Risks in therapeutic portacaval and splenorenal shunts.

Analyses of the records of 120 patients who underwent portacaval shunting (PCS, 57%) or splenorenal shunting (SRS, 43%) from 1966-1973 disclosed that patients in each group undergoing elective shunts had the same preoperative physical condition and postoperative mortality rates (approximately 20%). Although the post-operative death rate from emergency shunts was 48%, patients having these procedures were poorer risks. Long-term incidences of encephalopathy were the same, irrespective of the type of shunt (PCS, 46%; SRS 36%, P greater than 0.5). Despite comparisons of data most unfavorable for PCS, 5-year survival rates were also the same after either type of shunt (all PCS, 29 +/- 7.5%, SRS, 42.0 +/- 7.4%, P = 0.23). The survival rate after elective PCS was also the same as after SRS during the entire 5-year period. However, the survival after all elective PCS and SRS was significantly greater than after emergency PCS (P range = 0.005-0.038); the poorer results of emergency shunting could be partly attributed to the poorer condition of patients selected. A numerical score based on serum bilirubin concentrations, ascites, and urgency of shunting reliably predicts postoperative mortality. Long-term encephalopathy is predicted by a history of encephalopathy and the urgency of shunting.