Casualty departments could be harming rather than helping patients with serious injuries, experts claim.
Some emergency care treatments are not properly tested
Many treatments used are untested and potentially dangerous, says Professor Ian Roberts of the London School of Hygiene and Tropical Medicine.
The use of steroid drugs to reduce brain swelling was already known to cause "industrial slaughter", the BA Festival of Science in Norwich heard.
He also raised concerns about other common trauma treatments.
Professor Roberts said trauma care suffered because it mainly affected the poor.
"If you live in a poor and dangerous urban environment you're more likely to experience violence, get run over or have a dangerous job."
He said that when he started analysing trials of trauma treatments he ran into "an evidential black hole".
This led directly to his involvement in the Medical Research Council's Crash trial, which looked into the effectiveness of corticosteroids for treating traumatic brain injury.
The aim was to recruit 20,000 patients from around the world who would be randomly allocated corticosteroids.
But the results, published in 2004 in The Lancet medical journal, were so striking the trial was stopped after only 10,000 patients had been recruited from 239 hospitals in 49 countries.
According to the findings, corticosteroids raised the risk of death from 18% to 21%.
Professor Roberts said: "That's industrial slaughter because thousands of people get treated with this drug."
He said question marks also loom over other treatments for severely injured patients in intensive care units.
What testing the treatments had been subjected to involved very small numbers of patients, leading to enormous uncertainty about the results, he said.
Hyperventilation treatment to lower brain blood pressure, for example, had only been tested with one randomly controlled trial involving 77 patients.
Each day more than 300,000 people around the world are severely injured, of whom about 10,000 die. The leading causes are road traffic accidents and violence, from crime or war.
Professor Roberts said injury accounted for 7% of the total disease burden in the UK, yet attracted only 0.3% of research spending.
He acknowledged obtaining the necessary consent to conduct randomly controlled trials with severely injured patients was not easy, but said this was no excuse for not carrying them out.
But he added: "You either accept that patients go into randomly controlled trials without their own consent, or you say patients will always be exposed to untested treatment."
A Department of Health spokeswoman said: "As Professor Roberts indicated, there are great difficulties involved in studying patients with severe injuries.
"However, if there is mounting evidence of adverse drug reactions, the regulator has processes in place to investigate further."
And Martin Shalley, president of the British Association for Emergency Medicine, said: "A lot of the treatments that were postulated for the treatment of trauma have not been subject to gold standard trials.
"This is partly because a lot of them would be difficult to do, and if the patents have run out, companies wouldn't really wish to perform trials as there is nothing in it for them."