Sarah M Campbell1, Paul Corcoran1, Edel Manning1, Richard A Greene2. 1. National Perinatal Epidemiology Centre, Department of Obstetrics and Gynaecology, University College Cork, Ireland. 2. National Perinatal Epidemiology Centre, Department of Obstetrics and Gynaecology, University College Cork, Ireland. Electronic address: R.Greene@ucc.ie.
Abstract
BACKGROUND: The incidence of peripartum hysterectomy (PH) shows fifty-fold variation worldwide (0.2-10.5/1000 deliveries) and risk factors include advancing maternal age and parity, previous caesarean section (CS) and abnormal placentation. OBJECTIVES: In this first national study of PH in Ireland, our objectives were threefold: to describe the national trend in PH incidence over 15 years since 1999; to assess risk of PH associated with morbidly adherent placenta (MAP), placenta praevia and postpartum haemorrhage (PPH) during 2005-2013; and to describe the causes, interventions and outcomes of PH cases during 2011-2013. STUDY DESIGN: For the 15-year time-trend analysis, PH cases and denominator data were extracted from Ireland's Hospital In-Patient Enquiry database. Multivariate Poisson regression analysis assessed risk of PH associated with MAP, placenta praevia and PPH. In collaboration with the 20 Irish maternity units we carried out a three-year national clinical audit of severe maternity morbidity. PH was a notifiable morbidity and the audit included detailed review of MOH cases. RESULTS: In 1999-2013 there were 298 PH cases, a rate of 0.32/1000 deliveries. During the period 2005-2013, the PH rate was 50 times higher in deliveries involving PPH, 100 times higher with placenta praevia and 1000 times higher with MAP. During the clinical audit (2011-2013) there were 65 PH cases, a rate of 0.33/1000 deliveries, increasing with advancing age and parity. The reporting of abnormal placentation, primarily the co-occurrence of placenta praevia and MAP, was linked with previous CS. Fifty-six of the 65 cases suffered MOH, most commonly associated with placenta praevia, MAP and uterine atony. Prophylactic and therapeutic uterotonic agents were appropriately used in the majority of cases. CONCLUSIONS: The incidence of PH in Ireland has been consistently low over 15 years, averaging one case every 3000 deliveries. The recognised risk factors of MAP, placenta praevia and PPH were independently associated with PH, with MAP being by far the strongest predictor. The vast majority of PH cases in our clinical audit were associated with MOH. Some deficiencies were noted in antenatal care, in certain elements of treatment and clinical governance protocols but adherence to guidelines was generally high.
BACKGROUND: The incidence of peripartum hysterectomy (PH) shows fifty-fold variation worldwide (0.2-10.5/1000 deliveries) and risk factors include advancing maternal age and parity, previous caesarean section (CS) and abnormal placentation. OBJECTIVES: In this first national study of PH in Ireland, our objectives were threefold: to describe the national trend in PH incidence over 15 years since 1999; to assess risk of PH associated with morbidly adherent placenta (MAP), placenta praevia and postpartum haemorrhage (PPH) during 2005-2013; and to describe the causes, interventions and outcomes of PH cases during 2011-2013. STUDY DESIGN: For the 15-year time-trend analysis, PH cases and denominator data were extracted from Ireland's Hospital In-Patient Enquiry database. Multivariate Poisson regression analysis assessed risk of PH associated with MAP, placenta praevia and PPH. In collaboration with the 20 Irish maternity units we carried out a three-year national clinical audit of severe maternity morbidity. PH was a notifiable morbidity and the audit included detailed review of MOH cases. RESULTS: In 1999-2013 there were 298 PH cases, a rate of 0.32/1000 deliveries. During the period 2005-2013, the PH rate was 50 times higher in deliveries involving PPH, 100 times higher with placenta praevia and 1000 times higher with MAP. During the clinical audit (2011-2013) there were 65 PH cases, a rate of 0.33/1000 deliveries, increasing with advancing age and parity. The reporting of abnormal placentation, primarily the co-occurrence of placenta praevia and MAP, was linked with previous CS. Fifty-six of the 65 cases suffered MOH, most commonly associated with placenta praevia, MAP and uterine atony. Prophylactic and therapeutic uterotonic agents were appropriately used in the majority of cases. CONCLUSIONS: The incidence of PH in Ireland has been consistently low over 15 years, averaging one case every 3000 deliveries. The recognised risk factors of MAP, placenta praevia and PPH were independently associated with PH, with MAP being by far the strongest predictor. The vast majority of PH cases in our clinical audit were associated with MOH. Some deficiencies were noted in antenatal care, in certain elements of treatment and clinical governance protocols but adherence to guidelines was generally high.