Brazilian health officials say a Chinese coronavirus vaccine has demonstrated an overall efficacy rate of 50 per cent, far lower than originally indicated in a blow to one of the Latin American nation’s hopes for defeating the pandemic.

The São Paulo-based Butantan Institute said last week that the CoronaVac vaccine, developed by China’s Sinovac Biotech, was 78 per cent effective at preventing mild infections, and 100 per cent effective for moderate and serious cases that require hospitalisation or intensive care.

However, following criticism over a lack of transparency and details in the results, the institute announced fuller data on Tuesday that showed that CoronaVac’s general efficacy rate fell to 50.4 per cent when “very mild” cases were included.

A level of 50 per cent is widely considered the minimum threshold for a vaccine to be considered viable. While just above that, the data still casts doubt over a flagship Chinese initiative to tackle the Covid-19 pandemic.

Ricardo Palacios, clinical research medical director at the Butantan Institute, said the widest possible definition of symptoms had been applied in the tests and sought to resist comparisons with the results of other vaccine trials.

“People want to compare other studies, but that’s like comparing a person that runs a 1km race on a flat stretch and another doing it on a steep and obstacle-filled stretch,” he said. “Other studies did not include as a case [of infection] two days of a headache”.

Given the trials in Brazil were undertaken on health professionals with high exposure to coronavirus, Mr Palacios said he believed CoronaVac would have greater efficacy when deployed in the wider community.

The shot still needs to be approved by Brazil’s health regulator, Anvisa, which is also evaluating a request for emergency use of the Oxford-AstraZeneca vaccine. Together they are the main jabs that the government is relying on to immunise the country’s population of 210m.

CoronaVac uses a chemically inactivated version of the virus to provoke an immune response. It is delivered in two doses.

The Butantan Institute is partnering with Sinovac to manufacture the drug and ran trials involving almost 13,000 volunteers throughout Brazil. It said the vaccine generates an immune response after 28 days.

But the new information leaves many unanswered questions about the shot, which has produced diverging results in the different countries where it is being tested.

In Turkey, CoronaVac was deemed to be 91.25 per cent effective based on late-stage trials. Indonesia, which this week became the first country outside China to approve the vaccine, posted an efficacy rate of 65.3 per cent.

By contrast, BioNTech/Pfizer and Moderna have reported efficacy rates of 95 per cent and 94.1 per cent, respectively.

Lawrence Young, virologist and professor of molecular oncology at the University of Warwick, said the reporting of the Sinovac trial of CoronaVac was “very confusing”.

“It is difficult to interpret all this information without seeing the full data sets. This highlights the problem of issuing data by press release rather than publication in a peer-reviewed journal. It also emphasises the different approaches to requirements for regulatory approval,” he added.

The results will add to pressure on the government in Brazil, where more than 200,000 Covid-19 fatalities have been recorded, the second-highest for any country after the US.

Ministers in Brasília are scrambling to launch a stalled immunisation programme. But efforts have been marred by political infighting and an ambivalent attitude from President Jair Bolsonaro, who has expressed deep scepticism about the need for jabs and vowed not to take one himself.

Although the federal government has said that injections could begin anywhere between late-January and the start of March, authorities in São Paulo, the country’s most populous city, are pressing ahead with plans to start on January 25.

Additional reporting by Carolina Pulice