Background Postpartum haemorrhage (PPH) rates are a key indicator of maternal health. Massive PPH remains a severe morbidity and cause of maternal death. Rates may be changing with population demography and service provision, and are already established as quality measures of maternity services.

Aim To describe current PPH rates in a district general hospital and tertiary referral centre compared to reported levels, as part of the STOP (Surveillance and Treatment of PPH) study.

Methods Mixed methodology was used, including a prospective, observational study with weighted sampling representing 10 071 women delivering in two units in 1 year. Notes were reviewed in: all cases of electronic recording of blood loss ≥1000 ml, <24 ml or missing and a random selection of 1 in 6 cases 500–999 ml, 1 in 12 cases 25–499 ml.

Results There were significant differences between the units in demography, obstetric risk factors, computer systems and PPH rates. Total PPH rate (≥500 ml) was 34.7% (rates were 9.75, 4.12 and 2.08% using ≥1000, 1500 and 2000 ml respectively). Validation of electronic data with patient records showed 4.82% of cases needed re-categorisation (the majority had been underestimated).

Conclusion PPH rates are higher than previously reported. This newly described transcription error from paper to computer indicates an additional and significant underestimation of a serious obstetric complication with resultant implications for management and care of women. In the era of comparative performance management accurate data adjusted for individual risk factors must be used, to improve the quality of maternity services.