Development and Validation of Machine Learning Models to Predict Perioperative Transfusion Risk for Hip Fractures in the Elderly

Overview

Journal Ann Med

Publisher Informa Healthcare

Specialty General Medicine

Date 2024 Jun 20

PMID 38902847

Authors

Jiale Guo

Qionghan He

Yehai Li

Affiliations

Soon will be listed here.

Abstract

Background: Patients with hip fractures frequently need to receive perioperative transfusions of concentrated red blood cells due to preoperative anemia or surgical blood loss. However, the use of perioperative blood products increases the risk of adverse events, and the shortage of blood products is prompting us to minimize blood transfusion. Our study aimed to construct a machine learning algorithm predictive model to identify patients at high risk for perioperative transfusion early in hospital admission and to manage their patient blood to reduce transfusion requirements.

Methods: This study collected patients hospitalized for hip fractures at a university hospital from May 2016 to November 2022. All patients included in the analysis were randomly divided into a training set and validation set according to 70:30. Eight machine learning algorithms, CART, GBM, KNN, LR, NNet, RF, SVM, and XGBoost, were used to construct the prediction models. The models were evaluated for discrimination, calibration, and clinical utility, and the best prediction model was selected.

Results: A total of 805 patients were included in the study, of whom 306 received transfusions during the perioperative period. We screened eight features used to construct the prediction model: age, fracture time, fracture type, hemoglobin, albumin, creatinine, calcium ion, and activated partial thromboplastin time. After evaluating and comparing the performance of each of the eight models, the model constructed by the XGBoost algorithm had the best performance, with MCC values of 0.828 and 0.939 in the training and validation sets, respectively. In addition, it had good calibration and clinical utility in both the training and validation sets.

Conclusion: The model constructed by the XGBoost algorithm has the best performance, using this model to identify patients at high risk for transfusion early in their admission and promptly incorporating them into a patient blood management plan can help reduce the risk of transfusion.

References

Vamvakas E, Blajchman M . Transfusion-related mortality: the ongoing risks of allogeneic blood transfusion and the available strategies for their prevention. Blood. 2009; 113(15):3406-17. DOI: 10.1182/blood-2008-10-167643. View

Adunsky A, Lichtenstein A, Mizrahi E, Arad M, Heim M . Blood transfusion requirements in elderly hip fracture patients. Arch Gerontol Geriatr. 2003; 36(1):75-81. DOI: 10.1016/s0167-4943(02)00059-6. View

Ferraris V, Davenport D, Saha S, Austin P, Zwischenberger J . Surgical outcomes and transfusion of minimal amounts of blood in the operating room. Arch Surg. 2012; 147(1):49-55. DOI: 10.1001/archsurg.2011.790. View

Vivien B, Langeron O, Morell E, Devilliers C, Carli P, Coriat P . Early hypocalcemia in severe trauma. Crit Care Med. 2005; 33(9):1946-52. DOI: 10.1097/01.ccm.0000171840.01892.36. View

Polanco-Garcia M, Capielo A, Miret X, Chamero A, Sainz J, Revilla E . Effectiveness of a patient blood management protocol on reduction of allogeneic red blood cell transfusions in orthopedic surgery. Med Clin (Barc). 2018; 152(3):90-97. DOI: 10.1016/j.medcli.2018.04.021. View

Cho B, Serini J, Zorrilla-Vaca A, Scott M, Gehrie E, Frank S . Impact of Preoperative Erythropoietin on Allogeneic Blood Transfusions in Surgical Patients: Results From a Systematic Review and Meta-analysis. Anesth Analg. 2019; 128(5):981-992. DOI: 10.1213/ANE.0000000000004005. View

Riley R, Ensor J, Snell K, Harrell Jr F, Martin G, Reitsma J . Calculating the sample size required for developing a clinical prediction model. BMJ. 2020; 368:m441. DOI: 10.1136/bmj.m441. View

Farrow L, Brasnic L, Martin C, Ward K, Adam K, Hall A . A nationwide study of blood transfusion in hip fracture patients : linked analysis from the Scottish Hip Fracture Audit and the Scottish National Blood Transfusion Service. Bone Joint J. 2022; 104-B(11):1266-1272. DOI: 10.1302/0301-620X.104B11.BJJ-2022-0450.R1. View

Kim J, Shin H, You H, Park Y, Ahn K, Jung J . Effect of a Patient Blood Management Program on the Appropriateness of Red Blood Cell Transfusion and Clinical Outcomes in Elderly Patients Undergoing Hip Fracture Surgery. J Korean Med Sci. 2023; 38(8):e64. PMC: 9970785. DOI: 10.3346/jkms.2023.38.e64. View

10.

Blanco Rubio N, Llorens Eizaguerri M, Seral Garcia B, Burillo Fuertes P, Ranera Garcia M, Albareda Albareda J . [Preoperative intravenous iron as a strategy for blood saving in surgery for hip fracture]. Med Clin (Barc). 2012; 141(9):371-5. DOI: 10.1016/j.medcli.2012.09.029. View

11.

Magnotti L, Bradburn E, Webb D, Berry S, Fischer P, Zarzaur B . Admission ionized calcium levels predict the need for multiple transfusions: a prospective study of 591 critically ill trauma patients. J Trauma. 2011; 70(2):391-5. DOI: 10.1097/TA.0b013e31820b5d98. View

12.

Carson J, Grossman B, Kleinman S, Tinmouth A, Marques M, Fung M . Red blood cell transfusion: a clinical practice guideline from the AABB*. Ann Intern Med. 2012; 157(1):49-58. DOI: 10.7326/0003-4819-157-1-201206190-00429. View

13.

Rineau E, Chaudet A, Chassier C, Bizot P, Lasocki S . Implementing a blood management protocol during the entire perioperative period allows a reduction in transfusion rate in major orthopedic surgery: a before-after study. Transfusion. 2016; 56(3):673-81. DOI: 10.1111/trf.13468. View

14.

Beal E, Bagante F, Paredes A, Akgul O, Merath K, Cua S . Perioperative use of blood products is associated with risk of morbidity and mortality after surgery. Am J Surg. 2018; 218(1):62-70. DOI: 10.1016/j.amjsurg.2018.11.015. View

15.

Potter L, Doleman B, Moppett I . A systematic review of pre-operative anaemia and blood transfusion in patients with fractured hips. Anaesthesia. 2015; 70(4):483-500. DOI: 10.1111/anae.12978. View

16.

Sugand K, Ali R, Goodall R, Salciccioli J, Marshall D, Schuster-Bruce J . Trends in neck of femur fracture incidence in EU15+ Countries from 1990-2017. Injury. 2022; 54(2):645-651. DOI: 10.1016/j.injury.2022.11.038. View

17.

Beam A, Kohane I . Big Data and Machine Learning in Health Care. JAMA. 2018; 319(13):1317-1318. DOI: 10.1001/jama.2017.18391. View

18.

Wray J, Bridwell R, Schauer S, Shackelford S, Bebarta V, Wright F . The diamond of death: Hypocalcemia in trauma and resuscitation. Am J Emerg Med. 2021; 41:104-109. DOI: 10.1016/j.ajem.2020.12.065. View

19.

Warner M, Schulte P, Hanson A, Madde N, Burt J, Higgins A . Implementation of a Comprehensive Patient Blood Management Program for Hospitalized Patients at a Large United States Medical Center. Mayo Clin Proc. 2021; 96(12):2980-2990. PMC: 8649051. DOI: 10.1016/j.mayocp.2021.07.017. View

20.

Beard J, Officer A, Araujo de Carvalho I, Sadana R, Pot A, Michel J . The World report on ageing and health: a policy framework for healthy ageing. Lancet. 2015; 387(10033):2145-2154. PMC: 4848186. DOI: 10.1016/S0140-6736(15)00516-4. View