Validity of the Hewlett-Packard actograph in detecting fetal movements.

OBJECTIVE: To detect fetal movements during non-stressed cardiotocography to evaluate the validity of the Hewlett-Packard Doppler ultrasound actograph in detecting fetal movements. METHODS: This was a prospective, observational study. Thirty healthy pregnant women were divided into two gestational age groups (Group I, 31 + 0 to 34 + 0 weeks, n = 15; Group II, 37 + 0 to 40 + 0 weeks, n = 15). A Hewlett-Packard M-1350-A actocardiograph was used to make recordings of 45 min for each woman. Fetal movements were also detected sonographically by an observer and they were recorded independently by the pregnant woman. Ninety consecutive 30-s periods were analyzed to study agreement between the three techniques. RESULTS: The mean total agreement for fetal activity plus quiescence between ultrasound and the actograph was 63.7% for Group I, and 72.2% for Group II. In cases of fetal activity only, these values were 45.8% and 47.6%, respectively. Total agreements between ultrasound and the pregnant woman were 74.0% and 79.0%, respectively, for Groups I and II. Cohen's kappa values for total agreement between ultrasound and actograph were 0.31 (95% CI, 0.27-0.36), indicating 'fair' agreement in Group I, and 0.43 (95% CI, 0.38-0.47), indicating 'moderate' agreement in Group II. CONCLUSIONS: Use of the Hewlett-Packard actograph may offer some additional information during routine cardiotocography. However, our results were disappointing; the technique appears not to be suitable for detailed recordings of fetal behavior, and may even be misleading when difficult cardiotocographic patterns have to be interpreted.