EDITOR:

A 38-yr-old ASA I, gravida 2 para 1, with a history of previous lower segment Caesarean section (LSCS) for breech presentation was scheduled for an elective LSCS in this pregnancy. The latest ultrasound, which was performed at 32 weeks, demonstrated an anterior grade III placenta praevia. The patient was very keen to proceed under spinal anaesthesia with full understanding of the likelihood of increased risk of bleeding and the need for transfusion of blood and conversion to general anaesthesia.

The baby was delivered in good condition via the placenta. The placenta was found to be strongly adherent to the uterine wall and was removed in a piecemeal fashion. Torrential bleeding ensued, requiring rapid administration of colloid, blood and blood products via a rapid infuser along with pharmacological treatment to enhance uterine contraction and stop bleeding. During the course of resuscitation, general anaesthesia was induced with Ketamine and Suxamethonium and invasive monitoring was instituted with a view to proceeding to hysterectomy. The uterus was found to be adherent to the bladder and a subtotal hysterectomy was performed with repair of bladder. It was estimated that approximately 22 L of blood loss occurred. The patient required 29 units of packed red blood cells, 8 units of Fresh Frozen Plasma, 3 pools of platelets, 3 pools of cryoprecipitate, 2.5 L of crystalloid and 4.5 L of synthetic colloid (starch and gelatin) with no achievement of haemostasis. At this point, after discussion with the consultant haematologist, a bolus of 7.2 mg of rFVIIa was given intravenously despite the pH being 7.19 on arterial blood gas monitoring. The results were dramatic in terms of both successful haemostasis of the surgical field and ongoing fluid resuscitation requirements. Post operatively the patient was nursed in the intensive therapy unit (ITU) and required only one additional pool of platelets on ITU. She was electively ventilated overnight and was extubated the following morning. The patient made an uneventful recovery thereafter and was discharged from the hospital on the sixth postoperative day. Histology demonstrated placenta percreta.

Massive obstetric haemorrhage is a major cause of maternal mortality [1]. CEMACH (Confidential Enquiry in Maternal and Child Health) suggests risk of maternal mortality with low-lying placenta may be as high as 1 : 200. There is increasing experience of use of rFVIIa in cases of intractable haemorrhage from other causes. rFVIIa promotes clot formation by acting at a number of points in coagulation cascade [Reference Karalapillai and Popham2]. The data sheet for rFVIIa suggests potential thrombotic complications (0.6%); however, data from rFVIIa extended use have lack of thrombotic complications in acute bleeding episodes. Use of rFVIIa in major obstetric haemorrhage is ‘off label’ in the UK [Reference Karalapillai and Popham2] and hence it required approval from a consultant haematologist before use. The dose of rFVIIa ranges from 15 to 120 μg kg⁻¹. There is debate regarding the optimal timing of administration of rFVIIa. Loudon and Smith [Reference Loudon and Smith3] suggest that according to their hypothetical model in terms of cost effectiveness, the optimum time for administration of rFVIIa is after transfusion of 14 units of red cells as cost neutrality was maintained at this point even when two doses of rFVIIa were required. These findings need to be distributed widely as cost continues to be a contraindication for its use in some centres. Currently, there is lack of evidence-based guidelines on use of rFVIIa in major obstetric haemorrhage but this is an area that might merit further study.