key: cord-1018447-upsncn81 authors: Lozano, Manuel; Iftimi, Adina; Briz-Redon, Alvaro; Peiró, Juanjo; Manyes, Lara; Otero, María; Ballester, Mayte; de las Marinas, Dolores; Catalá, Juan Carlos; de Andrés, José; Romero, Carolina title: Clinical Characteristics of COVID-19 hospitalised patients associated with mortality: a cohort study in Spain date: 2022-04-22 journal: Infectious Medicine DOI: 10.1016/j.imj.2022.04.002 sha: 6c0d203443a2b1cec87415074b85c6f5e90f39fe doc_id: 1018447 cord_uid: upsncn81 The heterogeneity of patients with COVID-19 may explain the wide variation of mortality rate due to the population characteristics, presence of comorbidities and clinical manifestations. In this study, we analysed 5,342 patients' recordings and selected a cohort of 177 hospitalised patients with a poor prognosis at an early stage. We assessed during six months their symptomatology, coexisting health conditions, clinical measures and health assistance related to mortality. Multiple Cox proportional hazards models were built to identify the associated factors with mortality risk. We observed that cough and kidney failure triplicate the mortality risk and both bilirubin levels and oncologic condition are shown as the most associated with the demise, increasing in four and ten times the risk, respectively. Other clinical characteristics such as fever, Diabetes Mellitus, breathing frequency, neutrophil-lymphocyte ratio, oxygen saturation and troponin levels, were also related to mortality risk of in-hospital death. The present study shows that some symptomatology, comorbidities and clinical measures could be the target of prevention tools to improve survival rates. Since the outbreak of coronavirus disease disease began in December 2019, more than 149 million people have developed SARS CoV-2 infection, and more than 3 million have died worldwide. In Spain, by mid-2021 up to 3,5 million cases were reported causing more than 77,000 deaths [1]. There are several studies describing the clinical characteristics and outcomes of hospitalised patients with SARS-CoV-2. The heterogeneity of patients treated in China [2] , Italy [3] , UK [4] , USA [5] [6] [7] or Spain [8] [9] [10] may explain the wide variation of mortality rate due to the population characteristics, presence of comorbidities and clinical manifestations. The first confirmed case of COVID-19 in Valencia, Spain, was reported on February 19, 2020. The tertiary Hospital General Universitario of Valencia (CHGUV), that assists approximately 364,000 persons, was designated as a COVID-19 centre and described an infection rate with a heterogeneous distribution during the following 6 months. In this study, we analyse the symptomatology, coexisting health conditions, clinical measures and health assistance, in a selected cohort of patients with a poor prognosis at an early stage in hospitalised patients from Valencia during this period and assess the clinical characteristics associated with mortality. This observational prospective study was conducted at the Hospital General Universitario of Valencia (CHGUV), an academic public hospital that serves the largest area in the city, consisting of approximately 364,000 patients. The study was approved by the institutional review board, and the requirement for informed consent was waived. All consecutive patients who were tested for COVID-19 were included from February 19 to August 31, 2020. A total of 5,342 patients were treated during this period and 177 COVID-19 positive adults confirmed by PCR, admitted to the hospital due to clinical complications, with a World Health Organisation ordinal scale 4 (oxygen by mask or nasal prongs) or 5 (non-invasive ventilation or high-flow oxygen) [11], and followed up until recovery or death, were selected. We adhered to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines [12] . Data collected included patient demographics (residence, biological sex and age, in years); treatment reason (symptomatology, contact or other, categorised as yes/no); the recorded symptomatology during the observation period (yes, no) such as fever, asthenia, altered consciousness, headache, myalgia, arthralgia, eczema, nasal congestion, anosmia, sore throat, dyspnoea, cough, expectoration, pleuritic pain, haemoptysis, diarrhoea, and nausea; intensive care admission (yes, no); presence of smoking habit (current, former or no smoker) and pregnancy (yes, no); comorbidities (yes, no) such as hypertension, cardiovascular disease, diabetes mellitus, obesity, chronic obstructive pulmonary disease, asthma, oncological process, immunosuppression, stroke, kidney and liver failure, and deep vein or pulmonary thrombosis; as well as laboratory tests and clinical characteristics, including mean arterial blood pressure (mmHg), cardiac frequency (beats/minute), oxygen saturation (%), breathing frequency (breaths/minute), lymphocytes (units/µL), neutrophil lymphocyte ratio, platelets (units/µL), d-dimer (ng/mL), activated partial thromboplastin time (seconds), international normalized ratio, fibrinogen (mg/dL), bilirubin (mg/dL), lactic acid dehydrogenase (units/L), ferritin (ng/mL), creatinine (mg/dL), and troponin (ng/L). Outcomes observed were the length of stay until recovery or death. Clinical examinations included chest radiography if necessary (unilobar, multilobar, clean or not performed). Basic descriptive statistics were calculated for each collected covariate. Death in the whole sample was categorised (death status: yes, no) to run survival analyses as timedependent response variable. Demographic information, symptomatology, intensive care admission, smoking habit, pregnancy, comorbidities, and radiography results were considered as predictive covariates. Kaplan-Meier survival curves were estimated to compare the survival during the observation period, overall, and stratified for sex and age range (20-50 years (n=25); 51-60 years (n=22); 61-70 years (n=31); 71-80 years (n=47); and ≥81 years (n=52)). Simple Cox proportional hazards models were used to control sex and age as confounders, considering each covariate as predictive variable, with and without sex and age adjustment (Table S1 ). Multiple Cox proportional hazards models were built to identify the associated covariates with mortality risk. All multiple models were adjusted for sex and age. Each multiple model was built following three steps: 1) Obtaining a first multiple basal model by using all the symptomatology covariates previously associated with a p-value <0.2 in the simple analyses. Following a backward elimination procedure, all the symptomatology covariates associated with the mortality risk at a p-value level <0.1 in the likelihood ratio test were retained in the model; 2) comorbidities were added to this clinical symptomatology basal model individually and those with a p-value <0.2 were candidates to enter in the model. Following a backward elimination procedure, all these comorbidities candidate covariates associated with the mortality risk at a p-value level <0.1 were retained in the model; 3) The same procedure was repeated on the clinical symptomatology and comorbidities basal model using clinical covariates in order to obtain the final multiple model. Statistical analysis was carried out using R statistical software version 3.5.1 [13] . Kaplan-Meier curves were plotted and Cox regressions built by using the survival R package [13] . The final multiple model was validated by means of proportional hazards assumption testing based on weighted residuals [14] by using the cox.zph function. Influential outliers were assessed observing beta deviations with ggcoxdiagnostics function of survminer package [15] . Non-linearity was evaluated plotting the Martingale residuals and natural cubic splines with one or two internal knots were compared through Akaike (AIC) scores. Then, the lowest AIC non-linear model and linear model were compared using graphical examination and the Likelihood Ratio test (LRT). Significance level <0.05 was considered in all tests, although marginally significant effects (p-value <0.1) were also considered. Descriptive statistics of the study variables are displayed in Table 1 Performed thoracic X-rays showed multilobar outcome more frequently among nonsurvivors. Simple Cox proportional hazards models, considering each covariate as predictive variable, did not indicate differences between models with and without sex and age adjustment (Supplemental Table S1 In this Spanish cohort study, we assessed the relationship between the symptomatology, coexisting health conditions, clinical measures and health assistance, and mortality risk in a screening sample from COVID-19 positive adults with oxygen by mask, nasal prongs, non-invasive ventilation or high-flow oxygen, and followed until recovery (70.6%) or death (29.4%). Factors such as age and some clinical characteristics seem to play a role in this relationship. Overall, multiple model showed that patients who presented cough, specific comorbidities like Diabetes Mellitus, kidney failure and an oncologic process, as well as higher breathing frequency, neutrophil lymphocyte ratio and troponin levels, were related to an increased risk of in-hospital death. However, fever and oxygen saturation were associated with lower mortality risk. A relevant fact that has been elucidated in our results is the increased morbimortality detected in COVID-19 positive patients with chronic kidney disease almost tripled the mortality in our sample [16] . Perhaps postoperative creatinine could be a marker indicating the degree of severity of COVID-19 inpatients or perhaps it could be due to a possible direct involvement of the kidney by this coronavirus [17] . Until now, some authors have also established increased mortality in patients admitted for COVID-19 and those who developed acute kidney injury during the hospital stay [18] . The clinical characteristics of COVID-19 occur across a broad spectrum, ranging from asymptomatic infection to severe respiratory failure [19, 20] . The main symptoms include fever, cough, myalgia, and dyspnoea [19, 20] . Headache, diarrhoea, fatigue, sore throat, anosmia, ageusia, chest pain, haemoptysis, sputum production, rhinorrhoea, nausea, vomiting, skin rash, impaired consciousness, and seizure have been also observed [19] [20] [21] , but most severe are usually older patients showing dyspnoea, respiratory frequency ≥30/min, blood oxygen saturation ≤ 93%, some comorbidities (hypertension, Diabetes Mellitus, and cardiovascular disease), and abnormal chest imaging findings [22, 23] . The fact that the variable disorientation was statistically significant in our results, may be due in part to the fact that the most severe patients had increased breathing effort associated with hypoxeamia. This circumstance in turn corresponds to hypercapnia, which is a known contributing factor to disorientation and impairment of baseline neurological status. Compared with the surviving COVID-19 patients, the deceased had lower platelet levels and higher C-reactive protein and lactate dehydrogenase at admission, which have not been shown significant in the multiple model of the present study. These results are supported by another wide meta-analysis performed in 34 studies with 5,057 patients [25] . However, this second study also observed lymphopenia among dead patients (50.1%, CI 38.0-62.4), which has been shown to be associated with mortality in our study. On the other hand, other studies performed in COVID-19 non-survivors do show other coincident results with the present study. From a large Chinese study carried out in 1099 patients with laboratory-confirmed COVID-19, 67 died with fever and cough as the most common symptoms and lymphocytopenia was shown in 83.2% of them [19] . Lymphocytes decrease has been shown as the common clinical factor associated to an increased mortality risk in other studies [9, [26] [27] [28] . Several studies have found that tobacco smoke is a protective factor and that it influences the clinical course of patients affected by COVID-19 by decreasing the severity of the manifestations [29] [30] [31] . However, in our sample, we didn´t find any relevant role for tobacco smoke and clinical severity. To date, three large Spanish studies have been found assessing the associated factors with mortality risk. One of them reported the first 1255 adult cases in Madrid and also carried out multiple Cox models, observing some similar results regarding older age (HR 1.07, 95% CI 1.06-1.09), Diabetes Mellitus (HR 1.45, 95% CI 1.09-1.92) , and lymphocytopenia (HR 1.62, 95% CI 1.20-2.20) [9] . The second one was also performed in Madrid in 1828 patients during the same period with a fatality rate of 14.6%, although no associated factors to survival were assessed [10] . A third observational multicentre study described clinical characteristics of very old patients (≥80 years old) in 150 Spanish hospitals (2772 patients), observing similar associations between higher mortality risk and Diabetes Mellitus (25.6% of cases), oxygen saturation (<90%), unilateral-bilateral infiltrates on chest x-rays, neutrophils (≥7.5 x 10 3 /µL), and lymphocytes (<0.8 x 10 3 /µL). However, higher fever was related to an increased mortality risk by using logistic regressions [28] . Lymphocytopenia has been one of the most common clinical characteristic associated with COVID-19 patient's mortality across studies. An explanation for the relationship between the virus and the lower lymphocytes levels has been proposed by means of immune responses activation, which may overproduce pro-inflammatory cytokines, causing uncontrolled inflammatory responses in patients with severe COVID-19. This condition may lead to lymphopenia and lymphocyte dysfunction [32] . Among comorbidities, Diabetes Mellitus has been shown to be reiterative. Patients with severe COVID-19 and Diabetes Mellitus have the lowest lymphocyte counts compared with those with severe COVID-19 without Diabetes Mellitus, and those with non-severe COVID-19 with or without Diabetes Mellitus. Partially decreased lymphocyte subsets, age and Diabetes Mellitus were closely related to disease progression and prognosis [33] , since Diabetes Mellitus could lead to dysfunctional cellular immunity [34] . We assessed the relationship between symptomatology, coexisting health conditions, clinical measures and health assistance, with mortality risk in severe COVID-19 patients. We observed that cough and kidney failure triplicate the mortality risk and both bilirubin levels and oncologic condition are shown as the most associated with the demise, increasing in four and ten times the risk, respectively. Other clinical characteristics such as fever, Diabetes Mellitus, breathing frequency, neutrophillymphocyte ratio, oxygen saturation and troponin levels, were also related to mortality risk of in-hospital death. The present study shows that some symptomatology, comorbidities and clinical measures could be the target of prevention tools to improve survival rates. The data that support the findings of this study are available from CHGUV. Restrictions apply to the availability of these data, which were used under licence for this study. Clinical course and outcomes of critically ill patients with SARS-CoV-2 pneumonia in Wuhan, China: a single-centered, retrospective, observational study Baseline Characteristics and Outcomes of 1591 Patients Infected With SARS-CoV-2 Admitted to ICUs of the Lombardy Region Features of 20 133 UK patients in hospital with covid-19 using the ISARIC WHO Clinical Characterisation Protocol: prospective observational cohort study Presenting Characteristics, Comorbidities, and Outcomes Among 5700 Patients Hospitalized With COVID-19 in the New York City Area Characteristics and Outcomes of 21 Critically Ill Patients With COVID-19 in Washington State Characteristics of Hospitalized Adults With COVID-19 in an Integrated Health Care System in California Deploying unsupervised clustering analysis to derive clinical phenotypes and risk factors associated with mortality risk in 2022 critically ill patients with COVID-19 in Spain COVID-19 in hospitalised patients in Spain: a cohort study in Madrid Epidemiological characteristics of the COVID-19 outbreak in a secondary hospital in Spain The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies R: A language and environment for statistical computing. Viena: R Foundation for Statistical Computing Proportional hazards tests and diagnostics based on weighted residuals Drawing survival curves using COVID-19 and chronic kidney disease: a comprehensive review Covid-19 and kidney injury: Pathophysiology and molecular mechanisms Outcomes Among Patients Hospitalized With COVID-19 and Acute Kidney Injury Clinical Characteristics of Coronavirus Disease 2019 in China Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China. The Lancet Elsevier Cardiovascular Implications of Fatal Outcomes of Patients With Coronavirus Disease 2019 (COVID-19) A narrative review of coronavirus disease 2019 (COVID-19): clinical, epidemiological characteristics, and systemic manifestations Clinical Characteristics, Treatment, and Outcomes of Critically Ill Patients With COVID-19: A Scoping Review Clinical characteristics, laboratory outcome characteristics, comorbidities, and complications of related COVID-19 deceased: a systematic review and meta-analysis COVID-19 Clinical Characteristics, and Sex-Specific Risk of Mortality: Systematic Review and Meta-Analysis Clinical characteristics of 21 cases of COVID-19 induced death: A retrospective study COVID-19 patients' clinical characteristics, discharge rate, and fatality rate of meta-analysis Clinical Characteristics and Risk Factors for Mortality in Very Old Patients Hospitalized With COVID-19 in Spain Smoking and COVID-19: What we know so far COVID-19 and smoking: is nicotine the hidden link? Nations with high smoking rate have low SARS-CoV-2 infection and low COVID-19 mortality rate The contributory role of lymphocyte subsets, pathophysiology of lymphopenia and its implication as prognostic and therapeutic opportunity in COVID-19 Overall reduced lymphocyte especially T and B subsets closely related to the poor prognosis and the disease severity in severe patients with COVID-19 and diabetes mellitus Immunological mechanisms contributing to the double burden of diabetes and intracellular bacterial infections ☒ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.