Pre-hospital emergency care for people who are seriously injured is "incomplete, unpredictable, and inconsistent", experts have warned.
Many parts of the UK rely on volunteer doctors and people from charities who have "varying availability and capability".
Patients wait 55 minutes on average between being seriously injured and getting to hospital - time that is vital if lives are to be saved. Yet few parts of the UK have doctors on hand to treat patients before they arrive at hospital, and only some of those are properly trained in areas such as anaesthesia, the study said.
Furthermore, care for patients is "much worse" at night and there are concerns about how the UK would cope with a major incident, according to the experts from hospitals and universities across the UK, writing in the Emergency Medicine Journal.
They said the use of volunteer doctors to fill the gap in service provision "is considerable". They added: "The call for these doctors indicates that NHS ambulance services recognise a cohort of patients whose needs exceed their capabilities."
Between April and July 2009, the experts sent a freedom of information request to all 13 regional NHS ambulance services, 17 air ambulance charities and 34 organisations affiliated to the British Association for Immediate Care (Basics) operating in England, Wales and Northern Ireland. A telephone survey was also carried out with 215 major emergency departments.
The findings showed that, of the 13 ambulance services, only three knew how many doctors were available to support their service. Five ambulance services had recorded requests for doctors (3,720 and 4,651 incidents in 2007 and 2008, respectively).
Overall, nine services recorded doctors attending the scene while another 10 had no data.
All 34 organisations affiliated to Basics are funded by charities, with 33 of those saying they only deployed doctors if available. None have a guaranteed 24-hour service and only one operates routinely into the night as a scheduled service.
Of the 327 volunteer doctors working in these organisations, 45% had training in anaesthesia. Some 12% of services deployed critical care capability at every incident but 32% had no deployable critical care capability at all.