Computerised analysis of fetal heart rate recordings in patients with diabetes mellitus: the Dawes-Redman criteria may not be valid indicators of fetal well-being

Computerised cardiotocograph analysis has been used in our medical antenatal clinic for women with insulin dependent diabetes mellitus for five years. We had observed that many of the patients failed the computer criteria and wished to examine this finding in more detail. All cardiotocographs from 40 pregnant women with insulin dependent diabetes mellitus who delivered between 1992 and 1995 were reviewed. The number of traces failing the criteria were calculated. Outcome was compared between patients with a normal trace and those with an abnormal trace. 233 recordings were examined. 30.5 % of antenatal CTGs failed the criteria because of absent high episodes of variation (a parameter which is a specific marker of fetal well-bring). After excluding repeat traces on all patients a sample of 26 traces was examined. 34.6 % of these had failed because of absent high episodes of variation. Both these figures are significantly higher than the published normal values. There was no apparent difference in neonatal outcome between women whose trace had failed (9) and those with a normal trace (17). We suggest that the normal values of criteria used to analyse computerised CTGs may not be valid when applied to fetuses of women with diabetes mellitus.