Bacterial meningitis following combined spinal–epidural analgesia for labour

We wish to comment on the case reported by Bouhemad et al. (Anaesthesia 1998; 53: 292–5) describing a case of Streptococcus salivarius meningitis following combined spinal–epidural (CSE) analgesia for labour. The authors imply that the spinal component of the CSE technique led to the introduction of the organism into the CSF resulting in meningitis in the postpartum period. However, their theory of initial bacterial contamination of the epidural space by the epidural needle and subsequent introduction of bacteria into the CSF after dural puncture by the spinal needle cannot be proven. Do the authors then believe that if an epidural block alone had been instituted instead of a CSE, meningitis would not have occurred? Although a few cases of meningitis have been associated with a CSE technique in obstetrics [1, 2], meningitis has also been described after both spinal and epidural blocks performed as single procedures. The relative risk of meningitis after a CSE compared to either a spinal or epidural alone is unknown. Overall the incidence of this and other infective neurological complications is perceived to be extremely low. Ready and Helfer [3] reported two cases of meningitis after epidural analgesia for labour. Ngan Kee et al. [4] described a case of an epidural abscess after epidural block for Caesarean section. Conversely, Dripps and Vandam [5] prospectively reported no cases of meningitis after 10 098 spinal anaesthetics as did Phillips et al. [6] after 10 440 such cases. The latter studies included patients undergoing obstetric and urological operations which are known to be associated with peri-operative bacteraemia. Another potential route of bacterial transmission is, as the authors mention, through contamination of the local anaesthetic/opioid solutions used for the procedure. Since no details are given as to the method of mixing these solutions, we contend that bacterial contamination, presumably introduced by the anaesthetist at that time, could have occurred from either the spinal or subsequent epidural injections prior to delivery. Bouhemad et al. surprisingly comment further on the difficulty in obtaining sterile local anaesthetic solutions in obstetric analgesia practice. In our unit the pharmacy provides ready to use sterile double-packed syringes, as do many others in the UK, containing a low-dose mixture of bupivacaine and fentanyl which is routinely checked for bacterial contamination. We are now confident using these mixtures both for intrathecal and for epidural use for CSE labour analgesia. The authors omit other important details including the incidence of infective neurological complications in relation to the number of deliveries in their unit, the proportion of women who require regional anaesthetic intervention (epidural, spinal or CSE) and indeed whether the CSE technique is routinely used to provide labour analgesia. Finally it would be interesting to know if their local practice has changed since this incident.