The Idiosyncratic Classics

There are some clinical scenarios which can be considered classics of paediatric anaesthesia. Here’s Dr Andrew Weatherall with a little on tracheo-oesophageal fistula repair and a few things to consider when getting on with one. 

I distinctly remember being told that air goes in and out. I remember the blood goes round and round bit. I do not remember anything about bonus stomach joins being good. Nope, oxygen was supposed to be good. Except it might not be. But maybe it is. Well this got confusing quickly.

Now if I had my choice both in life and in work matters, I would generally prefer to keep my breathing bits pretty well separated from my gut bits. Well at least at the level below the head and neck. It seems inevitably close for the other bits.

It’s a situation of unexpected joined up bits that brings us one of the classic sorts of kids’ operations – tracheo-oesophageal fistula (yes we’re into the version with all the “O’s” down here – file that under the Coriolis effect for weird Antipodean things if you’re from North America I guess or just bask in the fact that this is the correct and right way to do it if you’re from other places).

The thing you might note when you ask about things that don’t happen very much is that they are often a spot where you find a fair bit of variance of practice. Where there’s no clear best way to do things and the literature is at most reports of a variety of ways to do things practitioners tend to combine things they’ve been shown, things they’ve seen happen or heard a story about and their own version of logic to come up with the best way forward.

I’ve seen or discussed a variety of approaches in my not-particularly-long career and most times I can see why people have ended up with the approach they describe. So I’ve seen cases start with rigid airway endoscopy and I’ve seen people just get on with it. I know anaesthetists who like to have a flexible bronchoscope to look at the airway (I’m one of them). Most people I’ve spoken with prefer not to give muscle relaxation before the fistula is secured but I’ve also heard of cases where the best option seemed to be to add muscle relaxation to produce conditions for the surgeon to get to it quicker.

One of these cases came up recently and as I did some planning and I thought it was probably time to have a re-look at what I tend to do and see how it matches up with the stuff described in the literature. Is there anything out there that can produce some sort of consensus? Then after achieving consensus amongst anaesthetists maybe we can turn to herding servals.

What are we talking about?

As a quick refresher (only for myself of course, all the clever people who can remember this consistently can skip ahead) there are numerous types of TOF (yes in the abbreviation we still use the “O’s”) but overwhelmingly a distal fistula is the one you’ll see.


Fine, just for this picture we’ll use the “E” version of the abbreviation.

The challenge for anaesthetists is pretty obvious where there is a continuity between the trachea and the lower oesophageal bit. That join heads to the stomach so the distal airways  have a good chance of getting visited by acidic secretions. Joins between the upper GIT and trachea have their own issues but the distal one also complicates the ventilation bit of what we do.

Add positive pressure to the trachea and the concern is that you’ll inflate the stomach. Lazy old air. Always trying to find the easiest way out of a tight spot. A lot of the decisions revolve about how to deal with that problem. The question I had when I went back to have a look was “are most people doing this one thing and where do I fit in?”

What the others do?

Alright not all others. The others described in this paper will have to do. The authors took the trouble to look back through the records of a 3 year period across four tertiary hospitals in Australia and New Zealand, picking up 106 cases with records to try and extract some sense from. They looked at a number of items to see how people went about their work and also looked out for ventilatory difficulty (defined by specific comments in the chart, severe gastric distension or the presence of a significant desaturation under 90% or a drop in saturation of 10% from baseline).

In the demographic breakdown they note a few things:

  • The mean birth weight was listed as 2557 g but that’s probably skewed by the tiny ones because 61.3% were over 2500 g.
  • 42% of the patients were premature.
  • A distal TOF was the most common version (this was the anatomy in 85.6%).
  • Overall the average timing for operation was 1.3 days after birth but those ones that either look like the ‘H’ or have the proximal join and the distal join tended to be operated on later (10.7 days on average).
  • In this series there were no thoracoscopic repairs.

Enough of the colour background, what about the anaesthetic?

Well the key bit we’re interested in is the induction and airway management. 20 of the patients turned up to theatres with the snorkel already placed and 5 had the version where there’s no join between trachea and oesophagus (straight oesophageal atresia) so they weren’t looked at.

So here are the key stats:

  • 32% had an IV induction and 68% an inhalational version.
  • At induction 26% of patients were given suxamethonium, and 10% had a non-depolarising muscle relaxant. I’ll confess I didn’t think it would be that high.
  • No one tried to occlude the TOF in any way.
  • Ventilation difficulty was noted in 6.9% of patients and 14.9% had significant desaturations intraoperatively but without significant sequelae. I couldn’t figure out from the paper whether some of those who had the significant desaturations were the same ones who had ventilatory difficulty.
  • Bronchoscopy was performed prior to thoracotomy in 40.5% of cases. In this study they couldn’t say whether this was the rigid version or the flexible one. In the discussion they make the point that practice varied greatly between institutions. At one place it was 4% while in another it was 90%.

So if that’s the story, how does it change the things I try to do? Well, not a whole lot. There’s not a whole lot in there that suggests anyone’s got the perfect answer.

So in an effort to stimulate healthy conversation (not conversation about healthy things necessarily but conversation with enough vigour to suggest it has an internal life) here’s the quick run down of things I’d rate as worth considering when a case like this next comes up (beyond the usual neonatal considerations).

1. A thorough assessment

OK this might seem pretty self explanatory but in that series described 57.5% of the kids had some kind of pretty reasonable comorbidity (and 42.5% of the total number of kids had a cardiac morbidity of some sort). So thorough assessment of the background conditions as well as getting as much understanding of the anatomy is pretty vital. Given the chance of soiling of the respiratory tract a particular focus on this sort of makes sense.

2. A chat with the surgeon

Look sometimes they make our lives difficult. There are a couple I’ve met who make chatting not entirely easy too. Mostly they are pretty good though and both before and during (and even after if you still like them) these cases require excellent communication.

Key things to establish include whether or not they plan to start with bronchoscopy and whether the plan is to undertake the procedure via a thoracotomy or thoracoscopically. They may have strong feelings about patient position and disposition afterwards. As one example if there is a long segment to repair for the oesophageal anastomosis they may well be keen to try and prevent tension on that anastomosis by maintaining a degree of flexion and preventing movement for a day or two.

This operation is also a lot easier if you really understand the many steps they’ll be working through. If you’re not into talking with the surgeon through the case this is the time to change it. It’s a true team effort. And you might just get a chance to fix their taste in music along the way.

Anyway, you need to chat.

3. The monitoring

Well, the standard stuff. Which for me would include two spots for monitoring saturations. This is something I’d do most times anyway because it’s not at all infrequent for people to bump up against your monitoring. Pre- and post-ductal saturations can be useful. There’s even a bit of correspondence out there describing a case where loss of the post-ductal oximetry when test ligation of the fistula was undertaken was the clue that told the team they’d clamped a right-sided aortic arch. I’d also be after invasive arterial monitoring.

4. The induction

My first choice has always been inhalational induction with sevoflurane and maintenance of spontaneous ventilation. Once the fistula is secured I’d switch on the muscle relaxation but up until then titrated doses of fentanyl with a bit of support from some intercostal blocks and ketamine has generally maintained an adequate plane of anaesthesia and analgesia.

5. The camera

The source paper I mentioned has an excellent section on all the benefits of bronchoscopy. You can confirm the size and location of the fistula. There were two cases where a second fistula was missed in their series and needed a second operation. Both had no bronchoscopy. For larger TOFs it’s feasible to place a catheter through the fistula to occlude it or help identification (probably easier with the rigid bronchoscopy version). There are reports in the literature of the flexible bronchoscope being used to show the surgeons the course of the fistula so they can complete the operation. So starting with one of these makes sense.

Regardless of whether it’s been done though I’d always choose to have a flexible bronchoscope present. Placing an endotracheal tube at the start even with rigid bronchoscopy isn’t a guarantee that it’ll stay in the right spot. We’re all aware that endotracheal tubes move every time you position kids and a centimetre would be a big move in this context. These kids are guaranteed to be repositioned after any initial look while supine so an ability to re-check that tube position can come in pretty handy.

Sometimes introducing a little positive pressure during flexible bronchoscopy can demonstrate the fistula effectively too, whereas it may be less clear if the patient is spontaneously ventilating through rigid bronchoscopy.

There are other potential advantages when it comes to complications. If ventilation changes it’s a lot easier to troubleshoot if you can check the position of the tube with respect to the airway and the fistula. Through the repair there will be periods of tension and distortion of the anatomy so tube position isn’t guaranteed throughout. There are cases where the endotracheal tube has slipped into the fistula and having a bronchoscope handy was a pretty useful tool.

There are reports of surgeons accidentally putting a suture through the end of the tube and a bronchoscope might help there. I’ve certainly had the experience of the endotracheal tube being pulled briefly out of the trachea and having bronchoscopy on a screen is quite persuasive for all in the room that things aren’t quite right.

There are a bazillion possibilities that might occur in which a flexible bronchoscope would be handy. Well, not a bazillion exactly but the number of times you really need it only needs to be (n = 1) for you to be glad it’s in the toolkit.

The challenge is not to go too nuts with it and delay the definitive part of the day. It turns out you are there for the surgery after all.

5. Two gastric tubes

This might seem obvious but it’s something you can forget. There’ll be two moments in the cases where the oesophagus is being patched up where this will come in handy. First you need to help the surgeons by bringing the upper pouch down towards the lower one. A slightly larger, more rigid one is ideal for this purpose when placed orally and easy to access.

They’ll also end up wanting a trans-anastomotic tube which will end up being pulled through and down by the surgeons and spend quite some time in place. This one needs to be a better long-term feeding option then and is worth discussing prior. This will be the one that takes up residence nasally for some time to come.

6. The after plan

As mentioned above the surgeon may have strong feelings about their joinery so they are part of the discussion. I’ve worked with a  few anaesthetists who talk fondly of the days when all the kids got a thoracic epidural and lost the snorkel at the end. This is an old story near as I can tell. Not many people I talk to seem to go down this road now, though paravertebral catheters may have a place in the kid who might get extubated within a couple of days. More often in my experience the patient will be intubated and ventilated for more than a couple of days, so the priority is setting the patient up for intensive care and positioning with the surgeons at the end to make sure the head and neck are in the desired position.


That’s just a start. Bet there’s something I’ve forgotten too. Is that close to what you do? Are there tips you could offer to get me doing it better?

Is there a chance we’re all in agreement on the vital bits of the case?

Because if that last one is a yes, I think I have to go and herd some wild cats.

"Malawi" (Serval) - Big Cats 2 (WHF, Kent)

Not that tough really. Silly old cheetahs.




First up I wasn’t kidding about sharing your tips and thoughts. The posts are always improved when people cleverer than the writer share their knowledge.

A note also that if you like the stuff on here you could maybe share it around. Also you should be able to find a spot on the page to sign up with your email if you’d like to get notified whenever a new post drops.

Right, here are some links to things mentioned along the way.

That main paper looking at how people go about it is this one:

Knottenbelt G, Costi D, Stephens P, Beringer R, Davidson A. An audit of anaesthetic management and complications of trachea-esophageal fistula and oesophageal atresia repair. Pediatric Anesthesia. 2012;22:268-74.

And now some case reports/letters that are a bit interesting:

Hack H, Raj N. An unknown second tracheoesophageal fistula: a rare cause of postoperative respiratory failure. Pediatric Anesthesia. 2006;16:479-83.

And another one with some use of the bronchoscope:

Gupta B, Ayub A, Bindra A, Gupta S, Mishra B. Distal tracheoesophagela fistula in pediatric patient – an anaesthetic challenge. Pediatric Anesth. 2014;24:886-8.

Here’s the one commenting on pre- and postductal saturation monitoring:

Dias R, Dave N, Munghate G, Parelkar SV, Garasia M. Monitoring pre -and postductal saturation – A must for neonates undergoing tracheoesophageal fistula repair. Pediatr Anesth. 2016;26:225-6.

Here’s a report on two different cases, one where a flexible bronchoscope was used to identify anatomy on a proximal fistula and one where the tube malpositioned during the case:

de Gabriele LC, Cooper MG, Singh S, Pitkin J. Intraoperative fibreoptic bronchoscopy during neonatal trachea-oesophageal fistula ligation and oesophageal atresia repair. Anaesth Intensve Care. 2001;29:284-7. 

Here’s one where they later figured out a suture was through the endotracheal tube (a bronchoscope didn’t help them on this one):

Kaur H, Kaur R, Goswami D, Sharma J, Jain A, Vajifdar H. Unusual complication in a neonate following tracheo-oesophageal fistula repair. Pediatr Anesth. 2011;21:1171-2.

That image of a serval was from the Creative Commons bit of flickr and is unchanged from Paul Drummond’s post.








Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s