Prediction of peripartum hysterectomy and end organ dysfunction in major obstetric haemorrhage

Abstract

Objectives The aims of this study are to determine the incidence and aetiology of major obstetric haemorrhage (MOH) in our population, to examine the success rates of medical and surgical interventions and to identify risk factors for peripartum hysterectomy and end organ dysfunction (EOD). Study design This prospective study from 2004 to 2007 was carried out in three Dublin maternity hospitals. Women were identified as having MOH if they received ≥5 units of red cell concentrate (RCC) acutely. Risk factors for hysterectomy or end organ dysfunction were calculated using logistic regression. Results One hundred and seventeen cases of MOH in 93,291 deliveries were identified (1.25/1000). The predominant cause was uterine atony. Haemostasis was achieved with medical therapy alone in 15% of cases. The hydrostatic balloon and the B-Lynch suture arrested bleeding in 75% and 40% of cases utilised respectively. Hysterectomy was required to arrest bleeding in 24% of women and 16% of women developed end organ dysfunction (11 had both). There was one maternal death. Independent risk factors for hysterectomy included the number of previous caesarean sections (OR 3.28, 95% CI 1.95–5.5), placenta praevia (OR 13.5, 95% CI 7.7–184), placenta accreta (OR 37.7, 95% CI 7.7–184), uterine rupture (OR 7.25, 95% CI 1.25–42) and the number of units of RCC transfused (OR 1.31, 95% CI 1.13–1.5). Independent risk factors for end organ dysfunction (EOD) were placenta accreta (OR 5, 95% CI 1.5–16.5), uterine rupture (OR 13.86, 95% CI 2.32–82), the number of RCC transfused (OR 1.31, 95% CI 1.13–1.5) and the minimum haematocrit recorded (OR 5.53, 95% CI 1.7–18). Conclusions MOH is complicated by hysterectomy in 24% and end organ dysfunction in 16% of cases. The risk of peripartum hysterectomy is increased with the number of previous caesarean sections, the aetiology of the bleed, namely placenta praevia/accreta or uterine rupture and the volume of blood transfused. Critically, failure to maintain optimal haematocrit during the acute event was associated with end organ dysfunction.