key: cord-0700437-62h6crla
authors: Chang, R.; Elhusseiny, K. M.; Yeh, Y.-C.; Sun, W.-z.
title: COVID-19 ICU and mechanical ventilation patient characteristics and outcomes - A systematic review and meta-analysis
date: 2020-08-18
journal: nan
DOI: 10.1101/2020.08.16.20035691
sha: 03012753c52e46ab9ec73d4b040a534d83f83ee3
doc_id: 700437
cord_uid: 62h6crla

Background Insight into COVID-19 intensive care unit (ICU) patient characteristics, rates and risks of invasive mechanical ventilation (IMV) and associated outcomes as well as any regional discrepancies is critical in this pandemic for individual case management and overall resource planning. Methods and Findings Electronic searches were performed for reports through May 1 2020 and reports on COVID-19 ICU admissions and outcomes were included using predefined search terms. Relevant data was subsequently extracted and pooled using fixed or random effects meta-analysis depending on heterogeneity. Study quality was assessed by the NIH tool and heterogeneity was assessed by I2 and Q tests. Baseline patient characteristics, ICU and IMV outcomes were pooled and meta-analyzed. Pooled odds ratios (pOR) were calculated for clinical features against ICU, IMV mortality. Subgroup analysis was carried out based on patient regions. A total of twenty-eight studies comprising 12,437 COVID-19 ICU admissions from seven countries were meta-analyzed. Pooled ICU admission rate was 21% [95% CI 0.12-0.34] and 69% of cases needed IMV [95% CI 0.61-0.75]. ICU and IMV mortality were 28.3% [95% CI 0.25-0.32], 43% [95% CI 0.29-0.58] and ICU, IMV duration was 7.78 [95% CI 6.99-8.63] and 10.12 [95% CI 7.08-13.16] days respectively. Besides confirming the significance of comorbidities and clinical findings of COVID-19 previously reported, we found the major correlates with ICU mortality were IMV [pOR 16.46, 95% CI 4.37-61.96], acute kidney injury (AKI) [pOR 12.47, 95% CI 1.52-102.7], and acute respiratory distress syndrome (ARDS) [pOR 6.52, 95% CI 2.66-16.01]. Subgroup analyses confirm significant regional discrepancies in outcomes. Conclusions This is the most comprehensive systematic review and meta-analysis of COVID-19 ICU and IMV cases and associated outcomes to date and the only analysis to implicate IMV's associtaion with COVID-19 ICU mortality. The significant association of AKI, ARDS and IMV with mortality has implications for ICU resource planning for AKI and ARDS as well as research into optimal ventilation strategies for patients. Regional differences in outcome implies a need to develop region specific protocols for ventilatory support as well as overall treatment.

Since its inception, the Coronavirus 2019 (COVID- 19) pandemic has placed unprecedented strain on healthcare globally with intensive care unit (ICU) care and ventilator availability being focal points of health resource allocation and planning. Understanding COVID-19 ICU and associated invasive mechanical ventilation (IMV) patient characteristics and outcomes as well as appreciating their regional variations is thus critically important for patient management as well as resource planning.

Despite an increasing number of studies relating to various aspects of severe COVID-19 and its ICU management; discrepancies in size, methodologies and research focus as well as regional differences in these studies calls for a comprehensive systematic review and analysis of reports to date.

Our review was conducted according to PRISMA guidelines [1] (S1 Table) . We searched Pubmed, Scopus, Embase, preprint servers bioRvix and medRvix and the Intensive Care National Audit and Research Center (ICNARC) website for publications through May 1 st 2020. A hand search through referenced published studies was additionally performed. Search terms used were ("COVID-19" OR "SARS-Cov-2") AND ("ICU" or "ventilator" OR "mechanical ventilation" OR "critical care" OR "intensive care" OR "hospitalized" OR "outcomes"). The study protocol was registered with the International Prospective Register of Systematic Reviews (PROSPERO CRD42020182482).

Studies reporting COVID-19 cases and their ICU admissions and IMV outcomes were included. Case reports, abstracts, or studies without outcomes were excluded. Two authors (RC and KME) independently screened titles and abstracts of retrieved studies and reviewed full-texts for further inclusion. Studies with larger outcome samples from same hospitals were analyzed to avoid patient Characteristics of studies and patients (Table 1) All included studies were observational and comprised 12,437 ICU patients admitted between December 2019 to May 1 2020. Of these, 6,875 patients were on IMV. Nine studies were from the USA [8] [9] [10] [11] 17, 18, 23, 25, 26] , 13 from China [12, 14, 16, [20] [21] [22] [27] [28] [29] [30] [31] [32] [33] , two from the UK [24, 32] , and one each from Italy [15] , Spain [13] , France [19] , and Mexico [35] .

Risk of bias assessment (Table 1)   7   162  163  164  165  166  167  168  169   170   171   172  173  174   7   . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 18, 2020. . https://doi.org/10.1101/2020.08. 16.20035691 doi: medRxiv preprint Fifteen studies were good and 13 were fair as assessed by the NIH tool. Follow-up duration was insufficient or unreported in seven [8, 10, 12, 14, 26, 29, 35] among which three had fair assessments. Of note, 14 studies [8, 9, [12] [13] [14] [15] [17] [18] [19] 21, 22, 24, 25, 34] . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted August 18, 2020. . https://doi.org/10.1101/2020.08. 16.20035691 doi: medRxiv preprint 0.34) (Fig 2a) and pooled ICU mortality rate of 12,437 patients from 20 studies [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [21] [22] [23] [24] [25] 31, 33, 35] was 28.3% (95% CI 0.27-0.36) (Fig 2b) , while the worst case scenario mortality is 60% (95% CI 0.49-0.69) considering 4,697 patients with unknown outcomes. ICU length of stay (LoS) from five studies [8, 10, 11, 15, 24] had a pooled mean duration of 7.78 (7.05-8.51) days. Substantial heterogeneity was observed with ICU outcomes (p-value <0.1, I 2 > 60%) and not explained by subgroup analysis. Eggers' test revealed no publication bias ( Figure 3 ).

IMV prevalence, duration and outcome. Pooled IMV prevalence and mortality was respectively 69% (95% CI 0.61-0.75) from 18 studies [8, [9] [10] [11] [12] [15] [16] [17] [18] [19] 21, 22, 24, 27, 31, 33, 35] with 10,240 cases (Figs 2c) and 43% (95% CI 0.29-0.58) from 12 studies [10, 11, 14, 18, 22, 23, 26, 28, 31, [33] [34] [35] with 2,212 cases based on a best case scenario, with worst case scenario mortality of 74% (95% CI 0.54-0.87) (Fig. 2d ).

IMV duration was pooled from four studies [9, [15] [16] [17] with a mean duration of 10.12 days (95% CI 7.08-13.16). There was significant heterogeneity among IMV mortalities (p-value < 0.001, I 2 >90%), which was not explained by subgroup analysis, and Egger's test revealed no publication bias (Fig 4) . is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted August 18, 2020. . https://doi.org/10.1101/2020.08. 16.20035691 doi: medRxiv preprint CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 18, 2020. . https://doi.org/10.1101/2020.08. 16.20035691 doi: medRxiv preprint CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 18, 2020. .

Prevalence of comorbidities and clinical features. The prevalence of comorbidities as well as presenting clinical features are listed in Table 2 with hypertension (HTN) 51%, obesity (BMI>30kg/m 2 ) 35% , diabetes (DM) 30% and fever 81%, cough 76%, dyspnea 75% the 3 most prevalent comorbidities and symptoms respectively. The pER's of select reported test findings were: bilateral infiltrates on chest radiography 83%, lymphopenia 78%, elevated alanine (ALT), aspartate aminotransferases (AST) 71%, 66.3% respectively, elevated troponin 22%. The pER for concurrent acute respiratory distress syndrome (ARDS) and acute kidney injury (AKI) were 84% and 32% respectively. . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 18, 2020. . 

. CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 18, 2020. . Association of patient characteristics with ICU and IMV mortality. We analyzed associated COVID-19 ICU patient characteristics and initial laboratory findings for potential association with ICU (Table 3 and Table 4 is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 18, 2020. . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 18, 2020. . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 18, 2020. Table 4 . Meta-analysis findings of the association between laboratory findings and COVID-19 ICU mortality. Abbreviation: ALT, alanine aminotransferase; AST, aspartate aminotransferase ; CRP, C-reactive protein; LDH, lactate dehydrogenase; WMD, weighted mean difference.

Regional differences in ICU and IMV outcomes. Subgroup analysis revealed the following regional 

This is the most comprehensive systematic review and meta-analysis of COVID-19 ICU cases and outcomes to date to our knowledge. Many reviews explored COVID-19 prognosis, with some reporting is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 18, 2020. . generally on all [36] or hospitalized cases [37] , or from a certain region [38] , or focused on specific aspects such as mortality [39] or outcomes [40] , with only one systematic review specifically addressing IMV mortality [41] . Here we attempt to provide a comprehensive assessment of ICUadmitted COVID-19 patient characteristics, regional discrepancies, as well as overall ICU and IMV specific outcomes and their associated factors.

Our pooled ICU best case mortality rate of 28.3% (95% CI 0.27-0.32) amongst COVID-19 patients is lower than earlier reports from Wuhan's 78% [20] and Seattle's 85% [8] , but in line with the more recent report of 31% from Atlanta [10] , as well as 25.7% and 41.6% reported in reviews by Quah et al. [39] and Armstrong et al. [40] respectively.

Our 43% (95% CI 0.29-0.58) pooled IMV mortality rate is also significantly lower than earlier reports of 86-97% from Wuhan [20.31] or 71-75% from Seattle [8.11] . The more recent Atlanta study revealed improved IMV mortality of 35.7% [10] , which is more consistent with the 35% to 46% mortality respectively reported for patients intubated with H1N1 influenza pneumonia [42] and other causes of ARDS [43] .

We found a pooled mean ICU LoS of 7.8 (7-8.5) days comparable to results from a review reporting a median LoS of 8 (5-13) days for patients from China and 7 (4-11) days from outside China [44] . As for IMV duration, we found a pooled mean duration of 10 days (95% CI 7-13), which is shorter than earlier reports from Wuhan and Seattle of 10-17 days [41] .

Taken together, our results are consistent with the recent finding by Armstrong et alof improved COVID-19 ICU outcome over time, which was attributed to increased ICU experience, and evolving admission criteria, treatments, and demographics [40] . We additionally hypothesize that aimprovements in reported mortalities over time and regions may arise as healthcare systems become less overwhelmed, and as more recent studies account for more complete patient outcomes. Significant differences between reporting geographic regions could additionally be due to different ICU admission CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 18, 2020. . https://doi.org/10.1101/2020.08. 16.20035691 doi: medRxiv preprint criteria and treatment protocols, diversity of ethnic genetic factors as well as mutational evolution of the virus itself [45] , but more studies are necessary to explore these aspects.

Systematic reviews and analyses of select patient demographics, comorbidities, selective laboratory findings and complications of the disease have been published [46, 47] , although there is no systematic review addressing all these aspects in COVID-19 ICU or IMV to date. We found that elderly COVID-19 patients with comorbidities have higher ICU and IMV fatalities; but this is not unique to COVID-19 as many studies have noted that ICU patients generally consists of elderly patients who have greater mortality [48] .

Interestingly, despite higher absolute prevalences of certain comorbidities and presenting features in COVID-19 ICU patients, the relative prevalence is the same as reported in less severe symptomatic or hospitalized cases. For example, a review of mostly Chinese studies similarly found hypertension to be the most common comorbidity in COVID-19 [49] ; and another review identified fever (88.7%), cough (57.6%) and dyspnea (45.6%) to be similarly most common manifestations [50] .

Our results also corroborated findings from meta-analyses where DM [51] , CVD [52] , abnormal liver function [53] have been found prevalent, associated and potentially prognostic in COVID-19.

For comorbidities, we found CVD, HTN and DM to be significantly associated with ICU mortality, which is consistent with findings by a review that identified these same comorbidities to be correlated with disease severity and ICU admissions [49] . Separately, Wang et al. also found HTM, DM, and CVD to correlate with COVID-19 severity and mortality [54] .

Of clinical findings: obesity, WBC, AST, D-dimer PT, CRP, lactate, LDH, and total bilirubin as well as lower PaO2/FiO2, lymphocyte count and albumin to be significantly associated with ICU outcome (Table 4) , which is again consistent with previous reports on associations of obesity, liver dysfunction, lymphocytopenia, coagulopathy, malnutrition, renal dysfunction, and cardiac injury with COVID-19 severity or mortality. CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 18, 2020. . While most clinical correlates we found above are consistent and noted by other reports to date, three correlates with ICU mortality deserve discussion. One is AKI (pOR 12.47, 95% CI 1.52-102.7), for which there are divergent reports on prevalence and significance. Thus, while some large COVID-19 systematic reviews excluded AKI [27, 37] and it is absent from some COVID-19 critical care management guidelines [55, 56] , its incidence in hospitalized COVID-19 has been reported to be as low as 0.5% [29] and some researchers even concluded that COVID-19 does not cause AKI [57] . However, we found a pooled AKI prevalence of 32% and a New York report cited 46% of hospitalized and 68% of ICU COVID-19 cases with AKI with 34% of those in ICU requiring replacement renal therapy (RRT), and with an OR of 20.9 (95% CI 11.7-37.3) for AKI vs non-AKI associated ICU mortality [58] .

Along with our analysis, this raises the concern that AKI maybe seriously under appreciated in critically ill COVID-19 patients.

ARDS is another significantly associated complication (pOR 6.52, 95% CI 2.66 -16.01) we found. Problematically, ARDS itself is a heterogeneous syndrome, with the six studies [8, 9, 22, [31] [32] [33] we analyzed variously applying the Berlin criteria [59] , WHO guidance [60] , or no criteria at all for its diagnosis. Then there is a notion that COVID-19 ARDS may be a separate entity and that current criteria are inadequate in guiding treatment [61] . Indeed, ARDS per the Berlin criteria defines a one week onset limit whereas COVID-19 related ARDS has a median onset of 8-12 days [20] .

Pathophysiologically, there are also differences between COVID or non-COVID ARDS and Gattinoni et al. have highlighted the non-uniformity of ARDS in COVID-19 based on elastance, ventilation/perfusion ratio, lung weight and recruitability [62] . Thus, further efforts are urgently needed in clarifying the pathophysiology and developing new stratifications of severity for COVID-19 specific ARDS to guide optimal ventilatory support for ICU COVID-19 patients.

Finally, the issue of defining ARDS in COVID-19 leads to applicability of IMV as its treatment. CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 18, 2020. . Early studies have reported ICU mortality rates as high as 86-97% for IMV in COVID-19 patients [18, 20, 31] , but no study to date specifically reported on the relative odds for IMV v. non-IMV in terms of outcome. Our finding that IMV itself is significantly associated with ICU mortality (pOR 16.46, 95% CI 4.37-61.96) coupled with subgroup analyses showing regions with higher ICU mortalities coincide with regions with higher IMV rates (USA, UK, Italy) are notable because it questions the benefits of IMV as well as the wisdom of public policies emphasizing ventilator availability in coping with the pandemic. The concern for potentially harmful overuse of IMV has been raised by past research, since IMV itself carries risks that may adversely affect survival [63] . Additionally, inappropriate timing of intubation too early or too late, inadequately trained or overwhelmed staff, as well as improper ventilation settings and IMV associated pneumonia can all potentially enhance mortality; and this should all be investigated more fully in COVID-19 cases.

From prior studies on ARDS, we know that other ventilation strategies such as non-invasive ventilation (NIV) may improve mortality over IMV [64] , and there are suggestions that non-rebreathing masks and prone positioning may improve outcome in hypoxemic COVID-19 patients [65] . The comparison of different ventilation strategies is beyond the scope of our review and there is yet no published study comparing ventilation strategies in COVID-19. However, our data supports a more conservative approach to initiating IMV, as well as the need for well designed trials to clarify the role of various ventilation strategies in COVID-19 associated ARDS.

Finally, our results highlight significant regional discrepancies in COVID-19 ICU and IMV rates and outcomes. As the pandemic evolves, regional differences of local health systems and their resources as well as genetic differences amongst populations and possible viral evolution over time can all account for such differences and deserves study. Regional differences also imply the need for flexible region-specific treatment protocols especially in resource-limited settings [65] instead of the CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 18, 2020. . https://doi.org/10.1101/2020.08. 16.20035691 doi: medRxiv preprint aim to set uniform international protocols for treatment.

There are multiple limitations with our study. We have issues with incompleteness of data where not all outcomes are accounted for in many studies, as well as underlying heterogeneity of data with small number of studies and studies with small numbers of patients. Moreover, all our retrieved reports were retrospective cohorts, which limits our ability to infer causality; and some of these were not peerreviewed. Furthermore, we could not adjust for confounders of potentially related variables in an analysis of survival vs. non-survival based on these studies. Finally, regional differences in health systems and their different treatment protocols may also confound results.

This is the largest and most comprehensive review and meta-analysis of ICU and IMV outcomes in COVID-19 ICU patients. Our findings parallel earlier reports on prevalences of associated comorbidities and clinical findings in COVID-19 patients implying largely the same set of factors is associated with severity and outcome regardless of stage of disease. However, we highlight the significant association of AKI and ARDS associated IMV in ICU outcomes, which deserve further research to refine diagnostic criteria and enable the development of optimally tailored treatment strategies, as well as better planning and allocation of critical care resources. Finally significant regional discrepancies in outcomes implies a need for further studies to allow a broader perspective of factors associated with ICU and IMV risks and outcomes under different healthcare systems in different populations. CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted August 18, 2020. . https://doi.org/10.1101/2020.08. 16 CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted August 18, 2020. . https://doi.org/10.1101/2020.08. 16.20035691 doi: medRxiv preprint 

Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement

Meta-analysis in clinical trials

Measuring inconsistency in metaanalyses

Operating characteristics of a rank correlation test for publication bias

Funnel plots for detecting bias in meta-analysis

Estimating the sample mean and standard deviation from the sample size, median, range and/or interquartile range

Characteristics and outcomes of 21 critically ill patients with COVID-19 in Washington state

Characterization and clinical course of 1000 patients with coronavirus disease 2019 in New York: retrospective case series

ICU and ventilator mortality among critically ill adults with coronavirus disease 2019

Covid-19 in critically ill patients in the Seattle region -Case series

Characterization of clinical progression of COVID-19 patients in Shenzhen

A cohort of patients with COVID-19 in a major teaching hospital in Europe

Baseline characteristics and outcomes of 1591 patients infected with SARS-CoV-2 admitted to ICUs of the Lombardy region

Biased and unbiased estimation of the average lengths of stay in intensive care units in the COVID-19 pandemic

Characteristics of hospitalized adults with COVID-19 in an integrated health care system in California

Presenting characteristics, comorbidities, and outcomes among 5700 patients hospitalized with COVID-19 in the New York City area

High prevalence of obesity in severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) requiring invasive mechanical ventilation

Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study

Clinical findings in critical ill patients infected with SARS-Cov-2 in Guangdong province, China: A multi-center, retrospective, observational study. medRxiv

Clinical characteristics of 34 COVID-19 patients admitted to intensive care unit in Hangzhou

Respiratory pathophysiology of mechanically ventilated patients with COVID-19: A cohort study

Intensive Care National Audit and Research Centre. ICNARC report on COVID-19 in critical care

Clinical characteristics of hospitalized covid-19 patients in New York City

Clinical characteristics of covid-19 in New York City

Hospitalization and Critical Care of 109 Decedents with COVID-19 pneumonia in Wuhan

Clinical features and short-term outcomes of 102 patients with coronavirus disease 2019 in Wuhan, China

Clinical characteristics of coronavirus disease 2019 in China

Clinical characteristics and outcomes of patients undergoing surgeries during the incubation period of COVID-19 infection

Clinical course and outcomes of critically ill patients with SARS-CoV-2 pneumonia in Wuhan, China: a single-centered, retrospective, observational study

Clinical features of patients infected with 2019 novel coronavirus in Wuhan

Clinical course and outcomes of 344 intensive care patients with COVID-19

Features of 16,749 hospitalised UK patients with COVID-19 using the ISARIC WHO Clinical Characterisation Protocol

COVID-19 Severe pneumonia in Mexico City -First experience in a Mexican hospital

2020) Incidence, clinical characteristics and prognostic factor of patients with COVID-19: a systematic review and meta-analysis

Prevalence and risk factors of mortality among hospitalized patients with COVID-19: A systematic review and meta-analysis

Predictive symptoms and comorbidities for severe COVID-19 and intensive care unit admission: a systematic review and meta-analysis

Mortality rates of patients with COVID-19 in the intensive care unit: a systematic review of the emerging literature

Outcomes from intensive care in patients with COVID-19: a systematic review and meta-analysis of observational studies

Mechanical ventilation utilization in COVID-19: A systematic review and meta-analysis

Critically ill patients with 2009 influenza A(H1N1) in Mexico

Epidemiology, patterns of care, and mortality for patients with acute respiratory distress syndrome in intensive care units in 50 Countries

COVID-19 length of hospital stay: a systematic review and data synthesis

Genetic diversity and evolution of SARS-CoV-2

Predictors of mortality in hospitalized COVID-19 patients: A systematic review and meta-analysis

Imaging and clinical features of patients with 2019 novel coronavirus SARS-CoV-2: A systematic review and meta-analysis

Mortality in critically ill elderly individuals receiving mechanical ventilation

Comorbid chronic diseases are strongly correlated with disease severity among COVID-19 patients: A systematic review and meta-analysis

Clinical, laboratory and imaging features of COVID-19: A systematic review and meta-analysis

Diabetes mellitus is associated with increased mortality and severity of disease in COVID-19 pneumonia -A systematic review, meta-analysis, and metaregression

Cardiovascular diseases and COVID-19 mortality and intensive care unit admission: A systematic review and meta-analysis

COVID-19 and liver dysfunction: a systematic review and meta-analysis of retrospective studies

Comorbid chronic diseases and acute organ injuries are strongly correlated with disease severity and mortality among COVID-19 patients: A systemic review and meta-Analysis

Intensive care management of coronavirus disease 2019 (COVID-19): challenges and recommendations

Surviving sepsis campaign: guidelines on the management of critically ill adults with coronavirus disease 2019 (COVID-19)

Coronavirus disease 19 infection does not result in acute kidney injury: An analysis of 116 hospitalized patients from Wuhan, China

Acute kidney injury in hospitalized patients with COVID-19

The Berlin definition of ARDS: an expanded rationale, justification, and supplementary material

Clinical management of severe acute respiratory infection when novel coronavirus (2019-nCoV) infection is suspected: interim guidance

The ARBSnet protocol may be detrimental in COVID-19

COVID-19 pneumonia: different respiratory treatments for different phenotypes?

Outcomes associated with invasive and noninvasive ventilation among patients hospitalized with exacerbations of chronic obstructive pulmonary disease

Respiratory support in COVID-19 patients, with a focus on resource-Limited Settings

Supporting Information S1 Table. Prisma Checklist S1 File. PROSPERO Registration