key: cord-0786656-26vw4icc
authors: Al-Hindawi, A.; Sokhi, J.; Cuddihy, J.; Lockie, C.; Christie, L.; Davies, R.; Singh, S.; Vizcaychipi, M.; Hayes, M.; Sisson, A.; Keays, R.
title: COVID-19 in London, a Case Series Demonstrating Late Improvement in Survivors
date: 2020-05-20
journal: nan
DOI: 10.1101/2020.05.16.20103853
sha: 8141320a3ac40ae52df834eec761a5c2b1a15d94
doc_id: 786656
cord_uid: 26vw4icc

Objective To determine whether the trajectories of survivors and non-survivors are different in patients admitted to intensive care in London. Design In this case series of 15 survivors and 16 non-survivors, data from admission to discharge was collected and aligned to lowest PaO2/FiO2 ratio where aggregation and trends were demonstrated. Setting Single centre case-series in London, Intensive Care. Participants All non-survivors were included (n=16). A biased set of survivors (n=15) who were demonstrated an unexpected and rapid recovery after a prolonged period of mechanical ventilatory support. Results Respiratory failure trajectories of survivors and non-survivors were similar once aligned indicating, from a respiratory function perspective, it is difficult to identify survivors from non-survivors with some survivors improving late in their disease (day 20 - 30 from symptom onset) Non-survivors are admitted earlier in their disease (p < 0.05) and had worse organ failure requirements prior to the nadir of their respiratory funciton (p < 0.05) compard to survivors. Conclusion Analysis of multiple factors fails to differentiate between survivors and non-survivors. Even when faced with multiorgan failure, perseverance until discharge must be advocated as late improvements do occur in survivors.

The SARS-COVID-2 virus giving rise to the syndrome of COVID-19 is a world-wide pandemic. Multiple case reports from China and America have highlighted the seriousness of the disease with increased mortality (49 -88%), length of stay and time of presentation [1] [2] [3] . In those publications, the median time from onset of symptoms to ICU admission was around 10 -11 days. The median time from illness onset to mechanical ventilation was 14.5 days, and the median time from illness onset to death was 18.5 days 1 . Zhou et al demonstrated that, in their retrospective cohort study, 32 Wuhan patients required invasive mechanical ventilation, of whom 31 (97%) died with Extracorporeal Membraneous Oxygenation (ECMO) was used in 3 patients who all died. The ICU length of stay in nonsurvivors was a median of 8 days (range 4-12 days). A more comprehensive analysis by the Chinese Centre for Disease Control and Prevention showed that the case fatality rate, across China, for patients admitted to intensive care was 49% (2087 admitted to ICU out of the 72,314 patients) 4 .

In New York, it was reported that the mortality in patients who received mechanical ventilation was 76.4% in 18-65-year olds and 97.2% in those over 65 years. This did not include the patients still remaining in hospital whose outcome was undetermined up to that point 2 . In Washington State, an ICU mortality of 67% has been reported whilst a further 24% of patients still remained on ICU with renal failure occurring more frequently in non-survivors 5 . None of the survivors had a serum ferritin greater than 2000 µg/L.

In the United Kingdom, as of the 8 th of May 2020, of the critical care outcomes reported (8250 patients), the mortality rate stands at just over 46.8%. The total number of patients admitted to ICU is 10758 and even if all the patients remaining in hospital were to survive the death rate would still be around 33%. Over 60% of these deaths occur within the first 10 days of admission to ICU and 92% of deaths have occurred by day 20. The median length of stay amongst non-survivors was 7 days (4 -13 days). Amongst patients who only required All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) 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 May 20, 2020. . https://doi.org/10.1101/2020.05.16.20103853 doi: medRxiv preprint 4 advanced respiratory support, mortality was 43.9% but if renal failure supervened mortality increased to 65.4% 6 .

It has been advised that triaging is necessary in the context of overwhelmed medical services: J. Phua et al state to 'Consider critical care triage to ration scarce resources in pandemic; prioritise patients who will benefit from ICU care' 7 . The concept of triaging during a pandemic and consideration of resource allocation has been implied at a UK National level, although there have been various iterations as time has gone by 8 .

The purpose of this paper is to describe the experience with COVID-19 patients mechanically ventilated on ICU who showed an unexpected and rapid recovery after a prolonged period of mechanical ventilatory support. Many of these patients required vasopressor infusions, multiple episodes of proning and some required renal replacement therapy culminating in a crude, unadjusted, mortality of 27%. We contrast the survivor trajectory with non-survivors demonstrating that the two groups are relatively indistinguishable until discharge. All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) 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 May 20, 2020. . https://doi.org/10.1101/2020.05.16.20103853 doi: medRxiv preprint 5 

The study was approved by the Chelsea and Westminster NHS Foundation Trust clinical governance team. As the report involves routinely collected non-identifiable clinical audit data, no ethical approval was required under the UK policy framework for health and social care.

Data collected SARS-CoV-2 positive patients addmitted to Chelsea and Westminster NHS Foudnation Trust Intensive Care Unit from 9 th of March till the 5 th of May 2020. All nonsurvivors were included in collection and analysis. A sample of survivors was chosen who had a prolonged course on ICU with features which pointed to a poor outcome but who went on to recover. This cohort of survivors thus represents patients who, given the current literature, withdrawal of care would have been considered.

Observational, pathological, and ventilatory values from patients were extracted from Cerner Electronic Patient Record (EPR) and analysed in R 3.6. The lowest PaO 2 /FiO 2 (PF) ratio of an arterial sample on each day was recorded.

To facilitate aggregation and identification of trends, patients' symptom days were aligned such that the lowest PF ratio on ICU constituted Day 0. Negative day numbers (as seen in Figure 1 and Figure 3 ) thus represent days patients were on ICU prior to the lowest PF ratio.

In Figure 2 , day 0 is defined as the first day of symptom onset; patient trajectories commence from ICU admission.

As this is a descriptive case-series, three patient-based variables were focused on due to their interest in COVID-19 literature; namely the PF ratio; lymphocyte count and ferritin. Due to the magnitude of ferritin values across patients, the natural-log is used.

For organ support variables, proning, vasopressors and renal replacement therapy were chosen. Vasopressor use was deemed if any of noradrenaline, vasopressin or metaraminol were delivered during that day. Proning episodes were extract from free text entries in Cerner EPR. All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) 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 May 20, 2020. Distributions in Figure 3 were tested for normality using Shaprio-Wilk test followed by an unpaired T-Test. All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) 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 May 20, 2020. . https://doi.org/10.1101/2020.05.16.20103853 doi: medRxiv preprint Results Survivors (n=15) and non-survivors (n=16) were analysed. Figure 1 demonstrates that, when patients are aligned from their lowest PF ratio day, there are similar trajectories in the survivors and non-survivors. Non-survivors start with a lower PF ratio than survivors, but PF nadir alone does not determine death as ratios recover in both groups. Survivor PF ratios improve to a greater extent than that of non-survivors. Lymphocytes also recover with little to be gleaned from absolute values whereas there appears a trend in ferritin levels that differentiate survivors from non-suvivors; ferritin levels are initially higher in non-survivors and plateau unlike survivors whose values recover from their zenith. (which was not certified by peer review) 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 May 20, 2020. . https://doi.org/10.1101/2020.05.16.20103853 doi: medRxiv preprint 8

The survivors, who were younger and had lower severity of disease at presentation (Table   1) , had higher PF ratios on admission to ICU. In the majority of survivors, presentation to hospital is later. Non-survivors presented earlier to hospital (p < 0.05), required intensive care therapy earlier in their illness course (p < 0.05), with worse respiratory function but ultimately had similar PF trajectories to survivors once aligned.

Descriptively, in the survivor group, the lymphocyte count recovers near the nadir of the PF ratio and coincides with the peak ferritin count. There doesn't seem to be a similar trend in ferritin values amongst non-survivors who present with a higher ferritin count and fail to fall.

Organ support and critical care intensity is different in the two groups -survivors rarely required renal replacement therapy prior to the nadir of their PF ratio but non-survivors presented earlier and required greater organ support. The timing of renal replacement therapy is statistically significant; non-survivors required renal support at mean day -8 compared to survivors on mean day +1 (p < 0.05) ARDSnet defined severe ARDS as a PF ratio < 13.3 kPa (100mmHg) with the PROSEVA trial's cut-off for proning being a PF ratio < 20 kPa (150mmHg) 9,10 . Taking these cut-offs into consideration, the majority of the survivors presented in this case series do not improve beyond severe ARDS criteria until day 20 -25 after symptom onset and do not improve beyond PROSEVA's proning cutoff until much later. In both groups, the amount of organ support required as their disease progresses increases and peaks at the day coinciding with the lowest PF ratio. In survivors this level of support is maintained and stabilises until approximately day 10 post lowest PF ratio. Interestingly, vasopressor usage (as defined as use of any noradrenaline, vasopressin or metaraminol) declines but renal replacement therapy increases post PF ratio trough. Non-survivors require intensive organ support early on in their admission to ICU. All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) 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 May 20, 2020. . https://doi.org/10.1101/2020.05.16.20103853 doi: medRxiv preprint 9 Based on this case series, there is little to differentiate survivors from non-survivors based on respiratory function and lymphocyte count until either recovery or death. Critical care needs, including vasopressors, proning and renal replacement therapy can extend beyond 10 -15 days after lowest PF ratio in survivors. This is in stark opposition to previous case series who fail to demonstrate trajectories of patients due to non-alignment of disease progression which is the novelty of this case-series.

Limitations of this case series include a small sample size in both survivors and nonsurvivors with selection bias occurring in the survivor cohort. All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. In this series, looking at multiple factors there was little to differentiate survivors from nonsurvivors upon which to base early prognostication.

Even when faced with multiorgan failure, perseverance until discharge must be advocated as late improvements do occur in survivors.

None applicable

None declared by all authors All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) 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 May 20, 2020. . https://doi.org/10.1101/2020.05.16.20103853 doi: medRxiv preprint 1 3 (which was not certified by peer review) 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 May 20, 2020. . https://doi.org/10.1101/2020.05.16.20103853 doi: medRxiv preprint

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

Presenting Characteristics, Comorbidities, and Outcomes Among 5700 Patients Hospitalized With COVID-19 in the New York City Area

Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China. The Lancet

Characteristics of and Important Lessons From the Coronavirus Disease 2019 (COVID-19) Outbreak in China: Summary of a Report of 72 314 Cases From the Chinese Center for Disease Control and Prevention

Characteristics and Outcomes of 21 Critically Ill Patients With COVID-19 in Washington State

ICNARC report on COVID-19 in critical care 08

All rights reserved. No reuse allowed without permission.(which was not certified by peer review) 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 May 20, 2020. All rights reserved. No reuse allowed without permission.(which was not certified by peer review) 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 May 20, 2020. . https://doi.org/10.1101/2020.05.16.20103853 doi: medRxiv preprint All rights reserved. No reuse allowed without permission.(which was not certified by peer review) 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 May 20, 2020. . https://doi.org/10.1101/2020.05.16.20103853 doi: medRxiv preprint All rights reserved. No reuse allowed without permission.(which was not certified by peer review) 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 May 20, 2020. . https://doi.org/10.1101/2020.05.16.20103853 doi: medRxiv preprint All rights reserved. No reuse allowed without permission.(which was not certified by peer review) 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 May 20, 2020. . https://doi.org/10.1101/2020.05.16.20103853 doi: medRxiv preprint