Characterisation and Outcomes of ARDS Secondary to Pneumonia in Patients with and Without SARS-Cov-2: a Single-Centre Experience
Total Page:16
File Type:pdf, Size:1020Kb
BMJ Open Resp Res: first published as 10.1136/bmjresp-2020-000731 on 30 November 2020. Downloaded from Critical care Characterisation and outcomes of ARDS secondary to pneumonia in patients with and without SARS- CoV-2: a single- centre experience Rahul Y Mahida ,1 Minesh Chotalia,1,2 Joseph Alderman,1,2 Chhaya Patel,3 Amber Hayden,4 Ruchi Desai,4 Emily Beesley,4 Louise E Crowley,1 Marina Soltan,1 Mansoor Bangash,1,2 Dhruv Parekh ,1,2 Jaimin Patel,1,2 David R Thickett1 To cite: Mahida RY, ABSTRACT Key messages Chotalia M, Alderman J, Introduction Acute respiratory distress syndrome et al. Characterisation and (ARDS) is the major cause of mortality in patients with This is the first UK study comparing the clinical outcomes of ARDS secondary SARS- CoV-2 pneumonia. It appears that development of ► characteristics and outcomes of acute respiratory to pneumonia in patients with ‘cytokine storm’ in patients with SARS- CoV-2 pneumonia and without SARS- CoV-2: distress syndrome (ARDS) secondary to pneumonia precipitates progression to ARDS. However, severity a single- centre experience. in patients with and without SARS- CoV-2. scores on admission do not predict severity or mortality BMJ Open Resp Res in patients with SARS- CoV-2 pneumonia. Our objective ► Are patients with SARS- CoV-2 ARDS clinically dis- 2020;7:e000731. doi:10.1136/ tinct to other patients with ARDS, therefore, requiring bmjresp-2020-000731 was to determine whether patients with SARS- CoV-2 ARDS are clinically distinct, therefore requiring alternative alternative management strategies? ► Additional material is management strategies, compared with other patients ► While the clinical syndromes of ARDS secondary to published online only. To SARS- CoV-2 and community- acquired pneumonia with ARDS. We report a single- centre retrospective study copyright. view, please visit the journal comparing the characteristics and outcomes of patients are similar, SARS- CoV-2 patients initially have a low- online (http:// dx. doi. org/ 10. with ARDS with and without SARS- CoV-2. er requirement for vasopressor support and require 1136/ bmjresp- 2020- 000731). Methods Two intensive care unit (ICU) cohorts of patients prolonged ventilation support. at the Queen Elizabeth Hospital Birmingham were RYM and MC are joint first analysed: SARS- CoV-2 patients admitted between 11 authors. March and 21 April 2020 and all patients with community- fulfilling the Berlin criteria for acute respira- acquired pneumonia (CAP) from bacterial or viral infection 1–3 Received 30 July 2020 tory distress syndrome (ARDS). Inten- who developed ARDS between 1 January 2017 and 1 Revised 15 October 2020 sive care unit (ICU) mortality rates of up November 2019. All data were routinely collected on the Accepted 19 October 2020 to 68% from SARS- CoV-2 ARDS have been http://bmjopenrespres.bmj.com/ hospital’s electronic patient records. 3 4 Results A greater proportion of SARS-CoV -2 patients reported. A multinational study undertaken were from an Asian ethnic group (p=0.002). SARS- CoV-2 prior to the SARS-CoV -2 pandemic found that patients had lower circulating leucocytes, neutrophils pneumonia was the underlying risk factor in and monocytes (p<0.0001), but higher CRP (p=0.016) on 59% of ARDS cases.5 Recently, a retrospec- ICU admission. SARS- CoV-2 patients required a longer tive cohort study undertaken in Wuhan, duration of mechanical ventilation (p=0.01), but had lower China found that 41.8% of adult patients vasopressor requirements (p=0.016). admitted with SARS- CoV-2 pneumonia devel- The clinical syndromes and respiratory Discussion oped ARDS.6 Risk factors for mortality from mechanics of SARS- CoV-2 and CAP- ARDS are broadly similar. However, SARS- CoV-2 patients initially have SARS- CoV-2 pneumonia include increasing a lower requirement for vasopressor support, fewer age, coronary heart disease, diabetes mellitus on October 1, 2021 by guest. Protected 7 8 circulating leukocytes and require prolonged ventilation and chronic kidney disease. Admission support. Further studies are required to determine whether CURB-65 scores do not predict severity or the dysregulated inflammation observed in SARS- CoV-2 mortality in patients with SARS-CoV -2 pneu- ARDS may contribute to the increased duration of monia, due to a rapidly progressing clinical © Author(s) (or their respiratory failure. 7 employer(s)) 2020. Re- use course. However, lymphopenia, eosinopenia permitted under CC BY. and elevated acute phase proteins are predic- Published by BMJ. tors of increased disease severity.7 9 10 For numbered affiliations see It appears that development of ‘cyto- end of article. INTRODUCTION kine storm’ in patients with SARS-CoV -2 Correspondence to SARS- CoV-2 pneumonia can progress to pneumonia is associated with progression Dr Rahul Y Mahida; hypoxaemic respiratory failure requiring to ARDS, however, the cytopathic effects r. mahida@ bham. ac. uk mechanical ventilation, with patients of the viral pneumonia may be just as Mahida RY, et al. BMJ Open Resp Res 2020;7:e000731. doi:10.1136/bmjresp-2020-000731 1 BMJ Open Resp Res: first published as 10.1136/bmjresp-2020-000731 on 30 November 2020. Downloaded from Open access important.9 11 12 Histological analysis of postmortem lung membrane oxygenation centre, and were managed as tissue from SARS- CoV-2 pneumonia patients has shown per previously published algorithms.18 All patients were diffuse alveolar damage (DAD).13 14 The presence of intubated, sedated and mechanically ventilated with DAD has previously been used to identify a subphenotype positive pressure ventilation. Baseline demographic, of ARDS with higher mortality.15 16 These findings suggest comorbidities, laboratory investigations, physiological that a similar pathological process occurs in patients with parameters and severity scores (Acute Physiology And ARDS with and without SARS- CoV-2. Chronic Health Evaluation II [APACHE II], Sequential Our objective was to determine whether patients with Organ Failure Assessment [SOFA] and Murray Lung SARS- CoV-2 ARDS are clinically distinct, therefore, Injury) were collected at ICU admission. Sequential phys- requiring alternative management strategies, compared iological and laboratory parameters were collected for 7 with other patients with ARDS.17 This retrospective study days whist on ICU. Sequential data are not available for provides clinical characterisation of ARDS patients with all patients due to deaths of 15 SARS-CoV -2 and 5 CAP- and without SARS- CoV-2 admitted to a single- centre ICU. ARDS patients within 1 week. Statistical analysis was performed using GraphPad Prism version 8.0. Data distributions were non- parametric METHODS and are presented as median with IQR for continuous This is a single-centre, observational, retrospective variables and number (percentage) for categorical vari- study from the ICU of the Queen Elizabeth Hospital ables. Differences between patient groups were analysed Birmingham, UK. All data were routinely collected on the using Mann- Whitney- U test for continuous data and Fish- hospital’s electronic patient records. Only data that were er’s exact test for categorical data. Two- sided tests were obtained as part of routine clinical care were collected used for all comparisons with p<0.05 considered statisti- for this study. All data were anonymised and entered by cally significant. the Local Clinical Care Team, without linkage to any patient identifiers, in accordance with national and local guidance. RESULTS Two ICU cohorts of patients were analysed: SARS-CoV -2 A total of 111 patients with SARS-CoV -2 ARDS and pneumonia patients admitted between 11 March and 29 patients with CAP-ARDS met the inclusion criteria 21 April 2020 (online supplemental figure 1) and all (table 1). Many patients (n=33) screened for CAP- ARDS copyright. patients with community-acquired pneumonia (CAP) were excluded, as pneumonia had developed >48 hours from bacterial or viral infection who developed ARDS after hospital admission. Patient demographic details between 1 January 2017 and 1 November 2019 (online are shown in table 1, with both groups being broadly supplemental figure 2). Patients who developed hospital- similar except for ethnic background. A greater propor- acquired pneumonia (HAP: defined as onset >48 hours tion of SARS- CoV-2 patients were of Asian/Asian British after hospital admission) were excluded. This was to iden- ethnicity (p=0.002), and a lower proportion were of tify a more relevant, directly comparable control group, White ethnicity (p=0.012), compared with CAP-ARDS in which infection was acquired in the community, and patients http://bmjopenrespres.bmj.com/ pneumonia was present at hospital admission. The caus- On ICU admission, SARS- CoV-2 patients had signifi- ative organisms and clinical course of CAP and HAP cantly lower APACHE- II and SOFA scores than CAP-ARDS also differ significantly. Patients with ARDS secondary to patients (see table 1: p<0.0001). SOFA scores remained others causes were also excluded. The sample sizes were lower in SARS-CoV -2 patients for 7 days following ICU determined pragmatically, to include all SARS-CoV -2 admission (figure 1A). pneumonia patients admitted to the ICU within the first SARS- CoV-2 patients had lower circulating leukocytes, 6 weeks of the pandemic. The sample size of the CAP- neutrophils and monocytes (p<0.0001 for all) than CAP- ARDS control group included all such patients within the ARDS patients on ICU admission. Leucocytes and neutro- 3 years preceding the pandemic, since all these patients phil counts remained lower in SARS- CoV-2 patients for 3 would have their hospital records within a rapidly acces- days following ICU admission, whereas monocyte counts on October 1, 2021 by guest. Protected sible electronic system, and would have received proto- remained lower for 6 days (figure 1B–D). Albumin was colised management similar to that received by the lower (p=0.003) while CRP (p=0.016) and platelet count SARS-CoV -2 ARDS patients.