Biliary Dyskinesia at a Borderline HIDA Result: Trial of Therapy or Straight to Surgery
A young patient with a borderline gallbladder ejection fraction but pain reproduced during the scan itself, and whether a time-limited medical trial or proceeding straight to cholecystectomy better fits what the test actually showed.
Trevor N., 27, works full standing shifts as a line cook, and has spent six months managing recurrent pain under his right ribs that shows up predictably after a big meal — bad enough now that he's started skipping the family-style dishes at work entirely rather than risk it during a shift. Ultrasound is unremarkable: no gallstones, normal gallbladder wall. A HIDA scan with cholecystokinin stimulation, ordered to look for functional gallbladder disease once stones were ruled out, returned a gallbladder ejection fraction of 38 percent — a number that sits inside a genuinely disputed gray zone in the literature, with different centers drawing the line for "abnormal" anywhere from 35 to 40 percent. What happened during the scan itself may matter more than the percentage: the infusion reproduced his exact, familiar pain.
Whether that reproduction predicts anything is itself contested, and the weight of the evidence has moved against it. Morris-Stiff's series is the one usually cited for the claim, reporting symptom recreation as the better predictor of medium-term outcome; larger and later work has not reproduced it. Edwards and colleagues found that neither pain provocation nor a low ejection fraction predicted symptom relief after cholecystectomy, Dave's hundred-patient series found reproduction of pain on CCK injection not significantly predictive of post-operative outcome, and at least one major payer now classifies pain provocation as unproven for precisely this purpose. So Trevor arrives with two findings rather than one signal and one distraction — and what neither of them is, on this evidence, is decisive. Trevor, for his part, has said he'd rather try something short before committing to an operation — as long as "short" actually means something specific, not an open-ended wait through more missed shifts. He's otherwise entirely well — no weight loss, no fevers, normal appetite between episodes — which is itself part of why the team has been careful not to treat this as anything more urgent than it actually is, even as his pain has genuinely started shaping how he works his shifts.
Reviewing the HIDA result together
His ejection fraction, 38 percent, sits inside the range plenty of centers treat as genuinely ambiguous rather than clearly abnormal — the traditional 35 percent cutoff isn't a bright line the literature actually agrees on. Before offering a 27-year-old surgery, a reversible medical trial — antispasmodic plus a structured low-fat diet, four to six weeks — is the lower-risk first step, and if it fails, nothing about that delay makes the surgical option any less available.
The thing that actually moved during his study wasn't the ejection fraction — it was his own pain, reproduced under CCK. Morris-Stiff put that ahead of ejection fraction at predicting medium-term relief, and it matches what he sees in clinic: the patients who recognize the pain during the scan are the ones who come back grateful.
Calling the ejection fraction "the ambiguous part" still treats a number the literature can't agree a cutoff for as the variable that decides this, when the thing that actually happened during the study is at least as informative.
Both of those findings are real, and neither one cancels the other out. The surgeon is leaning on Morris-Stiff, and that series is real — but Edwards and Dave went looking for the same signal in larger cohorts and didn't find it, so "CCK-reproduced pain predicts surgical benefit" is a live hypothesis rather than an established one, and a live hypothesis can't carry an irreversible operation in a 27-year-old by itself. What it can do is justify not dismissing him, which is a different and much smaller claim. A borderline ejection fraction with no medical trial attempted also has real value in confirming this pain is actually biliary before an operation whose relief rate, even in well-selected patients, isn't universal. A short, explicitly time-limited medical trial — three to four weeks, not open-ended — with cholecystectomy already scheduled as the next step if it doesn't meaningfully help, respects both signals instead of picking one and discarding the other. The condition for that to actually work is naming today, in the chart, what "the trial didn't work" means, so it doesn't quietly turn into six more months of the same pain.
Agreed to the three-to-four-week medical trial with an explicit written definition of failure (any return of his typical postprandial pain at the same or greater intensity) and cholecystectomy already scheduled provisionally for the following week if the trial doesn't hold.
Not agreed: the surgeon wanted the trial window shorter, two weeks rather than three to four, given how demanding Trevor's job is on his feet and how much a few more weeks of pain costs someone working full kitchen shifts; the gastroenterologist thought two weeks was too short to fairly judge a diet-and-antispasmodic trial. Split the difference at three weeks with Trevor's own agreement, rather than either physician's original number.