E. Conci1, M. Khatib2, F. Barakat3, P. Forouhi4, C. M. Malata2,5 1Cambridge University Medical School,Cambridge, , United Kingdom 2Cambridge University Hospitals NHS Trust,Plastic Surgery,Cambridge, , United Kingdom 3Barnet General Hospital,London, , United Kingdom 4Cambridge University Hospitals NHS Trust,Cambridge Breast Unit,Cambridge, , United Kingdom 5Anglia Ruskin University,Faculty Of Health Sciences,Cambridge, , United Kingdom
Introduction: Peri-operative blood loss contributes to the morbidity of patients undergoing immediate and delayed autologous free flap breast reconstruction post-mastectomy. The study aims to develop predictors for blood loss and transfusion requirements in microvascular breast reconstruction.
Methods:
A retrospective cohort study of autologous free flap breast reconstructions by a single plastic surgeon performed from January 2010 – December 2015 was conducted. Data from patient medical records were input into databases of electronic health records (EMR and EPIC). Data analysis was performed using STATA software.
Outcomes collected include haemoglobin drop (preoperative Hb – lowest Hb following surgery), estimated blood loss (EBL) and the total units of blood transfused (intra-operative and perioperative). Estimated blood loss (EBL) was the total weight of the swabs used during surgery minus their weigth prior to surgery while the estimated blood volume (EBV) was calculated using a validated formula [InBV= 70/√(BMI/22 )]. lnBV represents the indexed blood volume in ml/kg.
Variables studied pertain to the patient, namely age, BMI and chemotherapy status; the operation, namely timing, duration and extent (laterality, axillary clearance, mastectomy weight); and the flap transferred, namely its weight, number of pedicles and ischaemia time.
Results:
Of the 163 microvascular procedures, 133 were unilateral and 30 were bilateral.
The median estimated blood loss (EBL) was 640ml (IQR=407-1000) and the mean haemoglobin drop was 29.9g/L (±11.1). 44% of patients required a transfusion.
Multivariable linear regression analysis showed that immediate timing of reconstruction (p=0.016), concomitant lymph node dissection (p=0.01), increased duration of surgery (p<0.001) and higher mastectomy weight (p<0.001) were significantly associated with higher EBL.
Independent predictors of the likelihood of blood transfusion were bilateral reconstruction (p=0.02), lower estimated total blood volume (p=0.04) and higher mastectomy weight (p<0.001).
Haemoglobin drop was predicted by chemotherapy (including tamoxifen) within six months prior to surgery (p=0.006) and the duration of the operation (p=0.04).
Conclusion:
A significant association between higher EBL and transfusion requirements was found with several variables that denote the extent of the operation. This has provided empirical information to our unit in predicting perioperative blood loss and preoperative patient counselling. It has also enabled us to design measures to restrict blood loss in these “high-risk” patients and hence reduce patient morbidity and length of stay.