A world-first computer model that can predict when a bacterial infection will stop responding to treatment could offer a much-needed boost in the fight against one of the top global health threats.

Some 1.27 million deaths were linked to antimicrobial resistance in 2019, which occurs when bacteria, parasites, viruses and fungi stop responding to treatments.

Scientists are racing to develop new ways to predict when infections become resistant to medicines so that drug treatments can be used more effectively.

Monash University-led researchers have now developed a computer model that can predict when a certain antibiotic used for the sickest hospital patients, meropenem, will no longer work on strains of a specific infection.

It was successful for seven different strains of pseudomonas aeruginosa, which commonly becomes resistant to antibiotics and can kill immunocompromised people.

The findings were published in peer-reviewed journal Clinical Microbiology and Infection.

Co-lead author Associate Professor Cornelia Landersdorfer said the model accurately predicts when medicine resistance kicks in.

"The relationship between antibiotic, bacterial characteristics and resistance emergence during treatment is complicated, which means getting dosing regimens right can be tricky," the Monash Institute of Pharmaceutical Sciences researcher said.

Associate Professor Antonio Oliver from Spain's Investigación Sanitaria Illes Balears, also a co-lead author, believes the study to be a world first.

"To our knowledge, no previous studies have employed an integrated experimental and modelling approach," he said.

Prof Landersdorfer said the findings could be used to develop new ways to attack infections in the future.