Read a CT Like a Surgeon

Cut to the Chase: General Surgery Survival Series — Post 19

Two in the morning. The scan you ordered three hours ago finally loads, and the ED resident is at the desk asking what the plan is.

Your senior is asleep, and the rule nobody wrote down is that you keep them that way unless you have a reason. You scroll. You have a vague sense of what's in there, and not the answer this moment is demanding.

You think the problem is that you don't know enough radiology. It isn't.

The radiologist's read is sitting right there, a click away, and you could relay it up the chain without opening a single image yourself. But you're the one awake with this patient, and relaying is not the same as deciding.

A radiologist reads a CT to describe everything in it. You read it to answer one thing: does this person need an operation, and if so, does it happen now or can it wait until morning?

Open the images before you open the report. Ten minutes of scrolling with no idea what you're looking at is how you learn to see, and nobody is watching you do it at two in the morning.

You spent the last two years learning to read a scan the way a radiologist does, organ by organ, accounting for all of it. The job changed the day they handed you the pager, and nobody stopped to mention it.

So you stand at the workstation trying to out-radiologist the radiologist, and you see nothing at all.

That smaller question only has a few places to hide, which is exactly why you can train your eye to go there. You do it by reading every scan the same way, in the same order, so the routine carries you on the night you're too tired to think.

Before you scroll a single image, fix the clinical question in your head. A 20-year-old with right lower quadrant pain and fever: you're looking for free air, fat stranding, a dilated appendix, a fluid collection.

Then pull up the images and read them before you read the impression or relay anything up the chain.

And scroll the whole stack. The finding almost never lives on one slice, and a hundred of them are what a study is.

Then drop into the axial cuts and start at the top, every time. Free air rises, so the space under the diaphragm is the first place your eye lands. Pull up the lung windows while you're there, because a few bubbles of free air vanish into the gray on soft tissue and jump out in black against lung.

Free fluid does the opposite. It sinks into the pelvis and the paracolic gutters, so that's where you finish.

Free air rises and free fluid sinks. Hit the diaphragm and the pelvis first on every study, and you'll have the two findings that change tonight before you've looked at anything else.

Between those two points, you're hunting a short list. A loop of bowel that's dilated and full and then abruptly collapses to nothing. That flat segment is a transition point, and it's pointing straight at an obstruction.

Look at the fat. Clean fat is black. Inflamed fat goes hazy and gray, and that haze is often the first thing on the study that tells you where the trouble is.

And the finding that should make your stomach drop, because it has a clock on it: gas in the wall of the bowel, or air tracking back into the portal vein. That can mean the bowel is dying, and dying bowel doesn't wait for morning rounds.

The axial slices chop the bowel into a hundred little circles. The coronal reformats lay it out the way your eye actually wants to follow it.

When you're chasing an obstruction, switch to the coronals. If you still can't find where the caliber changes, say that out loud when you call, because not finding it is a finding.

And remember what the scan will never tell you, which is what to do about it.

A scan can read unremarkable on someone who is genuinely sick. Early ischemia hides, and so does early perforation. The clean scan on the patient you can't stop thinking about is the one you wake your senior up for anyway.

A clean scan on a patient who looks sick buys you nothing. Go back and put your hands on her again, then call anyway, and say the scan was clean while you're saying it.

What you want to be able to say is this: “I pulled it up. There's free air under the right hemidiaphragm and some fluid in the pelvis. I think it's a perforation.” Three sentences, and the person on the other end knows you were standing at the workstation.

Chief year I called my attending at two with the read off the report: partial small bowel obstruction, no transition point. I had not opened the images.

He came in anyway to see the patient, sat down at the workstation while I was still talking, and scrolled to the transition point in about a minute. It was a closed loop. He was scrubbed before I was.

The scan isn't a test you pass by naming everything on it. It's a question. Operate or wait, now or morning.

Some night this month a study will load at two in the morning and you'll go to the diaphragm first without deciding to.

Most of the time there will be nothing there, and you'll move on to the pelvis, and the ED resident at the desk will still be waiting on you.

---

What to Read Next: Post 20, “Where Do I Start With Abdominal Pain?” The scan is one input. When the patient is in front of you and nothing's been ordered yet, the question is where to even begin.