As of May 4, 2023, the department of Paris had the highest number of hospitalizations due to the coronavirus, with 952 patients. From a national perspective, roughly 12.7 thousand people were currently hospitalized with a COVID-19 diagnosis in France.For further information about the coronavirus (COVID-19) pandemic, please visit our dedicated Facts and Figures page.

An evaluation was conducted to predict the economic and clinical burden of vaccinating all immunocompromised (IC) individuals aged ≥30 years with mRNA-1273 variant-adapted COVID-19 vaccines versus BNT162b2 variant-adapted vaccines in Fall 2023 and Spring 2024 in France. The number of symptomatic SARS-CoV-2 infections, hospitalizations or deaths due to COVID-19, and long COVID cases, costs and quality-adjusted life years (QALYs) were estimated using a static decision-analytic model. Predicted vaccine effectiveness (VE) were based on real-world data from the original and BA.4/5 variant-adapted vaccines, suggesting higher protection against infection and hospitalization with mRNA-1273 vaccines. VE estimates were combined with COVID-19 incidence and probability of COVID-19 severe outcomes. Uncertainty surrounding VE, vaccine coverage, infection incidence, hospitalization and mortality rates, costs and QALYs were evaluated in sensitivity analyses. In an ideal situation where 100% coverage is achieved, the mRNA-1273 variant-adapted vaccine is predicted to prevent an additional 3,882 infections, 357 hospitalizations, 81 deaths, and 326 long COVID cases when compared to BNT162b2 variant-adapted vaccines in 230,000 IC individuals. This translates to €10.1 million cost-savings from a societal perspective and 645 QALYs gained. Results were consistent across all analyses and most sensitive to variations surrounding VE and coverage. These findings highlight the importance of increasing vaccine coverage, and ability to induce higher levels of protection with mRNA-1273 formulations in this vulnerable population.

BackgroundCOVID-19 infection is less severe among children than among adults; however, some patients require hospitalization and even critical care. Using data from the French national medico-administrative database, we estimated the risk factors for critical care unit (CCU) admissions among pediatric COVID-19 hospitalizations, the number and characteristics of the cases during the successive waves from January 2020 to August 2021 and described death cases.MethodsWe included all children (age < 18) hospitalized with COVID-19 between January 1st, 2020, and August 31st, 2021. Follow-up was until September 30th, 2021 (discharge or death). Contiguous hospital stays were gathered in “care sequences.” Four epidemic waves were considered (cut off dates: August 11th 2020, January 1st 2021, and July 4th 2021). We excluded asymptomatic COVID-19 cases, post-COVID-19 diseases, and 1-day-long sequences (except death cases). Risk factors for CCU admission were assessed with a univariable and a multivariable logistic regression model in the entire sample and stratified by age, whether younger than 2.ResultsWe included 7,485 patients, of whom 1988 (26.6%) were admitted to the CCU. Risk factors for admission to the CCU were being younger than 7 days [OR: 3.71 95% CI (2.56–5.39)], being between 2 and 9 years old [1.19 (1.00–1.41)], pediatric multisystem inflammatory syndrome (PIMS) [7.17 (5.97–8.6)] and respiratory forms [1.26 (1.12–1.41)], and having at least one underlying condition [2.66 (2.36–3.01)]. Among hospitalized children younger than 2 years old, prematurity was a risk factor for CCU admission [1.89 (1.47–2.43)]. The CCU admission rate gradually decreased over the waves (from 31.0 to 17.8%). There were 32 (0.4%) deaths, of which the median age was 6 years (IQR: 177 days–15.5 years).ConclusionSome children need to be more particularly protected from a severe evolution: newborns younger than 7 days old, children aged from 2 to 13 years who are more at risk of PIMS forms and patients with at least one underlying medical condition.

The COVID-19 pandemic caused an increase in the number of patients in French intensive care units. On August 29, 2022, the occupancy rate of coronavirus infected patients in resuscitation units stood at 16.8 percent in France. The occupancy rate in intensive care reached a critical point thrice since the pandemic began. The first peak was recorded in April 2020, the second one in November 2020, and the third one recently, as of April 20, 2021.

As of September 2022, the number of patients in intensive care due to the coronavirus in France stood at roughly 800. At that time, there were over 14 patients hospitalized due to COVID-19 in France.

To get further information about the coronavirus (COVID-19) pandemic, please refer to our dedicated Facts & Figures page.

Santé publique France's mission is to improve and protect the health of populations. During the health crisis linked to the COVID-19 epidemic, Public Health France is responsible for monitoring and understanding the dynamics of the epidemic, anticipating the different scenarios and implementing actions to prevent and limit the transmission of this virus on the national territory.

Daily hospital data relating to the COVID-19 epidemic by department and sex of the patient: number of hospitalized patients, number of people currently in intensive care or intensive care, cumulative number of people returned home, cumulative number of people who died.

For some patients, gender was not identified in the database. This can lead to a discrepancy between the H/F sum of an indicator and the total number of this indicator.

The region and iso 3166-1 codes of the zones have been added.

Warning: data under construction. May contain anomalies or missing data.

Laboratory measures on admission of confirmed Covid-19 hospitalized patients at Lyon University Hospitals, France.

Introduction: COVID-19 was found to be associated with an increased risk of stroke. This study aimed to compare characteristics, management, and outcomes of hospitalized stroke patients with or without a hospital diagnosis of COVID-19 at a nationwide scale. Methods: This is a cross-sectional study on all French hospitals covering the entire French population using the French national hospital discharge databases (Programme de Médicalisation des Systèmes d’Information, included in the Système National des Données de Santé). All patients hospitalized for stroke between 1 January and 14 June 2020 in France were selected. A diagnosis of COVID-19 was searched for during the index hospitalization for stroke or in a prior hospitalization that had occurred after 1 January 2020. Results: Among the 56,195 patients hospitalized for stroke, 800 (1.4%) had a concomitant COVID-19 diagnosis. Inhospital case-fatality rates were higher in stroke patients with COVID-19, particularly for patients with a primary diagnosis of COVID-19 (33.2%), as compared to patients hospitalized for stroke without COVID-19 diagnosis (14.1%). Similar findings were observed for 3-month case-fatality rates adjusted for age and sex that reached 41.7% in patients hospitalized for stroke with a concomitant primary diagnosis of COVID-19 versus 20.0% in strokes without COVID-19. Conclusion: Patients hospitalized for stroke with a concomitant COVID-19 diagnosis had a higher inhospital and 3 months case-fatality rates compared to patients hospitalized for stroke without a COVID-19 diagnosis. Further research is needed to better understand the excess of mortality related to these cases.

Objective The aim of this study was to develop an accurate regional forecast algorithm to predict the number of hospitalized patients and to assess the benefit of the Electronic Health Records (EHR) information to perform those predictions. Materials and Methods Aggregated data from SARS-CoV-2 and weather public database and data warehouse of the Bordeaux hospital were extracted from May 16, 2020, to January 17, 2022. The outcomes were the number of hospitalized patients in the Bordeaux Hospital at 7 and 14 days. We compared the performance of different data sources, feature engineering, and machine learning models. Results During the period of 88 weeks, 2561 hospitalizations due to COVID-19 were recorded at the Bordeaux Hospital. The model achieving the best performance was an elastic-net penalized linear regression using all available data with a median relative error at 7 and 14 days of 0.136 [0.063; 0.223] and 0.198 [0.105; 0.302] hospitalizations, respectively. Electronic health r..., Aggregated data from 2020-05-16 to 2022-01-17 regarding Bordeaux Hospital EHR. Bordeaux hospital data warehouse was used, during the pandemic, to describe the current state of the epidemic at the hospital level on a daily basis. Those data were then used in the forecast model including: hospitalizations, hospital and ICU admission and discharge, ambulance service notes and emergency unit notes. Concepts related to COVID-19 were extracted from notes by dictionary-based approaches (e.g. cough, dyspnoea, covid-19). Dictionaries were manually created based on manual chart review to identify terms used by practitioners. Then, the number and proportion of ambulance service calls or hospitalization in emergency units mentioning concepts related to covid-19 were extracted. Due to different data acquisition mechanisms, there was a delay between the occurrence of events and the data acquisition. It was of 1 day for EHR data, 5 days for department hospitalizations and RT-PCR, 4 days for weather, 2..., Data are stored in a .rdata file. Please use R (https://www.r-project.org/) software to open the data.

Dataset

This dataset was created by laura jezequel

Contents

Since December 2019, a new agent, the SARS-Cov-2 coronavirus has been rapidly spreading from China to other countries causing an international outbreak of respiratory illnesses named COVID-19. In France, the first cases have been reported at the end of January with more than 60000 cases reported since then. A significant proportion (20-30%) of hospitalized COVID-19 patients will be admitted to intensive care unit. However, few data are available for this special population in France.

We conduct a large observational cohort of ICU suspected or proven COVID-19 patients that will enable to describe the initial management of COVID 19 patients admitted to ICU and to identify factors correlated to clinical outcome.

IMPORTANT INFORMATION

This Thursday 4 November 2021, the format of the variable sursaud_cl_age_corona will be modified:

New coding: 0 = "All ages" 1 = "0-4 years" 2 = "5-14 years" 3 = "15-44 years" 4 = "45-64 years" 5 = "65-74 years" 6 = "75 years or more"

Former current: 0 = "All ages" A = "under 15 years of age" B = "15-44 years" C = "45-64 years" D = "65-74 years" E = "75 years or more"

Public Health France's actions

Public Health France's mission is to improve and protect the health of populations. During the health crisis linked to the COVID-19 epidemic, Public Health France is responsible for monitoring and understanding the dynamics of the epidemic, anticipating the different scenarios and implementing actions to prevent and limit the transmission of this virus on the national territory.

Description of the dataset

Two files are proposed:

• Daily data from SOS doctors and hospital emergencies by department, gender and age group of patients: number of emergency visits for suspected COVID-19, total number of emergency visits with an informed medical diagnosis, number of hospitalisations among emergency visits for suspected COVID-19, total number of medical procedures SOS Doctors for suspected COVID-19, total number of medical procedures SOS Doctors with an informed medical diagnosis.

• Daily data from SOS doctors and hospital emergencies by region, gender and age groups of patients: number of emergency visits for suspected COVID-19, total number of emergency visits with an informed medical diagnosis, number of hospitalisations among emergency visits for suspected COVID-19, number of medical procedures SOS Doctors for suspected COVID-19, total number of medical procedures SOS Doctors with an informed medical diagnosis. The region code corresponds to the INSEE code published by INSEE on www.data.gouv.fr.

Anomalies detected in files

Please note that some files may contain anomalies due to data collection difficulties.

Error reports are published daily by Etalab in the Community resources section: - Error report for the daily file; - Error report for the weekly file.

Please note that: - For some patients, the age group was not identified in the database. This may lead to a discrepancy between the sum of all age classes of an indicator and the total number of that indicator; - For some patients, gender was not identified in the database. This can lead to a discrepancy between the sum H/F of an indicator and the total number of that indicator.

Do not hesitate to report any other anomalies in the comments. These comments will be communicated to the team in charge of data collection and dissemination.

Other available data

Public Health France also publishes the hospital data relating to the COVID-19 outbreak.

COVID-19 cumulative outbreaks in nursing homes and cumulative incidence of inhabitant hospitalizations for COVID-19 in the Départements of Auvergne-Rhône-Alpes Region, France, March 1–July 31, 2020.

COVID-19 case, hospitalisation, death, seroprevalence, vaccination and population data, and age-dependent contact rate, severe burden risk and vaccine effectiveness parameter estimates, required to fit model and run simulations in article "Impact of vaccinations, boosters and lockdowns on COVID-19 waves in French Polynesia"

Estimated durations of the active periods of COVID-19 outbreaks in nursing homes, estimated delays and values of the outbreak probability peaks in the nursing homes and the hospitalization probability peaks for COVID-19 in the populations of Auvergne-Rhône-Alpes Region Départements, France, March 1–July 31, 2020.

1. Context

A variety of research is currently being developed in order to predict the future of the current covid-19 pandemic. Among all models, the SIR and other compartmental models are the main tool for explainable, extrapolation-robust predictions.

Our idea is to provide the community with an extensive set of socio-demographics (and health) indicators, with the aim of explaining/predicting the variability in the pandemic dynamics. There is already a lot of literature covering these kind of indicators, for instance people study contact matrices, which account for the typical structure of contacts that various segments of the population have (in short: who has contacts with who?).

However we haven't found systematic, Machine-Learning based studies of which indicators most impact the R0 (reproduction number) or other coefficients that are usually fitted to the (time-series) data of infected/hospitalized/deaths/etc. There are some nice initiatives similar to ours that have spawned on kaggle, and have been noticed (see the panel /datasets at https://www.kaggle.com/covid-19-contributions , in particular check out https://www.kaggle.com/jieyingwu/covid19-us-countylevel-summaries#counties.csv).

Given the variety of indicators we provide here, we expect one should be able to predict the department-to-department variations of the empirical coefficient R0, but also of other rates, such has the rate of the process (infected->hospitalized), and to some extent, the rates of (hospitalized->resuscitation), or the rates of ([various states]-> dead). This means the prediction deals both with the spread of the pandemic and the severity of its impact on people's lives and on the health system.

Other data sources, to complete this repo (including worldwide data): https://modcov19.math.cnrs.fr/publicdata/#numeric

2. Content

Here we provide a training+test set of 100 'examples': the 100 departments of France (we had to exclude Mayotte for lack of reliable/available data). They can be considered homogeneous in the sense that all indicators have been recorded in the same way (see below). Likewise, the time-series of hospitalized/in resuscitation/returned home/deaths are measured in a consistent way among the different French departments, since procedures and instructions are very similar everywhere. This is the advantage of remaining in a single country (here, France).

There are two kind of data we provide in the files.

The static data is an aggregation of socio-demographics and health indicators taken from the last couple of years (2016-2019). It is mostly curated by INSEE, the French national statistics institution, but INSEE itself is only a statistics-precessing place, and their data comes from other French agencies and from some surveys INSEE performs itself (like the census data). It comes as a single file, but it is actually the result of our concatenation of several databases (7 of them). Each original database has a separate original source that we provide in the metadata. Sometimes this source itself is a link to the INSEE website, which then details which agencies originally produced the data. For several of these files, we had to preprocess data from the city-level into the larger departmental level. All the original files coming from public institutions, and the codes that we used to pre-process them, are available at this gitlab: https://gitlab.inria.fr/flandes/covid-19-fr-socio-demographics.git

The time-series (or dynamic) data comes from Santé Publique France (also called Agence Nationale de Santé, ANS), one of the French Public Health agencies, which gathers data from hospitals and from the Regional Health Agencies (Agence Régionale de Santé, ARS). An additional time-series is the lockdown (confinement) level time series, that we produced ourselves (from reading the news, basically). This one is a bit particular, in the sense that it makes no sense to predict it, instead the level of lockdown (which decreases starting on May 5th) impacts the epidemic spread, and can be known in advance.

Note that the static data (many features) comes with feature names that start in a precise, regular way. We defined tags such that post-processing would be easy.

column names

• The first tag can be: Nbre, Pop, RateIncome, RateMedian, RatePoverty, Rate[whatever]
• The other tags are (in this order):
  • sex=all, sex=F, sex=H (F=Femme=Woman, H=homme=Man)
  • age=all, agemin=0_agemax=150 (or other values) (arbitrarily, the maximumof agemax is 150 years)
• The suffix is a description of the feature

There are many more features than there are examples in this data set, which means one has to be extremely cautious with over-fitting.

3. Acknowledgements

We thank INSEE and its partners for providing this wealth of data freely, and people wor...

Population and design: Monocentric cohort comparing COVID-19 patients exposed or not exposed to the treatment hydroxychloroquine (HCQ)+azithromycin (AZ) used as a standard of care in our center (Institut Hospitalo-Universitaire (IHU) Méditerranée Infection, Marseille, France). The inclusion period was from 2 March 2020 to 31 December 2021, with a follow-up period of 6 weeks. Inclusion and exclusion criteria: Data included were those of patients ≥ 18 years of age with PCR-proven COVID-19 disease regardless of symptoms (asymptomatic or symptomatic) and treated in our centre, i.e. having had a medical examination by one of the doctors in our centre (Institut Hospitalo-Universitaire (IHU) Méditerranée Infection, Marseille, France) either as outpatients or as inpatients, i.e. hospitalized on the day of the visit in our outpatient unit following evaluation or directly transferred from another medical ward except intensive care unit. The reasons for exclusion were an erroneous patient identification (identity surveillance and duplicates), lack of available medical data, lack of COVID-19 after checking the medical record (including patients without COVID-19 consulting for a post-COVID-19 syndrome), expression of opposition to the use of their medical data for research purposes (in accordance with European General Data Protection Regulation), and data from patients hospitalized in our center after intensive care. Data from COVID-19 outpatients left without medical advice were excluded. Methods Data source For inpatients: Between 03.03.20 and 12.03.2021, i.e. the first year of the epidemic, data were collected using the Electronic Patient Record. The DPI ‘”Dossier Patient Informatisé”) centralized all the medical information of a patient concerning his stays at the Assistance Publique - Hôpitaux de Marseille. During this input, a large amount of medical data concerning the treatment was collected (2799 patients). Between 13.03.21 and 31.12.2021, the data were extracted from the APHM administrative database (PASTEL). PASTEL does not contain medical data but only the following information: IPP (“Identifiant Permanent du Patient”) which is a unique identifier for each patient, age, sex, date of entry, the different services in which the patient may have been hospitalised and the date of discharge (1791 patients). For the outpatients, the medical cards of the patients who came to the outpatient unit filled in by the medical staff (11,725 patients). The content of these medical forms evolved during the epidemic, particularly with regard to the collection of patients' vaccination status and risk factors. At the outset, they contained information on prescribed treatments (29752 patients). The fusion of these databases was carried out via the IPP, and made it possible to obtain the total population of patients treated at the IHU with the following information: IPP, age, sex, outpatient care (Y/N), inpatient care (Y/N), date of entry and date of exit. If a patient had more than one episode of SARS-CoV-2 treated at the IHU, only the first stay was taken into account. If a patient was treated as an outpatient and then hospitalized, both pieces of information were kept. In the end, we obtain a database of 30,669 patients To this merged database were added only for the hospitalized patients the data concerning the treatment from the PHARMA database which is the pharmaceutical prescription file. Similarly, for each patient, the Sars-Cov2 variant was merged with the biology database (NEXLAB Medical Software) which lists all the samples and their associated genotype via an extraction. Finally, the information on death from all causes was obtained via an extraction from the Medical Information Department (DIM). Indeed, if the patient died during his inpatient care, the information system retrieves this information in the discharge mode. For patients discharged alive from the APHM stay, we queried the French National Death Registry (29) to find out whether they died within 42 days of their IHU treatment. All the data were quality controlled, looking for duplicates and aberrant data. Finally, the construction of the database and the quality controls carried out were verified by a bailiff.

In total 8 different datasets were included in this study; confirmed COVID-19 cases, hospitalization and deaths for Germany and France, as well as confirmed Influenza cases and confirmed hospitalizations due to severe acute respiratory infections (SARI). The COVID-19 datasets are available on both country- and regional level, which results in 16 (regional) + 1 (country) time series for each of the German datasets and in 13 (region) + 1 (country) for each of the French datasets. All datasets were split into N training and evaluation windows. Details are explained in section 2.4.

Hospital bed number per 100,000 people, per regions in France.

Health states, durations and disability weights for estimating the burden of COVD-19 in France [13].

The bold figures highlight the maximum correlation.