BACKGROUND: The measurement of the hepatic venous pressure gradient may identify a suboptimal response to beta-blockers in patients with varices at risk for bleeding. However, the cost-effectiveness of routine hepatic venous pressure gradient measurements to guide primary prophylaxis has not been examined. METHODS: We used decision analysis to evaluate two hepatic venous pressure gradient measurement strategies relative to standard beta-blocker therapy in a hypothetical cohort of patients with high-risk varices: (i) hepatic venous pressure gradient measurement 4 weeks after the initiation of beta-blocker therapy; and (ii) hepatic venous pressure gradient measurement prior to and 4 weeks after the initiation of beta-blocker therapy. The total expected costs, variceal bleeding episodes and deaths were calculated over a 1-year time horizon. RESULTS: Beta-blocker therapy was associated with total costs of $1464, seven variceal bleeding episodes, one variceal bleeding episode-related death and 15 deaths. One hepatic venous pressure gradient measurement was associated with total costs of $5015, four variceal bleeding episodes, one variceal bleeding episode-related death and 15 deaths. Two hepatic venous pressure gradient measurements were associated with total costs of $8657, four episodes of variceal bleeding, one variceal bleeding episode-related death and 15 deaths. Compared with beta-blocker therapy alone, the incremental costs per variceal bleeding episode prevented and death averted were, respectively, $108 185 and $355 100 (one hepatic venous pressure gradient measurement) and $202 796 and $719 300 (two hepatic venous pressure gradient measurements). The results were sensitive to the time horizon of the analysis, the probability of bleeding whilst on beta-blockers and the cost of hepatic venous pressure gradient measurement. CONCLUSION: Hepatic venous pressure gradient measurement to guide primary prophylaxis is an expensive strategy for reducing variceal bleeding or death, especially in patients with limited life expectancy, such as those with advanced, decompensated cirrhosis.
BACKGROUND: The measurement of the hepatic venous pressure gradient may identify a suboptimal response to beta-blockers in patients with varices at risk for bleeding. However, the cost-effectiveness of routine hepatic venous pressure gradient measurements to guide primary prophylaxis has not been examined. METHODS: We used decision analysis to evaluate two hepatic venous pressure gradient measurement strategies relative to standard beta-blocker therapy in a hypothetical cohort of patients with high-risk varices: (i) hepatic venous pressure gradient measurement 4 weeks after the initiation of beta-blocker therapy; and (ii) hepatic venous pressure gradient measurement prior to and 4 weeks after the initiation of beta-blocker therapy. The total expected costs, variceal bleeding episodes and deaths were calculated over a 1-year time horizon. RESULTS:Beta-blocker therapy was associated with total costs of $1464, seven variceal bleeding episodes, one variceal bleeding episode-related death and 15 deaths. One hepatic venous pressure gradient measurement was associated with total costs of $5015, four variceal bleeding episodes, one variceal bleeding episode-related death and 15 deaths. Two hepatic venous pressure gradient measurements were associated with total costs of $8657, four episodes of variceal bleeding, one variceal bleeding episode-related death and 15 deaths. Compared with beta-blocker therapy alone, the incremental costs per variceal bleeding episode prevented and death averted were, respectively, $108 185 and $355 100 (one hepatic venous pressure gradient measurement) and $202 796 and $719 300 (two hepatic venous pressure gradient measurements). The results were sensitive to the time horizon of the analysis, the probability of bleeding whilst on beta-blockers and the cost of hepatic venous pressure gradient measurement. CONCLUSION: Hepatic venous pressure gradient measurement to guide primary prophylaxis is an expensive strategy for reducing variceal bleeding or death, especially in patients with limited life expectancy, such as those with advanced, decompensated cirrhosis.