Learning from adverse clinical outcomes: major obstetric haemorrhage in Scotland, 2003-05.

OBJECTIVE: To describe a system for learning from cases of major obstetric haemorrhage.
DESIGN: Prospective critical incident audit.
SETTING: All consultant-led maternity units in Scotland, between 1 January 2003 and 31 December 2005. POPULATION: Women suffering from major obstetric haemorrhage (estimated blood loss > or = 2500 ml or transfused > or = 5 units of blood or received treatment for coagulopathy during the acute event).
METHODS: Hospital clinical risk management teams reviewed local cases using a standard, national assessment pro forma. MAIN OUTCOME MEASURES: Standard of care provided and learning points identified.
RESULTS: Rate of major haemorrhage was 3.7 (3.4-4.0) per 1000 births. Pro formas returned for 517 of 581 reported cases (89%); 41% were delivered by emergency caesarean section (compared with 15% of all Scottish births). Uterine atony was the most common cause (250 women, 48%); 32% had multiple causes. A consultant obstetrician gave hands-on care to 368 (71%) and a consultant anaesthetist to 262 (50%). Placenta praevia as a cause was independently associated with consultant presence. Central venous pressure monitoring was used in 164 (31%) women, and 108 (21%) women were admitted to intensive care. Parity, blood loss, and placenta praevia as a cause were independently associated with peripartum hysterectomy (performed in 62 women, 12%). Balloon tamponade and haemostatic uterine suturing were successful in 92 of 116 women (79%). Most cases were assessed as well managed, with 'major suboptimal' care identified in only 14 cases (3%).
CONCLUSIONS: It is feasible to identify and assess cases of major obstetric haemorrhage prospectively on a national basis. Most women received appropriate care, but many learning points and action plans were identified.