Thread by @aljabut

So folks wanted a #journalclub on the @JAMAPediatrics paper on #epidurals and #autism. Ok......here we go!

1. Let's review the intro first. Beyond conjecture, they offer scant data to justify their study which was an animal study (ref 5). They say 'standard clinical doses of LAs 👇produce neurotoxic effects & altered neurobeh devlpment in monkeys.'

These aren't standard doses. We use in fact substantially lower concentrations & doses of epidural local anesthetics. Plus, this small study in non-laboring monkeys included multiple behav outcomes. So high likelihood of chance finding. Also, anyone know if monkeys get ASD??

2. They reference 3 human studies to suggest that anesthesia drug exposure may be linked to ASD. Ref 3 report no data about this (ref 3). Ref 24 - only provide data for epidural meds lumped with oxytocin. So this statement is not supported by 2 citations.

3. The 3rd ref (ref 26) was a case-control study. 'Epidural-caudal anesthesia' had a 1.68 incr odds of autism. BUT - no adj for confounders & all delivery modes were included...so we don't know the indication for epidural. So the human data supporting any link is barely there.

4. Do we know if women had a spont or instrumental delivery? Errr...no. What about neonatal outcomes at/soon after delivery. Nope...

5. Any information about parental (genetic) factors? Nope.

6. Did they fully account for all potential confounders e.g., labor dystocia, prolonged labor, fetal status, fetal malposition before epidural request or match women with labor characteristics at the time of epidural placement to those not getting an epidural- Nope.

7. The categorization of epidural duration is arbitrary (4 hr increments). The lower point estimate for the linear trend is very low (1.05) with the lower 95th CI approaching the null (1.01). No mention of this in the discussion.

8. Do you really think that there isn't a residual confounder that may have a HR of 1.76? Err...nope.

9. They fail to describe other key physicochemical drug characteristics that influence placental transfer e.g., protein binding, pKA, lipid solubility. Amide LAs have a low umbilical vein: maternal vein blood concentration ratio. Any mention of this in the discussion? Nope.

10. They offer up some pretty incredible speculation to explain their findings, including a 'new multi-system homeostasis' from an epidural effect. This came outta nowhere.

11. There is no mention about the clinical implications of these findings for moms to be or #obanes who provide diligent care. Not sure why this didn't happen. #momsmatter #unintendedconsequence

12. The unadjusted absolute risk difference is very small (0.6%). Again....nothing mentioned about this anywhere in the paper. Urghh.

13. No hypothesis is provided...so this is - at best - an exploratory study that certainly isn't hypothesis confirming, with likely biased estimates of risk.

14. Very curious re the comments and responses during peer review. @JAMAPediatrics - an editorial response would be helpful.

I'm sure there are other issues folks have. But I'll stop there. #obanes #neonataologists #obstetricians - feel free to weigh in. @SOAPHQ @ASALifeline #stopbashingepidurals

PS: @MariaSheikhMD - remind you of our @Stanford_obanes JCs???

Oh...was their literature review comprehensive? Presume no coz they didn't cite or discuss this --- which showed no link between neuraxial labor analgesia and childhood learning disabilities 😯 …bi-nlm-nih-gov.laneproxy.stanford.edu/20736436/ #doyourhomework