Up to 200 Kiwis could be saved each year if a common medicine preventing stomach ulcers was avoided in hospitals' intensive care units (ICU).

That's according to the largest ever New Zealand-led randomised clinical trial involving 27,000 critically ill patients from five countries.

"Hundreds of New Zealand patients and tens of thousands of people in hospital intensive care units worldwide each year could be saved for essentially zero cost," said Dr Paul Young, lead author of the study and a Wellington Hospital intensive care specialist.

The prescribed medicine in question is Proton pump inhibitors (PPIs), such as Losec, which are given to about one in five patients who present to ICU. The medicine helps relieve symptoms of acid reflux and prevents bleeding from stomach ulcers.

The alternative is a histamine-2 receptor blockers (H2RBs) such as Zantac.

The study - published in the Journal of the American Medical Association (JAMA) - compared the risk of death between the two different types of ulcer-preventing medicines when used on patients in the ICU.

Researchers found that while PPIs were slightly better at preventing bleeding from ulcers, they more commonly led to deaths overall.

Out of the 13,413 patients who were treated with PPI, 2459 died at the hospital by day 90 - which equates to 18 per cent.

That's compared with 2333, out of 13,356 patients, who died after being treated with H2RBs - which equates to 17.5 per cent.

Young said the results of this trial were already making waves internationally.

"Initial indications are that the majority of intensive care specialists will stop using PPIs routinely to prevent ulcers in ICU patients."

He said the possibility that this small reduction in bleeding risk with the use of PPI medicines might come at the expense of a greater risk of death is of real concern.

"In particular, in this study, we showed that the risk of death was significantly higher with PPIs in patients admitted to the ICU after heart surgery and in those most gravely ill.

Young said they didn't know for certain how PPIs could increase the risk of death but suspected it related to the effect the medicines had on the patient's immune system.

"Both PPIs and H2RBs are established medicines with similar costs and this study is a great example of how, when comparing the effectiveness of treatments we've used for a long time like this, you never know what you're going to find."

Auckland City Hospital intensive care specialist Dr Colin McArthur said he thought the findings raised an important question that was likely to change doctors' approach to treating critically ill patients at risk of bleeding from ulcers in ICUs around the world.

It's a clinical judgment question, some patients may still benefit from PPIs because the risk of dying from bleeding out might be higher than other complications, McArthur said.

Medsafe group manager Chris James said the report will be useful for emergency medicine specialists caring for very unwell people needing complex care.

"Both medicines reported in the study are used in these critically unwell patients to protect against internal bleeding in the stomach, which can happen at a higher rate in people who are very unwell and needing care in an intensive care unit.

"A key finding of the study is that there was no significant difference between mortality rates for either of the two types of medicines - proton pump inhibitors and histamine-2 receptor blockers - which was the primary aim of the research."

Further research would be needed to determine whether the study's results are relevant to the use of proton pump inhibitors in standard use in the community, James said.

The three-year study included almost 27,000 patients from 50 ICUs in five countries: New Zealand, Australia, Britain, Ireland and Canada.

New Zealand hospitals that took part were Auckland City, North Shore, Middlemore, Waikato, Tauranga, Wellington and Nelson.