An algorithm for predicting blood loss and transfusion risk after total hip arthroplasty

Transfus Apher Sci. 2018 Apr;57(2):272-276. doi: 10.1016/j.transci.2018.03.006. Epub 2018 Mar 27.

Abstract

Introduction: Patients receiving blood transfusions after total hip arthroplasty have increased morbidity and longer lengths of stay compared to non-transfused patients. The aim of this study is to create an algorithm in order to identify patients at risk for transfusion after total hip replacement and define a safe point in hemoglobin levels after which the need for blood, transfusion can be excluded.

Methods: This retrospective study analyzed hemoglobin (Hb) levels for 5 days in patients undergoing total hip replacement. An algorithm was implemented to identify the critical trends of Hb levels in the first two postoperative days, trying to identify the patients at high risk of transfusion. Specificity, sensibility and efficiency were calculated in relation to the capability of the algorithm to correctly identify transfused patients.

Results: The algorithm found a pre-operative Hb ≥ 13 g/dl as a cut off between patients at low-risk or high-risk for transfusion. When parameters were calculated considering the best efficiency with the least number of false negatives, the algorithm showed a specificity of 84% and a sensitivity of 70% with an efficiency of 80.6%. Hb values >10 g/dl in the first operative day for low-risk patients and Hb level > 11 g/dl the second post-operative day for high-risk patients led to exclusion of the need for transfusion.

Conclusions: The algorithm suggested critical Hb levels to predict transfusion. In association with clinical data, the suggested critical values of Hb may be useful to schedule lab tests and a safe early discharge.

Keywords: Algorithm; Arthroplasty; Blood; Hip; Transfusion.

MeSH terms

  • Adult
  • Aged
  • Aged, 80 and over
  • Algorithms
  • Arthroplasty, Replacement, Hip / adverse effects*
  • Blood Loss, Surgical / prevention & control*
  • Blood Transfusion / methods*
  • Female
  • Humans
  • Male
  • Middle Aged
  • Retrospective Studies