Surgeons often underestimate the amount of blood loss in replacement surgeries

ObjectiveTo assess the accuracy of the clinically estimated blood loss (EBL) when compared with the actual blood loss (ABL) in replacement surgeries.MethodsThis prospective study was done in Sri Ramachandra Medical Centre from April 2011 to April 2013. Altogether 140 patients undergoing total hip replacement or total knee replacement were included with the inclusion criteria being patients with haemoglobin higher than 100 g/ml and coagulation profile within normal limits. Exclusion criteria were intake of anti-platelet drug or anti-coagulant, bleeding disorders, thrombotic episode, and haematological disorders. There were 65 men and 75 women. In this study, the consultants were free to use any clinical method to estimate the blood loss, including counting the blood-soaked mops and gauze pieces (estimating the volume of blood carried in all the mops and gauzes), measuring blood lost to suction bottles and blood in and around the operative field. The ABL was calculated based on a modification of the Gross’s formula using haematocrit values.ResultsIn 42 of the 140 cases, the EBL exceeded the ABL. These cases had a negative difference in blood loss (or DIFF-BL < 0) and were included in the overestimation group, which accounted for 30% of the study population. Of the remaining 98 cases (70%), the ABL exceeded the EBL. Therefore they were put into the underestimation group who had a positive difference in blood loss (DIFF-BL > 0). We found that when the average blood loss was small, the accuracy of estimation was high. But when the average blood loss exceeded 500 ml, the accuracy rate decreased significantly. This suggested that clinical estimation is inaccurate with the increase of blood loss.ConclusionThis study has shown that using clinical estimation alone to guide blood transfusion is inadequate. In this study, 70% of patients had their blood loss underestimated, proving that surgeons often underestimate blood loss in replacement surgeries.