Does Epidural Analgesia Predispose to Persistent Occiput Posterior?
In my January Science and Sensibility blog post, I answered the question “Can We Prevent Persistent Occiput Posterior?” but because it wasn’t relevant to the study that prompted the post, and the piece was already long, I didn’t look at the role of epidural analgesia. Let me now rectify that.
All five studies examining the relationship between epidural analgesia and persistent occiput posterior (OP) find an association between them. Three studies compared women with an epidural versus no epidural according to whether they had an OP baby at delivery and found that 4 to 10 more women per 100 having an epidural had an OP baby at delivery (Cheng, 2006; Lieberman, 2005; Sizer, 2000). The other two compared women with an OP baby at delivery according to whether they had an epidural and found that 13 and 27 more women per 100 with a persistent OP baby had an epidural (Fitzpatrick, 2001; Ponkey, 2003).
Their results, however, aren’t sufficient to convict epidurals because we can’t tell whether having an epidural led to persistent OP or more painful and prolonged OP labor led to having an epidural. Investigators in one of the five studies argued for OP labor coming first on the grounds that while epidurals were more common in women with an OP baby at delivery than in women with an OA baby at delivery (74% vs. 47%) at their institution, a rise in epidural use from 3% in 1975 to 47% in 1998 had, if anything, decreased the hospital’s overall rate of persistent OP (4% vs. 2%) (Fitzpatrick, 2001). This must mean that as epidurals became more freely available, women having difficult OP labors were more likely to opt for one. Epidurals were the result, not the cause, of persistent OP. On the other hand, we have some corroborating evidence for their guilt. For one thing, back pain is thought to be a major reason why women with OP babies are more likely to want epidurals, but it turns out that back pain isn’t unique to OP. Serial sonograms reveal that virtually identical percentages of women laboring with an OA baby report back pain (Lieberman, 2005). For another, three of the five studies took into account other factors associated with difficult labor such as labor induction, labor augmentation, and birth weight and still found that epidurals were an independent risk factor for persistent OP (Cheng, 2006; Lieberman, 2005; Sizer, 2000).
Nevertheless, evidence from observational studies isn’t strong enough to close the case. As I noted, observational studies can determine association but not causation. In addition, investigators may not be able to identify all the confounding and correlating factors that affect outcomes. For a more definitive answer, we need experimental studies. This brings us to randomized controlled trials (RCTs), in which participants are randomly allocated to one form of treatment or the other, and to meta-analysis of RCTs, in which statistical techniques are used to pool data from more than one trial.
The Cochrane systematic review of epidural versus no epidural in labor pools data from four RCTs (673 women overall) that reported on persistent OP (Anim-Somuah, 2011). Five more women per 100 assigned to the epidural group had a persistent OP baby, but meta-analysis found that the difference just missed achieving statistical significance. The risk ratio was 1.4, meaning a 40% increased risk of persistent OP in women assigned to the epidural group compared with women assigned to the no-epidural group, but the 95% confidence interval ranged from 1.0 to 2.0, meaning a 95% probability that the true value lies between no increase (ratio of 1:1) and double the risk (2.0). However, a problem with the RCTs of epidural versus no epidural is that substantial percentages of women assigned to the no-epidural group actually had epidurals, but, as is prescribed in RCT data analysis, they were kept in their original group. In two of the four trials (204 women), though, 10% or less of women assigned to the no-epidural group had epidurals. If we calculate the excess rate of persistent OP in these two trials, we find that the gap widens to 9 more women per 100 with epidurals having a persistent OP baby. We don’t know whether this difference would achieve statistical significance, but the fact that the excess is in the same range as reported in the observational studies (4 to 10 more per 100) gives confidence in its validity.
We also have two studies that suggest that the timing of the epidural may matter. One, of 320 women, reports that, after controlling for age, induction of labor, and birth weight, initiating an epidural at fetal station less than zero (above the ischial spines) resulted in 16 more women having a persistent OP or occiput transverse (OT) baby compared with initiation at 0 station or greater (at or lower than the ischial spines), an excess that rose to 20 more per 100 in first time mothers (Robinson, 1996). The other study analyzed outcomes in 500 first-time mothers according to whether an epidural was administered early (at or before 5 cm dilation), late (after 5 cm dilation), or not at all (Thorp, 1991). Seventeen more women per 100 in the early group had a persistent OP or OT baby compared with women in the late-epidural group, and 12 more had a persistent OP or OT baby compared with the no-epidural group, but rates were similar in women in the late and no epidural groups.
Taken all together, we may not have absolute proof of epidural culpability in predisposing to OP, but if I were on the jury, I would vote them “guilty as charged.”
Take home: Even without certainty, the precautionary principle dictates recommending to women desiring an epidural that they use other measures to cope with labor pain until they enter active labor and until it seems clear that positioning and activities are not putting a slow labor on track.
What do you tell your clients, students and patients about the impact on fetal positioning in labor and birth? Will having this information change what you say? Let us know in the comments section.
Anim-Somuah, M., Smyth, R. M., & Jones, L. (2011). Epidural versus non-epidural or no analgesia in labour. Cochrane Database of Systematic Reviews(12), CD000331.
Cheng, Y. W., Shaffer, B. L., & Caughey, A. B. (2006). Associated factors and outcomes of persistent occiput posterior position: A retrospective cohort study from 1976 to 2001. Journal of Maternal Fetal and Neonatal Medicine, 19(9), 563-568.
Fitzpatrick, M., McQuillan, K., & O’Herlihy, C. (2001). Influence of persistent occiput posterior position on delivery outcome. Obstetrics and Gynecology, 98(6), 1027-1031.
Lieberman, E., Davidson, K., Lee-Parritz, A., & Shearer, E. (2005). Changes in fetal position during labor and their association with epidural analgesia. Obstetrics and Gynecology, 105(5 Pt 1), 974-982.
Ponkey, S. E., Cohen, A. P., Heffner, L. J., & Lieberman, E. (2003). Persistent fetal occiput posterior position: obstetric outcomes. Obstetrics and Gynecology, 101(5 Pt 1), 915-920.
Robinson, C. A., Macones, G. A., Roth, N. W., & Morgan, M. A. (1996). Does station of the fetal head at epidural placement affect the position of the fetal vertex at delivery? American Journal of Obstetrics and Gynecology, 175(4 Pt 1), 991-994.
Sizer, A. R., & Nirmal, D. M. (2000). Occipitoposterior position: associated factors and obstetric outcome in nulliparas. Obstetrics and Gynecology, 96(5 Pt 1), 749-752.
Thorp, J. A., Eckert, L. O., Ang, M. S., Johnston, D. A., Peaceman, A. M., & Parisi, V. M. (1991). Epidural analgesia and cesarean section for dystocia: risk factors in nulliparas. American Journal of Perinatology, 8(6), 402-410.