In BMJ Open, Arlotto and colleagues reported the results of a retrospective single center study in Marseille, about the life-years lost by COVID-19 patients during the first pandemic wave from 1 March 2020 to 15 June 2020 (1). The authors concluded that a mortality due to COVID-19 lead to a limited number of years of life lost (YLL) due to both old age and preexisting comorbidities in the most vulnerable patients. As written in the conclusion, theses results should be interpreted “with caution”.
Crucially, we report several methodological bias:
-the Charlson Comorbidity Index (CCI) is used to assess what would have been the probability of death within 1 year of these patients in the absence of COVID-19. The 1-year mortality rates used are from the original study based on an inception cohort study of 604 patients admitted to the medical service at New York Hospital during 1 month in 1984. The comorbidity index was then tested for its ability to predict risk of death from comorbid disease in a cohort of 685 patients who were treated for primary breast cancer at Yale New Haven Hospital between 1962 and 1969 (2).

-Several studies have shown that the weighting of comorbidities is no longer appropriate. Moreover, medical progress since 1984 has not been taken into account at all, as the authors point out in the discussion (3).
A recent and exhaustive study has been conducted based on 6,602,641 subjects hospitalized in 2010, from medical, surgical, or obstetrical department, registered in the French National Health Insurance general scheme. One-year survival was 94.88%, decreasing from 98.41% for CCI of 0 to 71.64% for CCI of ≥5, far from the 15% one-year survival rate, used in Arlotto’s study for CCI of ≥5 (4).
-Moreover, the authors used mortality data from hospitalized patients, whereas without COVID-19, the patients would not have been.

The choice of the CCI can be discussed because other prognostic factors seem to be more relevant in a predominantly geriatric population. In a large population of patients admitted to hospital with COVID-19, disease outcomes were better predicted by frailty than either age or comorbidity (5).

In our opinion, methodological bias have flawed the study result, which differs from other studies on COVID-related YLL. Indeed, studies based on excess mortality rates seem to us to be of better methodological quality to measure the burden of disease. They show a number of YLL per patient ranging from 8 to 16, comparing to 3 in Marseille (6–9).
Given the limitations of this study, this figure does not allow to deduce that COVID 19 is responsible for a limited death toll, and cannot be interpreted as a sign of better care management in Marseille, as mentioned by the authors.

References
1. Arlotto S, Garès A, Giraud-Gatineau A, Lagier JC, Jimeno M-T, Peretti-Watel P, et al. Life-years lost by COVID-19 patients in public hospitals of Marseille (APHM-South-Eastern France): a limited death toll: a retrospective analysis. BMJ Open. 2021 Dec 1;11(12):e049475.
2. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. Journal of Chronic Diseases. 1987 Jan 1;40(5):373–83.
3. Choi JS, Kim M-H, Kim YC, Lim Y-H, Bae HJ, Kim DK, et al. Recalibration and validation of the Charlson Comorbidity Index in an Asian population: the National Health Insurance Service-National Sample Cohort study. Sci Rep. 2020 Aug 13;10(1):13715.
4. Bannay A, Chaignot C, Blotière P-O, Basson M, Weill A, Ricordeau P, et al. The Best Use of the Charlson Comorbidity Index With Electronic Health Care Database to Predict Mortality. Medical Care. 2016;54(2):188–94.
5. Hewitt J, Carter B, Vilches-Moraga A, Quinn TJ, Braude P, Verduri A, et al. The effect of frailty on survival in patients with COVID-19 (COPE): a multicentre, European, observational cohort study. The Lancet Public Health. 2020 Aug 1;5(8):e444–51.
6. Islam N, Jdanov DA, Shkolnikov VM, Khunti K, Kawachi I, White M, et al. Effects of covid-19 pandemic on life expectancy and premature mortality in 2020: time series analysis in 37 countries. BMJ. 2021 Nov 3;375:e066768.
7. Kontopantelis E, Mamas MA, Webb RT, Castro A, Rutter MK, Gale CP, et al. Excess years of life lost to COVID-19 and other causes of death by sex, neighbourhood deprivation, and region in England and Wales during 2020: A registry-based study. PLOS Medicine. 2022 Feb 15;19(2):e1003904.
8. Pifarré i Arolas H, Acosta E, López-Casasnovas G, Lo A, Nicodemo C, Riffe T, et al. Years of life lost to COVID-19 in 81 countries. Sci Rep. 2021 Feb 18;11(1):3504.
9. Ugarte MP, Achilleos S, Quattrocchi A, Gabel J, Kolokotroni O, Constantinou C, et al. Premature mortality attributable to COVID-19: potential years of life lost in 17 countries around the world, January-August 2020. BMC Public Health. 2022 Jan 9;22(1):54.

I’m prefacing this with the fact that I’m not an epidemiologist, virologist, medical scientist in any way, or healthcare worker. I’m therefore VERY hesitant to pop up and appear to be criticizing those who have spent decades in scientific research.

I am, however, a Councillor in Local Government in the UK with some minimal scientific training (two degrees – physics and military electronic systems engineering), and was therefore a bit involved in the local response to covid (For example, the seventh suggestion by the GBD for focused protection was to offer home delivery to vulnerable people; I assisted in setting up and funding local networks to do just that from March 2020, as did councillors across the UK), so the subject under discussion had immediate impact upon us. Should there have been acceptance of the GBD, I’d have been one of those involved in trying to carry it out (albeit only in a very minor way in comparison to many others). I'm also a qualified and experienced project manager who has had to translate strategies into practice many times.

It might therefore be of some use to give feedback from that side of things - the core of the question seemed to be "Was it Twitter that stopped implementation of the GBD rather than practical considerations." I can say that my own views of the GBD were not impacted by any Twitter publicizing (I only started reading Twitter some months ago), but because, when reading it with the view of “how would we do this,” it was rapidly apparent that it wasn’t a strategy but a slogan, and one that could not be carried out and would not be successful even if it could have been.

Our big issue was the pressure on healthcare systems, especially in ICU. By the time the GBD was published, it was clear that the majority of people in ICU were under the age of 60. A quarter were under 50. Only a small fraction had serious health issues.

Accordingly, “Focused protection” (which, reading through, consisted of 17 initiatives that were already routine, one that could be useful (free N95 masks), and one that was barking mad (split “vulnerable” people away from families and house them in hotel rooms for however-many-months-or-years)) would need to somehow isolate around 50% of the country (the JCVI Categories 1-9 (available here: https://www.gov.uk/government/publications/priority-groups-for-coronavir... as priorities for vaccination) would make this very clear to everyone, published in early January 2021). Of the 50% left, more than half would be children.

We would need to run the country (of 67 million people) with about 15 million adults. Including the health system, which would still certainly collapse with the number of people (of the “less vulnerable”) who would need treatment, unless spread out (somehow, without NPIs) over 2 years or more.

Notwithstanding that protection was voluntary, and even a small fraction of the more vulnerable would further swamp the system.

And as reinfection was already known, and the R0 of even the original strain was high enough that herd immunity from infecting the “less vulnerable” was impossible with the numbers involved, even if it had somehow been made to work, it wouldn’t have succeeded.

In short, the problems with the GBD were that minimal thought and analysis were needed to indicate that it was a slogan rather than a strategy, that the suggestions made were either routine or implausible (well covered by Gideon Meyerowitz-Katz here: https://gidmk.medium.com/focused-protection-from-the-great-barrington-de... ), that the numbers needed to protect were far beyond what was implied, that the healthcare system would have collapsed, that it would have needed to be applied for years rather than months, and at the end of it, there would be no herd immunity, anyway.

That was the key reason it wasn’t pursued; not Twitter publicity.

We write as a group of scientists who are concerned about ethical and methodological flaws that we have found in Professor John Ioannidis’ paper [1]. These include factual errors, statistical shortcomings, failure to protect the named research subjects from harm, and potentially undeclared conflicts of interest that entirely undermine the analysis presented. We have previously published our concerns about this study in 8 Rapid Responses authored by each of us as individuals [2-9], but since Prof. Ioannidis has declined to address them substantively, we are now writing collectively in the hope that he might consider our joint response more seriously.

We undertake this effort because, unfortunately, the author’s Rapid Responses to our concerns and criticisms have devolved into what are inarguably personal attacks on some of us. Thus, we wish to state the issues with the study simply and directly and, once again, ask Prof. Ioannidis to address them, but this time in a more objective fashion, free of deflection and personal hostility. There is no doubt that Professor Ioannidis has been one of the foremost scientists of our times. Indeed, he is the most cited living scientist. However, as scientists we believe that we should base beliefs on facts and science over opinions, even if those opinions come from an academic as eminent as Prof. Ioannidis.

We believe that the facts outlined below completely undermine the paper published in BMJ Open. While individual scientists are entitled to their own opinions – indeed diverse perspectives are required for science to prosper – they are not entitled to their own facts. That the facts contradict the published work is unfortunate, but could easily be remedied if Professor Ioannidis was interested in addressing our concerns and correcting the many mistakes in the paper:

1. Prof. Ioannidis’ paper uses a satirical index, the Kardashian index, as though it were a serious bibliometric index to imply that scientists who have been critical of the Great Barrington Declaration (GBD) are less impactful scientists as measured by citation metrics. The paper has led to targeted harassment of the named scientists, including junior researchers and women of color, which, not coincidentally, uses the language of Prof. Ioannidis’ paper. For example, activists (including at the American Institute for Economic Research [AIER]) have attacked, demeaned, and humiliated these scientists by calling them “science Kardashians,” while media outlets sympathetic to the GBD and AIER have parroted the same attacks [10,11], thus amplifying them. This has caused real emotional and professional harm to researchers, and is a direct consequence of the dubious ethics of Prof. Ioannidis’ publication.

2. Dr. Scott Atlas, President Trump’s chief coronavirus advisor, who embraced the GBD, has praised and thanked the GBD authors and their colleague Prof. Ioannidis for their support for his policies [12-14]. Atlas’ statements indicate that Prof. Ioannidis has a potential undeclared intellectual conflict of interest (CoI) that contradicts the supposed objectivity of the paper. His political writings elsewhere also suggest undeclared COIs [15]: Ioannidis describes COVID-19 mitigations as a dire threat to democracy, a position that strongly echoes that of GBD authors and their supporters [16], who have called mitigations like “lockdowns” and mandates requiring masks and vaccines “tyrannical” and “authoritarian” [17], comparing them to incipient fascism [18] and the Chinese Cultural Revolution [19]. Prof. Ioannidis’ views should be properly declared, as should any other advice or support he has given to the GBD authors in terms of COVID-19 policy and planning.

3. It is now well documented that the GBD received major support from the AIER [20,21]. Indeed, the key scientist involved, Martin Kulldorff, was recruited by the AIER, which invited him for a long weekend to socialize with like-minded people [20]. Soon after, Kuldorff and the AIER invited Jay Bhattacharya and Sunetra Gupta to a weekend conference at the AIER headquarters. There, AIER wined and dined the GBD authors, provided meeting space and lodging, arranged a meeting with journalists, and provided editorial feedback in drafting the declaration. The AIER also provided tools to amplify the message of the GBD, including professional videography of the interviews, social media expertise, and web services that allowed the rapid creation of a website for the GBD. Moreover, AIER is funded through an investment fund that itself owns shares in tobacco companies and many organizations that stood to lose enormous amounts of money if the United States were to have enacted further lockdowns in 2020 [21]. This represents a large intellectual and financial CoI for the AIER, which was closely involved in the GBD creation. This cannot and should not be ignored in any objective analysis.

4. The paper as published fails to include some critical basic scientific information required to evaluate whether its conclusions are supported by its evidence and analysis, the most glaring omission being a description of the statistical methodology used sufficiently detailed to allow others to evaluate and reproduce his analysis. This should be corrected.

5. Despite Prof. Ioannidis’ continued dismissal of our critiques, his social media analysis remains straightforwardly erroneous. This is not simply a matter of opinion. Twitter alone is not sufficient to draw conclusions about all of social media. Even his analysis of Twitter is very poorly done and captures no real measures of “impact” on that platform. That Professor Ioannidis has responded to this point by presenting simple counts of tweets from accounts – yet another completely uninformative metric given that people tweet about many things – mostly shows why social media analyses should include authors who actually use social media and, better yet, understand how to analyze social media. Clearly, Prof. Ioannidis does not meet either of these criteria. Indeed, he himself admits in the Competing Interests statement of his original paper that he has “no personal social media” [1] and elsewhere adds on his Stanford University webpage, “I admire people who can outpour their error-free wisdom in them [social media accounts], but I make a lot of errors, I need to revisit my writings multiple times before publishing, and I see no reason to make a fool of myself more frequently than it is sadly unavoidable” [22].

6. Professor Ioannidis concludes that the John Snow Memo (JSM) gained “superiority in news coverage,” but he did not study media coverage of the JSM and the GBD, so it is unclear how he can conclude anything on media coverage. In fact, based on the data, one could very plausibly argue that, since the GBD has received orders of magnitude more attention from the general public, media, and social media, his conclusion is entirely false. Indeed, in a recent study, Marco Zenone and colleagues analyzed media discourse on the GBD’s “natural” herd immunity strategy [23]. They found that “considerable media attention” was given to “a small, vocal, and heavily publicized coalition of scientists with prestigious credentials” and that “prominent government advisors promoted and legitimized the strategy.” They also found that “news coverage gave extensive attention to scientists promoting the strategy–notably, the Great Barrington Declaration and its authors.”

Based on the many errors in Prof. Ioannidis’ analysis listed above, we humbly suggest that his paper has utterly failed in its stated aims. There was no appraisal of social media aside from a thoroughly inadequate review of one metric from Twitter. Worse, the other metric from Twitter that he used, the Kardashian index, was originally intended as a satirical index designed to mock scientists who obsess over citation metrics. Indeed, Prof. Ioannidis’ paper itself makes a major error in that it does not review the scientific basis for either document. If Professor Ioannidis is truly interested in why some consider the GBD to be “nonsense,” he would be better served by reviewing the numerous factual errors and failed predictions in the GBD. To give just one example, the GBD mistakenly predicted that countries such as Sweden that followed a “Focused Protection” approach would reach “herd immunity” within 3-6 months [24]. A detailed review of the many errors in the GBD by Prof. Ioannidis would be a better use of his time than personally attacking scientists with whom he disagrees using poorly-researched metrics that are fundamentally inadequate to answer the question.
References

1. Ioannidis JP. Citation impact and social media visibility of Great Barrington and John Snow signatories for COVID-19 strategy. BMJ Open 2022;12:e052891
2. https://bmjopen.bmj.com/content/12/2/e052891.responses#conflicts-of-inte...
3. https://bmjopen.bmj.com/content/12/2/e052891.responses#is-this-incredibl...
4. https://bmjopen.bmj.com/content/12/2/e052891.responses#yamey-reply-to-io...
5. https://bmjopen.bmj.com/content/12/2/e052891.responses#response-to-prof-...
6. https://bmjopen.bmj.com/content/12/2/e052891.responses#scientific-inadeq...
7. https://bmjopen.bmj.com/content/12/2/e052891.responses#it-is-not-%E2%80%...
8. https://bmjopen.bmj.com/content/12/2/e052891.responses#on-weaponization
9. https://bmjopen.bmj.com/content/12/2/e052891.responses#scientific-inadeq...
10. https://www.telegraph.co.uk/news/2022/02/15/lockdown-debate-skewed-scept...
11. https://www.aier.org/article/science-kardashians-vs-the-great-barrington...
12. https://www.nytimes.com/2020/10/19/health/coronavirus-great-barrington.html
13. https://www.foxnews.com/politics/i-was-disgusted-dr-scott-atlas-unloads-...
14. https://www.foxnews.com/transcript/tucker-why-are-americans-puting-up-wi...
15. https://www.tabletmag.com/sections/science/articles/saving-democracy-fro...
16. https://brownstone.org/articles/who-is-served-by-emergency-powers/
17. https://brownstone.org/articles/the-truth-about-tyranny/
18. https://vinayprasadmdmph.substack.com/p/how-democracy-ends?s=r
19. https://brownstone.org/articles/the-covid-response-brought-the-cultural-...
20. https://www.bmj.com/content/374/bmj.n2268/rapid-responses
21. https://blogs.bmj.com/bmj/2021/09/13/covid-19-and-the-new-merchants-of-d...
22. https://med.stanford.edu/profiles/john-ioannidis
23. Zenone M, Snyder J, Marcon A, Caulfield T (2022) Analyzing natural herd immunity media discourse in the United Kingdom and the United States. PLOS Glob Public Health 2(1): e0000078. https://doi.org/10.1371/journal.pgph.0000078
24. https://www.thelancet.com/article/S2213-2600%2820%2930555-5/

I thank Helling, Meyerowitz-Katz (second comment) and Yamey and Gorski (third comments). I learned a lot in this exchange.
Meyerowitz-Katz is probably less critical of my work than I am myself. I acknowledged face and construct validity limitations in Twitter metrics. Indeed, number of followers are a surrogate. However, the difference between the two compared groups is extraordinary. Following his comments, I collected also data on number of tweets. Differences are immense: 20 JSM key signatories versus only 2 GBD signatories exceed 6000 tweets each. The top-10 (26900 to 177900 tweets per scientist) are all JSM key signatories. The JSM key signatories unleashed almost a million tweets, roughly 10-fold higher than GBD signatories. Even adding AIER and GBD accounts (25600 and 579 tweets, respectively), the difference remains enormous. Other social media platforms are worth exploring, but access to data is problematic, as Altmetric has also witnessed. Regardless, the stated primary aim of the paper was not to “evaluate the social media visibility of signatories, as denoted by Twitter followers”, but “to examine whether the prevailing narrative that GBD is a minority view among experts is true”: the paper provides strong evidence that this is untrue. Both GBD and JSM are supported by many stellar scientists. The Twitter analysis offers one potential explanation why GBD was dismissed as “unscientific nonsense” by Meyerowitz-Katz (https://twitter.com/gidmk/status/1344051607130890242) and others. Future research could add more insights on why, how and by whom thousands of scientists, including many with the highest citation impact worldwide, were summarily trashed.
We can all help diminish science weaponization. Researchers who publish more may suffer more weaponization. Gorski spends another 2064 words of talented writing offering examples, but weaponizations are millions. My COVID-19 papers have been cited about 5 thousand times in the scientific literature by tens of thousands of scientists and were discussed by millions of people. Many discussions were much better than my struggling work while others represented horrible misuse. The best way for any scientist to “control” use/misuse is to try to do better science, arriving at less biased estimates. I consider science communicators my heroes exactly because they can clarify important points, countering weaponization. They can explain why scientific papers or quotes are not aligned with conspiracy theories. Conversely, if science communicators lambast scientific papers or quotes as being aligned with such theories, they officially surrender science to conspiracy. We should dismantle conspiracy theories, not the science or scientists they misuse.
Ill-conceived guilt-by-association tactics damage public health efforts that require unity and tolerance. Science is poisoned by pigeonholing (“you are either with us, or against us”) and denouncing scholars. My track records shows my beliefs are (very) progressive, but I feel that science should not be shadowed by politics. I thank everyone who tried to help during this major crisis regardless of their political affiliation or lack thereof. Labeling scientists and organizations with whom one disagrees as anti-climate change, pro-tobacco, and anti-vaccine is harmful. The arguments raised this time sound as absurd as if someone were to attack Yamey as being pro-tobacco because his institution, Duke University, carries a tobacco magnate name. I am concerned when a scholar of Yamey’s caliber attacks colleagues in this fashion. Then, anti-science zealots are further empowered to claim that famous Stanford, Harvard, and Oxford professors as their own: Who says this? A famous Duke professor does.
I respect and cherish Yamey’s efforts on global health advocacy, but the attempt to link even me (!) to anything anti-climate change, pro-tobacco, and anti-vaccine makes me nauseated. My stance on these major health threats has always been uncompromising. In fact, during the pandemic the attack against my reputation in Greece and possibly also internationally was orchestrated (in part at least) because of my fierce repudiation of the tobacco industry. Just before the pandemic, I exposed a major alignment of the tobacco industry with media, social media and political magnates[1,2]. During the pandemic, Big Tobacco donated ventilators and received kudos from political leaderships for “social responsibility”(!)[3,4] – disgusted, I pushed for the elimination of the tobacco industry[5]. Several years ago, when I played a pivotal role in the approval of more restrictive EU regulations on tobacco advertisements, the tobacco industry struck back with vengeance[6]. During this pandemic, the reputational attack was even more malicious. Certainly, Yamey does not want to facilitate dark corporate plans of Big Tobacco – but unknowingly he does. Intellectual fratricide should stop. We need united efforts to solve urgent, major global health problems. This won’t happen if we only accept people who belong to only one clan.

References

1. Ioannidis JPA. Lethal news: The dexterous infiltration of news media by the tobacco industry agenda. Eur J Clin Invest. 2019 Jul;49(7):e13125. doi:10.1111/eci.13125.
2. Ioannidis JPA. Greece: Crisis, smoking and tobacco conflicts in social media. Eur J Clin Invest. 2017 Dec;47(12). doi: 10.1111/eci.12841.
3. Ioannidis JPA. How the pandemic has changes the norms of science. Tablet 2021, https://www.tabletmag.com/sections/science/articles/pandemic-science
4. https://www.forbes.com/sites/isabeltogoh/2020/03/30/big-tobacco-criticiz...
5. Ioannidis JPA, Jha P. Does the COVID-19 pandemic provide an opportunity to eliminate the tobacco industry? Lancet Glob Health. 2021 Jan;9(1):e12-e13. doi:10.1016/S2214-109X(20)30466-6.
6. Elaborated in my upcoming book “Variations on the art of the fugue and a desperate ricercar” to be published soon in English by Kedros publishing house in Athens (www.kedros.gr) as an e-book.