They brought a CT scanner to Oktoberfest. Here's what happened (and what the researchers missed).

Beer. Head and neck trauma. They kind of go together. I could tell you some stories from the ER. Too many, in fact.

Most years, Oktoberfest attracts around 390,000 people per day in Munich, Germany. The official event lasts over two weeks and takes place at a specific place called Theresienwiese. Who knew? This is important to us, though, because it means that, like other large events, a medical unit is deployed. (Event medicine is its own sub-field, by the way.)

Because alcohol consumption is obviously correlated with head and neck injuries, local hospitals there have come to expect an uptick in ER visits during the event. The increased demand for trauma services—most of which are mild injuries—can stress the healthcare system.

So, researchers decided to see what would happen if instead of bringing injured Oktoberfest patients to a CT scanner in a hospital, they brought a mobile CT scanner to Oktoberfest and plunked it down right there in the medical unit. The idea was that on-site imaging capability might decrease the need to send wounded attendees to the hospital by ambulance, thereby saving the system money and decreasing the use of limited medical resources. The goal was to rule out internal bleeding or a serious spine injuries right there on the scene, and then release the patients back “into the wild.” The results of this “experiment” were just published in the New England Journal of Medicine. Was this innovation a success? Depends who you ask. Because the researchers apparently overlooked a massive problem in the data. Can you guess the problem? Hint: it is sort of pandemic related, but sort of not.

Meanwhile, the study—good or bad—reminded me of a long-running but fascinating debate in the emergency medical field…

Let’s start with the research.

Here’s what the researchers think happened, as reported in the New England Journal of Medicine this week: Last year (2022), when CT scans were done right there at Oktoberfest, the number of hospital admissions was 14%-18% lower than in previous years (2015-2019) at the same point in the festival. Based on this, the researchers concluded that their idea worked and that a mobile CT scanner achieved the goal of decreasing local hospital admissions.

Here’s why we have no idea if the CT scanner had anything to do with it: The researchers used data from Oktoberfest 2015-2019 as the baseline for comparison. They omitted 2020-2021 because the event was cancelled both years due to the Covid pandemic, which makes sense. So far, so good.

But here’s the problem—and it’s a biggie.

The researchers provided no information about Oktoberfest attendance in 2022. It appears that attendance in 2022 was around 10% lower than it was in 2018-2019. That 10% decrease could account for the either higher or lower proportion of related trauma (or the same, I suppose), depending on the nature of that decrease. If decreased attendance was driven by fewer out-of-towners—people who traveled longer distances—the effect on traumas might have been higher than 10%, as those revelers might have been more likely to party a bit harder than locals for whom the event was less of a destination. So, a 10% decrease in attendance could have corresponded to, say, a 20% decrease in problematic drinking leading to head and neck trauma. Or the opposite could be true. The point is, we just don’t know because those data were not collected or presented.

Meanwhile, the cost of adding a mobile CT unit (and the clinicians and radiologists needed to run it) were not mentioned in the study, so we have no idea whether this was cost saving. We don’t even know whether fewer CT scans were done overall—that is, did the 205 CT scans conducted at Oktoberfest coincide with a decrease in similar scans at local hospitals? Or, paradoxically, did more scans happen because the scanner was right there just waiting to be used? Again, we are not given this information. It’s quite possible that the mobile CT unit led to more scans overall—scans which were not really necessary, despite the team’s efforts to use evidence-based algorithms. Doing more scans is not always good. There’s the cost, the small but real amount of radiation exposure, and the chance of false positive findings leading to “medical misadventure.” In my experience, if you hand a carpenter a hammer, they find a nail to hit. Sometimes, you wish they hadn’t.

Now, it’s still possible that the answers to these unknowns might still favor the mobile CT unit. But literally the opposite could be true. Given how many unknowns there are in this study, it strikes me as rather odd that the top medical journal in the world accepted this manuscript. (Sorry, my German Kollegen!). But, at the end of the day, the decision editors at major journals are just like the rest of us. “Ooh, they brought a CT scanner to Oktoberfest? Cool!”

An old debate in emergency care.

This topic is an iteration of an oft-discussed question in emergency care: Where should early care occur?

Imagine your parent or grandparent has a serious medical emergency. You call 911 and EMS shows up in an ambulance. Do you want the first responders to do a longer comprehensive evaluation and provide treatments on the scene? Or do you want the rescue workers to do the bare minimum—make sure your relative is alive and not about to deteriorate—and just rush them to the nearest appropriate hospital where all the specialists and fancy medical equipment are?

This debate boils down to a catchphrase: “Stay and play? Or scoop and run?”

There’s no universal right answer. It just depends. EMS workers can do a lot at the scene of an accident. Sometimes, the “stay and play” approach is the right one. Sudden cardiac arrest might be one of them. But there are some problems that require the human and material resources of a hospital. In those cases, “scoop and run" is the move. In reality, it just depends on the scenario, and factors ranging from the skills and equipment a particular EMS team has, to the distance to the nearest hospital (and how much better those resources happen to be)—and of course, the patient’s specific problem and present condition.

There are some obvious public health interventions that are offshoots of “stay and play” which make sense. Having an automated external defibrillator in airports, stadiums, and train stations is one example—though you need a lot of them to make it work, otherwise the devices are likely to be too remote to be accessible when needed.

Over time, researchers have tried moving hospital resources into ambulances: mobile stroke units to diagnose and treat strokes before reaching the hospital, intravenous antibiotics to treat suspected sepsis. The “earlier is better” hypothesis seems good in theory, but has not worked out as hoped—patient outcomes have either not changed much, or not at all—which is not good considering the expense and resources required to run these programs.

Research in this area tends to overplay the benefits of new toys—CT scanners that fit in the back of an ambulance—by following outcomes that are not “patient-centered.” Who cares if medications are given to patients 20 minutes sooner on average if the eventual outcomes are literally the same?

That’s why whenever I read studies like the Oktoberfest CT study (or the pre-hospital stroke and sepsis efforts), I’m careful to separate the outcomes along these lines. Do the interventions help patients in meaningful ways that they themselves care about? Do they save time and money? If the answer is yes to all three, you’ve got something. If not, you have to zoom out and ask whether the juice is worth the squeeze. Most of the time, it isn’t.

What are your questions and comments? Leave them below!