Baby’s Eyes and the Three Bears

This is the written post version of a talk by Dr Andrew Weatherall for the South African Society of Anesthesiologists Congress 2017. As invited faculty the organising committee covered registration, accommodation and travel to attend. 

There are a bunch of ways a talk involving three bears and baby’s eyes could go. I don’t think it’s likely anyone will get harmed in the commission of this thing.

This very brief version of a thing on retinopathy of prematurity and some considerations for anaesthetists aims only to cover a few things:

  • What’s going on?
  • What’s going to go on?
  • What are we going to do to let what’s going on go on?

A Lesson in Normality

Everyone cleverer than me probably remembers that retinal vascularisation kicks off around 15-18 weeks gestation. The nasal retina is done by about 36 weeks. The temporal retina turns out to be a little more sluggish, getting there at around 40 weeks.

So that kid who is born prematurely is at a stage where things are vulnerable. And for some kids that doesn’t go well.

The way things seek to go is this:

  1. An initial injury occurs that matters to the eye. Hypotension, hypoxia or hyperopia are potential candidates. Developing vessels are injured and angiogenesis is disrupted.
  2. Some patients will not end up with problems. Some patients go a different pathway and end up developing new and abnormal vessels. In that latter group it might be that their insulin-like growth factor-1 (which is an important thing interacting with Vascular Endothelial Growth Factor) is a bit low and when the injury happens there isn’t support of ongoing vessel growth. No growth of blood vessels leads to hypoxia of the retina. Later as the IGF-1 levels do start to rise neovascularisation gets kicked off and problems including development of fibrovascular tissue ensue.


Anyway, I seem to recall going through training and the message being “don’t give too much oxygen to the little babies” with exactly that much nuance. There seems to be a little more to it than that.

Simple Complex Oxygen

In early reports of retinopathy of prematurity (ROP) hyperoxia really looked like it might have been a culprit. Over time though the relationship between high inspired oxygen and ROP just didn’t seem as reliable as you might have thought.

However there is this phase where it looks like retinal hypoxia might be a key part of the pathophysiological hit. Which makes it sensible to have  look at whether avoidance of retinal hypoxia might be a sensible thing to try.

So in 2000 the STOP-ROP results were published. The plan in this had been to explore whether patients in whom evidence of neovascularisation had been picked up could have the progression of disease stopped by adding supplemental oxygen. Patients in the conventional group had their oxygen therapy managed to maintain saturations in the 89-94% range. Patients in the oxygen group had their therapy titrated to 96-99%.

Pleasingly they saw a trend to less progression to threshold disease (the sort of disease they want to intervene with). Not a statistically significant one but still, a trend. Except it also came with higher rates of pulmonary issues. Y’know lazy little things like pneumonia and the like.

What about prevention in the phase prior to this by not inducing a hyperoxic injury in the first place? Well Manja et al published a review in 2015 where they’d looked at prevention by maintaining saturations in the 85-89% range (instead of 91-95%). They showed a non-significant reduction in retinopathy.

However it came with an increased risk of in-hospital mortality. Damn. Plus there was no difference in severe disability at 18-24 months.

After more stuff on oxygen? Well there’s a fascinating read in the BOOST II publication which covers the trials and tribulations of a prospective observation of those low and high target ranges across the UK, Oz and NZ. Midway through the trial it emerged that the oximeter in use had a right-shifted algorithm in the 87-96% range. D’oh.

This pushed the measured saturation up 1 to 2 points which sort of matters here. A new algorithm was introduced by the company, accuracy apparently went up and then when they analysed those recruited with new oximetry algorithm in place they did find less ROP in the lower saturations group. But the death rate was also 23.1% vs 15.9%. That lower saturations group also had higher rates of necrotising enterocolitis leading to surgery or death.

So attempts at manipulating oxygen either to prevent or to stop progression are complicated by the pesky issue of more deaths and the rates of change in progression or demonstrated preventive effective aren’t overwhelming. What are we supposed to do with that?

Well the going theory in my general reading appears to be that the issue at the initial insult might not be purely be the high or low levels of oxygen, but sudden shifts in them in the presence of fragile developing retina and vessels that is the issue (I’m happy to be corrected if I haven’t quite landed on the right spot there). Babies eyes weren’t meant to be involved in surfing big waves I guess.

Which still leaves us with the question of what we should target. Well I think the first step is to ask the treating clinicians what they are aiming for.  But I’d expect them to say 91-95% is the current best option.

You want to do what?

If you take a step back the idea of firing lasers into tiny kid eyes is sort of breathtaking. The procedure is, I guess, relatively simple. You take an indirect ophthalmoscope, find an appropriate laser lying around and then dilate the pupils so you can focus the appropriate clever wavelengths on the avascular retina.

The treatment provides better long-term visual acuity than cryotherapy, which used to be a thing. In my experience to get a full treatment does take some time, probably needs some precision and does come with some safety consideration around the use of lasers for purposes rather than threatening global annihilation so you can get paid money by international governments.

I should mention in passing that there is an alternative treatment out there where anti-VGEF (that’s that vascular endothelial growth factor mentioned briefly above) is injected into the eye. As this would generally not be something involving anaesthetists so much I’m not going to go too nuts on detail. It sounds promising but the longer-term story after this treatment still isn’t known in so much detail.

The really crucial thing for the anaesthetist to know is that once Type I ROP is diagnosed (and nope, not going to spend heaps of time on diagnostic grading when there are clever ophthalmologists who will tell us where the kid is at) the aim has to be to get the chosen treatment done within 72 hours.

72 hours. That timeframe might really matter when we’re assessing what is the best way forward.

Which bear gets it just right?

Now I’ll confess that I haven’t not worked anywhere which does this any other way than a general anaesthesia in the operating suite, usually after hours when an ophthalmologist is available. This is the tremendous luxury of working in big hospitals in the city I live in.

The literature tells a different story though because evidently people who want to deliver care for precarious kids often have to make choices to get stuff done and seek alternatives. A quick search around reveals three options reported out there:


  1. Local Specialists
Denali National Park Bear

Look, I’ll confess I was really just looking for something that made people think of the number “3” and the 3 bears came to mind. Plus bears. Bears are cool.

There are a few reports out there in journal land where teams have explained how they go about this option, and it seems like it’s an option people consider when there is some barrier to management in operating theatres with an anaesthetist or sedation by someone comfortable with it. Sinha and Ranjan probably provided the most detail I could find on how they go about it:

  • After dilation of the eyes, provide local anaesthesia either via drops or by injection.
  • Apply ECG and pulse oximetry.
  • Wrap the baby more effectively than a historical baby floating in the reeds.
  • Grab a trusty assistant who “supports” the head and chin (now those are my quotation marks because I have formed a mental image of what that word means here but I could, of course, be mistaken).
  • Give a feed before and offer a dummy during.

They report in this letter that they can get it done. They’re not alone either. There are other letters or reports out there and in surveys it comes up as an option that people deploy.

2. The Sedation Station

Jethro Taylor Black Bear.jpg

I don’t know this black bear doesn’t seem that attentive to this topic to be honest.

A variety of sedation options (I had almost forgotten pentazocine was a thing) get reported in the literature. Sedation seems like a less popular option than a trip to the operating theatre but it does have a couple of advantages:

  • It is a technique very amenable to getting the whole thing done in the intensive care unit.
  • There is some work showing lower pain scores or pain behaviours when compared to local (yes it turns out scorching the expanse of the retina can be uncomfortable).
  • Some practitioners are clearly very comfortable providing sedation and leaving the patient to breathe away.

Were you to undertake this I imagine some of the steps mentioned in the local anaesthetic routine (particularly the wrap and having an assistant looking after the kid) would still help.

One thing that does make this area hard to get my head around is that the definitions of sedation are even broader than you might imagine. One paper mentioned the crucial of muscle relaxation as part of their sedation cocktail which is …. I um ….. I mean….

3. The General

Alaska US FWS Bear

I checked. This bear didn’t even try with that fur, it just does it all by itself. 

General anaesthesia is associated with less pain behaviour. Funny that. There is also some suggestion that you’re more likely to be be able to complete the treatment required. A single trip, all intervention done. The corollary being that the other techniques may have to be abandoned before the ophthalmologist is happy they are finished.

However we’re all very familiar with the logistics and resources required for neonatal anaesthesia. Plus you have that 72 hour window and sometimes it’s hard for a tiny kid to muscle out others to get into your theatre space.

Oh, and there’s that other thing…

A Quick Note on Neurotoxicity

There are important questions being asked about this topic and clearly a comprehensive review is beyond the scope of this little spot. The most recent bits of evidence coming out are at least a little reassuring that neurotoxicity related to anaesthesia might not be as big a clinical problem as our worst fears had entertained.

IN 2016 O’Leary et al. looked at 28366  kids who underwent surgery before 5-6 years old and matched them to 55910 non-exposed kids. There was a suggestion that those exposed had increased early developmental vulnerability (25.6% vs 25.0%). However kids who were above 2 years old when exposed had a higher risk of this while those under 2 had lower risk when exposed. They also had no increase in this vulnerability with increasing frequency of exposure.

This lack of logical biological association (less problems if younger, no association with increasing exposure) makes it seem less like the association is a huge problem.

Graham et al also looked at 3850 kids exposed to GA as well as 620 with two or more and compared them to 13586 non-exposed kids. They found no association with worse scores on a developmental assessment if the little person had their exposure before the age of 2. This was true even if they had multiple exposures. A single exposure between 2 and 4 years old was statistically significant.

So all theory about exposure at a critical stage of neural connection is …. not a thing? Maybe I’ll let the authors speak:

“These findings refute the assumption that the earlier the GA exposure in children, the greater the likelihood of long-term neurocognitive risk. The authors cannot confirm an association between multiple GA exposure and increased risk of neurocognitive impairment, increasing the probability of confounding to explain the results.”

[Gentle slap down acknowledged.]

Coupled with the findings of the GAS preliminary results and the reports from PANDA, I think we can be more confident when we seek to reassure parents on this front.

Flow and Pressure

Another thing we should definitely think about in little people is flow and pressure. Given that we’ve already said fluctuations aren’t OK, what pressures should we be targeting to make sure the good stuff gets to what we want to be a good brain.

Well we have some information that helps informs this a little, though it comes from kids between 0 and 6 months. Michelet and crew have published some work suggesting that a fall in systolic blood pressure of < 20% of baseline is probably OK when it comes to maintaining cerebral oxygenation under general anaesthesia.

Rhondali and crew (utilising transcranial Doppler in addition to cerebral oximetry) have done some work seeking to pick up where the limit of cerebral autoregulation might be and suggested that in the under 6 months child that critical point might be at around 38 mmHg (mean arterial pressure).

Subsequent analysis suggested that when MAP in this age group under 6 months fell below   45 mmHg you could see a fall in cerebral blood flow velocity but regional saturation stayed in a reasonable range, probably because cerebral metabolic requirements have been favourably changed.

Fall below 35 mmHg though and that situation changes and cerebral oxygenation is compromised (this was in the context of sevoflurane anaesthesia).

Now the kids having anaesthesia for ROP should probably not be 5 months and 28 days old so maybe we should try and be a little pragmatic. The first chat to have is probably with any associated neonatal clinicians to see what blood pressure they are targeting. Then it makes sense to set some mental limits:

  • Find out the child’s baseline blood pressure and figure out what a 20% fall in systolic blood pressure would like. Let’s not test that margin.
  • Unless the kid is still really premature, setting a limit on mean arterial pressure of 35 mmHg should make you feel comfortable.

On top of those limits it makes sense to try and keep any shifts in blood pressure slow. Little people don’t like big waves, right? There’s no benefit in rushing things in this setting.

The Best Bear

Getting back to the original question, which of the three bears gets it just right? Well I might have my preferences but sometimes it’s the situation you’re in that might force your choice.


  • If resources are not an issue, careful general anaesthesia seems to offer the best level of comfort for the patient and better operating conditions for the surgeon. At our end we can make sure they are warm, wet (meaning appropriately hydrated), pink (but not too pink) and sweet (because all the other parameters being superb doesn’t help if the sugar is low). Then avoid swings and roundabouts in the physiology and you’re in good shape.
  • If timing or practicality demands it, you might want to have a chat with the ophthalmology crew and neonatal team about conducting the procedure under sedation in the intensive care unit. This would be one of those cases where you’d want to deploy techniques you are pretty comfortable with though.
  • If resources don’t offer any sort of option otherwise then a chat should be had about getting things done under local anaesthesia. Sure there are compromises but eyes are also quite useful.


Whichever technique you choose there are some simple truths that always seem to apply:

  • Not all oxygen is good, so make your target sensible and recalibrate your own sense of what tone on the sats monitor sounds safe.
  • Apply appropriate monitoring so you can stay connected with the patient and their physiology.
  • Sudden shifts are not anyone’s friend.


And it turns out that all the bears probably get it right some of the time. Or at least that’s what I’d tell them because they’re bears. Who argues with a bear?



Did you get all the way here? Awesome. Please feel free to chip in with cleverer ideas or thoughts or experiences. Learning is a shared experience.

Also, if you happen to like stuff on here, there is a spot somewhere on the page where you can sign in to get an email each time a post gets up. You could do that. Or not, no pressure.

The images on here are all from flickr’s Creative Commons area. The first group of bears were posted by the Denali National Park people, the black bear by Jethro Taylor and the Kodiak was from the Alaska Fisheries and Wildlife types. None of them are altered.

Now, how about some reading because the sources never really have full justice done to them here.

The STOP-ROP paper is this one:

STOP-ROP Multicenter Study Group. Supplemental Therapeutic Oxygen for Prethreshold Retinopathy of Prematurity (STOP-ROP)) Randomized, Controlled Trial. I: Primary Outcomes. Pediatr. 2000;105:295.

The prevention review is this one:

Manja V, Lakshminrusimha S, Cook DJ. Oxygen saturation target range for extremely preterm infants: a systematic review  and meta-analysis. JAMA Pediatr. 2015;169:332.

BOOST II is right here:

The BOOST II UK, Australia and NZ Collaborative Groups. Oxygen Saturation and Outcomes in Preterm Infants. NEJM. 2013;368:2094-2104.

Now to techniques:

A general thing is here…

Sato Y, et al. Multicenter observational study comparing sedation/analgesia protocols for laser photocoagulation treatment of retinopathy of prematurity. J Perinat. 2015;35:965-9.  

The thing describing a practical approach to doing local anaesthetic approaches is here:

Sinha R, Ranjan B. Laser treatments for retinopathy of prematurity under topical anesthesia – prospective from our experience. Pediatr Anesth. 2013;22:376.

Now, on to neurotoxicity:

O’Leary JD, et al. A population-based Study Evaluating the Association Between Surgery in Early Life and Child Development at Primary School Entry. Anesthesiol. 2016;125:272-9.

Graham MR, et al. Neurodevelopmental assessment in kindergarten in children exposed to general anesthesia before the age of 4 years. Anesthesiol. 2016;125:667-77.

Now that blood pressure and flow stuff:

Michelet D, et al. Intraoperative changes in blood pressure associated with cerebral desaturation in infants. Pediatr. Anesth. 2015;25:681-8. 

Rhondali O, et al. Impact of sevoflurane anesthesia on cerebral blood flow in children younger than 2 years. Pediatr. Anesth. 2013;23:946-51.

Rhondali O, et al. Sevoflurane anesthesia and brain perfusion. Pediatr. Anesth. 2014;25:180-5.



One response to “Baby’s Eyes and the Three Bears

  1. Pingback: Baby’s eyes and the three bears | Prehospital and Retrieval Medicine - THE PHARM dedicated to the memory of Dr John Hinds·

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