Is Extraneous Imaging the New Normal In Trauma Radiology?

One of the hazards of living too long and working in the same field is that you eventually find yourself saying things like “Back in my day … .” Earlier in one’s career, there may be an initial vow to avoid that particular phrasing because you thought your elders sounded ridiculous when they said it back then.

It’s not too hard to imagine why. Starting out, we see the experienced ones who show us the ropes as authorities on how things are done. At that point, we have no other frame of reference. Subsequently, better methods might come along and occasionally a mentor turns out to be a complete turkey, but we regard most of our instruction as a gold standard.

However, as times change and entropy eventually has its way, deviating things from what we had learned to be best practices, it’s easy to have a negative bias and, usually having little or no ability to stem the tide, gripe about it.

Feel free to write today’s blog off as a mid-career “back in my day” rant.

I did my radiology residency between 2000 and 2004. It was at a “level 1 trauma center,” a designation that seemed to matter a lot more to some folks than others. From our perspective, that just meant that we got exposed to a lot of trauma cases. However, years later, it seemed to me that a lot of non-level-1 facilities for which I worked did just as much trauma imaging if not more.

We were near an intersection of two big interstates, which provided an endless supply of motor vehicle accidents. Our vast catchment area also included a number of ski slopes and plenty of rural territory for folks to wreck their ATVs and snowmobiles. We had a helipad on our roof for trauma copters, and routinely received transfers from smaller hospitals when they got patients with sufficiently severe injuries.

Imaging for these patients had a certain routine to it, guided by the clinical evaluation of ER staff and the trauma team. As a radiology resident, I had a certain sense of how bad a case was before looking at a single image, just from what had been ordered. An X-ray series covering the c-spine, chest, and pelvis was not necessarily a big deal. If abnormalities turned up, they might prompt additional XR views, or progression to targeted CT.

A head, c-spine, and chest/abdomen/pelvis CT right off the bat meant that the patient was pretty badly off, and virtually guaranteed to have a bunch of abnormalities. There could still be further imaging. For instance, if there was a spinal fracture, a dedicated CT or even MR of the relevant area might happen.

There was almost always a sense that things were being done for a reason although, depending on who was riding herd in the ER on a given day, the quality of reasoning might fluctuate.

Fast forward to the current day and that reasoning seems a lot harder to come by. Being in telerad, I can’t tell you which, if any, of the hospitals I cover is a “level 1” facility, but they all regularly send me trauma studies, and there is pretty much zero variability. Everybody gets a supine chest and pelvis XR, usually horribly limited, immediately followed by a head-through-pelvis CT. Further, they have the CT techs reformat the data to generate dedicated thoracic and lumbar spinal series for dictation as separate studies (before anybody’s even looked at the original scan to determine whether the spine looks abnormal).

They don’t bother with the c-spine XR anymore. I saw that evolve during the past 20 years at other facilities. Maybe the ER/trauma people decided that they were always proceeding to neck CTs anyway, so why bother with the extra step? I have to imagine the same thing will happen with the chest and pelvis XR sooner or later.

All of this would make sense to me in the case of sufficient injuries. If a patient comes in looking like he or she was in a war, you don’t want to dither.

That’s not what I’m seeing though. The vast majority of these pan scans are for patients who did things like fall out of a wheelchair or roll out of bed the wrong way, and they’re getting more comprehensive scanning than my multi-trauma cases of yesteryear. Back then, if that extent of imaging was ordered, you knew the images would be full of abnormalities. Now, the vast majority have zero acute findings.

The attending radiologists back in my residency knew at least some of what was coming down the road. They were only half-joking when they said that ERs would soon be using CT as a triaging tool. I don’t specifically recall them predicting that non-physicians would gain the ability to order these pan scans, but they probably foresaw that too. Why not? If you’re going to scan everything regardless of the patient’s clinical presentation, you might as well have a clerk write the order.

What’s the problem with all of this? Isn’t it better that everybody gets evaluated with the finest-toothed comb possible? The word has gotten horribly overused and invested with entirely too much sanctimonious power in recent years, but isn’t it safer this way? Don’t you want patients to be safe?

I’ve got a few issues with it, not the least of which being the extra radiation dosage. Remember when we tried to minimize exposure? There were reasons for that. However incremental it may be, it’s not nothing.

There is another factor that I think is being largely ignored. Radiologists are mere mortals. Crank up the number of negative studies you throw at them, and sooner or later, their index of suspicion regarding any given case is going to drop. It can’t help but instill a certain sense of “why are we doing this?” Then there is just plain old fatigue, both physical and mental. If you don’t believe that, you might want to think about just where the big, bad radiology shortage of recent years has come from.

This also doesn’t occur in a vacuum. Entities outside of the ER are far from oblivious to what’s going on. I’m kind of surprised that audits don’t happen when hospitals claim RVUs for dedicated spinal scans on every single negative C/A/P CT. If such audits don’t eventually turn up, CMS may “bake in” the expected cost by reducing the value of each scan component, an action likely to be swiftly followed by all other insurers. Heck, maybe they already have.