A final word from the Editor in Chief

I have reached the end of my tenure as Editor in Chief of the Journal of the Intensive Care Society (JICS) and am preparing to hand over to my successor. So much has happened during the past 7.5 years at the helm that I’d like to reflect on this time before looking ahead to the future. Being Editor in Chief of JICS has been a thoroughly enjoyable rollercoaster ride, and one I am glad to have undertaken. With relatively little editorial experience, the there was a steep learning curve at the beginning, aided by a considerable amount of behind the scenes assistance from my predecessor. For anyone thinking of undertaking a position in a scientific journal, I can highly recommend it. It provides a way to shape and improve the way research and education is disseminated and an opportunity work with some incredible people.

In 2018 when I was appointed as Editor in Chief, I hadn’t put much thought into what might lie ahead of me during my time in the role. With only a few years’ experience as an Associate Editor under my belt, I wondered whether I’d have what it takes to run a journal. Looking back on those early days, I thought the greatest challenge ahead of me was whether we could bring our society journal in line with the somewhat controversial Plan S. ¹ However, events began to unfold around the world that completely overshadowed this. A pandemic, new wars, our planet over-heating, a global fuel crisis, the increasing cost of living, unimaginable changes on the political landscape, and the escalating influence of artificial intelligence (AI). As all good stories should consist of three parts, I’d like to delve into a trio of those events in the immortal framework of the good, the bad and the ugly.

Firstly, the good. Climate change is obviously not good. It is real, it is out of control, and it is likely to be the undoing of our species. Whilst turning this ship around may be impossible, we at least need to do whatever we can to slow it down. At the journal, our small but significant contribution to this during my tenure (our good) has been to transition our quarterly publications from paper (wrapped in plastic) to an electronic format. This change came with a predictable push back from society members. They may well have valued carrying a copy of the journal in their bag as they travelled to work, however, the reality was more likely to have been stacks of unopened packages accumulating behind letterboxes around the country. According to my new best friend (but maybe not forever), Chat GPT (which we will come to shortly), producing and posting a single copy of our journal within the UK probably resulted in the production of 0.6–0.7 kg of carbon dioxide. If we say there are 3000 members of the Intensive Care Society, with four editions per year, that was roughly 7200-8400 kg of carbon dioxide a year. Although a miniscule proportion of total human carbon dioxide production, we must all do our bit to bring this situation under control. Additionally, like most of the rest of the world after the COVID-19 pandemic (something else we will come to shortly), meetings were switched to online, further reducing our carbon footprint. Why we all thought it was necessary to regularly trek across the country to sit round a table and talk is anyone’s guess.

The bad. Whilst we learnt a great deal about human pathophysiology, the limits of the healthcare systems we work in and how as a team we can tackle the greatest of challenges, there is no part of COVID-19 that is good. Over 7 million people died as a result of the virus and the overall number of excess deaths is probably double that. ² This is neither the time nor the place to think back to those dark days of 2020–21 but it is worth considering how the pandemic affected medical publishing. Estimates vary but somewhere in the region of half a million scientific articles were published on the topic of COVID-19 in a very short space of time. ³ My best friend says that if you stacked all these articles on top of each other, the pile would be roughly the same height as the tallest building on Earth. This makes COVID-19 one of the most studied diseases ever. How much of the information contained within these manuscripts was useful or even accurate is unclear. The number of submissions to this journal rose steadily through late 2020 and early 2021. We would usually welcome an increase in submissions, however, as an intensive care medicine journal, virtually everyone on the editorial board became involved in the front-line care of critically ill patients. This meant they (and our usual pool of external peer reviewers) were largely unavailable for the evaluation of manuscripts. Tricky times for us. For months, in fact years, we were faced with trying to determine the value of small, single centre, retrospective studies of critically ill patients with COIVID-19. Whilst well intentioned, very few, were likely to have had any impact on patient outcomes. The way in which research was disseminated also changed during the pandemic, with pre-printing becoming the norm. In the early days of COVID-19, non-peer reviewed preprints dominated the landscape, with outlets such as MedRxiv and BioRxiv hosting nearly 40% of all English language scientific articles related to COVID-19. ⁴ Once the vaccination programme rolled out and the intensive care units slowly returned to business as usual, the submissions continued to pour in. Now that surge of activity has settled, we are experiencing far fewer submissions to the journal than before the pandemic. This could be due to a multitude of reasons: ‘regular’ research was put on hold during the pandemic and some of it was abandoned afterwards due to lack of sufficient funding; research active clinicians may have chosen not to return to academia; and perhaps we all had a wakeup call about our work-life balance. Research is also becoming a very expensive activity, with universities driving up overheads, which may be preventing progress and reducing outputs.

And finally, the ugly. As a journal editor, nothing during my tenure has frightened me more than the almost overnight acceptance of AI into our lives. Much like the introduction of the internet at the turn of the millennium, AI has infiltrated our lives in such a way that it is already hard to remember life before it. I don’t think you need an ageing and increasingly tech non-savvy person like me to list what AI can help you with. I am what I would consider to be a ‘casual’ user of AI and have no doubt that I am only scratching the surface of what can be achieved with these systems if you really put your mind to it. Online large language model (LLM) AI systems like Chat GPT have devoured colossal amounts of information that have been fed to them and are learning at an incredible rate. I appreciate that what AI produces for us is not always 100% accurate (so called AI hallucinations), but who amongst you never makes a mistake? Surely inaccuracy is a trait we all display from time to time, even if we don’t want to admit it. Is everything we read in a manuscript or textbook true? How ever you look at it, AI is an incredibly powerful tool, that used correctly can save enormous amounts of time by gathering and summarising information. The ugly part comes when one starts to experiment with Chat GPT and ask it to write a report or perhaps an article for you. If you push a little further it will design grant applications, trial protocols, and entire datasets for make-believe trials along with a neat manuscript full of figures to accompany it. According to one study, you can detect the ‘fingerprints’ of LLMs in 13.5% of abstracts on PubMed. ⁵ How accurate this prediction is remains unclear, because the technique employed to detect LLM influence is a subtle change in language style within articles. Regardless, the use of LLMs to aid writing is here and as editors it is difficult to know what to do about it. Publishers produce guidance on the topic ⁶ but as an editor or reviewer faced with a manuscript that doesn’t feel right, when assurance was provided by the authors that AI was not used to create it, what does one do? The guidance advises that the Editor in Chief should be informed. Yikes, that’s me! What training have I had on this? None. The internet did not exist when I went to school. Yes readers, can you imagine a world without social media? I promise you it was truly wonderful. But in all seriousness, I am not trained to detect subtle or even overt AI use in a manuscript, or how to confront author about this. At the beginning of my tenure, we dealt with some research fraud, but that was, by comparison, very easy to manage. Extraordinarily, we witnessed authors submitting carbon copy replicas of already published work, just with a different set of authors. I suspect that type of fraud is now vanishingly rare because LLMs can just invent a study, its results and the accompanying manuscript in a matter of moments. It is truly terrifying.

As I draw this final editorial article as Editor in Chief to a close, I feel it is important to think ahead and to consider what the next 5–10 years might look like for JICS. I hope that my successors continue to build on everything we have achieved to date and guide the journal on an ascent of the league tables, to establish JICS as the go to place to publish intensive care research not only in the UK, but around the world. At the same time, they shouldn’t lose sight of the fact that this is a society journal that needs to serve the society’s members, readers and authors alike. No doubt there will be unexpected encounters along the way, but the bigger and more resilient the journal becomes, the more likely it is to weather future storms.

I would like to thank several groups of people, without whom the journal couldn’t function, and have made my job possible for the last few years. The editorial board for all their expertise, enthusiasm, support, advice and hard work; our network of peer reviewers who provide such shrewd feedback on manuscripts; the team at the Intensive Care Society for supporting the journal on its bumpy journey; the team at Sage Journals for all their guidance; the authors who have submitted articles; yourselves, the readers; and my family, most of whom weren’t even born when I embarked upon this editorial adventure. My 4 year old son recently said to me, ‘Daddy, do you know that the heart pumps blood and blood contains oxygen’. I feel my work on this planet is now complete.