Epidemiologist John Ioannidis of Stanford University on the errors of restrictive measures in the pandemicat 30.12.2021
Prof. John P.A. Ioannidis  published two studies  in December 2020, conducted in collaboration with two teams of medical and statistical specialists from the United States and Australia, evaluating the effectiveness of non-pharmaceutical pandemic interventions. These measures (mandatory social distance, travel restrictions, self-isolation, prohibiting public events, school closures, and even total quarantine) were the primary decision-makers levers in crisis management to reduce infections by reducing contact between people. The purpose of the two studies was to answer a question that we are all concerned about today: Are restrictive measures essential to end the pandemic?  "Strictly speaking, they did not demonstrate their effectiveness, but rather caused significant long-term damage."
The first study, "Effects of non-pharmaceutical interventions on COVID-19: A Tale of Three Models" , was published on December 10, 2020, and evaluates the three mathematical SIR models for predicting COVID-19 infection. The first two models were created by the Imperial College of London and were instrumental in containing the global pandemic. Each mathematical model predicts the rate of infection and the evolution of the epidemic using a different equation and other factors, with a different interpretation of the actual data. As a result, the first forecast model (Model 1) ignores population mobility when calculating the number of infections, whereas the second (Model 2) incorporates only population mobility into the equation.
Regarding the evaluation of restrictive measures (non-pharmaceutical interventions), each model has a different approach. Thus, according to Model 1, quarantine measures are the most effective, while according to Model 2, the same actions have shown little or no benefit. In the logic of Model 3, simply banning public events is considered beneficial, while quarantine has no relevant impact. Based on the Bayesian sequencing method, Model 2 was better supported by actual data than Models 1 and 3 for both time horizons analyzed (until May 5, 2020, and July 12, 2020, respectively). The third model (Model 3) envisions a combination of restraint measures, including population mobility.
Specifically, Ioannidis' team wanted to evaluate Flaxman et al.'s statements from an article published in the prestigious journal Nature , which claimed that Model 1, which justified the quarantine policies, would have resulted in the avoidance of a large number of deaths. According to Flaxman et al., total quarantine would have reduced the number of infections by 80%, preventing 3.1 million deaths in 11 European countries.
Professor Ioannidis et al.'s study demonstrates that non-pharmaceutical interventions' effects have proven ineffective in practice, regardless of how promising the predictions were. These predictions are dependent on the specifications of the statistical model, the assumptions, and the data used in the equations of those models and thus have significant errors. As a result, different conclusions have been drawn about the effectiveness of quarantine measures in reducing the epidemic wave and the number of deaths.
The second study , conducted by Professor Ioannidis in collaboration with a team of Stanford University researchers in medicine and statistics, examines the same non-pharmaceutical interventions, comparing the effects of more restrictive measures on infection rates with the impact of less restrictive measures. According to the study, there is no conclusive evidence that more stringent measures contributed significantly to lowering the infection curve in England, France, Germany, Iran, Italy, the Netherlands, Spain, or the United States in early 2020. However, the study's authors do not deny that these measures play a specific role in preventing and reducing infection. Still, they cannot find an additional benefit of more restrictive measures (isolation at home, cessation of business, and public life) than the less restrictive measures regime.
Ioannidis et al. emphasize the importance of a comprehensive vision that considers both the long-term benefits and harms to the population. Thus, some measures, such as avoiding public meetings, may work; others, on the other hand, do not and may even increase the number of deaths. School closures, for example, can increase the risk of vulnerable relatives becoming infected by children, who are more likely to encounter various viruses and thus become carriers.
On the other hand, quarantine resulted in a slew of "collateral losses," the most significant of which was the health system's inability to function correctly, as it was almost entirely focused on combating Coronavirus while ignoring the proper treatment of other diseases. More restrictive non-pharmaceutical measures, which have limited population mobility and access to health services for patients without COVID, have jeopardized the satisfactory operation of thousands of hospitals. As a result of the quarantine conditions, many patients with acute treatable diseases avoided seeking medical care, resulting in disease aggravation and even unjustified deaths under normal conditions. Quarantine imposed during high viral activity has also forced infected people to spend more time in cramped quarters with vulnerable relatives. Furthermore, such measures put economically at greater risk disadvantaged people and essential workers, homes became infection hotspots, whereas wealthy and healthy citizens could stay at home without risk. Another factor is the stress caused by quarantine, exacerbated by the media, that affects our immune responses to respiratory infections.
In addition, prolonged quarantine exacerbates the economic crisis, resulting in mass unemployment, and the unemployed may lose access to health care. Quality of life and mental health can deteriorate across entire populations. For example, in the United States, arms sales have increased dramatically since the quarantine's inception, with unforeseeable consequences. Disadvantaged populations are disproportionately affected. The number of people suffering from malnutrition worldwide due to the pandemic's economic impact has already surpassed one billion. Suicide risk has increased, as has the risk of domestic violence and child abuse. The disintegration of society is worsening, resulting in chaotic outcomes such as riots and wars.
The conclusion is that the mathematical models' predictions on the evolution of the pandemic, which offered the reasons for imposing the restrictions, were not confirmed by accurate data. Nevertheless, Prof. Ioannidis et al. do not dispute the role of these interventions in protecting public health. Still, in the meantime, they don't find a significant benefit of more restrictive measures (isolation at home and blocking economic and public events) in eradicating the pandemic, which would justify substantial damage in many other areas of health and social life.
 Stanford Prevention Research Center, Department of Medicine, Stanford University, USA; Department of Epidemiology and Population Health, Stanford University, USA; 6Department of Biomedical Data Sciences, Stanford University, USA; Department of Statistics, Stanford University, USA; Meta-Research Innovation Center at Stanford (METRICS), Stanford University, USA.
 https://www.medrxiv.org/content/10.1101/2020.07.22.20160341v3, https://onlinelibrary.wiley.com/doi/full/10.1111/eci.13484