What is the success/failure rate of epidurals?

small_25645404This question was posed to me by a group of childbirth educators:

“I wonder if you can help us find up-to-date data on how likely an epidural is to be successful?  Woman sometimes ask about this because a friend had a failed epidural, and the figures that we have found are quite variable.”

This is one of those questions where I feel I need to give you the boring caveats which explain why it’s hard to answer (and perhaps why those figures are variable) first, but I promise I will then try to give you a half-decent answer!  It is hard to answer this question definitively for a few reasons.

1. Success rates differ quite a bit between people and places and studies, and even between the same anaesthetist at different times, so we are not going to be able to pinpoint an exact and universal number.  The variables that affect success are not just to do with the anaesthetist, the setting or the equipment, but some are also to do with the individual woman (for instance body shape and size, and the ability to get and stay in certain positions, which is not always easy when one is in labour!)

2. Because pain and pain relief are subjective and people have different expectations, the definition of ‘success’ will vary between people, including researchers.  To put it into research terms, there is no universal outcome measure for this.  Does an epidural have to remove every last smidgen of sensation to be successful, or is it still a success if it removes 90 per cent of pain but the woman is OK with that?  What if it is successful for the first hour but then the catheter moves and it no longer works?  There are some big questions here about what constitutes success when it comes to epidurals, and (as I imagine you know) this can be an interesting question/topic for groups to discuss!

So, with those points in mind, let’s look for the figures you’re after…

One of the more obvious places to start is the Cochrane review which compared epidural with non-epidural or no analgesia in labour (Anim-Somuah et al 2011).  This found that, compared to those other possibilities, epidural was more effective overall, although it did have downsides including increased length of second stage, increased need for oxytocin and increased likelihood of instrumental delivery.  Unfortunately, though, because it was comparing epidural to other possibilities, the review did not specifically focus on success or failure rates of epidural in itself.

surgeon_penguinLuckily, a few other groups of researchers have specifically looked at this question.  Some of the better examples include a (freely available) review by Hermanides et al (2012) which considered a number of studies of epidural, not just in childbirth, and concluded that the overall failure rate was around 30%.  This isn’t necessarily the most helpful figure though, because it doesn’t just apply to women in labour, but to all people having epidurals, which means a lot of people who are very ill and having surgery.  Two retrospective studies that were included in that review did focus on evaluating epidurals in labouring women, and these found failure rates of 13.1% (Eappen et al 1988) and 14% (Pan et al 2004), although (as above) they each used different criteria for measuring success.

Those figures have been challenged more recently by a study by Thangamuthu et al (2013) who, in consultation with a group of consultant anaesthetists, created a standardised definition (thus addressing some of the problems mentioned above, although some uncertainty will always remain) and used this to measure the epidural failure rate within a large UK hospital Trust.  Their standardised tool used several means of measuring epidural success/failure which included asking women for their own assessment at a follow-up visit.  (This is a great idea, though it is interesting to note Thangamuthu et al’s (2013) concern that even women who have to have their epidurals resited several times sometimes still describe it as successful).

Overall, the epidural failure rate in Thangamuthu et al’s research was 23%, which they acknowledge is at the upper end of published figures, but they also note that their standardised definition is being used in other hospitals with similar results.  So far, this is the best research in this area that I could find and, although I think it is important to share some of the uncertainties and previous data in this area with women so that they can take this into account in their decision-making, I think Thangamuthu et al’s (2013) epidural failure rate of 23% is probably the best data we have on this topic at the moment.


Anim-Somuah M, Smyth RMD, Jones L. Epidural versus non-epidural or no analgesia in labour. Cochrane Database of Systematic Reviews 2011, Issue 12. Art. No.: CD000331. DOI: 10.1002/14651858.CD000331.pub3.

Eappen S, Blinn A, Segal S (1998). Incidence of epidural catheter re-placement in parturients: a retrospective chart reviewInternational Journal of Obstetric Anesthesia 7: 220–5.

Pan PH, Bogard TD, Owen MD (2004). Incidence and characteristics of failures in obstetric neuraxial analgesia and anesthesia: a retro-spective analysis of 19,259 deliveries.  International Journal of Obstetric Anesthesia 13: 227–33

Thangamuthu A, Russell IF and Purva M (2013). Epidural failure rate using a standardised definition. International Journal of Obstetric Anesthesia 22(4):310-315

photo credit: archibald jude via photopin cc

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.