The decision for red cell transfusion is usually made empirically by clinicians. We aim to perform a two-year retrospective audit of obstetric red cell transfusions at a tertairy women's hospital to evaluate the (1) incidence of obstetric red cell transfusion, (2) indications for transfusion and (3) haemoglobin triggers and targets of transfusion. Obstetric patients who were given red cell or whole blood transfusion during their pregnancy and who delivered a live or still birth between 1 Jan 2014 to 31 Dec 2015 are included in the analysis. Findings of this study would provide insights into transfusion practice and appropriateness of obstetric red cell transfusion at a tertiary obstetric hospital in Singapore.

Studies suggest an inter-hospital variation in transfusion rates, despite the availability of transfusion guidelines. As part of an initiative to implement patient blood management at our institution, we conducted this retrospective cohort study with the aims of defining the incidence of red cell transfusion among our multiethnic obstetric population and evaluating transfusion practice, with regard to indications and appropriateness.

Ethics approval has been obtained from the institutional review board. Patients who fulfil study criteria would be identified from the hospital's delivery and blood bank registries. Further information would be extracted from the medical records and blood bank database. The delivery registry would also be accessed to obtain the total number of parturients who delivered for the study period.

The following data would be collected and a pro-forma completed:

• Demographic data: age, body mass index, gravidity, parity
• Obstetric data: gestation age at pregnancy booking, occurrence of obstetric conditions associated with bleeding, mode of delivery, estimated blood loss at delivery and antenatal prescription of iron therapy
• Transfusion data: indications for red cell transfusion, number of transfusion events, haemoglobin (Hb) levels at different time points of pregnancy, location of transfusion, adverse events associated with transfusion, activation of massive transfusion protocol

For the purpose of the study, a "transfusion event" is defined as the time interval between prescription of red cells and complete administration to patient. If multiple units were prescribed, all units are considered part of the single event if the time of prescription between consecutive units does not exceed 72 hours.

The incidence of transfusion would be defined as the percentage of patients ('cases') requiring red cell transfusion antenatally and up to 6 weeks postnatally among ALL patients who delivered a live or still birth in the period 2014-2015.

Statistical analysis would be performed using SPSS for windows, version 19 (IBM, Armonk, NY). Data would be presented as mean (+/- SD), median (IQR) or count (%), as appropriate. Missing data would be excluded from the analysis.

Inclusion Criteria:

• Any parturient who had delivered a live or still birth in the period 1 January 2014 to 31 December 2015 AND administered red cell or whole blood transfusion anytime during pregnancy and up to 6 weeks postnatally

Exclusion Criteria:

• Intrauterine foetal blood transfusions
• Pregnancies that resulted in miscarriages or stillbirth before 24 weeks of gestation

Parturients who delivered in the period 1 January 2014 to 31December 2015 (as identified in the hospital's delivery database) and who were administered red cell or whole blood transfusion anytime during pregnancy and up to 6 weeks postnatally (identified using hospital blood bank database) would be included in the study.