Abstract

Perinatal mortality rates are an important index of the performance of perinatal health care services, but comparisons are confounded by variations in the prior risk status of the clienteles of different districts and different maternity units. A method of allowing for these differences has been devised. It is based jointly upon the exclusion of certain classes of birth, and on indirect standardisation for birthweight and a number of modifying factors. The method is described, tested, demonstrated, and proposed for more general use.