Ventriculoatrial or ventriculoperitoneal shunts in the treatment of hydrocephalus in children?

The data on all 881 primary or revision shunt operations performed on 158 paediatric patients treated in Gothenburg, Sweden from 1967 to 1984 and 101 patients treated in Oulu, Finland from 1968 to 1983 were pooled for the purpose of comparative evaluation of the function of ventriculoatrial (VA) and ventriculoperitoneal (VP) shunts. Ventriculoperitoneal shunting was the method of choice in Gothenburg and ventriculoatrial shunting in Oulu. The results of the 723 operations (305 VA and 418 VP shunts) were evaluated as the other 158 operations were for ventriculostomas, shunt removals and other procedures. 80 children had exclusively VA shunts and 133 children had exclusively VP shunts. Irrespective of the method of analysis the VP shunts were more frequently infected. The estimated relative risk for obstruction of the shunt (Meyer-Kaplan method) was shown to be significantly higher in VA shunts, but only at a low level of statistical significance (p less than 0.1). All other shunt complications were distributed uniformly in both groups. There was, however, a trend towards a higher mortality among children with exclusively VA shunts. Therefore it was concluded that despite the higher risk for infection in VP shunts, these still should be considered a safer choice, as the complications of VA shunts present greater risks.