[Skip to Navigation]
Sign In
Citations 0
May 2011

Prediction and Statistical Analysis—Reply

Author Affiliations

Author Affiliations: CHU Hautepierre, Centre de Chirurgie Visc[[eacute]]rale et de Transplantation, Strasbourg, France.

Arch Surg. 2011;146(5):633-634. doi:10.1001/archsurg.2011.83

In reply

We very much appreciate the thoughtful comments of Dr Braillon regarding our article on predicting mortality rate after major hepatectomy. Postoperative liver failure is the main cause of death after extended liver resection, and several attempts at defining predictors of this lethal postoperative complication have been made in the past.

Unfortunately, all previously published studies have analyzed the occurrence of postoperative liver failure on the basis of postoperative biological parameters, once liver resection has already been performed. Our study attempts to go beyond postoperative assessment by identifying preoperative factors to truly predict and avoid mortality after liver resection in patients in which risk may exceed benefits. Dr Braillon takes issue with the size of our study. We would like to point out that our study is the first to show that a combination of easily obtainable biological parameters, measurable preoperatively in a simple blood sample, can predict the mortality rate in patients requiring right-sided major or extended major liver resection with extra-hepatic bile duct resection. The purpose was to show that the principle of predicting outcome in these patients up front is valid in a retrospective cohort of patients. Such a study intentionally will have a limited number of high-risk patients. It is reassuring to note that the results of our “short series” have recently been confirmed in a larger study by Breitenstein et al.1

Add or change institution