Friday, June 19, 2020

Science and Politics

Our Prime Minister, like other national leaders, has said that all the decisions made about managing the coronavirus were based on the best scientific advice by medical experts. This is a little misleading. There are several problems with relying on science in a situation like this.

  • The scientific process is slow, so science does not provide instant answers. Science works by independent checks by peer reviews and confirmation of results by other researchers. This process takes time.

  • This means that early results will often be exposed as incorrect. With regard to coronavirus, researchers are publishing results without waiting for peer review. Some have had to be withdrawn when peer review by independent experts has exposed flaws. When more repeat studies are completed by other independent researchers, some early results that have been published will be proven wrong. There is nothing wrong with this, because it is the way that science works. However, it is a problem when decisions have to be made quickly.

  • Testing medical hypotheses is difficult. Scientists like to do double-blind experiments where a potential treatment is applied to a sample of people and a randomly selected control group are left without treatment, as this is a good way to test if a treatment is effective. This type of experiment is often not possible with medicine, as it is immoral to exclude treatments from people who might benefit.

  • Medical studies often used statistical methods to assess treatments. These methods can identify correlations between outcomes and personal characteristics or practices, but establishing a causal relationship is much more difficult. And because many characteristics and events are not always well defined or recorded accurately, statistical studies can produce misleading results. These problems will often be exposed by peer review or more thorough repeat studies, but that will take time.

  • Robust experiments are difficult to design. Even when samples are selected carefully, bias can creep in if the results were affected by factors that the designers of the study missed, and therefore failed to control. These problems will be gradually ironed out by effective peer review and repetition of studies, but that takes time, so it is not surprising that the first results to be released are later proven to be wrong.

  • Each epidemic is different. Therefore, relying on what was effective during the previous epidemics can lead to mistakes.

  • There are no perfect models (except on the catwalk), so modelling always bring in uncertainty. The models used by epidemiologists to estimate death rates when the coronavirus first emerged were flawed. Subsequent expert reviews revealed that faulty assumptions and inappropriate parameters were used. There was no allowance for the effect of the virus on people in aged care facilities. These flaws resulted in estimates that were unreliable. They have now quietly been forgotten.

  • On medical issues, there will always be disagreement. Put three specialists in a room, and they will often disagree on the cause of the problem, the best treatment, and the possible outcome for the patient. This disagreement is normal in medical science. The best way to resolve this problem when big decisions have to be made is to gather as many experts as possible and thoroughly debate the issues. These debates will not always produce an agreement, but it will at least expose the flawed arguments.

  • Sometimes science has not answered the questions that are most important for policymakers. There is no science that indicates that a lockdown of an entire economy is more effective for dealing with an epidemic than quarantine of people who are sick. Quarantines have been applied in the past, so they have a known track record, but that evidence is not always applicable to other viruses. The lockdown solution has not really been used in the past, so there is very little evidence about the economic costs and the effectiveness of such a remedy. The outcome will only be evident when annual excess death statistics are released in a year's time, but even then, there will still be debate.

  • Science cannot answer spiritual questions. If the coronavirus had a spiritual cause, science has to ignore it, because it can only assess physical cause and effect.

  • Six months after the coronavirus was first discovered, it is fair to say that more is not known about the way it behaves than is known. That is not a failure. It is the normal situation when something new has been discovered. Most of the uncertainties will be resolved as the scientific process continues, but we should not expect medical science to have all the answers straight away. We should not be surprised if hypotheses are to be proved, or if later studies expose flaws in earlier ones. This is the normal scientific process at work. However, it does mean that we should not expect immediate answers to all questions when a new disease emerges. Assuming that you can know things that can't be known is dangerous.

Because the medical science is at best uncertain, and at worst, non-existent, may of the decisions about responding to the coronavirus mostly had to be political decisions. This is right because the ultimate responsibility for the outcome rests with them. Of course, the politicians like to claim that were relying on medical and scientific advice, but they are really just propagating an untruth to give their decisions more credibility with the wider population.

No comments: