Aim: Ten-year outcome analysis of all pregnancies in diabetic women in a population of 1.5 million people.
Methods: Ascertainment of patients through the regional obstetrical computer, and by direct contact with each obstetrical unit. Retrospective assessment of early miscarriage of pregnancy from hospital records. Data are presented for the six smallest obstetrical units, the four smaller district hospitals, two larger teaching hospitals and for the regional referral centre.
Results: Nine hundred and eighty-six fetal outcomes were identified, 753 in mothers treated with insulin before the pregnancy, 131 in mothers in whom insulin was started for the first time during the pregnancy and 102 in mothers treated by diet only. Overall perinatal mortality rates were 35.8 per 1000 for those mothers booked and delivered at a local maternity unit, 28.9 per 1000 for those booked and delivered at the regional centre, but 75.0 per 1000 for those who had booked locally but were transferred to the centre mid-pregnancy. Information on blood glucose control before and during pregnancy was relatively poorly documented. For the available data at the regional centre, only 160 of the 416 mothers had an identifiable preconception HbA1c measurement (mean 7.9%, range 3.3-16.8%): at booking 360 of these mothers had a mean HbA1c of 7.5% and by the third trimester mean HbA1c was 6.3% (range 3.3-13.2%).
Conclusions: The outcome of pregnancy in a diabetic mother in Northern Ireland remains a higher risk than for the general population. There is evidence that results in the regional centre are better, but problems arise when transfers occur mid-pregnancy. Measurement and recording of blood glucose control at all stages before and during pregnancy is incomplete. Diabet. Med. 18, 546-553 (2001)