Bilious Vomiting at Five Weeks: Which Study, Whose Hands, and How Urgent
A 5-week-old term boy is brought to the emergency department at 03:00 with three episodes of bilious emesis over 12 hours and one episode of lethargy. He was breastfed and thriving until yesterday. Heart rate is 168, capillary refill 3 seconds, weight 4.3 kg. The abdomen is soft, non-distended, but mildly tender in the right upper quadrant with no palpable mass. Lactate is 2.4 mmol/L, white cell count 15.8 x 10^9/L, sodium 133 mmol/L and bicarbonate 18 mmol/L. Abdominal radiograph shows a dilated stomach with a paucity of distal bowel gas and no pneumatosis. The surgical registrar asks whether an ultrasound can replace an upper gastrointestinal series because the family is anxious about radiation, and whether a contrast enema would be preferable as a 'one and done' test. There is no on-site paediatric radiologist until 08:00; a general radiographer staffs the fluoroscopy suite overnight.
1. What is your diagnostic pathway and why? Address the role of ultrasound, the contrast enema, and the timing relative to the missing paediatric radiologist.
Read the worked answer
Bilious vomiting in a neonate is malrotation with midgut volvulus until excluded, and the bowel is at risk within hours. The pathway is therefore urgent and should not wait for daylight.
Why not ultrasound alone:
- A positive ultrasound (inverted superior mesenteric artery/superior mesenteric vein relationship, 'whirlpool' sign, dilated fluid-filled duodenum) is highly specific and useful, but sensitivity is operator-dependent and a normal study does not exclude malrotation. In an unstable or equivocal case it cannot be the only test.
- Ultrasound is excellent as a first look in a stable infant in a centre with expertise, and it avoids radiation, but tonight the on-call skill set is general.
Why not a contrast enema:
- Enema diagnoses and often treats intussusception, and shows a microcolon in distal atresia. A normally sited caecum does not exclude malrotation, and a normal enema is falsely reassuring. It is not the test for volvulus.
The right study:
- Upper gastrointestinal series with a nasogastric tube (not oral feeding in a vomiting infant, to avoid aspiration), low-volume water-soluble low-osmolar contrast, pulsed fluoroscopy with last-image hold, and images including a true lateral.
- Normal: duodenojejunal junction to the left of the pedicle, posterior, at or above the level of the duodenal bulb. Abnormal: corkscrew duodenum, inferiorly displaced junction, spiral passage of contrast.
Practicalities:
- Call the paediatric radiologist and the surgeon in simultaneously; a general radiographer can run the screening under telephone or teleradiology supervision.
- If the infant deteriorates, has peritonitis or becomes acidotic, proceed straight to theatre without imaging.
- Resuscitate with fluid first; bicarbonate of 18 and lactate of 2.4 indicate early compromise.
2. The family asks about the radiation risk. What dose are you quoting, how do you minimise it, and how do you explain it?
Read the worked answer
An upper gastrointestinal series in a 5-week-old delivers an effective dose of the order of 0.3-1 mSv depending on screening time and field size - roughly a few weeks to a couple of months of natural background radiation at sea level (about 3 mSv per year).
Dose reduction measures:
- Pulsed fluoroscopy at the lowest frame rate that is usable, with last-image hold rather than continuous screening.
- Tight collimation to the upper abdomen, no magnification, shortest possible screening time, and a single experienced operator at the pedal.
- Limit the number of overhead images; store and review rather than repeat.
- Do not use fluoroscopy to 'watch' for a few minutes; use timed spot images after small aliquots of contrast.
How to frame it for the family:
- The risk of a missed volvulus is catastrophic and time-critical: ischaemic bowel can develop within hours, with short bowel syndrome or death. This is not a risk that can be traded against a sub-millisievert radiation exposure.
- There is no meaningful increased cancer risk demonstrated at these doses; the linear no-threshold model is used for prudence in health physics, but the individual absolute risk at under 1 mSv is very small.
- Explain why an enema is not a substitute and why an ultrasound alone is not sufficient, in plain language, and document the conversation.
3. The upper GI series shows the duodenojejunal junction positioned to the left of the spine at the level of the bulb, but the duodenum does not cross the midline and contrast reaches the jejunum slowly. What do you do?
Read the worked answer
A junction that is not clearly left of the pedicle and at or above the bulb, with a duodenum that fails to cross the midline, is abnormal and should be treated as nonrotation/malrotation until proven otherwise - do not discharge on the basis of a partially reassuring film.
Interpretation points:
- The duodenum should sweep to the right of the spine, cross the midline to the left, and reach the junction in a fixed retroperitoneal position. Failure to cross is a recognised sign of malrotation.
- Slow transit is nonspecific (hypomotility, dehydration, obstruction) but in this clinical setting raises the question of intermittent volvulus or a coexisting duodenal web or stenosis.
- Obtain delayed images at 10-20 minutes; review the true lateral to place the junction posteriorly; consider repeating after nasogastric decompression.
- Remember the false-negative rate of upper GI series is not zero (reported around 3-5% even in expert hands), and ultrasound is a useful adjunct rather than a rescue test.
Action:
- Discuss immediately with the paediatric surgeon and proceed to diagnostic laparoscopy or laparotomy based on clinical status; a normal-looking junction on one view does not justify observation in a symptomatic infant.
- At surgery, Ladd procedure if malrotation is confirmed; if not, look for a web, annular pancreas or atresia.
- Document the findings and the reasoning, and keep the family informed of the possibility of a normal operative finding.