In terms of innovation at the frontline, one of our great local achievements was the deployment of a new electronic health record solution in just a matter of days, amidst the first surge of patients and three months before the expected roll-out date. Doing this was critical from an infection control point of view, but also to facilitate remote reviewing and prescription for our patients.
I also admired the dramatic increase and effectiveness of how we are all now communicating. I have never interacted this much with colleagues and the teams, inside and outside my hospital. A huge proliferation of WhatsApp groups have helped in a myriad of ways such as coordinating the transfer of patients and equipment within our local network, teaching non-specialists to use anaesthetic machines, or keeping our local group of clinicians up-todate with literature and protocols. On a wider scale, open and fast communication on an international level e.g. via publications and webinars meant that we were able to learn from the experience of other countries and prepare before the first surge hit us. This newly acquired community spirit is one of the many positive things to come out of this pandemic and I sincerely hope that it survives way beyond this crisis.
I also must acknowledge the great work of scientists and clinicians around the world who have continued to generate research output while dealing with their day-to-day clinical responsibilities. The evidence generated has undoubtedly saved many lives by, for example, refining COVID-19 treatment guidelines with regards to indications and modes of mechanical ventilation, anticoagulation, antiinfective agents or steroids.
Following the surge in patients, we have seen an accompanying proliferation of literature about physiological hypotheses and interventions, in an attempt to help us understand, predict and manage the COVID-19 disease. For example, the Medrxiv and Biorxiv repositories list no less than 2,500 articles deposited in the first 10 weeks of the epidemic. The issue of isolating the signal from the noise has been acute, and the temptation to try a vast array of interventions with little or no supportive evidence has been high. As stated by Matthew Siuba and colleagues in their paper, ‘Treading Lightly In a Pandemic’: “Facing an unprecedented pandemic, intensive care clinicians feel obligated to intervene in order to prevent, modulate, or reverse COVID-19 disease and avert death. Human nature is to err in favor of intervention rather than thoughtful inaction, commission over omission.”
However, generating robust clinical evidence in the context of an acute epidemic is extremely challenging. The time it takes to set up a new project, include patients (in overwhelmed hospitals), collect and analyse data and report the results in a peer-reviewed publication to inform practice is nowhere near compatible with the timeframe of an epidemic, certainly not the first wave.
So what are we left with? Most of the current evidence comes from (usually small) retrospective analyses and case series. From a data analysis standpoint, we must highlight the role of causal inference methods, to try and disentangle true effect and confounding. But any secondary analysis of medical data relies on the availability of large datasets, which is another challenge. Among the many projects that have been launched, the large real-time open access patient level dataset released by the nCoV-2019 Data Working Group must be complimented and has yielded many interesting findings already [see: www.nature.com/articles/s41597-020-0448-0].
Together with a few academic clinicians, we have set a national registry to collect patient data twice daily, as part of a National Service Evaluation [see: www.imperial.ac.uk/artificialintelligence/research/covid/covid-icu/]. Analyses performed on these kinds of datasets have heavily influenced clinical practice, for example, by identifying risk factors for poor outcomes (e.g. Wu JAMA Int Med 2020 https://jamanetwork.com/journals/jamainternalmedicine/fullarticle/2763184), highlighting a high prevalence of thromboembolic events in COVID-19 patients (e.g. Klok FA et al. Incidence of Thrombotic Complications in Critically Ill ICU Patients with COVID-19. Thrombosis Research) or confirming the relationship between heparin treatment and survival (e.g. Tang N et al. Anticoagulant Treatment is Associated with Decreased Mortality in Severe Coronavirus Disease 2019 Patients with Coagulopathy. J Thromb Haemost 27 March 2020).
It should be pointed out that the impact of AI and advanced technology on the outcome of patients and the crisis in general has been rather limited. For example, no AI is available today to guide us about which patient will most benefit from being admitted to intensive care or being put on mechanical ventilation. But what has made the most impact is simple communication tools, large scale national policies, the dedication of key workers, and the discipline of each and every citizen, enforced by the law when needed.
Which brings me to a topic close to my heart, and the most important issue of our time: the climate crisis and the collapse of ecosystems. All the points I’ve made above are applicable to the climate change, except that climate change is a much bigger threat than this pandemic, so our response must be even greater. And here again, salvation won’t come from some fancy technology like carbon capture, fusion reaction or geo-engineering. Rather, it will be vigorous policies enforced by law, and the discipline and sacrifice of individuals to accept a temporary decline in their living conditions and right to consume for the greater good.
Matthieu Komorowski, MD PhD is a Consultant in Intensive Care at Charing Cross Hospital, London and Clinical Senior Lecturer at Imperial College London