Wednesday, January 09, 2008

Evidence Based Obstetrics

One of the great things about blogs is how a post about one topic can quickly morph into a good discussion about another. I bet early meetings of Ben Franklin's Junto (or later the American Philosophical Society) were similar to what happens when a good discussion gets hijacked by a barely related intriguing idea.

Anyway, Drugmonkey posted an informative update about the current presidential candidates' positions on science. If there is one topic that has been wrongly overlooked by candidates this election season, it is a proper airing of their perceptions about science and health research funding, the role that science should play in public policy decisions and how science education can be improved. Head over to this DM post for a digest of Science magazine's handling of the candidates' positions on science. Within the comments thread, is a sub-discussion about the lack of scientific evidence for the choices obstetricians make during delivery. The specific question that I argue has no scientific answer to is: Do fetal cardiac decelerations necessitate Caesarian section? I admit: the left turn was kind-of my fault,but my appeal to you, dear reader, is to find your way to this thread and chime in about the role of evidence in Western obstetrical care. (Read about the candidates' views, too!)

I have a love-hate relationship with the current emphasis on 'evidence based medicine.' It is certainly indicated, even needed in many cases. Enough of the studies are poorly designed or in conflict of interest, that it is hard to change a policy because some perspective made it through peer review. I suppose I would advocate a slow transition to evidence-based practice. Meta-analysis means more to me as a mode for medical decision making.

But bringing this full circle, what does it say about the role of science in society that one of the areas of life that is perceived of as being influenced the most by science (medicine), is actually not very scientific? If science cannot be practically applied to medicine, or more specifically birth, can it really be applied to policy decisions? It's not like there's any shortage of research material.

1 comment:

Lindsay said...

One must be careful to avoid oversimplifying complex physiologic and medical events. Saying "all decels require a C-section" is just as wrong as saying "no decels require a C-section". First of all, these statements ignore the fact that there are 3 main categories of decels, and within those categories, a huge range of severity based on amplitude, duration and frequency. Decisions in obstetrics aren't made based on a single blip in the fetal heart tracing, they take into account the overall pattern of the FHR and the contractions, as well as details of obstetrical history and maternal health. Decels are just one small part of the overall picture.

I don't think there's really much controversy in obstetrics regarding emergent C-section for a truly BAD fetal heart tracing (eg. long, deep, frequent late decels or fetal bradycardia). Our understanding of fetal physiology tells us that these tracing represent fetal hypoxia, and that's not something anybody is just going to sit by and watch for the minutes or hours that it might take for a woman to deliver vaginally. Of course, this is not "evidence based" in the sense of having been studied in a randomized, controlled trial. Can you imagine trying to recruit pregnant women for that trial? (Not to mention getting ethics approval in the first place.) Some things are simply not amenable to testing by RCT, which means we're stuck making decisions based on our best understanding of anatomy, physiology and past experience (personal and collective). That's life, and medicine, and that's the reason why our practices will never be (nor should they be) 100% "evidence based".

Incidentally, my understanding of the current state of things is that there's no evidence for routine continuous fetal monitoring in labour, which is to say that every woman in labour does NOT need to have every single fetal heartbeat monitored from the moment she arrives in hospital until she delivers. Doing this does indeed increase the rate of intervention, without having any effect on overall outcomes. But that doesn't mean we don't care about the fetal heart rate at all. The current recommendations suggest intermittent auscultation (ie. listening once in a while, particularly right after a contraction) in low risk births, and continuous monitoring only if there is a good reason for it. But no matter how you choose to monitor, if you find signs of fetal distress, you're obliged to at least consider intervention.