Now Reading
COVID-19: How Evidence-Based Medicine Eliminates Bias

COVID-19: How Evidence-Based Medicine Eliminates Bias

The COVID-19 pandemic has killed over 360,000 people to date. In the face of a novel pathogen, fear has reigned. And with no effective treatment and an influx of struggling patients, despairing doctors have turned to treatments designed for other illnesses to fight off this one.

Two in particular come to mind. Chloroquine has been used to treat malaria since 1934; its analogue, hydroxychloroquine (HCQ), has been used to treat rheumatoid arthritis. The latter is in the spotlight today. A controversial laboratory study indicated that HCQ has subtle anti-viral properties. However, laboratory results frequently don’t translate into human effectiveness.

But regardless of this truism, a hungry international media exaggerated claims of HCQ’s efficacy, aided by reckless acts on the part of politicians like US President Donald Trump. India soon jumped on the bandwagon, and the Indian Council of Medical Research endorsed HCQ has prophylaxis among healthcare workers and contacts of patients.

Predictably, good-quality clinical research has since shown that HCQ has no clear benefits.

Other treatments that have found favour include remdesivir, ivermectin and convalescent plasma therapy. However, clinical trials involving thousands of participants1 have found no evidence that these treatments work as well as they are hoped to. The most promising outcome was seen with remdesivir, in the form of marginal benefits.

Patients may indeed have to wait a long time for a cure. This is not at all unprecedented: influenza has been around since at least 1580 AD and has no effective cure.

This said, the hype surrounding HCQ prompts one question: What should dictate medical treatment – emotion or evidence?

Emotional influences, in the form of personal intuition, experience, traditions and patient- and peer-pressure often guide medical decisions but they also lead to deleterious effects. Smoking tobacco was once thought to be able to fight off the plague, a consideration that arose more recently in the context of the ongoing pandemic as well.

In a pandemic, emotion is even more perilous. Emotion-driven treatments that are otherwise inefficacious may contribute to a false sense of security, unwarranted side-effects, divert resources and delay research into treatments that may actually work.

Such risks are best countered with randomised clinical trials. In such trials, researchers compare the effects of consuming a drug in one group of people with the effects in another group that didn’t receive the drug. For example, in the first reported trial, conducted in 1747, James Lind divided sailors with scurvy into six groups. One group was given citrus fruits and the other five were given other foods. At the end of his study, Lind proved that eating citrus fruits got rid of scurvy. However, as an indication of how hard it is to change minds and their opinions, it was 42 years before lemons were acknowledged as medicines.

Clinical trials aren’t free of flaws, of course. Biases may undermine them. They include selection bias – preferentially choosing some patients for a treatment; performance bias – treating one group better than the other; and reporting bias – reporting positive results more than negative ones.

Randomisation mitigates selection bias by dividing trial participants entirely by chance into treatment and non-treatment groups. Blinding helps further: in single-blinded trials, patients don’t know if they’re in the treatment or in the non-treatment group; in double-blinded trials, the doctors don’t know either. The best clinical trials are randomised, controlled and double-blinded, which is also why they take so long to perform.

Unsurprisingly, most studies that touted HCQ in the early stages of the pandemic were low-quality unblinded studies.


Also read: The Wire Science’s investigative series on why hydroxychloroquine has found favour with ICMR


Today’s best medical practices use evidence-based medicine (EBM). EBM is defined as the conscientious, explicit and judicious use of best available evidence to make medical decisions. It consists of reviewing and assessing all the available information to determine what the best treatment could be.

Not all evidence is equally good or valid either. Instead, they range from expert opinion to the highest level, meta-analysis: the systematic pooling and appraisal of all research data, followed by statistical analysis. The outcomes inform the best practices to be followed by all clinicians.

The EBM pyramid. Source: Author provided

In effect, EBM empowers clinicians to reasonably correct approaches even when working with uncertainty, healthcare systems to foster guidelines, and governments to make policy and deliver services. EBM also encourages doctors to share decision-making with their patients, together with clear risk-benefit assessments and consensus on desirable outcomes.

This is particularly useful in India, where 80% of care is provided in unsupervised environments. And even as EBM is gaining ground in India, there is some reluctance to adopt it as clinicians feel its precepts challenge their autonomy and institutional hierarchy. This is perhaps reflected in the paucity of EBM literature from India.

A further challenge is that fully a fifth of India’s medical practitioners practice Ayurveda, Unani and homeopathy. EBM principles applied in these fields could help integrate these practitioners into mainstream medicine if they are also able to identify efficacious remedies.

Patients faced with the prospect of a debilitating infection, and possibly death, want treatments. But this conflicts with the time required to establish the evidence base for any specific treatment. In this scenario, clinicians could offer the best supportive treatment (including oxygen, treating complications, intensive care, etc.) even as they recruit willing patients into high-quality clinical trials. India has a large volume of patients so there is an opportunity here to test different treatments to qualify the most suitable for use by all patients.


Also read: Will COVID-19 Change AYUSH Research in India for the Better?


More importantly, if the government promoted research and EBM during the pandemic, it could help change attitudes forever.

All doctors and medical students could be trained in EBM with access to free online government resources with access to the latest version of the scientific literature. Second, national expert networks could develop India-specific evidence-based guidelines for various conditions. Third, a truly independent regulatory body could monitor clinical practice judged against the national guidelines. Fourth, a public education campaign could make Indians aware of the importance of EBM. This will raise patient expectations as well as standard of care.

In sum, the pandemic is an opportunity to abandon emotion and anxiety in the search for treatments and to adopt EBM. This will allow both Indian doctors and patients to think critically, and develop expectations of optimal care in every clinical scenario..

Dr Saif Razvi, MD FRCP, is a consultant neurologist in the UK. The views expressed in the article are the author’s own.


  1. Insofar as they have been conducted

Scroll To Top