Clinician Overconfidence in the Visual Estimation of the Post-Hepatectomy Liver Remnant Volume: A Proximal Source of Liver Failure after Major Hepatic Resection?

Abstract

Introduction: Post-hepatectomy liver failure (PHLF) remains a significant source of morbidity and mortality in patients undergoing major hepatic resections. The accuracy of visually estimating the future liver remnant (FLR) without formal CT liver volumetry (CTLV) is unknown. Methods: Twenty physicians in diagnostic radiology, interventional radiology, and hepatopancreatobiliary surgery reviewed 20 CT scans of patients without underlying liver pathology. We evaluated clinician accuracy to estimate the FLR for three proposed major hepatic resections: left hepatectomy (LH), right hepatectomy (RH), and right trisectionectomy (RT). The percent-difference between the mean and actual CTLV volume was tested along with specialty differences using mixed effects regression analysis. Results: The actual FLR (based on CTLV) remaining after LH ranged from 59-75% (physician estimated range: 50-85%), 23-40% for RH (estimated range: 15-50%), and 13-29% for RT (estimated range: 8-39%). For LH, the mean FLR was visually underestimated by 95% of physicians (MPD FLR: 8%; range: 2-16%; P < 0.0001). For RH, the mean FLR was overestimated by 95% with a MPD of 22% (range: 6-45%; P < 0.0001). For RT, 90% overestimated the FLR by a MPD of 25% (range: 6-50%; P < 0.0001)(Figure). HPB surgeons overestimated the FLR for proposed RH and RT by a MPD of 25% and 34%, respectively. Conclusion: Visual estimation alone of the FLR is potentially dangerously inaccurate as compared to CT liver volumetry. Overestimation is most apparent in a small FLR where the risk of PHLF is greatest.