Management of the univentricular connection: are we improving?

OBJECTIVE To assess the impact of the bidirectional cavopulmonary shunt, total cavopulmonary connection, and baffle fenestration on outcome of the Fontan operation in our unit. METHOD We reviewed 123 bidirectional cavopulmonary shunts and 264 Fontan operations performed from 1980 to 1995. Analysis of pulmonary artery size (right and left main and lower lobe branches) before and after bidirectional cavopulmonary shunt was performed. Outcome of the Fontan operation was analysed in various time periods to assess the effect of prior bidirectional cavopulmonary shunt, use of the lateral tunnel, and fenestration. RESULTS Operative risk for the bidirectional cavopulmonary shunt was 4% (CI = 2-10%) with a survival of 89% (CI = 83-95%) at 36 months. Probability of conversion to Fontan operation at 36 months was 49% (CI = 38-61%). Pulmonary artery size (Nakata and lower lobe indices) fell after bidirectional cavopulmonary shunt (P = 0.0006). Fontan risk dropped from 8.5% (1980-1987) to 1.8% (1988-1995) (P = 0.02), coinciding with the use of the bidirectional cavopulmonary shunt. There was no further risk reduction after introduction of the lateral tunnel and baffle fenestration, although these comparisons are limited by relatively small numbers. Duration of hospital stay related to pleural effusions was lowest for patients with a fenestrated lateral tunnel operation (P < 0.05). CONCLUSION The bidirectional cavopulmonary shunt is a suboptimal stimulus for pulmonary artery enlargement, but may reduce the risk of Fontan operation in selected children. Fenestrated lateral tunnel operations have reduced the duration of postoperative pleural effusions.