Editor—The finding of a persistently poor outcome of pregnancy in women with insulin dependent diabetes in two (northern) English regions is an important statement of the problem.1 2 Both studies provide figures and show outcomes that are no different from those widely reported in the past. Unfortunately, neither give evidence of any degree of centralisation of obstetric or diabetic care, with on-site neonatal intensive care, although this is a proved means of improving the outcome of pregnancy for diabetic mothers.3 The St Vincent declaration guidelines on the outcome of pregnancy, referred to in the accompanying editorial (p 263), are based on the Scandinavian reports held up as examples of good practice and state that “an interdisciplinary team should provide centralized diabetic pregnancy care in a hospital treating at least 20–30 cases a year. Pregnant diabetic patients should regularly visit the centre, before, during, and after the pregnancy.”

The combined diabetes pregnancy clinic at the Royal Maternity Hospital in Belfast has existed for over 40 years, and outcome audit has shown the value of this approach.4 In an audit of over 800 pregnancies in diabetic mothers identified in Northern Ireland over the past 10 years the perinatal mortality in mothers cared for throughout at this centralised clinic during 1985–95 was 27/1000, compared with 70/1000 for patients referred later in pregnancy and 33/1000 for those mothers cared for in other maternity hospitals in Northern Ireland.

Overall perinatal mortality for the whole population of Northern Ireland (1.5 million) during this decade was 9.3/1000 total births, so that even at a centralised clinic there is still an increased risk in diabetic pregnancy. Centralisation of care and improved cooperation among the obstetricians and diabetes physicians within a health region will lead to a better outcome.5

Editor—Casson et al make several misrepresentations …