key: cord-0684692-d4icqm34
authors: Arlotto, Sylvie; Legueult, Kevin; Blin, Alice; Cortaredona, Sebastien; Giraud-Gatineau, Audrey; Bailly, Laurent; Jimeno, Marie-Thérèse; Delorme, Léa; Brouqui, Philippe; Lagier, Jean-Christophe; Million, Matthieu; Dellamonica, Jean; Colson, Philippe; Carles, Michel; Raoult, Didier; Pradier, Christian; Gentile, Stéphanie
title: Who Were Hospitalized Deceased Patients from COVID-19 During the First Year of Pandemic? Retrospective Analysis of 1104 Deceased Patients in South of France
date: 2022-04-29
journal: J Epidemiol Glob Health
DOI: 10.1007/s44197-022-00039-3
sha: ace3ea88545398248b24e136b99ebecb22fd6ebf
doc_id: 684692
cord_uid: d4icqm34

INTRODUCTION: Following the first year of the COVID-19 pandemic, a complete analysis of the characteristics of the deceased hospitalized patients was performed, to identify factors related to premature mortality and to compare patient profiles according to the epidemic periods. METHODS: Retrospective analysis of 1104 deceased patients in two University Hospitals in South-eastern France, between March 1, 2020 and March 12, 2021 from Hospital’s electronic medical records was performed. RESULTS: Mean age was 80 years (± 11.1) and 10% of the deceased were younger than 65 years with specific comorbidities, e.g., genetic conditions, metastatic cancer, or massive obesity. Among the three clusters identified, two clusters (75% of deceased patients) include very elderly patients with numerous comorbidities, and differ by their proportion of dependent institutionalized patients. The third cluster is made up of younger patients with fewer but severe comorbidities. Deceased patients’ profiles varied according to the epidemic periods: during the first period (March–June 2020), more patients were institutionalized. The second period (September–December2020) coincided with a higher mortality rate. CONCLUSIONS: This study confirmed that most patients hospitalized and dying from COVID-19 were frail, i.e., elderly and/or highly comorbid and that the small proportion of young patients had severe comorbidities. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s44197-022-00039-3.

The World Health Organization (WHO) declared coronavirus disease 2019 (COVID-19) a pandemic on March 11, 2020 [1] . By March 12, 2021, COVID-19 had spread to 119 M people, resulting in 2.7 M deaths worldwide [2] .

The first case of COVID-19 in France was diagnosed on January 24, 2020; by March 12, 2021 , there were 4 M COVID-19 cases and 90,207 deaths.

As in the rest of Europe, France has experienced successive epidemic periods corresponding to different SARS-CoV-2 variants. French regions were not equally impacted during these different periods. Indeed, clear spatial heterogeneity of in-hospital COVID-19 incidence and mortality rates were found, following the spread of the epidemic according to various factors as well as population age structure, urban development and population density, economic level, health system, climatic and meteorological factors, and anticontagion policies and practices [3] . Only two regions (Ile de France and Grand Est) had a higher mortality rate due to COVID-19 compared to other regions, but only during the first epidemic period [4] . The South-Eastern region of France was particularly affected, with 35,053 patients hospitalized for COVID-19, including 6140 deaths [5] .

However, little is known of the medical and sociodemographic profile of deceased COVID-19 patients in Europe. As the available data are mainly derived from hospital information systems and/or death certificates, their quality is generally incomplete notably for sociodemographic characteristics. To our knowledge, no comprehensive study has focused on deceased COVID-19 patients in France, during the first year of pandemic.

Indeed, out of the 7 published studies that have investigated the profile of deceased patients, 5 are Chinese studies on the Wuhan epidemic [6] [7] [8] [9] [10] , one is an Ethiopian study [11] and finally one study was conducted by an Italian team [12] All of these studies focused on the first months of epidemic, with limited study periods, from 1 to 2 months. Moreover, cohorts of COVID-19 deceased patients were small, between 14 and 320 patients.

However, all these studies already revealed that deceased patients were older and suffered more often from underlying comorbidities, particularly cardiovascular disease and diabetes. The Chinese studies focused mainly on the cause of death as well as on biological markers of disease severity.

Since then, other studies have investigated the risk factors for death in hospitalized COVID-19 patients and report that advanced age, as well as the presence of multiple comorbidities such as hypertension or diabetes, is uniform criteria regardless of the epidemic period and the associated variant [13] [14] [15] [16] .

For these reasons, the objective of this paper is to present a complete sociodemographic and clinical profile of patients who died in the two University Hospitals of Southeastern France, in Marseille and in Nice during the first year of COVID-19 pandemic. The secondary objectives were to identify the profiles related to premature mortality, i.e., before 65 years of age [17] and to compare deceased patients' characteristics according to the different epidemic periods.

This was a retrospective analysis of all in-hospital deceased patients with a SARS-CoV-2 PCR-positive test in the University Hospitals of Marseille and Nice between March 1, 2020, and March 12, 2021. Among the 32 university hospital centers in France, only 2 are located in the Provence Alpes Côte d'Azur region: in Nice and Marseille. University hospitals are public institutions that are located in a major city and have highly specialized departments as well as research and teaching units. Then, they treat ordinary patients and COVID-19 patients and also have intensive care units.

Patients with positive PCR test but for whom death was not attributable to COVID-19 were excluded.

We collected data from the medical records: sex, age, lifestyle profile, i.e., living environment, level of autonomy and bedridden state, the care pathway, i.e., circumstances of admission, transfer to the intensive care unit and the comorbidities.

All medical record data were collected anonymously. Patient and Public Involvement: no patient was involved. The number of hospitalizations of patients with COVID-19 was extracted from the hospital's information system.

We collected the date of the first positive PCR test for each patient to compare deceased patients' profile according to the epidemic period [18] . The first epidemic period (Wuhan-Hu-1 SARS-CoV-2) lasted from March 1, 2020 to June 15, 2020; the second one from September 1, 2020 to December 31, 2020 (B.1.160 variant predominated) [19] ; and the third one from January 2021 to the end of June 2021 (B.1.1.7 variant predominated) [20] . Variants were named according to the Pangolin classification [21] .

In Marseille, a small outbreak between July and August 2020 involved a B.1.416 variant with few deaths which were excluded from the comparison [19, 22] .

An illustration of the different COVID-19 epidemic period and predominant variants as a function of time in South-eastern France is available in supplementary Fig. SF1. 

Age was grouped into the following classes: < 65 years; 65-74 years; 75-84 years; 85 years and over.

All data on deceased patients were compared between the different epidemic periods as defined above.

Categorical variables are presented as n (%) and continuous variables as mean (std) (or median and interquartile range in those with no criteria of normal distribution). We used the Wilcoxon Mann-Whitney test, Student's t test, χ 2 test, or Fisher's exact test to compare differences between groups of patients.

To allocate patients to groups as suggested by the recorded data, we performed a multiple correspondence analysis (MCA) combined with an agglomerative hierarchical cluster analysis (AHCA) [23] .

To identify factors specific to patients who died before age 65 (i.e., premature mortality), we performed a multivariable logistic regression. Two regression models were constructed, one with the obesity variable (BMI ≥ 30), the other with the massive obesity variable (BMI ≥ 40). Any statistically significant factor in univariate analyses (p < 0.1) was selected as a potential candidate for the multivariable logistic model.

All analyses involved two-sided p values, with statistical significance defined by p ≤ 0.05. Statistical analyses were performed with SPSS statistical software version 20. The MCA and AHCA were performed using the R 3.5.3 software and the FactomineR package [24] .

Between March 1, 2020 and March 12, 2021, 10,565 patients were hospitalized for COVID-19 in the University Hospitals of Marseille and Nice (7464 and 3101, respectively), of which 1173 died (793 and 311, respectively) ( Fig. 1 ).

Of these deaths, 69 were not attributable to COVID-19 and were excluded. These had PCR-positive test but no related COVID-19 symptoms. Deaths were attributed to metastatic cancer (27 patients), trauma (11 patients), septic shock (9 patients), stroke (9 patients), cardiac decompensation (7 patients) or aspiration pneumonia (6 patients).

A total of 1104 (94%) deaths were, therefore, attributable to COVID-19, i.e., a mortality rate of 10.5%. (10% for Nice and 10.6% for Marseille).

All deceased patients' characteristics for each of the two University Hospitals are presented in supplementary Table  ST1 .

The average age of the deceased patients was 80 years; the proportion of patients under 65 years of age was higher in Marseille. Two-thirds of deceased patients were men, and half of the deceased patients had loss of autonomy.

Two-thirds of the deceased patients were hypertensive, one-third were diabetic, and one in six was a patient with obesity. The prevalence of comorbidities was similar in both University Hospitals except for heart disorders, diabetes and genetic conditions. Deceased patients in Nice had a higher Over half (58%) of the deceased patients came from home, 21% from an institution and 21% from a previous hospitalization. Fewer deceased patients in Nice had been admitted in intensive care.

Deceased Patient Clusters Three clusters of deceased patients were identified by AHCA (Fig. 2) .

Each cluster includes patients with common sociodemographic characteristics and comorbidities. A comparison of the deceased COVID-19 patients' characteristics according to the 3 clusters is presented in Table 1 . Figure 2 displays an age gradient on the X axis and a comorbidity frequency gradient on the Y axis. Thus, elderly and comorbid patients are shown on the bottom left of the factorial plane and the young people without comorbidities on the top right.

The first cluster accounted for 23.8% of the deceased patients, (red dots in Fig. 2 ), mainly including institutionalized patients (90.5%). Most of the patients in this cluster were very elderly (84.9 years on average) and bedridden (43.3%), more than half of them were women (54%) while there were mostly men in the other two clusters. These patients were more likely to die within the first 24 h of admission (18%), and more often in the emergency department (16.7%). For these patients, the most prevalent comorbidities were neurological (67.3%) and psychiatric (36.5%) conditions.

The second cluster (in green) accounted for 50.9% of the deceased patients and mainly includes elderly patients (83.8 years on average) with loss of autonomy but not institutionalized and with several underlying conditions (4 comorbidities on average), e.g., active tumors, heart and respiratory disorders.

Finally, the third cluster accounted for 25.2% of the deceased patients (blue dots in Fig. 2 ). These patients were younger (67.7 years on average). They were more likely to be diabetic (38.5%), obese (27.3%), of which 19.7% with massive obesity, and to suffer from pre-existing respiratory disease (29.5%). They were more likely to smoke (30.9%). A small percentage had specific comorbidities: genetic conditions (5.4%), liver disease (9.7%) and autoimmune disorder (8.3%). Their clinical pathway was different: most of them were admitted to the intensive care (85.3%), 40.6% came from other hospitals in the region, as well as the 6 patients who died without diagnosed comorbidity. In this cluster, patients were more numerous during the third epidemic period. 

The epidemic periods (P1, P2, and P3) differed in terms of duration and of number of cases and deaths. The hospital mortality rate was higher in P2. Deceased patients' profile also varied between the different epidemic periods (Table  ST2 ). During P1, there were more institutionalized patients among the deceased. During P2, deceased patients were younger and were more often transferred to an intensive care. Patients who died during the P3 had more heart disorders and more often hypertensive but had fewer neurological conditions.

Of the 1104 deceased patients, 9.5% were under 65 years. Table 2 presents the characteristics of these deceased patients.

Among these 105 who were under 65, 21 had fewer than two comorbidities (1.9% of all deceased patients): 16 were male and only 2 patients were under 40 years. All had been admitted to the intensive care, except for two terminally ill patients with a condition other than COVID-19. Among the patients admitted to the intensive care, half of them were admitted within the first 24 h of hospitalization.

Two multivariate analysis models were performed to identify the characteristics of these younger patients ( Table 2 ). The first model, using BMI > 30, showed that they were 6 times more likely to suffer from a genetic condition and 4 times more likely to have a metastatic cancer and to be hospitalized in intensive care. The second model, using BMI ≥ 40, yielded the same factors but found that patients who died before the age of 65 were 3 times more likely to have massive obesity.

Our study shows that 90% of deceased patients during the first year of the epidemic were at least 65 years old. In addition to their very poor state of health related to age, most of them suffered from severe chronic conditions. This finding is consistent with data already published in the literature showing that the risk of death due to COVID-19 strongly depends on patients' age and previous health status [6] [7] [8] [9] [10] [11] 25] .

Among the three distinct clusters, the first two (3/4 of the deceased patients) are very old and comorbid patients: one of the two cluster is represented by institutionalized patients, often bedridden and demented, with neurological and/or psychiatric conditions; the second concerns elderly patients living at home who have lost their autonomy. These elderly patients suffer from heart disorders, hypertension, respiratory diseases, diabetes and malignancy, according to literature review [6, 12, [26] [27] [28] [29] [30] . They are, therefore, frail patients who were particularly at risk of dying from an acute event such as COVID-19 [31] , a situation already described during influenza epidemics [32] , supporting the recommendation to target these populations for vaccination.

Finally, the third cluster includes younger patients, with fewer comorbidities but severe, i.e., genetic diseases and metastatic cancers, and/or massive obesity, testifying to their frailty, who are opportunely targeted by current vaccination campaigns [33] .

In our study, deceased patients under 65 years of age represented 10%, similar to that reported by Biagi et al. [12] .

Only 2% of patients under 65 years of age had fewer than two comorbidities, suggesting potentially undiagnosed comorbidities. Several articles already focused on this subject and showed that the infection appears to be more severe in certain patients, suggesting that genetic or epigenetic factors are at play [34] . Surprisingly, although univariate analysis revealed that patients younger than 65 years had more autoimmune and liver disease, multivariate analysis did not identify these factors as being related to premature mortality. In addition, although patients younger than 65 years were significantly more obese, the multivariate analysis did not identify obesity (BMI > 30) as a factor related to premature mortality. Obesity (BMI > 30) and autoimmune and liver disease appear to be a factor in hospitalization and severity, but not necessarily a factor related to premature mortality.

Since the elderly are those most at risk of death, the role of geriatricians in the management of these patients should be considered. In the literature, several authors deplore the fact that the context of a health emergency has given rise only to acute management without multidisciplinary [35] , as is their tolerance to treatment, notably to corticosteroids [36] . In addition, the isolation to which the elderly were exposed during the pandemic and the resulting loneliness may have increased their physical and psychological frailty [37] . If the epidemics of COVID-19 were to continue, frailty should be addressed to provide the most effective therapy and multidisciplinary and geriatric management should be anticipated [36] . In addition, to avoid the negative consequences of isolation and loneliness, multicomponent programs with adequate strategies [38] . The analysis of the profile of the deceased patients according to the different epidemic periods shows some differences in profile even if globally the elderly and with comorbidities subject remains the first victim. During the first epidemic period, we observed a higher proportion of patients coming from institutions for elderly people. This can be explained by the lack of knowledge of the disease, the lack of means to test patients and of personal protective equipment in addition to the lockdown of France during this first period, which led to the fact that only public hospitals were able to take care of patients with COVID-19.

The second epidemic period lasted much longer and was more deadly. The mortality rate of hospitalized patients was also higher and the proportion of young patients was higher as well the rate of admission to intensive care, despite improved medical management and better knowledge of COVID-19 disease. These results show the virulence of this variant [39] . In view of such variation, epidemiological surveillance of deceased patient should continue to better anticipate a possible change of target.

A limitation of this study is that our analysis focused on deaths in two reference hospitals for complex cases requiring high-level resuscitation, which certainly led to overestimate the proportion of "young" subjects. Our patients represent only 18% of all deceased patients hospitalized in our region [40] .

As the peak of the third epidemic period was reached on April 12, 2021 and the epidemic ended at the end of June 2021 [20] , we cannot conclude on the impact of the B.1.1.7 variant on the profile of deceased patients. Since the end of this work, a fourth period has taken place in France with a new variant (B.1.617.2). The data on the profile of the patients are not known.

This study confirms that most patients hospitalized and dying from COVID-19 were elderly and/or highly comorbid. In addition, premature mortality concerns 10% of our population of deceased patients, with specific profiles, e.g., autoimmune and genetic conditions. This reinforces the arguments for the vaccine strategy including these at-risk populations as a priority and suggests the need for multidisciplinary care, involving a geriatrician, for elderly and frail patients. 

WHO declares COVID-19 a pandemic

COVID-19 public data

Factors associated with the spatial heterogeneity of the first wave of COVID-19 in France: a nationwide geo-epidemiological study

Parcours hospitalier des patients atteints de la Covid-19 lors de la première vague de l'épidémie

Available from: https:// geodes. sante publi quefr ance. fr/# bbox= 29024 8,56830 30,74938 3,41533 5&c= indic ator&f= 0&i= covid_ hospit. rad&s= 2021-03-12& selco dgeo= 93&t= a01& view= map1

Clinical characteristics of 113 deceased patients with coronavirus disease 2019: retrospective study

Clinical characteristics of 14 COVID-19 deaths in Tianmen, China: a single-center retrospective study

Analysis of 92 deceased patients with COVID-19

Clinical characteristics and immune responses of 137 deceased patients with COVID-19: a retrospective study. Front Cell Infect Microbiol

Comparison of clinical characteristics among younger and elderly deceased patients with COVID-19: a retrospective study

Clinical characteristics of COVID-19 related deaths in ethiopia

Clinical and epidemiological characteristics of 320 deceased patients with COVID-19 in an Italian province: a retrospective observational study

A systematic review and meta-analysis of inpatient mortality associated with nosocomial and community COVID-19 exposes the vulnerability of immunosuppressed adults

Clinical characteristics and risk factors for mortality in COVID-19 inpatients in Birjand, Iran: a single-center retrospective study

Trends and risk factors of in-hospital mortality of patients with COVID-19 in Germany: results of a large nationwide inpatient sample

Mortality by COVID-19 Before Vaccination -One Year Experience of Hospitalized Patients in Madrid

Indicateurs de mortalité 'prématurée' et 'évitable

Clinical outcomes in COVID-19 patients infected with different SARS-CoV-2 variants in Marseille, France. Clin Microbiol Infect

Emergence and outcomes of the SARS-CoV-2 'Marseille-4' variant

Coronavirus : circulation des variants du SARS-CoV-2

A dynamic nomenclature proposal for SARS-CoV-2 lineages to assist genomic epidemiology

Introduction into the Marseille geographical area of a mild SARS-CoV-2 variant originating from sub-Saharan Africa: An investigational study

Findings Groups in Data: An Introduction to Cluster Analysis

Wiley Series in probability and Statistics

FactoMineR: An R Package for Multivariate Analysis

Risk factors for intensive care unit admission and inhospital mortality among hospitalized adults identified through the US coronavirus disease 2019 (COVID-19)-associated hospitalization surveillance network (COVID-NET)

Factors associated with clinical severity in emergency department patients presenting with symptomatic SARS-CoV-2 infection

COVID-19 and autoimmune diseases

COVID-19 and Down syndrome

Impact of sex and metabolic comorbidities on coronavirus disease 2019 (COVID-19) mortality risk across age groups: 66 646 inpatients across 613 US hospitals

The effect of age on mortality in patients with COVID-19: a meta-analysis with 611,583 subjects

Frailty hinders recovery from influenza and acute respiratory illness in older adults

Obesity and mortality among patients diagnosed with COVID-19: results from an integrated health care organization

Pathological features in 100 deceased patients with COVID-19 in correlation with clinical and laboratory data

Clinical characteristics of older patients: The experience of a geriatric short-stay unit dedicated to patients with COVID-19 in France

COVID-19 and the elderly: insights into pathogenesis and clinical decision-making

COVID-19 in the geriatric population

Exercise, nutrition, and medication considerations in the light of the COVID pandemic, with specific focus on geriatric population: a literature review

Aflak Kattar M, Bleistein L, El Rais H. Parcours hospitaliers des patients atteints de la Covid-19 de mars 2020 à janvier 2021

COVID-19) en France du 1er mars 2020 au 1er juillet 2021, selon la tranche d'âge

Acknowledgements A. Gares, S. Tumelero, Erasmus students.