Saturday, March 23, 2013

MRCP: Stop Already.

A typical general surgery scenario plays out like this.  Patient comes into ER at 1AM with RUQ pain, gallstones on US, elevated WBC and mildly abnormal LFT's.  The patient is admitted, hydrated and started on antibiotics with the plan to re-assess in the morning.  I see my ICU patients, do a 7:30 case and then check the overnight admit's morning labs.  The WBC is now normal and the bilirubin is slightly improved (down to 1.4 from 1.9).  I go to see her but she's gone.  I query the nurse.  Nurse says she's off to MRI.  I sit down in patient's room, watch TV for a bit, perhaps eat one or two of the sausage links off the plate of her sleeping roommate.  No, I don't do that.  I just go drink some more sour free coffee and see other people, silly.

I tell the nurse to text page me when she gets back.  She does so.  I race back between cases.  The patient "feels much better".  Her abdominal exam is rather unremarkable.  We shoot the shit for a while re: biliary pathophysiology (involving crudely drawn anatomic diagrams on the back of patient satisfaction survey forms; I like the way I draw the stomach but my gallbladder/biliary ductal rendition looks rather like the way a kindergartner would draw equine procreation).  There's an opening in the afternoon.  She doesn't want to ever experience the sort of pain she had last night.  I commiserate.  I can imagine, I say,  although I can't, having never experienced a gallbladder attack myself so to reproduce empathy I imagine that one time I spilled lye on my leg working at a metal treatment plant in the summer and the lye ate through my jeans, my epidermis, the dermis and bits of the subcutaneous fat.   

In the OR I do a cholangiogram, AS PER USUAL.  The c-gram is completely fine.  No filling defects.  Patient grateful and happy.  She is discharged that evening.  As I type all the mindless useless discharge info into the EMR I notice the MRCP report has finally gone live.  The MRCP is normal.  I fall to my knees and thank the Lord.  I tell the nurses who eye me suspiciously that I am falling to the floor and praying in an ironic fashion, as a way to amuse my internal voice.  The patient never asks about it.  I muse fleetingly about who actually ordered it.

From the American Journal of Surgery is a retrospective review of the accuracy of MRCP when correlated with intraoperative cholangiogram findings.  The results are sad sad sad:


Four hundred twenty patients who underwent IOC were reviewed and met criteria for the study. Seventy patients had preoperative MRCP. Accuracy of MRCP when compared with IOC was 70%.


MRCP has a high rate of false normal results compared with IOC and is not as accurate as more invasive techniques. There is no need for preoperative MRCP in patients with suspected choledocholithiasis caused by stones.
MRCP (magnetic retrograde cholangiopancreatography)  is a costly imaging modality (although one would have no idea how much it costs due to pricing opacity and lack of published data---I spent 30 minutes googling "how much does an MRCP cost" without finding a reliable estimate, try it yourself).  I have found it to be one of the most overused studies in modern American medicine.  Typically a GI doc or a surgeon will order one when there is pre-op suspicion of choledocholithiasis (stone in the common bile duct).  The idea is that by getting the MRCP, and it is positive, then one can proceed to ERCP/stone extraction with a certain degree of confidence.  Similarly, a negative result gives the surgeon a mandate to proceed straight to the OR for lap chole.  The problem, as the above study shows, is that MRCP studies are highly unreliable.  I can't tell you how many times I have obtained intra-operative cholangiograms on patients with ostensibly normal bile ducts (based on a pre-op MRCP) who were found to have....stones in the CBD.

The problem is that not enough surgeons perform routine intra-operative cholangiography.   Those who don't  do it on every case will miss a certain percentage of CBD stones (overall incidence of 10-15% per the literature).  So they try to make up for it by getting these pre-op MRCP's in order to assuage their consciences or whatever. It's wasteful and useless.  Do a cholangiogram in the OR.  Every time.  You'll leave the OR feeling confident every single case.

I mean, if you feel strongly that the patient has a CBD stone (jaundiced, dilated CBD on US) then just do the damn ERCP.  If the liver tests are trending down and the patient is doing well clinically, it's likely he passed the stone.  So take him to the OR and do the lap chole/grams.  If the cholangiogram is positive, then you know for certain the ERCP will be efficacious.  It really ought to be a lot simpler than we make it out to be...


Paracelsus said...


Anonymous said...

What's your practice when you see a possible "incidental" stone in the common duct?

Uzer said...

What if they can't cannulate the duct post op? Back to OR? Has that happened much to you?

Jeffrey Parks MD said...

I'm not sure what you mean by incidental stone. If you mean an IOC that shows questionable filling defect, then follow patient clinically, check LFT's in AM.

ERCP failure usually occurs <1-5% of the time. I have yet to seen it. I have seen patients with multiple large CBD stones that were unable to be cleared after multiple ERCP's and I had to take back for open CBDE a few times.

Also, don't forget laparoscopic transcystic duct CBD stone clearance. Any time you have a seemingly wide caliber cystic duct, always consider sliding as fogarty balloon downstream. It's worked for me maybe a dozen times in my career.

Anonymous said...

This study was performed at our institution and argues for the selective use of IOC as compared to routine. It's retrospective data, but I think it was a well-done study and it actually caused some of our surgeons to change their practice patterns (from routine to selective use of IOC), mostly based on our surprisingly high percentage of false-positives on routine IOC. What do you think?

Rahul Singh said...

I have found MRCP to be most useful when ever we are suspecting a stone in the CBD. I must confess that I have not been used to IOC in routine cholecystectomies.
MRCP if done on a high resolution machine gives excellent idea about the presence/absence of CBD stone as well as the so called "road map".