Coronavirus: Correlation does not mean causation, doctors say about BCG vaccine study

B8D422FC-4A0D-4F40-8ED4-50425F4958BD

Protection conferred by BCG vaccine against novel coronavirus is an interesting study and a plausible hypothesis but has been done early in the course of an epidemic and when the settings are not the same in all countries. We need more data from trials to be able to say anything with confidence.

A study posted on a preprint server medRxiv suggest that BCG vaccination offers some level of protection against infection by novel coronavirus (SARS-CoV-2) and even reduce mortality in COVID-19 patients. A preprint is yet to be peer-reviewed and published in a scientific journal.

The study by Gonzalo H. Otazu and others from the New York Institute of Technology, New York found an association between countries that have a universal BCG vaccination and reduced coronavirus cases and even deaths. However, association is different from causation and till protective effect of BCG vaccine against COVID-19 is tested one cannot be certain. The authors write: “The correlation between the beginning of universal BCG vaccination and the protection against COVID-19 suggests that BCG might confer long-lasting protection against the current strain of coronavirus.”

“This study has major limitations,” Prof. Madhukar Pai, a TB expert at McGill University says in a tweet.

This study looks at vaccination at the country/population level to protective effect at the individual level. “In this study, low quality evidence observed at the population level is used to make sweeping inferences about BCG’s effectiveness on an individual level,” says Emily MacLean from McGill University in Nature Microbiology.

The preprint looked at countries that have a universal BCG vaccination policy with countries that still do not have such a policy in place; the study has not included India. According to the study, middle- and high-income countries that have a current universal BCG vaccination policy have reported fewer deaths than countries that do not have such a policy. While it acknowledges that there is increased mortality with age, it concludes that elderly population in countries that had in place a universal BCG vaccination much earlier appear to have been protected by the vaccine.

“The protection offered by BCG is not robust and long-lasting. Children younger than five years have 80% reduction in acute disseminated TB such as meningitis and miliary TB due to BCG vaccination,” says virologist Dr. Jacob John, formerly with CMC Vellore. “But the vaccine does not protect against adult TB. It does not even protect children from getting infected.”

No uniform protection

“The protection offered by BCG vaccine is not uniform across the world. In India, the vaccine does not offer protection to adults. But in UK, there is 70% reduction in all forms of TB disease,” Dr. John says. The vaccine was included in India’s universal immunisation programme only in 1978. As a result, many, if not most, people older than 45 years in India have not received BCG vaccination as a child.

But BCG vaccine does have non-TB specific protective effect and confers protection against common respiratory infectious diseases. “This is manifested not at an individual level but is seen at the population level. At the population level, three-four years after vaccination, children who have not received BCG vaccine have higher mortality than vaccinated children in many countries,” says Dr. John.

Plausible hypothesis

“BCG given early in life does improve the immune system. The vaccine can prevent intracellular infections. So the protective effect of BCG against COVID-19 is a biologically plausible hypothesis,” says Prof. Gagandeep Kang, executive director of the Translational Health Science and Technology Institute (THSTI), Faridabad.

The study is based on COVID-19 data as on March 21 when the global caseload had just crossed 3,00,000. On March 21, India and South Africa had just 195 and 205 cases, respectively; on April 4 India and South Africa had 3,450 and 1,505 cases, respectively. “India will see an avalanche of cases in the next two-three weeks,” says Dr. John. Therefore, the conclusions drawn by the study based on constantly changing data is premature.

“The problem with this study is that it is done early in the course of an epidemic and the settings are not the same in all countries,” says Prof. Kang. “Testing has not been enough in many countries including India and many countries may be at an earlier stage in the epidemic.”

A few studies and experiences have shown that people may be infected but remain asymptomatic. So the protective effect of the vaccine will be known only when actual number of people infected becomes known through aggressive testing.

Low levels of testing

According to the study, there is evidence that BCG vaccination can be correlated with reduced mortality. While acknowledging the concerns about the number of tests done by each country varies, they still write that BCG vaccination has “reduced the number of reported COVID-19 cases in a country”. They have also used mortality rate as a measure of disease severity in each country. “Mortality rates are a robust measure that has less dependence to the levels of COVID-19 testing… BCG vaccination seems to significantly reduce mortality associated with COVID-19,” they write. `

With 96 deaths and 3,450 cases as on April 4, the mortality rate is 2.7%, which is high. “That suggests we are underestimating the true prevalence of COVID-19 in the population. We are probably testing only the very sick, or the surely infected, and not those with mild symptoms. Relying on the mortality rate when testing is not high is not a good way to measure the effectiveness of the vaccine,” says Dr. John.

Across countries the death rate is higher in people older than younger people. An estimated 13% of COVID-19 patients 80 and older die compared with 1.25% of those in their 50s and 0.3% of those in their 40s. Also, 85% of cases do not require hospitalisation while only 5% require ICU admission. COVID-19 patients with co-morbidities have higher mortality risk. The quality of healthcare even within India varies. Two old COVID-19 patients — 93-year-old man his wife who is 88 years old — were recently discharged from a hospital in Kerala, while younger patients in other States have died. To say that BCG vaccination “significantly reduces COVID-19 mortality” at this stage of the epidemic may be premature. It ignores the main risk factors that vaccination has no control over.

“It’s an interesting study and a plausible hypothesis. But correlation — two things that go together — does not mean causation,” says Prof. Kang. “We need more data from trials to be able to say anything with confidence.”

“There are ongoing trials using BCG vaccine on healthcare workers. We need data from such randomised control trials. Or there should be case-control studies to understand the protective effect of BCG vaccination,” she adds.

Published in The Hindu on April 4, 2020