According to a new study, drug resistant infections will kill an extra 10 million people a year worldwide – more than currently die from cancer – by 2050 unless action is taken.

They are currently implicated in 700,000 deaths each year.

The analysis, presented by British economist Jim O’Neill, said the costs would spiral to $100 trillion.

He was appointed by UK’s Prime Minister David Cameron in July to head a review of antimicrobial resistance.

The reduction in population and the impact on ill-health would reduce world economic output by between 2% and 3.5%.

The analysis was based on scenarios modeled by researchers Rand Europe and auditors KPMG.

They found that drug resistant E. coli, malaria and tuberculosis (TB) would have the biggest impact.

In Europe and the United States, antimicrobial resistance causes at least 50,000 deaths each year, they said. And left unchecked, deaths would rise more than 10-fold by 2050.

Jim O’Neill is best known for his economic analysis of developing nations and their growing importance in global trade.

He coined the acronyms BRIC (Brazil, Russia, India and China) and more recently MINT (Mexico, Indonesia, Nigeria and Turkey).

He said the impact would be mostly keenly felt in these countries.

The review team believes its analysis represents a significant underestimate of the potential impact of failing to tackle drug resistance, as it did not include the effects on healthcare of a world in which antibiotics no longer worked.

Joint replacements, Caesarean sections, chemotherapy and transplant surgery are among many treatments that depend on antibiotics being available to prevent infections.

The review team estimates that Caesarean sections currently contribute 2% to world GDP, joint replacements 0.65%, cancer drugs 0.75% and organ transplants 0.1%.

This is based on the number of lives saved, and ill-health prevented in people of working age.

Without effective antibiotics, these procedures would become much riskier and in many cases impossible.

The review team concludes that this would cost a further $100 trillion by 2050.