Six prospective cohort studies (n=527, range three to 334) were included. None of the studies reported enrolling consecutive patients. All studies provided an adequate description of the spectral analysis and used valid outcome measures. All studies except one provided a clear definition of foetal distress. All studies reported that the same clinical data was available as would be available in practice. None of the studies explained reasons for withdrawals. Three studies provided data on uninterpretable/missing results.
Antepartum (two studies, n=29):
Both studies found that a decrease in spectral energy in the low frequency band was significantly associated with foetal distress (pO2 and pH in umbilical vein).
Intrapartum (four studies, n=197):
Three of the studies reported that spectral energy in the LF and HF bands decreased in the case of foetal distress (foetal hypoxia or acidaemia). Only one study provided sufficient data to calculate measures of diagnostic accuracy. This study reported a sensitivity of 97.5% and a specificity of 86.1% at an LF value of <0.0013 for the detection of foetal acidaemia.