Complications and Predictors of Morbidity for Hip Fracture Surgery in Patients with Chronic Liver Disease

Background: Despite the fact that patients with chronic liver disease (CLD) are at increased risk of complications after a fracture of the hip, there remains little information about the risk factors for acute postoperative complications and their overall outcome. The aim of this study was to describe inpatient postoperative complications and identify predictors of postoperative morbidity. Methods: Patients with CLD who had been treated for a fracture of the hip between April 2005 and August 2019 were identified from a retrospective search of an intramural trauma registry based in the Northeastern United States. Medical records were reviewed for baseline demographics, preoperative laboratory investigations, and outcomes. Results: The trauma registry contained 110 patients with CLD who had undergone surgery for a fracture of the hip. Of these, patients with a platelet-count of ⩽100,000/µL were 3.81 (95% CI, 1.59–9.12) times more likely to receive a transfusion than those with a platelet-count of >100,000/µL. Those with a Model for End-stage Liver Disease (MELD) score of >9 were 5.54 (2.33–13.16) times more likely to receive a transfusion and 3.97 (1.06–14.81) times more likely to develop postoperative delirium than those with a MELD score of ⩽9. Of patients without chronic kidney disease, those with a creatinine of ⩾1.2 mg/dL were 6.80 (1.79–25.87) times more likely to develop acute renal failure (ARF) than those with a creatinine of <1.2 mg/dL. In a multivariable model, as MELD score was increased, the odds of developing a composite postoperative complication, which included transfusion, ARF, delirium, or deep wound infection, were 1.29 (1.01–1.66). Other tools used to assess surgical risks, Charlson Comorbidity Index, Elixhauser, and American Society of Anesthesiologist scores, were not predictive. Conclusions: Patients with CLD who undergo surgery for a hip fracture have a high rate of postoperative complications which can be predicted by the preoperative laboratory investigations identified in this study and MELD scores, but not by other common comorbidity indices.