Uterine rupture and epidural analgesia: a spurious risk factor?

Authors


Sir,

We were interested to read of the study by Zwart et al.1 assessing factors involved in uterine rupture in the Netherlands.

In 182 of the 210 cases of uterine rupture they reported, the uterus was scarred by a previous caesarean delivery. The authors infer from this that the best prevention is to reduce the primary caesarean delivery rate. Zwart et al. state that the caesarean delivery rate in the Netherlands (13%)2 is ‘among the lowest in the world’. This comment is astonishing. The third world accounts for 88% of births worldwide, and in rural sub-Saharan Africa, for example, caesarean delivery is unavailable to many, and the overall rate is estimated to be around 1%, but uterine rupture is relatively common. In an audit of 8070 caesarean deliveries in Malawi, uterine rupture was present in 4.1%; 74% of the ruptures occurred in an unscarred uterus.3 It follows that worldwide the major cause of uterine rupture is the absence of caesarean delivery.

Among the possible risk factors for uterine rupture epidural ‘anaesthesia’ is listed.1 This probably refers mainly to epidural analgesia for labour. This dangerous implication is likely to be seized upon by the vocal anti-epidural lobby. It is, however, grossly misleading because, as the authors acknowledge, they were unable to adjust their risk factors for the many obvious confounding variables. The epidural rate in the Netherlands is given as 5.9%, but the rate in the high-risk LEMMoN population is 40.1%, affirming that epidural analgesia is used preferentially, and appropriately, in high-risk women. No causal relationship between epidural analgesia and uterine scar rupture can be inferred.

When the use of epidural analgesia began to grow in the 1970s, it was suggested that a scarred uterus was a contraindication, as the signs of impending rupture might be obscured. The many years of experience since have tended to dispel this fear. As Zwart et al. confirmed, pain is not a reliable indicator of uterine rupture: it does not always persist between contractions or cause localised tenderness. In 1976 Crawford described several patients in whom epidural analgesia allowed the transmission of pathological pain from placental abruption or scar dehiscence, but blocked the pain of uterine contraction.4 He termed this phenomenon the epidural sieve. Experienced anaesthetists have all seen cases in which impending rupture causes pain that is indistinguishable from contraction pain, but breaks through a functioning epidural. Errors have arisen when such pain has prompted ever-increasing epidural doses rather than an investigation of the cause. It must be acknowledged that converting epidural analgesia to anaesthesia with a large epidural bolus of local anaesthetic, with or without opioid, will obtund any pathological pain, such as, for example, taking a scalpel and incising the lower abdomen. The possibility of impending scar rupture should always be considered when a parturient complains of abdominal pain despite a working epidural and two warm dry feet. Such a block, combined with careful evaluation of the fetus, can therefore assist rather than obscure the diagnosis of uterine rupture.

Ancillary