A third of covid-19 patients admitted to UK hospitals die
BMJ 2020; 369 doi: https://doi.org/10.1136/bmj.m1794 (Published 30 April 2020) Cite this as: BMJ 2020;369:m1794Read our latest coverage of the coronavirus pandemic
All rapid responses
Rapid responses are electronic comments to the editor. They enable our users to debate issues raised in articles published on bmj.com. A rapid response is first posted online. If you need the URL (web address) of an individual response, simply click on the response headline and copy the URL from the browser window. A proportion of responses will, after editing, be published online and in the print journal as letters, which are indexed in PubMed. Rapid responses are not indexed in PubMed and they are not journal articles. The BMJ reserves the right to remove responses which are being wilfully misrepresented as published articles or when it is brought to our attention that a response spreads misinformation.
From March 2022, the word limit for rapid responses will be 600 words not including references and author details. We will no longer post responses that exceed this limit.
The word limit for letters selected from posted responses remains 300 words.
Dear Editor
There is a common underlying circumstance among people with an increased risk of severe COVID-19: pre-existing endothelial dysfunction. Hypertension, diabetes, obesity, cardiovascular disease, cancer, chronic kidney disease, chronic liver disease, Alzheimer’s disease, aging, male sex, and black race increase the risk for a severe SARS-CoV-2 infection. Each of these conditions can be associated with endothelial dysfunction (1). Of note is the fact that vascular endothelial dysfunction is regarded as a primary phenotypic expression of normal human aging (2).
Endothelial cells have very separate and singular functions that are predominant in vascular biology. Blood vessel tone, fluid filtration, hemostasis, neutrophil recruitment, hormone trafficking, and other functions are all under control of endothelial cells. When endothelial dysfunction is present, the ability to perform one or more of these functions results in reduced vasodilation, a proinflammatory state, and prothrombic properties (1).
All endothelial cell function can be greatly altered by virus infection (3). Cytokine storm during viral infection is a prospective predictor of morbidity and mortality and endothelial cells, as central regulators of cytokine storm, have been identified (4).
Elevated levels of inflammation-inducing cytokines in the blood of hospitalized COVID-19 patients have been shown, and disease severity is probably driven by this out of proportion inflammatory response to the virus (5). There is evidence of direct viral infection of the endothelial cell and diffuse endothelial inflammation (6), secondary dysregulated immune response resulting in excessive activation of the endothelium by enhancing antibodies and memory T cells could result in a cytokine “tsunami” (7). Endothelium appears to be the battlefield of COVID-19.
Blood clot forming is only one of several complications of endothelial dysfunction. It is recognized that clotting is a major feature of severe COVID-19 and subtle clotting might begin early in the lungs, perhaps due to an inflammatory reaction in their fine web of blood vessels, which could set off a cascade of proteins that prompts blood to clot and prevents it from getting properly oxygenated (8). However, a report on 3200 COVID-19 positive deceased patients in Italy reported atrial fibrillation as one of the most common comorbidities (9); consequently, many patients at the beginning of COVID-19 were using oral anticoagulant and nevertheless they died. Prior anticoagulation was not protective unless the cause of death was independent by their coagulation state in all deceased patients. Clotting is one important severity risk factor in COVID-19, but systemic inflammatory vasculitis secondary to endotheliitis appears to be the most feared complication in SARS-CoV-2 infection. This explains why anticoagulant therapy even started at the beginning of the disease does not always result in a striking effect on mortality.
We speculate that the degree of pre-existing endothelial dysfunction could be a condition sine qua non for developing moderate-severe SARS-CoV-2 infection. During the COVID-19 pandemic, the deaths of so many older adults could be caused by a common underlying circumstance: the progressive decline of endothelial function with age (2) further complicated by other coexisting comorbidities associated with endothelial dysfunction such as hypertension, cardiovascular diseases, diabetes or cancer.
An effective endothelial dysfunction treatment might change the disease course.
References
1) P Rajendran, T Rengarajan, J Thangavel, Y Nishigaki, D Sakthisekaran, G Sethi, and I Nishigaki. The Vascular Endothelium and Human Diseases. International Journal of Biological Sciences. 2013; 9(10):1057-1069. doi: 10.7150/ijbs.7502.
2) EM D Herrera, C Mingorance, R Rodríguez-Rodríguez, MA de Sotomayor. Endothelial dysfunction and aging: An update. Ageing Research Reviews. 2010.V.9;Iss.2:142-152. DOI: org/10.1016/j.arr.2009.07.002.
3) ER Mackow, IN Gavrilovskaya. Hantavirus regulation of endothelial cell functions. Thromb Haemost. 2009 Dec; 102(6):1030-41. DOI: 10.1160/TH09-09-0640.
4) M Wadman, J Couzin-Frankel, J Kaiser, C Matacic. How does coronavirus kill? Clinicians trace a ferocious rampage through the body, from the brain to toes. Science. Biology Coronavirus. Apr. 17, 2020 DOI:10.1126/science.abc3208.
5) J R Teijaro, KB Walsh, S Cahalan, DM Fremgen, E Roberts, et al. Endothelial Cells Are Central Orchestrators of Cytokine Amplification during Influenza Virus Infection. Cell. 2011;Vol.146,ISS.6,p.980-91. DOI: 10.1016/j.cell.2011.08.015.
6) Z. Varga, A J Flammer, P Steiger, M Haberecker, R Andermatt, A S Zinkernagel, et al. Endothelial cell infection and endotheliitis in COVID-19. The Lancet, Vol. 395, No. 10234, p1417–1418. Published: April 20, 2020. DOI: 10.1016/S0140-6736(20)30937-5.
7) P Rajendran, T Rengarajan, J Thangavel, Y Nishigaki, D Sakthisekaran, et al. The vascular endothelium and human diseases. Int J Biol Sci. 2013 Nov 9;9(10):1057-69. DOI: 10.7150/ijbs.7502.
8) M Wadman, J Couzin-Frankel, J Kaiser, C Matacic. A rampage through the body. Science 24 Apr 2020:Vol.368, Issue 6489, pp. 356-360. DOI: 10.1126/science.368.6489.356.
9) Istituto Superiore di Sanità. Characteristics of COVID-19 patients dying in Italy. Report based on available data on March 20th, 2020. https://www.epicentro.iss.it/coronavirus/bollettino/Report-COVID-2019_20....
Competing interests: No competing interests
Dear Editor
This research study, although awaiting full publication, provides covid-19 UK hospital admission data analysis up until 18 April 2020. The study outcome of 16,749 patients shows about half (49%) of these patients discharged and almost half (47%) had no co-morbidity. It is not clear what percentage in the first half discharged and the second half remaining admitted, had comorbidities respectively.
However, interestingly the co-morbidity data shows, chronic cardiac disease was found in almost a third (29%) of patients and only about 20% had non-cardiac co-morbidity, given that almost half the originally admitted patients had no co-morbidity (47%). It is not clear out of 33% deaths in hospital, how many had co-morbidity, either cardiac or non-cardiac?
Although 33% death figure is alarming, there are many issues could be looked at in further detail. Also the Covid-19 case incidence and death figures are fast changing on a daily basis so as the hospital admission outcomes and accordingly from 18 April to 5 May 2020 there had been a further vast change in data and counting those might generate different figures now and in future. But as this study already says, this is a rapid pre-print publication, I assume the data will be further updated accordingly.
If we take account of the world statistics, covid-19 incidence and death rates are hugely different in different parts of the world. For example, in the USA case incidence and deaths were 1 133 069 and 66 385 respectively, whereas in India 39 980 and 1 301 respectively, and in the UK 182 260 and 28 131 respectively (ECDC data 4 May 2020). Obviously these death figures are not about hospital admitted patients nor fully reliable, so it is difficult to compare those death rates with the death rate as quoted in this pre-print research study- https://www.medrxiv.org/content/10.1101/2020.04.23.20076042v1
It is therefore not possible or realistic for me to provide a whole statistical analysis of the world data, in this rapid response, so I briefly presented a snapshot data comparison here.
This is a new horizon for medical science and everyday a new information is emerging about the impact of covid-19 and so the real outcome is yet to be seen, given the reliability, consistency, accuracy and completeness of any current data are subject to change and variation. It is therefore perhaps early to predict how Covid-19 outcome will actually compare with the outcome of other previously seen pandemics, which only time will tell.
Competing interests: No competing interests
Dear Editor
I have very grave concerns about this survey that makes no mention of Hypertension as a comorbidity being made. It would seem likely unfortunately that the high percentage 47% of no comorbidity record may be a reflection of poor history recording (see BMJ blog covid19 and hypertension: risks and management). It seems a great pity that this wasn't looked at more closely.
One way of addressing this lack of information may have been to record all the medications patients were on when admitted and look at this to identify a link. If the team still has access to the information perhaps they could do this.
It also might be possible to see from this why there is a difference in mortality between the men and women. It is well known that women have received different medications to men after having been admitted to hospital for cardiac events.
Competing interests: No competing interests
Creating Misinformation: how a BMJ headline about covid-19 spread virally
The spread of the novel coronavirus has been accompanied by an equally viral spread of misinformation. In these unprecedented times, with a third of the world population currently in lockdown, there is an insatiable hunger for information. Because there is significant scientific uncertainty about the biology and epidemiology of the coronavirus, every piece of emerging evidence is likely to attract widespread attention, especially if it is published in a credible source. Translating and summarising research findings for the lay public should therefore be done with extra care. Failing to provide proper context could lead to misinformation, whereby factually correct numbers are interpreted in the wrong way. This BMJ news piece serves as an interesting example.
The BMJ posted the title of this news piece on Twitter and added ‘putting the fatality rate on par with Ebola’. This comparison came from a quote by the senior author on the preprint. The tweet quickly gained traction and was retweeted more than 350 times within 48 hours – much more often than the average tweet of BMJ. Just the word ‘Ebola’ is enough to catch public attention. The haemorrhagic fever has a grim and bloody reputation, due to recent devastating outbreaks in Africa and apocalyptic Hollywood depictions. Not surprisingly, news media quickly picked up BMJ’s news post with headlines ranging from ‘Scientists say virus as deadly as Ebola for hospital victims’ in the Times [1] to ‘Experts warn coronavirus ‘as dangerous as Ebola’ in shocking new study’ in Express [2]. But is this really the case?
There are two statements in the news piece that shocked and served as attention-magnets: 33% of hospitalised covid-19 patients died and the claim that this statistic makes covid-19 as deadly as Ebola. Both claims deserve further investigation. The findings reported in the preprint were based on an analysis of the medical records of 16,749 people with covid-19 across 166 UK hospitals between the 6th of February and the 18th of April. The median hospital stay of those patients was 7 days, with an interquartile range of 4-12 days. To be able to calculate how many covid-19 patients have died in those 166 hospitals, the authors limited the sample to 6628 patients – only those who were admitted latest on April 4th, 14 days before the actual end date of the data collection of the study. The authors note that the 166 hospitals accounted for 28% of all admitted covid-19 patients in the UK. Of the 6628 covid-19 patients for which they investigated the outcome (death, discharged alive or still in hospital), they found that 2212 patients had died – a whopping 33%. A simple extrapolation to the rest of the clinical system would suggest that around 8848 cases would have died in hospitals all over the country. However, as of April 4th, 4461 deaths were reported in the UK by official sources [3]. Of course, these numbers come with considerable uncertainty, but the gap cannot be explained by imprecision alone. So what does this mean? It can be that the 166 UK hospitals are more specialized hospitals, treating the most severe covid-19 patients, who are at higher risk of dying. Unfortunately, no information about the characteristics of these hospitals is given in the preprint, and there is no explanation or discussion of why these hospitals account for roughly half the covid-19 deaths in the UK at that moment in time.
Then the claim that the hospital fatality rate is on par with Ebola’s hospital fatality rate. Let’s zoom out a bit. The case fatality rate – the ratio between confirmed deaths and confirmed cases – is currently around 7% for covid-19 on a worldwide average. Keep in mind that his is about confirmed cases; given the lack of testing capacity in almost every country in the world, the number of cases is much higher, meaning that the true infection fatality rate is much lower. Germany, South Korea and Iceland have been able to implement widespread testing, their apparent CFRs are 4.3%, 2.4% and 0.6% respectively [4,5]. For Ebola, the CFR is on average 50% - every second confirmed Ebola patient is likely to die from it [6]. This is an enormous difference with covid-19. But this study did not investigate the CFR, it looked at hospital fatality rates that are estimated for the subset of infected people who are sick enough to need hospital care. For covid-19, that is a small share of all confirmed cases. And all confirmed cases to date are an unknown share of all cases. So we are talking about the death rate among a small, very selected group of very sick patients. For Ebola, almost all patients become very sick and are in need of hospital care: the CFR for Ebola is about the same as the hospital fatality rate. Putting Ebola and covid-19 in the same sentence is comparing a small fraction of one population with the majority of another population. It is like saying that one rainy day in otherwise sunny Florida means the weather is same as in rainy UK. In other words, it is a comparison that should not be made.
The important distinction between case fatality rate and hospital fatality rate is quickly lost in the never-ending news cycle. Combining Ebola and covid-19 in a single sentence is an effective way to get public attention, but is not an effective strategy to communicate science, or clarify risk. Omitting important information on how to understand and interpret the high hospital fatality rate in 166 UK hospitals could lead people to inaccurate conclusions, and to think that they have a high risk of dying when they are admitted with covid-19. The World Health Organization called the pandemic (mis)information overload an ‘infodemic’, whereby misinformation is circulating at high speed [7]. Misinformation does not always come from bad intentions, it is often a technically correct fact that is taken out of context, creating a narrative on its own [8]. No matter the intention, misinformation can have serious consequences by influencing people’s knowledge and practices around a disease. Every piece of emerging evidence around covid-19 is likely to attract attention and thus deserves to be reported responsibly – especially if it is published by a credible source such as a leading medical journal. This means thinking like a chess master: a few steps ahead to how a headline might be perceived by journalists and lay public.
1. Blakely R. Scientists say virus as deadly as ebola for hospital victims. The Times [Internet]. 2020; Available from: https://www.thetimes.co.uk/article/covid-19-death-rates-are-comparable-t...
2. Kaidan M. Experts warn coronavirus is ‘as dangerous as Ebola’ in shocking new study. Express [Internet]. 2020 Apr 30; Available from: https://www.express.co.uk/life-style/health/1275700/ebola-elderly-patien...
3. Our World In Data. Coronavirus (COVID-19) Deaths [Internet]. 2020 [cited 2020 May 7]. Available from: https://ourworldindata.org/covid-deaths?country=GBR
4. Our World In Data. Case fatality rate of the ongoing COVID-19 pandemic [Internet]. 2020 [cited 2020 May 7]. Available from: https://ourworldindata.org/grapher/coronavirus-cfr
5. Lipsitch M. Estimating case fatality rates of COVID-19. Lancet Infect Dis [Internet]. 2020;3099(20):30245. Available from: http://dx.doi.org/10.1016/S1473-3099(20)30245-0
6. World Health Organization. Ebola virus disease - Key Facts [Internet]. 2020 [cited 2020 May 7]. Available from: https://www.who.int/news-room/fact-sheets/detail/ebola-virus-disease
7. Zarocostas J. How to fight an infodemic. Lancet [Internet]. 2020;395(10225):676. Available from: http://dx.doi.org/10.1016/S0140-6736(20)30461-X
8. Southwell BG, Niederdeppe J, Cappella JN, Gaysynsky A, Kelley DA, Oh A, et al. Misinformation as a Misunderstood Challenge to Public Health. Am J Prev Med. 2019;57(2):282–5.
Competing interests: No competing interests