Objective : To determine the optimum surgical operation for the prevention of recurrent esophageal varices bleeding in terms of quality-adjusted survival at 10 years (QAS 10 yr) from the choice of 4 types of surgical operations.Design : Decision analysis modeling, and review of the literature.Methods : A decision tree for the management of recurrent esophageal varices bleeding and the occurrence of post-operative complications was constructed, with surgical options of porto-systemic shunt (P-S shunt) small caliber mesocaval shunt (H-graft) distal spleno-renal shunt (DSRS) and esophageal devascularization (non - shunt). Probabilities of 30-days mortality, 5 or 10-year total survival post operative rebleeding and liver failure rates at 10 years, and mortality rates for re-bleeding and liver failure were abstracted from the world literature via MEDLINE from which the quality-adjusted survival at 10 years was derived. The quality adjustment or utility was obtained from the Euro QOL classification system. Special scenarios in which 30-day operative mortality and later mortality probability were fixed were also considered, as a form of sensitivity analysis.Results : DSRS was found to be the procedure with highest QAS 10 yr, based on the available literature, but clinically the advantage may not be relevant. Other scenarios revealed other surgical options to be superior (non-shunt and H-graft operation) but also by small margins.Conclusion : The available literature reviewed here can not provide sufficient information to make an equivocal decision regarding the optimal operation for the prevention of recurrent esophageal varices bleeding. Economic approaches (efficiency studies) to this problem may yield plausible solutions.

Archives of Gynecology & Obstetrics ปี 2544, พฤษภาคม ปีที่: 9 ฉบับที่ 10 หน้า