Antibiotic prophylaxis for caesarean section: exchanging one risk for another?


  • No external funding and no competing interests declared. Published with the patient’s written consent.

In April 2011, our unit changed the timing of prophylactic antibiotic administration at caesarean delivery from after delivery to before incision, in accordance with published recommendations [1–3]. To reduce the risk of drug error, and since antibiotic and thiopental substitutions are more common in obstetric anaesthesia [4], we started mixing co-amoxiclav in a 30-ml syringe, to distinguish it visually from the 20-ml syringe of thiopental. Despite this measure, we have had a serious obstetric drug error.

Following a fetal scalp pH of 7.19 at 9 cm cervical dilatation, a decision was made to perform either a caesarean section or an instrumental delivery. The anaesthetic plan was to top up the well-functioning epidural, and convert to general anaesthesia if the block proved inadequate. General anaesthetic drugs (not routinely pre-prepared in our unit) were drawn up and placed in a separate tray on the anaesthetic worktop in theatre. The syringe of thiopental was appropriately labelled in a 20-ml syringe and lay next to the empty thiopental vial with syringes of suxamethonium and atracurium. The antibiotic had not been drawn up.

Following examination in theatre, the obstetrician decided to attempt a vaginal delivery. At about the same time, the anaesthetist gave an intravenous dose of what was thought to be co-amoxiclav, but was actually 500 mg thiopental. The error was rapidly recognised, cricoid pressure applied, suxamethonium given and the airway secured. The maternal oxygen saturation did not drop below 95%. The obstetric team proceeded immediately to a caesarean section, and the baby was delivered promptly. The baby has made a full recovery.

Our concern about antibiotic administration unfortunately proved justified. Fortunately, there was no regurgitation or aspiration in our case, and adverse outcomes for mother and baby were limited, though both the mother and her partner have understandably found the experience traumatic. It is possible that a caesarean delivery might have been avoided had the error not occurred.

We also have concerns about anaphylaxis to antibiotics administered pre-delivery; a maternal death from this cause has been described [5]. Whilst serious after delivery, there is only one patient to consider.

Following a serious case review, we now keep drugs for induction of general anaesthesia in obstetric theatres in a box with a lid to introduce a further barrier to accidental administration. Thiopental syringes are now triple-labelled to aid identification and all antibiotics are now double-checked with an anaesthetic assistant before administration.

We are also considering changing from thiopental to propofol in our obstetric practice. However, this will require training for all grades of anaesthetic staff. Our hospital is facing a sudden unavailability of thiopental, which may force us to use propofol anyway.

We have considered giving antibiotics by infusion rather than injection, but this might lead to accidental administration of phenylephrine, which is currently prepared in 100-ml bags of 0.9% sodium chloride.

We would be interested to hear what strategies other units have adopted to prevent this drug error occurring. Have we been too hasty in changing our practice? We feel that further discussion and surveillance are merited.