Patience Carter of Philadelphia, a victim of the Pulse nightclub shootings, speaks at a news conference at Florida Hospital Orlando on June 14 alongside physician Brian Vickaryous. (Phelan M. Ebenhack/Associated Press)

For more than three hours after the shooting began, Christopher B. Colwell stayed outside the building to triage victims, sending those with massive gunshot injuries straight to the emergency department and assessing others who had sprained an ankle, hit their heads or suffered shrapnel wounds in the scramble to escape the hail of bullets.

Only when the scene was deemed safe was the 34-year-old ER physician able to enter with a small SWAT team and a couple of police officers to perform the ritual of pronouncing people dead.

He pressed gloved fingers against victims’ necks — careful not to confuse the throb of his own adrenaline-fueled heartbeat with the reassuring whoosh he hoped to feel of blood rushing up carotid arteries.

Most had been killed where they had tried to take cover. But one man had staggered away from where he had been hit and was still conscious for much of the time that Colwell was saving lives outside. When Colwell reached William “Dave” Sanders, the 47-year-old business teacher and basketball coach at Columbine High School in Colorado had bled to death.

In the 17 years of mass shootings and stalled debates about gun control that separate Columbine from the recent massacre in Orlando, another debate has evolved among medical professionals and first responders about how to prevent deaths like Sanders’s. It hovers over the decision in Orlando to wait three hours after Omar Mateen began shooting before breaching the Pulse nightclub where he was holding hostages and where unknown numbers were wounded.

“Scoop and run” — the idea of moving victims to a trauma hospital as quickly as possible — is a mantra of modern U.S. emergency care. Internal hemorrhage can be handled only in an operating room. But stanching bleeding from arms and legs often needs to happen even sooner.

These days, law enforcement officers routinely enter active shooting scenes to stop the bloodshed rather than waiting for SWAT teams, as they did at Columbine. One of the first police officers to arrive in Orlando described being told by commanders to stay put rather than pursue Mateen into a bathroom; paramedics remained even further back, never coming within 100 yards of the building during the attack. The medical examiner’s office in Orlando has not revealed whether any of the dead at Pulse could have been saved if the hostage standoff had ended and they had received medical care sooner.

A controversial question is what role paramedics and even physicians play in those dangerous, pre-hospital settings, where minutes can make the difference between life and death.

“The philosophy about how to handle these mass shooting events is much in flux,” said David R. King, a trauma surgeon and active member of the Special Operations community at Massachusetts General Hospital.

Some advocate keeping medical teams away from the scene until it is declared safe. Others embrace creating an intermediate “orange zone,” where paramedics dressed in protective gear can provide initial treatment. Still others favor a far more aggressive approach, sending in tactical medical personnel who are trained, like King, to handle guns as well as tourniquets to stop bleeding and needles to decompress punctured lungs.

Nobody knows the answer, King says. There is no standard procedure recommended by the American College of Surgeons. And even if there were, he says, the decision of whether to send medical experts into an active “red zone” probably would rest with the mayor or police chief of whatever unsuspecting community had been assaulted, such as Orlando, by a shooter armed with a high-velocity weapon and many rounds of ammunition.

The severe loss of blood, or exsanguination, has long been a leading cause of death for troops on the battlefield. And it can happen quickly. A gunshot wound to a major vessel, such as the femoral artery in the thigh, causes blood to spurt out rapidly under pressure. Death takes less than five minutes. Slower bleeding also can be perilous, leading to a “bloody vicious cycle” when the blood loses its ability to clot.

From Columbine to San Bernardino, here's a look at some of the notable U.S. mass shootings since 1999. (Alice Li/The Washington Post)

One study found that 25 percent of battlefield deaths were “potentially survivable,” and that of those, 90 percent were bleed-outs.

Hence the reintroduction of the tourniquet, a device that has been around since Roman times but fell out of favor because of concerns that its prolonged use caused limb loss. Today, with swift evacuation times, the tourniquet is a lifesaver for troops. Every soldier carries one. Often two.

A report released June 17 by the National Academies said that up to 20 percent of U.S. trauma deaths could be prevented and urged the translation of wartime lessons to civilian systems, particularly in pre-hospital settings.

Tourniquets are available in many ambulances. And emergency treatment priorities have been flipped, says Eric Goralnick, an emergency medicine physician and former Navy officer at Brigham and Women’s Hospital in Boston. The ABCs of emergency care — airway, breathing and circulation — are now CAB, Goralnick says, putting the circulatory system first.

Less clear is who provides pre-hospital care — and where. Michael Neeki, an attending physician at Arrowhead Regional Medical Center in California who volunteers with a SWAT team, advocates “trying to introduce more [medical] capabilities into the field.” Neeki headed to the scene of the December shooting in San Bernardino, Calif., carrying a handgun, 10 tourniquets and sophisticated medical equipment to join the search for the shooters in the Inland Regional Center where 14 people died.

There is also a move to train members of the public. Immediately after the 2013 Boston Marathon bombing, King, the Massachusetts General surgeon, raced to the hospital to find victims arriving with tourniquets fashioned from belts and T-shirts, from shoelaces and pieces of rubber tubing. Most improvised tourniquets do not work well, he says. They cannot be pulled tight enough to prevent arterial bleeding and can cause damage by stopping the flow of blood back to the heart through veins.

Still, King emphasizes the critical role bystanders can play. He contributed to the 2015 Hartford Consensus, a call to action spurred by the shootings at Sandy Hook Elementary School in Newtown, Conn., where 20 children and six staff members were killed in December 2012.

A joint effort by physicians, nurses, law enforcement officials and emergency services, the consensus reclassifies bystanders as “immediate responders” and advocates for tourniquets to be made available — much like defibrillators — in shopping malls and museums, in schools and theaters, and in train stations and airports.

As a result, in October, the Obama administration began a campaign to educate the public on hemorrhage control.

It is called “Stop the Bleed.”

“The biggest lesson since Columbine has been the value of the tourniquet,” says Colwell, now chief of emergency medicine at Denver Health. As well as speed.

He has a “pretty good idea” that some wounds he saw that day in April 1999 “would have been survived under different circumstances.”

Seeking to offer something positive from his experience, Colwell began lecturing about what he learned at Columbine and then in 2012, when he was working in the ER after a gunman opened fire in a movie theater in Aurora, Colo. Similarly, Goralnick is in Europe to help identify best practices in after-attack treatment, by comparing strategies used in Boston, Brussels and Paris, where French doctors performed extensive care in makeshift facilities close to the site of the November terrorist attacks.

Those sessions are part of an increasingly urgent attempt to ensure that the medical response to mass attacks continues to improve as it has done in the 17 years since Columbine.

“We ended up waiting much longer to access the scene than we would today,” Colwell recalls. “But I don’t think that was necessarily a bad decision based on the knowledge we had then.”