Analysis of Mortality of Critically Ill Patients With COVID-19

Introduction The World Health Organization (WHO) has recently declared SARS-CoV-2 (COVID-19)
disease as an international alarm public health emergency. This ongoing pandemic of COVID-19
disease is devastating, despite widespread implementation of control measures, which is
creating a challenge for national healthcare. Since the beginning of the epidemic until the
date of this project, more than 800,000 cases have been detected in the world, of which
almost 214,000 have been detected in Spain.

The results of the recent analysis of clinical characteristics in a selected cohort of 1,099
COVID-19 patients throughout China have shown that up to 15% (173 / 1,099) developed severe
disease according to the American Thoracic clinical criteria for severe community pneumonia.
Society, and of these seriously ill patients, 20.6% presented as admission to Intensive Care
Units (ICU) (33/173, 19%), or the use of mechanical ventilation, both invasive and
non-invasive. invasive (81/173, 46%) or death (14/173, 8%).

In a retrospective study that included 52 critically ill patients with COVID-19 admitted to
the Intensive Care Unit (ICU) from the Wuhan Jin Yin-tan Hospital (Wuhan, China), a 28-day
mortality rate of 61% was found, although the causes of death of the patients included were
not reported.

Nowaday, the crude mortality in Europe is 5.4% of reported cases, although it is important to
be noted that this calculation is subject to multiple notification biases and due to the
different diagnostic test policies in the countries, and probable differences in healthcare
and demographic models between countries. Similarly, in Spain, mortality with 213,024 cases
notified to the National Surveillance Network is 10.4%. Using modeling, a mortality has been
estimated among hospitalized cases that could reach 14% (95% confidence interval [CI] 3.9%
-32%). However, the mortality and causes of death of critically ill patients with COVID-19 in
Spain are still unknown.

To date, no specific antiviral treatment that has been shown to be effective for the
2019-SARS-CoV-2 infection is recommended. Symptomatic and supportive care, along with
respiratory and contact isolation, are recommended for infected individuals.

Selective digestive decontamination (SDD) is a prophylactic treatment for critically ill
patients, mainly requiring invasive mechanical ventilation, based on an oropharyngeal paste
and an enteral suspension containing antimicrobials (usually tobramycin, colistin, and an
antifungal) and an intravenous antibiotic (usually a cephalosporin. second generation),
administered during the first 4 days of ICU treatment. The goal of SDD is to prevent or
eradicate, if present, the abnormal, oropharyngeal and intestinal transport of potentially
pathogenic microorganisms, such as aerobic gram-negative bacilli (AGNB),
methicillin-sensitive Staphylococcus aureus, in patients at risk of nosocomial infections. In
critically ill patients, it has been demonstrated with a high level of evidence, that SDD
prevents serious infections and reduces mortality. In an observational study in two hospitals
in China, 51% of the patients developed a hospital-acquired infection and this incidence was
associated with an increase in mortality. Therefore, the preventive strategy of SDD could be
postulated as an alternative to reduce mortality and the incidence of infections in
critically ill patients with COVID-19. However, this hypothesis has not been evaluated at the
moment.

At the present time, observational studies focused on the clinical characteristics of
critically ill patients with COVID-19 and the factors associated with their mortality have
small sample sizes, making knowledge of the epidemiology of critically ill patients clearly
insufficient. And what is more relevant, this scenario of lack of knowledge, when
extrapolated to the situation of the current outbreak in Spain, places us in the need to
describe the epidemiological manifestations of the COVI-19 in Spain in order to handle with
safety and efficiency the challenge of the control and treatment of the outbreak in Spain of
the COVID-19.

Specifically, the study will focus on critically ill patients with COVID-19 for two reasons:

They represent the cohort of patients with the highest use of material and professional
resources in the health system.

Given the current moment of pandemic and healthcare challenge, it is therefore pertinent to
conduct an epidemiological surveillance of COVID-19 disease to reliably define the incidence
of mortality and morbidity in this group of patients, as well as the risk factors for
infection by SARS-CoV-2 associated with mortality.

Objective The main objective is to evaluate the mortality of critically ill patients with
COVID-19 admitted to the ICU, as well as to analyze the factors associated with mortality.

The secondary objectives:

To analyze the epidemiological characteristics of critically ill patients with confirmed
COVID-19 in Spain.

To describe the usual clinical practice in the ICU of critically ill adult patients with
COVID-19, to identify risk patterns and possible effectiveness or adverse effects of the
treatments received.

To compare the mortality of critically ill patients with COVID-19 treated with SDD compared
with those who do not received SDD.

To describe the cause of death of critically ill patients with COVID-19 received invasive
mechanical ventilation.

Method Design: An observational, prospective and multicenter study will be carry out,
envolving of 12 ICUs in Madrid (University Hospital of Getafe, as coordinating center;
University Hospital Puerta de Hierro, University Hospital del Henares of Coslada, University
Hospital of Torrejón, Hospital Universitario Ramón y Cajal, Hospital Infanta Cristina,
Hospital del Tajo, Hospital Quirón Madrid Sanchinarro, Hospital Universitario Severo Ochoa,
Hospital Universitario Gregorio Marañón, Hospital Universitario de Móstoles, Hospital
Universitario La Princesa, Hospital Universitario La Paz, Hospital Universitario Doce de
Octubre, and Hospital Santa Creu i Sant Pau in Barcelona). Inclusion criteria: patients over
18 years of age who are admitted to the ICU with the confirmed diagnosis of COVID-19.

Recruitment period: The study will start once the study has been approved by the Ethics and
Research Committee of the Hospital Universitario de Getafe with an expected date of April
2020, with an estimated duration of 3 months.

Case definition: Any critically ill patients with a positive respiratory tract sample result
(tracheal aspirate or nasopharyngeal exudate) using the PCR technique for the SARS-CoV-2
virus and requiring admission to the ICU at trial will be considered a confirmed case of
COVID-19 of the physician in charge of the patient.

Data collection: Demographic variables: age, sex, reason for admission to the ICU.

Comorbidities: severity at ICU admission (SAPSIII); Daily SOFA until day 7, diabetes
mellitus, high blood pressure, chronic obstructive pulmonary disease, other chronic lung
diseases (interstitial lung disease), chronic kidney failure, previous liver disease
(cirrhosis), immunosuppression, previous heart disease (ischemic, valvular), chronic
arrhythmias (atrial fibrillation); previous cerebrovascular disease (ischemic or
hemorrhagic).

Noninvasive respiratory support prior to ICU admission: non-invasive ventilation, or high
flow therapy with nasal cannula, and duration of non-invasive respiratory support.

Invasive respiratory support parameters during the first 7 days: tidal volume, level of
positive end-expiratory pressure, plateau pressure, inspiratory fraction of oxygen.
Biochemistry values and arterial blood gases, blood count and coagulation parameters on
admission to the ICU during the first 7 days from ICU admission. Pharmacological therapies
received for COVID-19 disease during the ICU stay: azitromycin, corticosteroids,
hydroxychloroquine, antiretrovirals (lopinavir/ritonavir), tocilizumab, oseltamivir,
darunabir, remdesivir, baricitinib.

Bacterial / fungal infections acquired in the ICU (pneumonia, bacteriemia): isolated
microorganisms in diagnostic samples, isolated microorganisms in surveillance samples
(oropharynx and rectum if surveillance samples are available); antibiotic treatment.

Ventilatory management: prone positioning, administration of inhaled nitric oxide,
performance of tracheostomy, date of tracheotomy, modality of tracheotomy (surgical,
percutaneous).

Classification of predefined causes of death:

Refractory hypoxemia (peripheral oxygen saturation measured by pulse oximetry less than 80%,
or PaO2 <60 mmHg of arterial blood sample) despite maximum ventilatory support (FiO2 100%,
PEEP level> 5 cm H2O) with maintenance MAP> 60 mmHg in the 3 hours prior to exitus lethalis.

Refractory shock (mean arterial pressure less than 60 mmHg despite vasoactive support with
norepinephrine at> 2 mcg / kg / min and / or dobutamine at> 20 microgr / kg / min, and serum
lactate level> 2 mmol / L in the 3 hours prior to exitus lethalis), associated or not with
organ failure, but with peripheral oxygen saturation measured by pulse oximetry greater than
90%, with no evidence of documented infection in the previous 48 hours.

Septic shock. A refractory shock which fulfill previous criteria plus the presence of a
documented infection in the 48 hours prior to exitus lethalis.

Others: Pulmonary thromboembolism (defined based on clinical suspicion, associated with any
of the classifications [Wells or Geneva] according to the low, intermediate or high
probability and the determination of the high D-dimer (> 500 μg/L) in the 24 hours prior to
exitus lethalis; Incohercible bleeding; Yatrogenic respiratory complications (accidental exit
of the orotracheal tube with impossibility of obtaining an artificial airway, fatal
obstruction of the orotracheal tube); Arrhythmias: sudden unanticipated onset of malignant
arrhythmias (irreversible ventricular fibrillation), or asystole. Arrhythmias appearing
during the last minutes before exitus lethalis will be excluded.

Main outcomes. ICU mortality, 28-day mortality, and hospital mortality. Secondary outcomes:
appearance of nosocomial infections during ICU admission (ventilator-associated pneumonia ,
nosocomial pneumonia, catheter-related bacteremia, bacteraemia); development of complications
during ICU admission (development of acute respiratory distress, barotrauma, development of
organ failure [cardiovascular, renal, hepatic, hematological]), duration of mechanical
ventilation, length of ICU stay, length of hospital stay, withdrawald of life support, and
cause of mortality.

Statistic analysis. Given the estimated mortality of 40% of the cases hospitalized in the
ICU, a convenience sample size of 600 participants will be estimated. Considering a loss of
15% for the outcomes, 750 subjects will finally be included.

To evaluate the risk factors associated with the mortality of critically ill patients with
COVID-19, a logistic model adjusted for baseline variables (age, severity at ICU admission),
empirical treatment received, failure of non-invasive respiratory support will be performed.

The effectiveness of empirical treatments for COVID-19 will be performed through logistic
analysis adjusted for age, severity at admission and complications during ICU admission.

In addition, a stratified cluster analysis will be proposed for ICUs that apply the SDD
strategy and units that do not routinely use SDD.

Data quality. The data will be reviewed and approved by the Steering Committee of the study
and validated by the collaborating researcher responsible for each center. All data will be
centrally reviewed. Alerts will be sent for lost data, aberrant data or discrepancies
thereof.

At each center, local study investigators will review data from 5% of randomly selected
participants, and all participants about whom evaluators may have questions about
inconsistent data.

The data collection of all included cases will be verified by an independent research
coordinator hired through an external CRO and who will evaluate the quality of data
collection independently of the main researcher. Data collection will be re-evaluated until
the forms and all results are completed.

Full data will be required to include subjects in the final analysis, and only subjects
without missing data will be accepted for the primary variables or patients with missing data
less than 10% for the secondary variables, if these cannot be retrieved by the investigators.
All bizarre data will be sent to the reference centers for verification. The presence of
missing data in the main variables will be reason for the exclusion of the participant in the
analysis.

Ethical-legal aspects. The study will be carried out in accordance with the updated Helsinki
declaration in its latest version. The study protocol, together with the data collection
sheets and Annexes will be sent for approval to the Ethics and Research Committee of the
University Hospital of Getafe, and will ask them for a waiver consent due to the
characteristics of the study design (non-intervention study for statistical purposes, action
plan to improve the quality of care, in a health emergency situation), as provided in the
spanish laws.

This study does not include any control group. Participation in this study does not imply any
risk for the included patients since it is based on the usual clinical practice in common use
in all hospitals, which are already applied at the Getafe University Hospital.

The study will be based on the Spanish legislation for biomedical research, and in accordance
with national laws of Protection of Personal Data and guarantee of them that the identity of
the subject will be kept confidential and anonymous throughout the study, in such a way that
no personal data will be collected that allows the identification of the subjects included
and only the data necessary to achieve the study objectives will be collected.

Confidentiality The use, disclosure, publication or dissemination of the document is not
authorized in any way without the written consent of the Steering Committee. Given that the
study proposes the analysis of the usual clinical practice, it is considered not pertinent to
contract a civil liability insurance, since it is a low-intensity clinical study.

During the period of inclusion of the participants, the database to which only the study
coordinating team (principal investigator and co-investigators) will have access will be
generated anonymously, with irreversible dissociation of personal data.