Jessica Y Liu1,2, Ryan J Ellis3,4, Q Lina Hu3,5, Mark E Cohen3, David B Hoyt3,6, Anthony D Yang4, David J Bentrem4, Clifford Y Ko3,5, Timothy M Pawlik7, Karl Y Bilimoria3,4, Ryan P Merkow3,4. 1. Division of Research and Optimal Patient Care, American College of Surgeons, Chicago, IL, USA. jessica.liu@facs.org. 2. Department of Surgery, Emory University, Atlanta, GA, USA. jessica.liu@facs.org. 3. Division of Research and Optimal Patient Care, American College of Surgeons, Chicago, IL, USA. 4. Surgical Outcomes and Quality Improvement Center, Department of Surgery, Northwestern University, Chicago, IL, USA. 5. Department of Surgery, David Geffen School of Medicine, University of California Los Angeles, Los Angeles, CA, USA. 6. Department of Surgery, University of California Irvine, Irvine, CA, USA. 7. Department of Surgery, The Ohio State University Wexner Medical Center, Columbus, OH, USA.
Abstract
BACKGROUND: Post hepatectomy liver failure (PHLF) is associated with significant perioperative morbidity and mortality. A tool to identify patients at risk for PHLF may allow for earlier intervention to mitigate its severity and help clinicians when counseling patients. Our objective was to develop a PHLF risk calculator. STUDY DESIGN: Patients who underwent hepatectomy for any indication from 2014 to 2017 were identified from ACS NSQIP. A multivariable logistic regression model was developed that included preoperative and intraoperative variables. Model fit was assessed for discrimination using the C-statistic, and calibration using Hosmer and Lemeshow (HL) Chi square. Validation of the calculator was performed utilizing tenfold cross validation. RESULTS: Among 15,636 hepatectomy patients analyzed, the overall incidence of clinically significant PHLF was 2.8%. Preoperative patient factors associated with increased PHLF were male gender, preoperative ascites within 30 days of surgery, higher ASA class, preoperative total bilirubin greater than 1.2 mg/dl, and AST greater than 40 units/l. Disease related factors associated with PHLF included histology, and use of neoadjuvant therapy. Intraoperative factors associated with PHLF were extent of resection, open surgical approach, abnormal liver texture, and biliary reconstruction. The calculator's C-statistic was 0.83 and the HL Chi square was 10.9 (p = 0.21) demonstrating excellent discrimination and calibration. On tenfold cross validation, the mean test group C-statistic was 0.82 and the HL p value was 0.26. CONCLUSION: We present a multi-institutional preoperative and early postoperative PHLF risk calculator, which demonstrated excellent discrimination and calibration. This tool can be used to help identify high-risk patients to facilitate earlier interventions.
BACKGROUND: Post hepatectomy liver failure (PHLF) is associated with significant perioperative morbidity and mortality. A tool to identify patients at risk for PHLF may allow for earlier intervention to mitigate its severity and help clinicians when counseling patients. Our objective was to develop a PHLF risk calculator. STUDY DESIGN:Patients who underwent hepatectomy for any indication from 2014 to 2017 were identified from ACS NSQIP. A multivariable logistic regression model was developed that included preoperative and intraoperative variables. Model fit was assessed for discrimination using the C-statistic, and calibration using Hosmer and Lemeshow (HL) Chi square. Validation of the calculator was performed utilizing tenfold cross validation. RESULTS: Among 15,636 hepatectomy patients analyzed, the overall incidence of clinically significant PHLF was 2.8%. Preoperative patient factors associated with increased PHLF were male gender, preoperative ascites within 30 days of surgery, higher ASA class, preoperative total bilirubin greater than 1.2 mg/dl, and AST greater than 40 units/l. Disease related factors associated with PHLF included histology, and use of neoadjuvant therapy. Intraoperative factors associated with PHLF were extent of resection, open surgical approach, abnormal liver texture, and biliary reconstruction. The calculator's C-statistic was 0.83 and the HL Chi square was 10.9 (p = 0.21) demonstrating excellent discrimination and calibration. On tenfold cross validation, the mean test group C-statistic was 0.82 and the HL p value was 0.26. CONCLUSION: We present a multi-institutional preoperative and early postoperative PHLF risk calculator, which demonstrated excellent discrimination and calibration. This tool can be used to help identify high-risk patients to facilitate earlier interventions.
Authors: Lily V Saadat; Debra A Goldman; Mithat Gonen; Kevin C Soares; Alice C Wei; Vinod P Balachandran; T Peter Kingham; Jeffrey Drebin; William R Jarnagin; Michael I D'Angelica Journal: J Am Coll Surg Date: 2021-05-21 Impact factor: 6.532