This study aims to describe the epidemiology of COVID-19 patients in a Swiss university hospital.
This retrospective observational study included all adult patients hospitalized with a laboratory confirmed SARS-CoV-2 infection from March 1 to March 25, 2020. We extracted data from electronic health records. The primary outcome was the need to mechanical ventilation at day 14. We used multivariate logistic regression to identify risk factors for mechanical ventilation. Follow-up was of at least 14 days.
145 patients were included in the multivariate model, of whom 36 (24.8%) needed mechanical ventilation at 14 days. The median time from symptoms onset to mechanical ventilation was 9·5 days (IQR 7.00, 12.75). Multivariable regression showed increased odds of mechanical ventilation with age (OR 1.09 per year, 95% CI 1.03–1.16, p = 0.002), in males (OR 6.99, 95% CI 1.68–29.03, p = 0.007), in patients who presented with a qSOFA score ≥2 (OR 7.24, 95% CI 1.64–32.03, p = 0.009), with bilateral infiltrate (OR 18.92, 3.94–98.23, p<0.001) or with a CRP of 40 mg/l or greater (OR 5.44, 1.18–25.25; p = 0.030) on admission. Patients with more than seven days of symptoms on admission had decreased odds of mechanical ventilation (0.087, 95% CI 0.02–0.38, p = 0.001).
This study gives some insight in the epidemiology and clinical course of patients admitted in a European tertiary hospital with SARS-CoV-2 infection. Age, male sex, high qSOFA score, CRP of 40 mg/l or greater and a bilateral radiological infiltrate could help clinicians identify patients at high risk for mechanical ventilation.
Citation: Regina J, Papadimitriou-Olivgeris M, Burger R, Le Pogam M-A, Niemi T, Filippidis P, et al. (2020) Epidemiology, risk factors and clinical course of SARS-CoV-2 infected patients in a Swiss university hospital: An observational retrospective study. PLoS ONE 15(11): e0240781. https://doi.org/10.1371/journal.pone.0240781
Editor: Chiara Lazzeri, Azienda Ospedaliero Universitaria Careggi, ITALY
Received: May 6, 2020; Accepted: October 2, 2020; Published: November 13, 2020
Copyright: © 2020 Regina et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The author(s) received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
Abbreviations: CCI, Charlson Comorbidity Index; AKI, Acute Kidney Injury; ARDS, Acute Respiratory Distress Syndrome; LUH, Lausanne University Hospital; EHR, Electronic Health Records; CRP, C-Reactive Protein; PCT, Procalcitonin; MV, Mechanical ventilation; SOFA, score Sequential Organ Failure Assessment score; qSOFA score, quick SOFA score; NEWS, National Early Warning score; CRB-65 score, Confusion/Respiratory rate/Blood pressure/age ≥ 65 years score; WHO, World Health Organization
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) first emerged in Wuhan (China) in December 2019. WHO, in February 2020, named the resulting disease COronaVirus Disease 2019 (COVID-19) . There is a wide spectrum of severity ranging from asymptomatic presentation to severe pneumonia requiring ventilator support and death . Since March 2020, the pace of SARS-CoV-2 spread around the globe increased as the epidemic evolved to a pandemic . In Europe, following a diminished caseload during summer 2020, new cases and deaths are increasing since September 2020 .
In the present study, we report the epidemiological and clinical characteristics of patients hospitalized for COVID-19 in a Swiss university hospital during the early phase of the pandemic as well as risk factors for progressive respiratory failure requiring mechanical ventilation.
Study design and participants
In this retrospective observational study, we included all adult patients consecutively hospitalized with a confirmed SARS-CoV-2 infection from March 1, 2020, to March 25, 2020. Patient’s initial physicians used local guidelines to decide on admission (only patients with risk factors for severe disease or needing medical care were hospitalized). For all patients, we ensured a follow-up of 14 days or more during hospital stay or up to discharge or death if they occurred first.
This study took place in Lausanne University Hospital (LUH), a one-thousand-bed tertiary university hospital in Lausanne, Switzerland. LUH serves as a primary-level community hospital for Lausanne (population circa 300’000 inhabitants) and as a referral hospital for Western Switzerland (population circa 1–1.5 million inhabitants). LUH increased its outbreak response capacity by setting up new intensive care units for the management of COVID-19 patients.
LUH electronic health record (EHR) provided data on epidemiological, clinical, radiological and laboratory data.
Epidemiological data included age, sex, height, weight, and relevant comorbidities, including the Charlson Comorbidities Index (CCI). We collected data on clinical presentation, SARS-CoV-2 treatments, concomitant treatments, non-pharmacological interventions and clinical course within LUH.
We recorded radiological findings from reports of chest radiography or computed tomography (CT). We defined healthcare workers as professionals having direct contact with patients (nurses, physiotherapists, physicians, etc.) or patient samples.
Laboratory data included full blood count, D-dimers, creatinine, highly sensitive cardiac T-troponin, C-reactive protein (CRP), procalcitonin (PCT), ferritin, liver function tests, blood type and real-time PCR to detect SARS-CoV-2 .
We entered all data in an electronic clinical report form (eCRF) using the REDCap® platform (Research Electronic Data Capture v8.5.24, Vanderbilt University, Tennessee, USA) . Fellows of the Infectious Diseases, Hospital Preventive Medicine and Internal Medicine Services at LUH entered the data and two of the authors (JR, MP) verified their integrity.
Treating physicians made all decisions regarding supportive care. Specialists in infectious diseases reviewed all SARS-CoV-2 treatment decisions according to the local recommendations. These included protease inhibitors (ritonavir-boosted lopinavir or atazanavir), hydroxychloroquine or remdesivir. Selected critically ill patients with high inflammatory markers (CRP, D-dimers, PCT and ferritin) received tocilizumab. The choice of treatment depended on drug availability, safety profile drug-drug interactions.
Treating physicians discussed advance planning of care and do not resuscitate orders with all patients. Limitations of care were decided upon admission according to the patient’s values and goals and the treating physician’s appreciation. Treating physicians documented all inpatients’ limitation of care in LUH EHR. We categorized these limitations into two levels: 1) no limitation; 2) limitation to the best supportive care provided in non-monitored wards or intermediate care units but without mechanical ventilation (MV).
We defined a confirmed SARS-CoV-2 infection as a positive test for SARS-CoV-2 using real-time polymerase chain reaction (qPCR) technology in any respiratory sample.
We defined obesity either as a body mass index (BMI) of 30 kg/m2 or higher, or, when missing anthropometric data, a medical diagnosis of obesity.
We used the Berlin definition for Acute Respiratory Distress Syndrome (ARDS) . We defined shock as refractory hypotension requiring infusion of vasopressors. Acute kidney injury (AKI) was identified and classified according to the 2012 Kidney Disease Improving Global Outcome guidelines . We defined liver injury as a 3-fold or greater increase in transaminase levels.
We calculated quick Sequential Organ Failure Assessment (qSOFA) score, Confusion/Respiratory rate/Blood pressure/age ≥ 65 years (CRB-65) score and National Early Warning Score (NEWS) were assessed according to their original descriptions [9–11].
We defined MV as invasive respiratory support through a laryngeal or a tracheostomy tube
The primary outcome was the use of MV for respiratory failure attributed to SARS-CoV-2 pneumonia, within 14 days after admission.
For the descriptive analysis, we first described patient’s characteristics. We summarized categorical variables as numbers (percentages), normally distributed continuous variables as mean ± standard deviation (SD) and continuous variables with a skewed distribution as median [interquartile range (IQR)]. We then tested for associations between patient’s characteristics and MV. We used Pearson’s chi-square test for binary characteristics, Student t-test for normally distributed variables and Mann-Whitney-Wilcoxon test for continuous variables with a skewed distribution.
We used a general linear model based on univariate logistic regression to calculate odds ratio for MV. We finally performed a multivariate logistic regression with a LASSO (Least Absolute Shrinkage and Selection Operator) penalisation. We used LASSO regression analysis, as the number of MV cases was low and did not allow us to obtain a multivariate regression model with more than four predictors. LASSO provides the most parsimonious model for high number of covariates and small sample size . In addition, it accounts for collinearity. For multivariate analysis, we selected relevant clinical, biological and radiological parameters as well as patients characteristics previously identified in the literature and after authors consensus (clinical knowledge). To identify the optimal penalization parameter (lambda), we used a bootstrap resampling procedure: 1000 bootstrap samples of the same size of the full database (lambda = 4). We retained as the model the one with the lowest AIC (Akaike Information Criterion).
We excluded from the analysis, patients whose care was limited to the best supportive care and patients already mechanically ventilated on admission. As we did not have treatments starting date, we did not include SARS-CoV-2 specific treatments as variables in the model.
For the inflammatory biomarkers, we converted continuous variables to categorical variables using cut-off values previously identified in the literature: 1000 ng/ml for D-dimer, 40 mg/L for CRP, 0.5 μg/l for procalcitonin and 300 μg/l for ferritin [13–15].
We did not impute any values for missing data.
Statistical analyses were performed using R software v3.6.2 (R Foundation for Statistical Computing; www.r-project.org).
This project was conducted in accordance with the Declaration of Helsinki, the principles of Good Clinical Practice and the Swiss Human Research Act (HRA). The project received approval from the Ethics Committee of canton Vaud, Switzerland (2020–00657) that waived the need for informed consent. All data were anonymized before analysis.
Overall, 200 patients with confirmed SARS-CoV-2 infection were hospitalized at LUH during the study period. In 54 (28.5%), care was agreed to be limited to best supportive care on admission, these patients were older and with more comorbidities (Table 1).
We excluded 54 patients with limitation of care and one patient already mechanically ventilated of admission and included 145 (72.5%) patients in the statistical analysis (Fig 1).
Median patient age was 62.0 years [IQR 52.0–74.0], ranging from 20.0 to 89.0 years. Eighty-one (55.8%) of all inpatients were aged 65.0 years or younger. Median BMI was 26.99 [IQR 23.84–31.84].
One hundred and nine (76.2%) patients had at least one comorbidity, with a median Charlson Comorbidity Index of 3.0 [IQR 1.0–5.0]. Table 2 summarizes the characteristics of this population whose most frequent comorbidities were hypertension (39.3%), obesity (33.8%), diabetes (21.4%), coronary artery disease (13.1%) and chronic kidney disease (10.3%).
Thirty-one (22.0%) patients were treated with angiotensin-converting enzyme inhibitors (ACEI) or angiotensin II receptor blockers (ARBs). Nine (6.2%) were treated with immunosuppressive drugs of which six (4.1%) patients were transplant recipients (5 solid-organ and 1 haematopoietic stem-cell transplantation). Sixteen (11.0%) patients had an active malignancy at the time of admission.
Clinical characteristics on admission
Table 3 describes clinical characteristics of patients on admission. The mean duration of symptoms preceding SARS-CoV-2 RT-PCR test was seven days [IQR 4.0–10.0]. The most frequent symptoms at the time of testing were fever in 104 patients (71.7%), cough in 100 (69.0%), fatigue in 72 (49.7%) and dyspnoea in 77 (53.1%).
Overall, 100 (69.0%) patients had a radiological exam and presented new lung infiltrates which were bilateral for 82 (56.6%) of them. Table 3 describes vital signs, qSOFA score, NEWS and CRB-65 scores on admission.
Table 4 describes the median value of commonly measured inflammatory parameters (white blood cell count, CRP, procalcitonin, D-dimers and ferritin).
97 (66.9%) of all patients received SARS-CoV-2 treatment (Table 5). The most frequently prescribed medication were protease inhibitors in 86 patients (59.3%) and hydroxychloroquine in 70 patients (48.3%). Sixty-seven patients (46.2%) received two or more concomitant SARS-CoV-2 treatments. Fifty-six (38.6%) patients received antibiotics. One hundred and five (72.4%) patients required supplementary oxygen during the follow-up period.
Table 6 summarizes patient’s clinical course. At the end of the follow-up, 27 (18.6%) patients were still hospitalized, 87 (60.0%) patients were discharged, 8 (5.5%) patients were transferred to a rehabilitation centre and 9 (6.2%) to another acute care hospital. Fourteen (9.7%) patients died during hospitalization.
Overall, 36 (24.8%) patients required mechanical ventilation after a median of two days since admission [IQR 0.00–3.00]. Median time from symptom onset to mechanical ventilation was 9.5 days [IQR 7.00–12.75]. Regarding patients requiring MV, 26 (72.2%) had a least one session of prone positioning, 24 (66.7%) received a vasopressor, 11 (30.5%) were eventually weaned from ventilator support and 11 (30.5%) died. The median duration of MV was six days [IQR 5.00–11.00]. Twenty-two of the 36 patients (61.1%) requiring MW were admitted to new dedicated COVID-19 ICUs.
Table 7 describes complications during follow-up. The most frequent complications were ARDS (n = 41, 28.3%), acute kidney injury (n = 23, 15.9%), hospital-acquired pneumonia (HAP) (n = 21, 14.5%), acute confusional state (n = 16, 11.0%) and rhythm disorder (n = 16, 11.0%).
Risk factors for mechanical ventilation
S1 Table summarizes unadjusted odds of MV at 14 days. Unadjusted odds of MV at 14 days were greater in males (odds ratio 2.65, 95% confidence interval 1.15 to 6.72) and in overweight patients (odds ratio 2.46, 95% confidence interval 1.11 to 5.81). None of the comorbidities increased the unadjusted risk MV at 14 days. Unadjusted odds of MV at 14 days were greater for patients presenting with dyspnoea (odds ratio 4.29, 95% confidence interval 1.86 to 10.85) on admission.
NEWS score ≥7 (odds ratio 9.18, 95% confidence interval 3.66 to 26.55), qSOFA score ≥2 (odds ratio 6.86, 95% confidence interval 2.68 to 18.47) or a CRB-65 score ≥2 (odds ratio 4.64, 95% confidence interval 2.07 to 10.65) increased the unadjusted odds of MV.
The presence of a radiological infiltrates increased the odds of mechanical ventilation (odds ratio 11.0, 95% confidence interval 3.14 to 70.45) as did the presence of a bilateral infiltrates (odds ratio 7.05, 95% confidence interval 2.75 to 21.87). Acute kidney injury on admission (odds ratio 4.27, 95% confidence interval 1.87 to 9.91), D-dimers of 1000 ng/ml or greater (odds ratio 3.28, 95% confidence interval 1.37 to 8.25), CRP of 40 mg/l or greater (odds ratio 6.79, 95% confidence interval 1.51 to 18.58) and PCT of 0.5 μg/l or greater (odds ratio 5.99, 95% confidence interval 1.52 to 29.66) increased the unadjusted odds of mechanical ventilation.
Table 8 summarizes adjusted odds of MV for patients with complete dataset. Multivariable regression showed increased odds of mechanical ventilation with age (OR 1.09 per year, 95% CI 1.03–1.16, p = 0.002), in males (OR 6.99, 95% CI 1.68–29.03, p = 0.007), in patients who presented with a qSOFA score ≥2 (OR 7.24, 95% CI 1.64–32.03, p = 0.009), with bilateral infiltrate (OR 18.92, 3.94–98.23, p<0.001) or with a CRP of 40 mg/l or greater (OR 5.44, 1.18–25.25; p = 0.030) on admission. Patients with more than seven days of symptoms on admission had decreased odds of mechanical ventilation (0.087, 95% CI 0.02–0.38, p = 0.001).
Our study identified several risk factors for unfavourable disease progression leading to MV in patients admitted with COVID-19 to a Swiss university hospital.
A quarter of the patients for which there were no limitation of care eventually required MV. MV occurred early during the course of hospitalization and the median duration of MV was shorter than previously reported [16, 17]. This effect is likely due to the limited duration of follow-up and could also result from a selection bias towards patients without limitations of care.
As infection with SARS-CoV-2 may cause an excessive host immune response, leading to ARDS and death , we would expect biomarkers of inflammation to be associated with unfavourable outcomes. In this study, CRP >40 mg/L on admission was associated with higher odds of MV, suggesting that an unfavourable course is more frequent in patients with a severe inflammatory response. Several studies have identified an increased risk of mortality in COVID-19 patients with elevated CRP [15, 19]. We believe that CRP is an ubiquitously measured biomarker whose result could potentially help clinicians assess the risk of MV in patients with COVID-19. Its use could be easily scaled up, and it is available as a point-of-care test.
In our study, the risk of mechanical ventilation increased with higher score values for NEWS, CRB-65 and qSOFA. qSOFA has been proven a useful predictor of mortality among patients with suspected infection , mainly of bacterial aetiology , but also influenza [21, 22]. We opted for including qSOFA in our final multivariate analysis, since it is widely used by clinicians in our institution, and is quick and easy to apply. A higher Sequential Organ Failure Assessment score (SOFA) has been previously linked to increased mortality due to COVID-19 . Data for its calculation are not routinely collected for all patients outside the ICU, making it a less pragmatic tool to quickly evaluate the risk of MV in this patient population.
Age as a categorical variable was not significantly associated with MV in our univariate model but was significantly associated with MV when included as a continuous variable in our multivariate model. Numerous studies have linked age to mortality and to MV for SARS-CoV-2 patients [2, 13, 23]. The lack of significant association in the univariate model could be due to the small size of our sample and to limitations of care agreed to in older patients.
We additionally identify male sex as a predictor of unfavourable outcome in patients with COVID-19, as previously described [24–26]. A study recently underlined different immune response in male and female SARS-CoV-2 patients, which could explain more severe evolution in male patients .
Several factors such as obesity, pregnancy, healthcare worker status, previous ACE inhibitors or ARB II treatment and immunosuppressive drugs before admission were not associated with severe disease in this study.
Our study has several limitations to be acknowledged. First, due to its very nature the sample size is limited, which could lead to observation bias in the analyses, with some findings likely to evolve over time. The follow-up period was also limited and several patients were still hospitalised at the time of data analysis. Finally, due to the constantly evolving nature of the epidemic, the clinical care of patients likely evolved during their hospitalisation as changes were made to the recommendations for treatments administered at LUH. The single centre nature of the study limits the generalisability of the results.
However, this study gives some insights in the epidemiology and clinical course of patients admitted in our institution with SARS-CoV-2 infection. We found that age, male sex, bilateral SARS-CoV-2 pneumonia, elevated CRP and qSOFA equal or greater to two increased the risk of mechanical ventilation. The timely identification of these patients could help us target treatment and better manage the attribution of resources.
- 1. World Health Organization: Novel Coronavirus (2019-nCov) Situation Report-22. World Health Organization, 2020, February 11, 2020. Available: https://www.who.int/docs/default-source/coronaviruse/situation-reports/20200211-sitrep-22-ncov.pdf?sfvrsn=fb6d49b1_2
- 2. Yang X, Yu Y, Xu J, Shu H, Xia J, Liu H, et al. Clinical course and outcomes of critically ill patients with SARS-CoV-2 pneumonia in Wuhan, China: a single-centered, retrospective, observational study. Lancet Respir Med. 2020;0. pmid:32105632
- 3. World Health Organization: Novel Coronavirus (2019-nCov) Situation Report-22. World Health Organization, 2020, March 11, 2020. Available: https://www.who.int/docs/default-source/coronaviruse/situation-reports/20200311-sitrep-51-covid-19.pdf?sfvrsn=1ba62e57_10
- 4. World Health Organization: Coronavirus disease 2019 (COVID-19): Weekly Epidemiological Update. World Health Organization, September 21, 2020. Available: https://www.who.int/docs/default-source/coronaviruse/situation-reports/20200921-weekly-epi-update-6.pdf?sfvrsn=d9cf9496_6
- 5. Corman VM, Landt O, Kaiser M, Molenkamp R, Meijer A, Chu DK, et al. Detection of 2019 novel coronavirus (2019-nCoV) by real-time RT-PCR. Eurosurveillance. 2020;25: 2000045. pmid:31992387
- 6. Harris PA, Taylor R, Minor BL, Elliott V, Fernandez M, O’Neal L, et al. The REDCap consortium: Building an international community of software platform partners. J Biomed Inform. 2019;95: 103208. pmid:31078660
- 7. Acute Respiratory Distress Syndrome: The Berlin Definition. JAMA. 2012;307: 2526–2533. pmid:22797452
- 8. Definition AKI. Kidney Int Suppl. 2012;2: 19–36. pmid:25018918
- 9. Singer M, Deutschman CS, Seymour CW, Shankar-Hari M, Annane D, Bauer M, et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA. 2016;315: 801–810. pmid:26903338
- 10. Bauer TT, Ewig S, Marre R, Suttorp N, Welte T, CAPNETZ Study Group. CRB-65 predicts death from community-acquired pneumonia. J Intern Med. 2006;260: 93–101. pmid:16789984
- 11. McGinley A, Pearse RM. A national early warning score for acutely ill patients. BMJ. 2012;345. pmid:22875955
- 12. van Smeden M, Moons KG, de Groot JA, Collins GS, Altman DG, Eijkemans MJ, et al. Sample size for binary logistic prediction models: Beyond events per variable criteria. Stat Methods Med Res. 2019;28: 2455–2474. pmid:29966490
- 13. Zhou F, Yu T, Du R, Fan G, Liu Y, Liu Z, et al. Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study. The Lancet. 2020;0. (20)30566–3 pmid:32171076
- 14. Yan L, Zhang H-T, Goncalves J, Xiao Y, Wang M, Guo Y, et al. A machine learning-based model for survival prediction in patients with severe COVID-19 infection. medRxiv. 2020; 2020.02.27.20028027.
- 15. Gong J, Ou J, Qiu X, Jie Y, Chen Y, Yuan L, et al. A Tool to Early Predict Severe 2019-Novel Coronavirus Pneumonia (COVID-19): A Multicenter Study using the Risk Nomogram in Wuhan and Guangdong, China. medRxiv. 2020; 2020.03.17.20037515.
- 16. Ling L, So C, Shum HP, Chan PKS, Lai CKC, Kandamby DH, et al. Critically ill patients with COVID-19 in Hong Kong: a multicentre retrospective observational cohort study. Crit Care Resusc J Australas Acad Crit Care Med. 2020.
- 17. Karagiannidis C, Mostert C, Hentschker C, Voshaar T, Malzahn J, Schillinger G, et al. Case characteristics, resource use, and outcomes of 10 021 patients with COVID-19 admitted to 920 German hospitals: an observational study. Lancet Respir Med. 2020;8: 853–862. pmid:32735842
- 18. Xu Z, Shi L, Wang Y, Zhang J, Huang L, Zhang C, et al. Pathological findings of COVID-19 associated with acute respiratory distress syndrome. Lancet Respir Med. 2020;8: 420–422. pmid:32085846
- 19. Lu J, Hu S, Fan R, Liu Z, Yin X, Wang Q, et al. ACP Risk Grade: A Simple Mortality Index for Patients with Confirmed or Suspected Severe Acute Respiratory Syndrome Coronavirus 2 Disease (COVID-19) During the Early Stage of Outbreak in Wuhan, China. Rochester, NY: Social Science Research Network; 2020 Feb. Report No.: ID 3543603. Available: https://papers.ssrn.com/abstract=3543603
- 20. Maitra S, Som A, Bhattacharjee S. Accuracy of quick Sequential Organ Failure Assessment (qSOFA) score and systemic inflammatory response syndrome (SIRS) criteria for predicting mortality in hospitalized patients with suspected infection: a meta-analysis of observational studies. Clin Microbiol Infect Off Publ Eur Soc Clin Microbiol Infect Dis. 2018;24: 1123–1129. pmid:29605565
- 21. Chang S-H, Yeh C-C, Chen Y-A, Hsu C-C, Chen J-H, Chen W-L, et al. Quick-SOFA score to predict mortality among geriatric patients with influenza in the emergency department. Medicine (Baltimore). 2019;98: e15966. pmid:31169726
- 22. Papadimitriou-Olivgeris M, Gkikopoulos N, Wüst M, Ballif A, Simonin V, Maulini M, et al. Predictors of mortality of influenza virus infections in a Swiss Hospital during four influenza seasons: Role of quick sequential organ failure assessment. Eur J Intern Med. 2020;74: 86–91. pmid:31899057
- 23. Hua J, Qian C, Luo Z, Li Q, Wang F. Invasive mechanical ventilation in COVID-19 patient management: the experience with 469 patients in Wuhan. Crit Care. 2020;24: 348. pmid:32546258
- 24. Shi Y, Yu X, Zhao H, Wang H, Zhao R, Sheng J. Host susceptibility to severe COVID-19 and establishment of a host risk score: findings of 487 cases outside Wuhan. Crit Care. 2020;24: 108. pmid:32188484
- 25. Du Y, Tu L, Zhu P, Mu M, Wang R, Yang P, et al. Clinical Features of 85 Fatal Cases of COVID-19 from Wuhan: A Retrospective Observational Study. Am J Respir Crit Care Med. 2020 [cited 9 Apr 2020]. pmid:32242738
- 26. Lagi F, Piccica M, Graziani L, Vellere I, Botta A, Tilli M, et al. Early experience of an infectious and tropical diseases unit during the coronavirus disease (COVID-19) pandemic, Florence, Italy, February to March 2020. Eurosurveillance. 2020;25. pmid:32372754
- 27. Takahashi T, Ellingson MK, Wong P, Israelow B, Lucas C, Klein J, et al. Sex differences in immune responses that underlie COVID-19 disease outcomes. Nature. 2020; 1–9. pmid:32846427