This pandemic is frightening for many people especially those at the highest risk and the need for effective prevention and treatment is pressing.

When there is no direct evidence about what is best to do with a new disease, we need to create it. Clearly, that means recognising that some seemingly sensible interventions may not produce the intended effects. All of medicine is a bit uncertain (unlike maths, where proof is immutable). We can be very certain about some things (drinking bleach is harmful) and less certain about others (could vitamin D be useful?) Admitting uncertainty is something medicine historically finds difficult. But if we dont recognise the gaps in our knowledge, we wont research them. And good research has to be the way out of this.

There are now thousands of research papers on Covid-19, but many of such poor quality that we shouldnt rely on them. This means extra care and critique is needed. And on that basis, we simply dont know if vitamin D, for example, will help with Covid-19 (although all UK residents get a recommendation either to take or consider taking vitamin D the evidence for benefit is either absent or small with the exception of asthma). To know if it will help against Covid-19 needs rapid research for instance, assessing the vitamin D levels of people with severe Covid-19 against similar people who have less severe Covid-19 or avoid getting it altogether, or asking a group of people at high risk to diligently take supplements, and monitoring what happens.

As far as hydroxychloroquine goes, initial trials of this drug which is typically used to prevent malaria (and has been promoted by Donald Trump as a possible cure for all things Covid-19) were initially of poor quality, tiny, and with no control groups. Yet they were widely reported as showing it helped, while better quality trials that followed showed it didnt. Research is now focused on whether it might work as a preventive measure. Knowing what doesnt work is vital. It means that we can stop wasting resources on things that dont work and avoid unnecessary side-effects. Its also clear that hydroxychloroquine has cardiac side-effects, a fact that needs care in monitoring. There is no such thing as an intervention that is free of side-effects.

This doesnt just go for drug treatments. It should hold true for non-drug treatments. There is a clamour for everyone in public to wear a face covering. Does it work? Some research papers insist it does offering reasons as to why fabric will stop the passage of viral particles from the mouth and nose. But what really matters is whether this would result in fewer people being infected. So far, three reviews of the evidence (which have yet to be peer-reviewed) have found no or slender evidence, at best, for any benefit. But a bigger concern is a lack of investigation of unintended consequences. Would people wearing face coverings manage not to touch their faces or would they subconsciously take risks, which results in more infections, not fewer? What does this mean for deaf or lipreading people, or those with cognitive difficulties? Hazards for some may be accepted if there are outweighing benefits for everyone but, when the evidence is so slender, we need rational deliberation.

Would policymakers assume face coverings work, and make decisions on, for instance, encouraging the use of public transport on this basis without high-quality evidence to tell us this is safe? The truth is that we dont know, because we dont have the studies. The counter-argument is its a global pandemic, we cant wait for the results of research. I would argue that because it is a global pandemic, we need good, rapid research even more: the results will be applicable to far more people than usual when a drug or potential preventive measure is tested, hence the need to get better evidence for everyone. And yes, the need is urgent but this pandemic has no clear end in sight.

We could do fast, pragmatic, real-world studies, supplying masks to a geographical area with encouragement and instruction on use, and monitor for comparative changes in infection rates, together with carrying out field studies to observe peoples behaviour. Difficult, yes; but impossible, no. The World Health Organization has made it clear that we need research during pandemics (and drug trials have got off to an amazingly fast start), but this needs to hold true for things like face coverings too. What seems a sensible and obvious solution may not be. Harms and unintended consequences are everywhere, but are only reckoned with properly if we look for them.

Medicine is now in the era of Covid-19. This could be a turning point: where we take the uncertainty we are facing and, rather than assuming that our interventions will work, everyone citizens, patients, researchers and healthcare staff comes together to reduce it.

Margaret McCartney is a GP who writes about evidence-based medicine; she is the author of The State of Medicine: keeping the promise of the NHS

Read more:

There are few certainties in coronavirus medicine research is our best weapon - The Guardian