Understanding the Role of Science in Society's Health
Written on
In November last year, a new virus surfaced in China and rapidly became a global concern. Initially, information about this virus was scarce, and even now, despite an intense wave of scientific inquiry, much remains unclear. Nevertheless, health experts have continued to provide public health guidelines. Previous pandemics have taught us how dire situations can become, necessitating prompt decisions without awaiting complete data.
Given the swift transmission of the virus, we relied on the limited information available and our experiences with prior outbreaks to make informed predictions. Naturally, many of our predictions were likely to be incorrect.
Now, as we gain more insights, a critical question arises for scientists and health officials: Are we prepared to accept our mistakes?
Science and medicine are ideally objective fields guided by evidence. We pose questions, create hypotheses, and conduct experiments to gain insights into our complex world. However, science is not conducted in isolation, nor is it devoid of human influence. Like any human endeavor, science is subject to the same limitations and biases that affect other sectors. Our nature guarantees that errors will occur.
One of science's most commendable features is its ability to self-correct.
Throughout history, scientists and health professionals have held numerous incorrect beliefs. Yet, when new evidence emerges that contradicts the prevailing view, corrections occur, and paradigms shift—though such changes often meet resistance from established figures in the field. Aspiring scientists often seek to uncover findings that challenge conventional wisdom, which is fundamental to scientific progress and has led to advances in health, technology, and our understanding of the universe.
As a scientist, I understand the frustration that arises when conspiracy theorists and those opposed to established science dismiss the conclusions drawn by dedicated experts. I know what it is like to have my reputation tied to being correct and to experience the sting of being wrong.
However, acknowledging mistakes is integral to the scientific process. While it can be painful, recognizing when we are wrong allows us to adjust our course. This doesn’t undermine science’s reliability; rather, it fortifies it. Healthy scientific discourse thrives on disagreement, discussion, and debate, which can only occur in an environment where mistakes are acknowledged and improvements sought.
The challenge, however, is that science does not always operate this way. Sometimes, pride and political considerations impede intellectual honesty and scientific rigor. During the coronavirus pandemic, for instance, public health officials made statements and predictions that, when proven unfounded, led to a reluctance to retract or revise them.
When politics enter the equation, scientific inquiry and self-correction are obstructed as public perception begins to influence our conclusions. We may start to see what we wish to see, opting for rhetoric over empirical evidence to defend our positions.
I wish to emphasize three pandemic-related issues that illustrate how pride and politics can adversely impact scientific inquiry. I am not advocating for or against any specific public health measures, but rather highlighting how the fear of being wrong has become a significant intellectual burden for scientists and health officials, potentially jeopardizing the scientific enterprise.
The Asymptomatic Spread Debate
During a press briefing on June 8th, the World Health Organization’s (WHO) technical lead stated: “From the data we have, it still seems to be rare that an asymptomatic person actually transmits onward to a secondary individual.” This statement sparked immediate backlash on social media and from other health professionals, prompting the WHO to retract it the following day. What was the issue?
The original comment stemmed from a sincere evaluation of existing literature. The retraction was not due to a shift in expert opinion regarding asymptomatic transmission but rather in response to the backlash it provoked. In fact, their interim guidance later reiterated the same findings more explicitly:
> Comprehensive studies on transmission from asymptomatic individuals are difficult to conduct, but the available evidence from contact tracing suggests that asymptomatically-infected individuals are much less likely to transmit the virus than those who develop symptoms.
Why is the asymptomatic spread debate so contentious? Much of the public policy has hinged on the likelihood of asymptomatic transmission. Why are face masks mandated? Because we cannot identify asymptomatic carriers. Why close schools? Partially due to uncertainty about asymptomatic individuals. Why enforce strict lockdowns? The same reasoning applies.
This concept has been central to our social guidelines, making any softening of our stance on asymptomatic spread unpopular.
To clarify, I do not dispute the possibility of asymptomatic transmission. We know asymptomatic carriers exist, and it is reasonable to assume they could transmit the virus. However, the intense reactions from experts regarding the reduced likelihood of asymptomatic spread are not primarily grounded in abundant experimental or observational evidence.
The literature concerning asymptomatic COVID-19 transmission is complex. Many studies assume asymptomatic carriers significantly contribute to virus spread, confuse asymptomatic with presymptomatic transmission, or rely on theoretical models rather than direct observational data.
Some papers suggest extensive documentation of asymptomatic spread yet cite minimal evidence from specific family clusters.
The idea of asymptomatic COVID-19 transmission is a valid hypothesis and initially justified public policy. However, as more data emerges, this hypothesis is increasingly contested.
It could be that asymptomatic transmission is overstated (i.e., it is rare), or it might simply be incorrect. Are we ready to accept that possibility?
The Mask Mandate Inquiry
The debate surrounding the necessity for the public to wear masks to mitigate virus transmission is currently a focal point among health officials.
Many claim it is a scientific fact that masks reduce transmission. But is there solid evidence to support this? According to the WHO's interim guidance:
> At present, there is no direct evidence (from studies on COVID-19 and in healthy people in the community) on the effectiveness of universal masking of healthy people to prevent infection with respiratory viruses, including COVID-19.
Yet, this has not deterred assertions that mask-wearing is crucial in curbing transmission. I have observed unsubstantiated claims circulating on social media about specific percentages of risk reduction associated with mask-wearing.
The rationale begins with a plausible hypothesis: If everyone wears a mask, asymptomatic carriers are less likely to spread the virus due to the barrier it provides. However, this hypothesis has been promoted to a scientific certainty without direct evidence.
As previously mentioned, it is unclear how frequently asymptomatic carriers transmit the virus. Furthermore, cloth masks (the type typically recommended) primarily block larger droplets from coughs or sneezes but are less effective against smaller aerosolized particles that may carry the virus. Notably, a 2015 study indicated that cloth masks could be ineffective in preventing infections and might even increase the risk due to moisture retention and improper use.
Official recommendations regarding mask-wearing have varied, likely due to the lack of evidence supporting their effectiveness in healthy populations. Do these cloth masks offer any protection? Perhaps, especially in preventing symptomatic individuals from expelling droplets. However, if someone is coughing, they are likely symptomatic, which suggests a more effective strategy would be to keep symptomatic individuals home or require them to wear masks when necessary.
I am not claiming that masks lack value or that public mask recommendations are unreasonable. I argue that the rationale is not founded on direct evidence, and it is conceivable that the assumption of their significant protective effect could be incorrect. Are we prepared to admit that?
Instances of Scientific Confirmation Bias
Lastly, I want to briefly address examples of confirmation bias in scientific literature. I have mentioned the tendency for journal articles to exaggerate the impact of asymptomatic transmission to support their modeling and mask recommendations. Such misleading claims are likely unintentional manifestations of confirmation bias, where researchers seek evidence that aligns with preconceived conclusions.
A recent study exemplifies this, with one author claiming:
> Our study establishes very clearly that using a face mask is not only useful to prevent infected coughing droplets from reaching uninfected persons, but is also crucial for these uninfected persons to avoid breathing the minute atmospheric particles that infected people emit.
What’s problematic here?
The paper does not convincingly support its claims. In fact, the data suggests that infection rates were declining before mask mandates were enacted. The authors make several assumptions, including compliance and appropriate mask usage, and begin measuring effectiveness immediately after the mandate, neglecting the virus's incubation period.
Moreover, they assume that any deviation from predicted infection rates is solely attributable to the mandate, ignoring natural transmission dynamics.
The data presented does not conclusively support the case for mask usage, regardless of their effectiveness. Ironically, the authors conclude with a call for sound science in public health decision-making.
Following publication, an epidemiologist from Johns Hopkins noted that her colleagues requested a retraction of this paper. This, too, is a part of the scientific process: confronting flawed research.
A more striking instance of confirmation bias appears in a recent article in Nature, where the authors argue that lockdown measures significantly reduced virus transmission in several European countries.
While evaluating public health measures is essential, we must scrutinize claims asserting that data validates prior interventions. The data might support such assertions, but skepticism toward these claims is both healthy and necessary.
I identify two significant flaws in this paper.
Firstly, the authors utilized a model incorporating hard-to-estimate assumptions about transmission rates and intervention effectiveness. They acknowledged this limitation upfront.
The more critical issue is the absence of a control in their analysis.
The researchers assumed their model accurately predicted deaths that would have occurred without lockdowns, then compared their estimates to actual COVID-19 death rates. By demonstrating a disparity between projected and actual deaths, they confidently claim lockdowns significantly slowed transmission.
However, they overlooked an existing control within their data. Their model suggested that only a complete lockdown would yield significant effects on transmission. To test this, we can examine Sweden, which did not impose a full lockdown. Their model predicted 28,000 deaths by May 4th, while actual deaths in Sweden totaled only 2,769. This is a substantial discrepancy. Interestingly, their model predicted 2,800 deaths under public health interventions—accurately reflecting actual deaths despite Sweden's lack of a lockdown.
I am not suggesting that these scientists intend to produce flawed research; rather, confirmation bias can obscure objective evaluation, especially on politically charged issues such as lockdown effectiveness. There is immense pressure to validate decisions made. But what if we were incorrect? Are we ready to confront that?
I do not argue for or against specific measures implemented to combat the virus. In fact, I believe many precautions were justified based on the limited data available. What I contend is that for science to flourish in revealing the complexities of our world, we must be open to the possibility of being wrong and allow space for error.
Factors such as political motivations and pride can hinder this essential aspect of the scientific process, diminishing scientific rigor, fostering confirmation bias, and ultimately eroding public trust in scientific claims. Even amidst intense scrutiny and pressure, we must strive for intellectual honesty and the humility to accept when we might be incorrect. Society relies on this commitment.