Fresh from a look at different classic, maybe it’s time for a refresher on olive-stomach children. Dr Andrew Weatherall wanders through some papers.
The kid with pyloric stenosis presents an exam question classic. Probably because everyone should be able to reminder a small child with an olive-type thing somewhere in the epigastrium. Or maybe it’s because the presentation of a small child casually projectile vomiting and then looking like they want more food is just creepy enough when you read it in a book as a student that the concept sticks.
Pyloric stenosis is certainly a classic though. And not just for those clinical signs. It’s a good one for wrestling with real world options: now or when later? RSI every time? Well maybe except for when the answer is no.
So having spent the last post saying “RSI isn’t really the old RSI”, maybe it’s time to check in with a few thoughts on perioperative stuff for kids with gut nubbins.
The Medical Bit
This is not the post to go through the polygenic mode of inheritance and modifying environmental factors. Or the 4-5 times higher incidence in males. Or spend more time on the progressively worsening projectile vomiting picture.
And it’s probably not the post to mention that the classical “olive” in the upper abdomen which is actually only present in 48% of patients at admission.
I don’t want to avoid the chance to snuggle deeply within the warm blanket offered by a cliché though so here’s the one I’ll make sure we hit: fixing the patient with pyloromyotomy is not a surgical emergency, it’s a medical one.
There. Felt good, right?
The Medical Bit
Here’s the quick version:
- Vomiting happens.
- The stuff that is vomited doesn’t have the alkaline secretions from the small intestine because there is that whole blockage at the gastric outlet thing happening.
- The vomit is all gastric juices with hydrogen and chloride ions (and a bit of sodium and potassium thrown in).
- The patient can therefore end up dehydrated and with a metabolic alkalosis marked by low potassium and low chloride.
So let’s assume that the patient turns up with a degree of dehydration and the classic picture of hypochloraemia, hypokalaemia and metabolic alkalosis. Yeah we’ll assume that even though that picture is probably a little less usual now that diagnosis with ultrasound has made pick-up a bit quicker. The review article I found most useful when looking back over this topic quotes rates in recent reviews of 25% hypochloraemia (and < 10% for severe hypochloraemia with levels < 85 micromol/L).
There’s a theme around the things you read and that’d be “don’t rush”. The aim is correction of dehydration and getting the electrolytes back into the acceptable range, with a general agreement that a serum chloride over 100 is the magic marker (along with a patient who is looking rehydrated and passing urine).
There are a few different approaches out there though. Miozzari et al looked at 139 patients and described a relationship between chloride administration and changes in circulating bicarbonate. They indicate that that administering around 10 mmol/kg in their series reduced plasma bicarbonate around 3 mmol and they were pretty happy with that. So as Hartmann’s solution has has a bit less chloride than “normal” saline (10.9 mmol/100 mL vs 15.4 mmol/100 mL) you need a little more Hartmann’s to produce this. They came up with 90-100 mL/kg of Hartmann’s to correct the alkalosis vs 60-70 mL/kg of 0.9% NaCl (not all as one push).
More recently Dalton et al came up with a slightly different approach. In their retrospective look at 505 patients, 202 of the patients had electrolyte abnormalities that required fluid resuscitation, which sort of highlights that the days of the shocked small person at presentation are maybe mostly in the past. Of those 202 patients, 64% of them required multiple boluses of 20 mL/kg (they were using normal saline and their aim for acceptable normalisation of electrolytes was a chloride > 100, bicarbonate < 30 and potassium between 3.4 and 5.2).
After their study they also made a plan to change practice. Patients with chloride under 97 will now apparently get an initial bolus of 40 mL/kg of normal saline for initial resuscitation and a continuous fluid rate at 1.5 times maintenance (for which they’re using 0.45% saline + 5% dextrose). In severe hypochloraemia (< 85 mmol/L) they’ll now kick off with 60 mL/kg in 3 divided doses (plus the maintenance bit).
But what about if the chloride is normal? Then they’ll look at bicarbonate and the boluses of 40 mL/kg and 60 mL/kg would be used if values hit 33 mmol/L or 40 mmol/L for the bigger bolus.
The other note here is that they actually give the boluses in divided doses. So that’s 20 mL/kg, wait an hour, the next 20 mL/kg then wait… well you get the idea.
The obvious next question then is what’s the trigger to say “we’re there”?
Well the patient should be rehydrated. There’s no need to rush and it’s safer to know your patient is hydrated and passing urine. On biochemistry the papers I read all seem in agreement that a chloride above 100 mmol/L is the aim of the game, along with a bicarbonate under 30 mmol/L.
I have read one earlier review suggesting that an even better way of being sure that the ECF has been restored along with electrolyte balance being achieved is actually to check urinary chloride and aim for > 20 mmol/L. If you went down this road you’ve have the advantage of not requiring multiple blood tests. The amount of times I’ve seen it done? Zero.
Emptying that Stomach
So let’s assume the big moment has come. But we want to be sure that really empty. But these kids all have a gastric tube in there so surely there’s nothing left to drain, right? Well Cook-Sather and colleagues have explored just this in 72 kids. In kids who had undergone preoperative suction of a gastric tube, the tube itself was removed and aspirated. After induction they passed a 14 French orograstric catheter and aspiration was repeatd. On average they got out 4.8 mL/kg (though it came with a +/- 4.3 mL/kg). The average volume of gastric contents in fasting kids is 0.4 +/- 0.45 mL/kg.
That is a big difference right there and led them to suggest aspiration with a 14 gauge feeding tube before going to sleep, regardless of what you’d done before. Seem over the top? Well 5 mL/kg of residual gastric fluid is also pretty over the top.
To sleep but how?
Well I’ve always been taught that this is a pretty fair indication for RSII, while being very mindful of the issues of doing that in a neonate, particularly the risk of airway distortion or compression with any form of cricoid pressure
But generally they come with a drip. Use the drip. But do that RSII with a bit of common sense applied. Hmmm… if only someone had written a thing about RSII recently…
There are bits in the reviews that touch on the older belief that it might be a good idea to try out awake intubation. No. Nope, nope, nopity nope.
This is only likely to make life difficult. More airway injuries, lower success rates with intubation, longer to intubate, and similar other complication rates.
Plus they’re awake. Sheesh.
Now this assumes that the cannula works. What about the patient who turns up with a cannula that just can’t cut it anymore? That might be a more interesting chat.
Scrimgeour and crew reported on the alternate approach – inhalational induction. They looked back from 2005 to 2012 at 269 patients, of whom 252 had inhalational induction. They aspirated the nasogastric that was in place and went ahead and kicked off with inhalational induction with sevoflurane, nitrous oxide and oxygen. And lo and behold they had no cases of aspiration and declared it safe.
Except of course reporting that something is safe because you didn’t observe it in a series with less numbers than the known incidence rate is pretty gutsy. And the accompanying editorial made that point too. It even used words like “perplexed”, “counterintuitive”, “inadvisable” and … well, it is pretty strong stuff for a journal about gassing little babies.
It makes you wonder if the paper they published doesn’t really cover how confident they are in the technique because they’ve done it for much longer and really, really know it works out. Of course you don’t get much backstory in a journal.
I guess what this does suggest is that if you are denied a chance to use an intravenous route then inhalational induction is still out there and it is most likely you will get there fine, as long as you go about it carefully. You could always place an intraosseous needle but this isn’t a practice I’ve seen that much.
Or not to sleep?
There is an alternative here and that would be a regional approach. That review article by Kamata et al goes a little into this with some small case series and it’s sort of an intriguing read because I have no direct experience of this. Kachko et al have a total series of 60 kids and 24 had spinal anaesthesia for their open operation, with patients given 0.7-0.8 mg/kg of 0.5% isobaric bupivacaine (the plain stuff). A quarter of the patients needed a dose of IV midazolam because of restlessness but overall their time in the operating room and the listed emergence time were shorter.
Somri et al had 23 patients having spinal anaesthesia in their series and 22% of the kids (5 of the cohort) needed intraoperative doses of propofol while in two of the children LP was unsuccessful.
Of course you might say that as more and more surgeons go for a laparoscopic approach (because of factors like shorter duration of stay), spinal isn’t really an option. Except that Islam et al have actually reported on a comparison of kids having spinal or GA for laparoscopic procedures. Time from end of procedure to out of theatre was shorter, though 25% needed conversion to GA.
There are also reports of using single shot epidurals for the procedure or a big dose via the caudal space. The success rates look pretty impressive although the doses of bupivacaine makes my eyes widen a little (3.6 mg/kg and 4 mg/kg in the two series).
I honestly wouldn’t have thought of choosing a regional approach but maybe I was being too narrow-minded. That said, if a technique doesn’t quite get there in one-fifth to a quarter of patients meaning I don’t spare them exposure to other agents (if avoiding the potential for any form of neurotoxicity is part of the game), and if the non-spinal techniques require doses of local anaesthetic approaching toxicity then I’m not convinced I want to change things up.
The Bit In The Middle
That’d be maintenance. Classically these kids are described as having a risk of apnoea because the metabolic alkalosis induces changes in cerebral pH that make patients less likely to respond to the usual CO2 triggers for respiration. Waiting until bicarbonate is corrected should help reduce this problem, though the CSF changes lag behind serum correction.
If using GA though it would appear to make sense to use short-acting things, particularly if the kids are presenting younger and still in the risk phase for postoperative apnoea. So my personal practice is to use remifentanil during the case and minimal amounts of desflurane as a volatile agent (one of the very few instances where I’d reach for this).
Analgesia afterwards just doesn’t seem like a huge problem with either an open or laparoscopic technique if the surgeons are able to add some local anaesthetic to the picture, so why use things that linger?
The Pain of It All
Which brings us to the pain management element. Most of these patients do perfectly well with local infiltration plus simple nonopioid options postoperatively. This doesn’t mean people haven’t tried a variety of things to make good even better.
There are reports of caudals, ultrasound-guided rectus sheath block and paravertebral blocks all out there. And while I personally love a good regional technique I can’t quite figure out what I’d be adding for my patients who seem to do perfectly well without me spending the extra time in the operating theatre performing ultrasound-guided blocks either before or after their half-hour operation.
So after going through all of this stuff, what’s a practical list of things to consider when presented with one of these kids needing an anaesthetic?
There’s no rush. Wait until they’ve had their fluid replacement and their electrolytes are back in the acceptable range (both on the chloride and bicarbonate side).
- Empty that stomach
The NG they have isn’t enough. I do use that big gastric tube before starting the anaesthetic and if you haven’t tried it you might be surprised to know it’s really well tolerated. And I’d go with 4 quadrants as part of the deal. Left, right, back and front. I was shown all of that as a skeptical trainee and saw enough come up when they were allegedly “empty” that I was turned around.
- Use that cannula
I do use that cannula. I mean why wouldn’t you? And I do treat it as a controlled RSII with a low threshold for relaxing on the cricoid pressure. That said, if the cannula fails and access is difficult inhalational induction is there as an option.
- Use short-acting things
For the case itself there’s a role for opioids but the afterwards doesn’t involve a high analgesic requirement. So my personal preference is to use remifentanil with volatile maintenance with either sevoflurane or desflurane.
- Simple analgesia
Yeah. There a lot of interesting blocks you could do. I’m just not sure you’d ever know if you were really making a difference that matters to the patient. In which case. Why?
And that’s a, well, not-so-quick story about the olive-stomach children. I thought it would be simpler. I guess it’s a pretty good example of how much you can pack into a 30 minute case.
Actually doing a regional thing here isn’t as cool as a porcupine in a tree.
The images here are from flickr.com Creative Commons and come from Sean Davis (Zero) and the USFWS-Mountain Prairie (the porcupine).
You could always consider adding clever thoughts in the comments. Or signing up to get emails when posts hit.
Now, the useful papers.
Here’s a link to the good review thing:
Here’s the first thing on fluids:
And then the other one:
Here’s the review article that mentions the urinary chloride thing.
Here’s that thing on gastric fluid volumes:
Here’s the thing on inhalational induction:
With the editorial:
Now the thing on spinals:
And the other one:
Now the one with laparoscopic pyloromyotomy and spinals:
Interested in the single shot epidural anaesthesia thing?
Willschke H, Machata AM, Rebhandi W, et al. Management of hypertrophic pylorus stenosis with ultrasound guided single shot epidural anaesthesia – a retrospective analysis of 20 cases. Pediatr Anesth. 2011;21:110-5.
Or a big caudal?
Moyao-García D, Garza-Leyva M, Velásquez-Armenta EY, Nava-Ocampo AA. Caudal block with 4 mg/kg (1.6 mL/kg) of bupivacaine 0.25% in children undergoing surgical correction of congenital pyloric stenosis. Pediatr Anesth. 2002;12:404-10.
What about looking at rectus sheath block…
And lastly, maybe a paravertebral discussion…
Mata-Gómez J, Guerrero-Domínguez R, García-Santigosa M, Ontanilla A. Ultrasound-guided paravertebral block for pyloromyotomy in 3 neonates with congenital hypertrophic pyloric stenosis. Braz J Anesthesiol. 2015;65:302-5.
Did you get this far? Wow. Maybe you should take a few moments to enjoy things underwater in a feeding frenzy then. It comes with bonus Attenborough baked right in.