breathing was confirmed by hyperventilation provocation test in 12% of patients, a finding that to our knowledge has not previously been described. Dysfunctional breathing is most likely not specific to COVID-19. Nevertheless, it might be the sequela of the dissociation between dyspnea and the severity of hypoxemia described in COVID-19 patients at the acute phase.^28,29

In the present cohort, cognitive problems, reported by telephone assessment and confirmed at the outpatient clinic, were frequent: 21% of the patients reported at least 1 cognitive symptom, and cognitive impairment was confirmed in 38% of patients who were subsequently evaluated. The underlying mechanisms are unknown, but these symptoms might be the sequala of central nerve system injury by SARS-CoV-2, as occurs during other viral infections such as glandular fever.³⁰

Psychological testing was conducted only on patients who returned for an ambulatory assessment, making it difficult to determine the true prevalence of these findings. Symptoms of anxiety were found in 31% of patients and symptoms of depression in 21%. Conversely, the prevalence of psychological symptoms in ICU patients, systematically evaluated at the outpatient clinic, was meaningful. These symptoms appear to have been less common in ICU patients than in the overall assessed population. The psychological consequences of an ICU stay have been well described.³¹

Although echocardiographic evaluation of cardiac function was performed only in ICU and symptomatic patients, systolic left ventricular dysfunction was rare and found exclusively in ICU patients. A critical limitation is that the pre–COVID-19 cardiac function was not known, but these findings suggest that possible heart damage from COVID-19 may not have frequent sequelae. However, patients with severe cardiac injury leading to death within 4 months would have been excluded from this study.

This study has several limitations. First, this investigation was an uncontrolled cohort study, which precludes comparison of prevalence of findings with patients not experiencing COVID-19. Second, this study was performed during the first months of the epidemic when corticosteroids, anticoagulation at higher doses, and treatment with other immunomodulators were not systematically used. Third, the absence of a non–COVID-19 control group or even pre–COVID-19 assessments on the same patients limits the ability to conclude that findings after 4 months are temporally related to COVID. Fourth, many patients who were invited to participate declined both the telephone assessments and the subsequent ambulatory assessments. It is possible that patients who did not participate had fewer symptoms than those who did.

Four months after hospitalization for COVID-19, a cohort of patients frequently reported symptoms not previously present, and lung-scan abnormalities were common among those who were tested. These findings are limited by the absence of a control group and of pre–COVID-19 assessments in this cohort; further research is needed to understand longer-term outcomes and whether these findings reflect associations with the disease.

Corresponding Author: Xavier Monnet, MD, PhD, Service de Médecine Intensive-Réanimation, Hôpital de Bicêtre, 78, rue du Général Leclerc, 94 270 Le Kremlin-Bicêtre, France (xavier.monnet@aphp.fr).

Accepted for Publication: February 22, 2021.

Published Online: March 17, 2021. doi:10.1001/jama.2021.3331

Correction: This article was corrected on November 9, 2021, to add a supplement listing the nonauthor collaborators of the COMEBAC study.

Authors/Writing Committee for the COMEBAC Study Group Luc Morin, MD, MSc; Laurent Savale, MD, PhD; Tài Pham, MD, PhD; Romain Colle, MD, PhD; Samy Figueiredo, MD, PhD; Anatole Harrois, MD, PhD; Matthieu Gasnier, MD; Anne-Lise Lecoq, MD, PhD; Olivier Meyrignac, MD, PhD; Nicolas Noel, MD, PhD; Elodie Baudry, MD; Marie-France Bellin, MD; Antoine Beurnier, MD; Walid Choucha, MD; Emmanuelle Corruble, MD, PhD; Laurent Dortet, PharmD, PhD; Isabelle Hardy-Leger, MA; François Radiguer, MA; Sabine Sportouch, MA; Christiane Verny, MD; Benjamin Wyplosz, MD, PhD; Mohamad Zaidan, MD, PhD; Laurent Becquemont, MD, PhD; David Montani, MD, PhD; Xavier Monnet, MD, PhD.

Affiliations of Authors/Writing Committee for the COMEBAC Study Group: Université Paris-Saclay, AP-HP, Service de Réanimation Pédiatrique et Médecine Néonatale, Hôpital de Bicêtre, Santé de l'Enfant et de l'Adolescent, Le Kremlin-Bicêtre, France (Morin); Université Paris-Saclay, AP-HP, Service de Pneumologie et Soins Intensifs Respiratoires, DMU 5, Thorinno, Inserm UMR_S999, Hôpital de Bicêtre, Le Kremlin-Bicêtre, France (Savale, Montani); Université Paris-Saclay, AP-HP, Service de Médecine Intensive-Réanimation, Hôpital de Bicêtre, DMU 4 CORREVE Maladies du Cœur et des Vaisseaux, FHU Sepsis, Le Kremlin-Bicêtre, France (Pham, Sportouch, Monnet); Université Paris-Saclay, AP-HP, Service de Psychiatrie, Hôpital de Bicêtre, DMU 11 Psychiatrie, Santé Mentale, Addictologie et Nutrition, Équipe MOODS, INSERM U1178, CESP (Centre de Recherche en Epidémiologie et Santé des Populations), Le Kremlin-Bicêtre, France (Colle, Gasnier, Choucha, Corruble); Université Paris-Saclay, AP-HP, Service de Réanimation Chirurgicale, Hôpital de Bicêtre, DMU 12 Anesthésie, Réanimation, Douleur, Le Kremlin-Bicêtre, France (Figueiredo, Harrois, Radiguer); Université Paris-Saclay, AP-HP, Centre de Recherche Clinique Paris-Saclay, DMU 13 Santé Publique, Information Médicale, Appui à la Recherche Clinique, INSERM U1018, CESP, Le Kremlin-Bicêtre, France (Lecoq, Becquemont); Université Paris-Saclay, AP-HP, Service de Radiologie Diagnostique et Interventionnelle, Hôpital de Bicêtre, Le Kremlin-Bicêtre, France (Meyrignac, Bellin); Université Paris-Saclay, AP-HP, Service de Médecine Interne et Immunologie Clinique, Hôpital de Bicêtre, DMU 7 Endocrinologie-Immunités-Inflammations Cancer-Urgences, Le Kremlin-Bicêtre, France (Noel, Hardy-Leger); Université Paris-Saclay, AP-HP, Service de Gériatrie Aiguë, Hôpital de Bicêtre, DMU 1 Médecine Territoire Gériatrie, Le Kremlin-Bicêtre, France (Baudry, Verny); Université Paris-Saclay, AP-HP, Service de Physiologie et d'Explorations Fonctionnelles Respiratoires, Hôpital de Bicêtre, DMU 5 Thorinno, Inserm UMR_S999, Le Kremlin-Bicêtre, France (Beurnier); Université Paris-Saclay, AP-HP, Service de Microbiologie, Hôpital de Bicêtre, DMU 15 Biologie-Génétique-PUI, INSERM 1193, Le Kremlin-Bicêtre, France (Dortet); Université Paris-Saclay, AP-HP, Service des Maladies Infectieuses et Tropicales, Hôpital de Bicêtre, DMU 7 Endocrinologie-Immunités-Inflammations Cancer-Urgences, INSERM U1018, CESP, Le Kremlin-Bicêtre, France (Wyplosz); Université Paris-Saclay, AP-HP, Service de Néphrologie Transplantation, Hôpital de Bicêtre, DMU 4 CORREVE Maladies du Cœur et des Vaisseaux, Le Kremlin-Bicêtre, France (Zaidan).

Author Contributions: Dr Monnet had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Morin, Savale, Pham, Colle, Figueiredo, Harrois, Gasnier, Lecoq, Meyrignac, Noel, Baudry, Bellin, Beurnier, Corruble, Dortet, Hardy-Leger, Sportouch, Wyplosz, Zaidan, Becquemont, Montani, Monnet.

Acquisition, analysis, or interpretation of data: Morin, Savale, Pham, Colle, Figueiredo, Harrois, Gasnier, Lecoq, Meyrignac, Noel, Bellin, Beurnier, Choucha, Corruble, Dortet, Radiguer, Sportouch, Verny, Wyplosz, Zaidan, Becquemont, Montani, Monnet.

Drafting of the manuscript: Morin, Savale, Pham, Colle, Figueiredo, Harrois, Meyrignac, Noel, Baudry, Bellin, Beurnier, Corruble, Dortet, Hardy-Leger, Sportouch, Wyplosz, Zaidan, Becquemont, Montani, Monnet.

Critical revision of the manuscript for important intellectual content: Morin, Savale, Pham, Colle, Figueiredo, Harrois, Gasnier, Lecoq, Meyrignac, Noel, Baudry, Bellin, Beurnier, Choucha, Corruble, Dortet, Radiguer, Verny, Wyplosz, Zaidan, Becquemont, Montani, Monnet.

Statistical analysis: Morin, Pham, Colle, Figueiredo, Noel, Dortet, Radiguer, Wyplosz, Monnet.

Obtained funding: Savale, Baudry, Wyplosz, Becquemont, Monnet.

Administrative, technical, or material support: Morin, Pham, Figueiredo, Harrois, Lecoq, Baudry, Bellin, Beurnier, Sportouch, Wyplosz, Becquemont, Montani, Monnet.

Supervision: Morin, Savale, Pham, Figueiredo, Noel, Baudry, Choucha, Corruble, Dortet, Sportouch, Wyplosz, Zaidan, Becquemont, Montani, Monnet.

Other: Sportouch.

Conflict of Interest Disclosures: Dr Harrois reported personal fees from LFB Laboratory and Edwards, and fees from Mindray Fees to his institution, outside the submitted work. Dr Lecoq reported nonfinancial support from Pfizer SAS and from Sanofi Aventis France outside the submitted work. Dr Noel reported personal fees from MSD and from Biogen outside the submitted work. Dr Beurnier reported personal fees from Sanofi outside the submitted work. Dr Becquemont reported grants from Sanofi Genzyme for a presentation concerning Gaucher disease, and pharmacogenetics and drug interactions with Eliglustat outside the submitted work. Dr Montani reported personal fees from Actelion, GSK, Pfizer, MSD, Chiesi, Boehringer, and Incyte Biosciences France; grants from Bayer and MSD; and nonfinancial support from Acceleron, outside the submitted work. Dr Monnet reported personal fees from Getinge Pulsion Medical and Baxter outside the submitted work. No other disclosures were reported.

Funding/Support: This study was funded by Assistance Publique-Hôpitaux de Paris.

Role of the Funder/Sponsor: The funder was not involved in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. The sponsor did not have the right to veto publication or to control the decision regarding to which journal the paper was submitted.

Group Information: The COMEBAC study investigators are listed in Supplement 2.

Data Sharing Statement: The data sets generated during the current study are available from the corresponding author on request.

Additional Contributions: We thank Thomas Piraino, RRT, FCST (Department of Anesthesia, Division of Critical Care, McMaster University), for English editing. He did not receive financial compensation for his contribution. We thank Jackie Abdel Nour, Younes Boujana, Marie Deloncle, Tatiana Fouasse, Nassim Hamidat, Mario Hindie, Yasmine Iftene, Jean-Vincent Laqua, Jules Madjarian, Dalil Sadki, and Sabrina Safer, all from Paris-Saclay Medical School, for their help in data collection. They received financial compensation for their contribution. We thank Cassandre Sainrose, Paris 8 University, for her help in data collection. She did not receive financial compensation for her contribution. We thank the patients who participated in this cohort, as well as the physicians, psychologists, nurses, care helpers, biologists, pharmacists, other allied health professionals, and the administrators of the Bicêtre Hospital.