Postpartum Haemorrhage
: Defining Incidence and Modelling Risk Factors to Predict Different Thresholds of Blood Loss

Student thesis: Doctoral ThesisDoctor of Philosophy


Background: Postpartum haemorrhage (PPH) remains a major cause of maternal mortality and morbidity, and in recent years there has been a temporal increase in the incidence of PPH and associated morbidities in resource rich countries. Individual risk factors for PPH have long been identified but the relative importance of each has been under explored and thus the potential for preventative strategies is unknown.

Aim: The aim of this study was to, i) ascertain the incidence of PPH at various thresholds in a South of England population, ii) identify the relative importance of predictor variables associated with PPH at different blood loss thresholds and iii) identify the independent and cumulative association of prepregnancy, pregnancy acquired and intrapartum variables on estimated blood loss following birth.

Methodology: A prospective observational study was undertaken in two maternity services. Estimated blood loss data for all women (n=10,213) were imported from NHS electronic summary records. A representative sample of cases (n=1897) was selected for review, using a weighted sampling strategy. Univariate analysis identified variables associated with mean estimated blood loss and PPH at various thresholds. Multivariate regression modelling assessed the association of sequentially acquired variables with PPH ≥500 ml, ≥1000 ml, and ≥1500 ml.

Results: The incidence of PPH ≥500 ml, ≥1000 ml ≥1500 ml ≥2000 ml and ≥2500 ml was 33.9% (95%CI 31.4 to 36.5), 9.4% (95%CI 8.5 to 10.4), 4.0% (95%CI 3.4 to 4.6), 2.0% (95%CI 1.6 to 2.4) and 0.8% (95%CI 0.7 to 1.0) respectively. Incidence of PPH ≥1000 ml was investigated by mode of birth. The incidence for spontaneous vaginal birth (SVD) was 4.75% (95%CI0.37 to 3.7) and instrumental vaginal birth, 12.1% (95%CI 9.3 to 14.6). The incidence following abdominal birth was 18.2% (95%CI 15.8 to 20.7); elective CS 11.8% (95%CI8.9 to 14.5), emergency CS 22% (95%CI18.6 to 25.4). Multiple regression analyses identified different independent variables associated with overall PPH at different thresholds. Novel independent variables resulting from this study associated with PPH at varying levels, were Black African ethnicity (≥500 ml and ≥1000 ml) OR 1.68 (95%CI 1.23 to 2.28) and OR 1.50 (95%CI1.13to 1.98), assisted conception (≥500 ml) OR 3.80 (95%CI1.69 to 8.57), antenatal attendance feeling ‘generally unwell’ (≥500 ml) OR 2.03 (95%CI1.18 to 3.49), antenatal steroid administration for fetal reasons (≥1500 ml) OR 2.00 (95%CI 1.17 to 3.41). In addition some previously known variables were confirmed. These were the impact per unit of BMI (Kg/m2) OR 1.04 (95%CI 1.01 to 1.04); previous PPH (≥500, ≥1000, ≥1500) 2.75 (95%CI1.40 to 5.44) 1.88 (96%CI 1.13 to 3.11) 2.39 (95%CI1.33 to4.28) multiple pregnancy (≥1000, ≥1500) 2.33 (95%CI1.23 to 4.41) 2.60(95%CI1.27 o 5.38) retained placenta (≥1000) 7.51 (4.08 to 13.8), interval to suturing (≥1000 ml) 1.74 (95%CI1.46 to 2.08). There was also a liner association with maternal temperature in labour and level of PPH.

Conclusion: This study found higher rates of PPH at all thresholds and, with all modes of birth. Which is not fully explained by rising Caesarean section rates. Prepregnancy and pregnancy acquired variables are commonly mediated through intrapartum events, and previous pregnancy management can impact on blood loss in subsequent pregnancies. Novel variables found in this study require further investigation, particularly the impact of Black African ethnicity, assisted conception techniques, antenatal steroid administration, and, feeling “generally unwell”. Modifiable risk factors include preconceptual weight loss, expedient suturing of genital tract trauma and regular recording of maternal temperature in labour, which may alert staff to higher risk of PPH.

Date of Award2014
Original languageEnglish
Awarding Institution
  • King's College London
SupervisorJane Sandall (Supervisor) & Rachel Tribe (Supervisor)

Cite this