Deepika Mohan, MD, did not set out to make a video game. She set out, first, to survive a brutal Sunday on call.

This was 2008. Mohan was a fellow in critical care and trauma training at UPMC Presbyterian in Pittsburgh. It was a warm spring day, the kind that sends half the city outside and the other half, somehow, into ambulances. The admissions kept coming. Because UPMC Presbyterian receives much of its trauma volume as transfers from other hospitals, Mohan spent the night fielding case after case from outside emergency departments. Then two patients arrived back to back, and the contrast between them lodged in her brain like a splinter.

The first was a young girl who had rolled her father's SUV. She was pan-scanned at an outside hospital and found to have no injuries, but because she was hysterical and tachycardic, the emergency physician felt uneasy and sent her by helicopter to Pittsburgh. Centralized image sharing was not routine back then, so the scans didn't come with her. She had to be imaged all over again. She still had no injuries. But by then, it was 2 a.m., and she was admitted for observation anyway, at enormous expense.

The second patient was a woman in her 70s who'd been rear-ended at low speed on her way to church. She went to a local emergency department with severe abdominal pain, but because the crash seemed minor, the physician kept her for imaging rather than transferring her quickly. Hours later, a fax came back from a radiology group saying there was free fluid in the abdomen and a hollow viscus injury could not be ruled out. By the time she finally reached Pittsburgh, Mohan said, she was hypotensive, peritonitic, and "basically ready to die."

One patient did not need the helicopter ride at all. The other needed help far sooner than she got it. That lopsided night eventually led Mohan to create Night Shift, a trauma triage video game designed to help emergency physicians spot severe injury in older adults before it slips past them. 

A randomized clinical trial, published April 20 in JAMA, found that physicians who trained with the game were less likely to under-triage severely injured older patients than their medical peers who received standard education.

The Morning After

The next morning in 2008, Mohan met with her research advisor, Matthew Rosengart, MD, expecting to talk about her plans for the year ahead. But she was still too worked up from the night before to focus on anything else. She walked him through the two cases that had been bothering her. One patient got an expensive, unnecessary transfer by helicopter. The other needed urgent transfer and got delay, uncertainty, and a fax machine. 

Rosengart didn't let the conversation end with venting. "He demanded that I use my frustration to inform my research agenda," Mohan said.

That question led Mohan to one of trauma care's most stubborn blind spots. Older adults with severe injuries are under-triaged far too often. Part of the problem, Mohan said, is that they often don't look especially sick, even when they are. She likes to make the point with a comparison that's hard to shake: An older person who falls and breaks four ribs can face a mortality risk similar to that of a young person shot in the liver.

Why do older patients get missed so often? Mohan says a lot of it comes down to heuristics, the mental shortcuts doctors rely on every day. In the emergency department, those shortcuts are just part of the job. Nobody has time to stop and carefully rethink every case from scratch. Physicians have to make quick calls, using the information they have in front of them, about who is truly sick, who can safely wait, and who needs help right away.

"Often, that allows us to get through the day," Mohan said. The trouble is that older patients don't always fit the pattern those shortcuts expect. "I fell in the bathroom" does not sound like "some dude just shot me while I was walking down the street," as she put it.

Age also muddies the usual physiologic cues. A patient with chronic hypertension may already be in shock at a blood pressure that looks acceptable on paper. A patient taking beta-blockers may not become tachycardic until the situation is nearly catastrophic. 

Mohan said doctors in community emergency departments may go a long time between truly severe trauma cases. And once those patients leave, the story usually leaves with them. The doctor who made the initial call may never know how things turned out, which makes it much harder to sharpen those instincts over time.

The Science of Decision

To think through that gap between knowledge and action, Mohan turned to Baruch Fischhoff, PhD, a Carnegie Mellon professor of engineering and public policy who studies judgment and decision-making. Fischhoff describes decision science as a field that compares two worlds: the ideal world where people have all the time and information they need, and the real one where they do not. The goal, he said, is to narrow that gap. 

He's quick to add that this doesn't mean emergency physicians are making reckless decisions. "Emergency department physicians are incredibly hardworking and well trained, making few mistakes under often harrowing circumstances," Fischhoff said. The challenge, he added, is to reduce those mistakes "without destabilizing what is basically working incredibly well."

That framing mattered to Mohan. She wanted to reinforce clinical judgment where it most often slips. By the time she was digging into the problem, she had also run into an uncomfortable fact. Standard continuing medical education was not doing much to change real-world behavior. If heuristics were driving nonadherence to trauma triage guidelines, then more lectures and reminders were unlikely to solve the problem. 

The issue was getting information to stick and surface at the right moment, not simply delivering more of it in fluorescent-lit rooms with weak coffee.

So Mohan went idea shopping outside medicine. She borrowed narrative engagement from preventive health, analogical encoding from organizational science, deliberate practice from expertise research and chess, and group-based interventions from the intelligence world. Then she found a Nature paper describing video game training that improved cognitive control in adults.

"And I thought, 'why not?'" Mohan said.

Enter the Night Shift

This was not the secret ambition of a lifelong gamer. Mohan describes herself as "a Luddite who is deeply skeptical about technology." She said she grew up before video games were common and added that her Indian parents wouldn't have been thrilled anyway. But she needed a way to reach physicians that was engaging, scalable, and consistent. A video game began to look like an unexpectedly practical delivery system.

The result was Night Shift, created in 2016 with Schell Games. The game puts players into the role of a young emergency physician making trauma triage decisions in a high-stakes, emotional setting. It uses story, quick puzzles, and feedback. Players make choices with limited information under time pressure, which is a fair summary of emergency medicine on its best and worst days.

From Fischhoff's perspective, the game is designed to strengthen how physicians already think. He said Mohan's intervention complements clinicians' ordinary heuristics with a kind of "meta-heuristic" running in the background and asking a simple question: Is this patient in unusual trouble? When that answer lights up yes, the clinician can stop and ask why.

Mohan is careful not to claim the mechanism is fully understood. "That's somewhat speculative," she said. But she suspects the game may help physicians absorb and retain information in a way standard classroom instruction often does not. 

"It's not that people cannot use what they learn in the classroom," she said. "The problem is that often, that informing doesn't make it from the working to long-term memory."

By using emotion, storytelling, and puzzles, she said, the team hoped to help clinicians make that leap.

The trial results suggest something meaningful may be happening. In the yearlong study, Mohan and colleagues followed 800 physicians staffing emergency departments at non-trauma centers across the United States. Half were assigned to play the game for two hours and then for 20 minutes every quarter. The other half received standard continuing education. 

The doctors who played Night Shift did better. Their under-triage rate for severely injured older adults was 49 percent, compared with 57 percent in the control group. And they weren't just transferring more people across the board. Over-triage stayed the same, which suggests the game may have helped them make better calls, not just more cautious ones.

Mohan also found that adherence to trauma triage guidelines was strongest within 30 days of gameplay, then faded until physicians played again. Quarterly exposure, she said, may not be the ideal dose. Shorter, more frequent refreshers, perhaps even a 90-second weekly microdose, may work better.

For Mohan, the broader lesson reaches beyond trauma and beyond games. Good training should sharpen clinicians' judgment and draw them more fully into the work of decision-making.

"Effective training makes the physician, or clinicians more generally, a part of a solution rather than bypassing them or excluding them from the decision-making," she said. "You need to engage the learner, allowing them to hone their judgment."

She sees obvious applications in stroke, sepsis, and chest pain. She also imagines similar methods being used in less cinematic but equally consequential situations, including dementia diagnosis in primary care and communication with families of critically ill patients.

"Lots of horizons to conquer," she said.

Disclosure information for study authors is available in the original study publication.