Paramedics should give cardiac arrest patients uninterrupted chest compressions to improve their chances of survival, a US study has suggested.
Researchers say 200 uninterrupted chest compressions work best
Researchers in Arizona found survival rates trebled when a technique which emphasises the importance of non-stop compressions was introduced.
Very few people who suffer a cardiac arrest reach hospital alive.
But in the Journal of the American Medical Association, researchers insisted these rates could be improved.
Minimally interrupted cardiac resuscitation (MICR) involves providing 200 uninterrupted compressions, administering adrenaline early and waiting a little longer to insert a tube into the trachea to ventilate the lungs.
The practice was taught to paramedics in two metropolitan cities.
In the period examined prior to instruction, some 218 patients were attended after heart attack, of whom four - or 1.8% - survived long enough to reach a hospital.
After paramedics had been taught the technique, that figure increased to 5.4%. Thirty-six of 668 patients reached hospital.
"During resuscitation efforts, the forward blood flow produced by chest compressions is so marginal that any interruption of chest compressions extremely harmful," Dr Bentley Bobrow from the Mayo Clinic in Arizona wrote.
"Excessive interruptions of chest compressions by pre-hospital personnel are extremely common. Therefore, MICR emphasizes uninterrupted chest compressions."
The UK's Resuscitation Council said guidelines had moved in this direction, but that this model placed even more emphasis on chest compressions from the outset, and spent less time initially on ventilation.
In this country at present, some 30 compressions are recommended before ventilation begins.
"It is a significant study, but has to be interpreted very cautiously. It is not strong enough for us immediately to change what we are doing," said Jerry Nolan, the body's chairman, arguing that a more tightly controlled study was needed.
He added that one important caveat was the very low rate of survival to begin with in the Arizona study.
"We would be looking at rates of five to 10% with our methods, so it is unclear whether the Arizona model would have such a dramatic impact as it has in the study."
Judy O'Sullivan, cardiac nurse at the British Heart Foundation, said the study was "interesting" but also stressed the need for more investigation.
"It doesn't give us enough evidence to change current practice in the UK."