Basic principles of the foetal heart rate during delivery without hypoxia and acidosis.

BACKGROUND: Using naked-eye evaluation of foetal heart rate (FHR) patterns remains difficult and is not complete. Computer-aided analysis of the FHR offers the opportunity to analyse the FHR completely and to detect all changes due to hypoxia and acidosis. In order to better understand these changes FHR patterns in non-acidotic foetuses should be studied by first separating FHR into (i) basal FHR (baseline) and (ii) all decelerations.

METHODS: The FHR signals (i.e., R-R intervals) of 637 fetuses were recorded by a computer. To enter the study all foetuses must have been delivered by the vaginal route - in consequence without a significant loss of FHR signals. During forceps/vacuum delivery recordings were continued. If necessary a new electrode was inserted. Recordings of foetuses with chorioamnionitis and tracings of malformed neonates and tracings shorter than 30 min were excluded. No drugs were given to the mother during the time of recording. Thus 484 recordings were left. In this study only the last 30 min of each record were analysed using our own programmes written in MATLAB. 3 parameters were determined electronically: (i) the mean frequency (FRQ, bpm), (ii) the number of turning points (N/min), which we called 'microfluctuation' (MIC) and (iii) the oscillation amplitude (OA, bpm) (see Fig. 2). Computer analysis of the FHR offers the opportunity to separate baseline FHR from deceleration patterns using appropriate algorithms rearranging and sequencing all baseline segments (or all decelerations) to a new file. Therefore each of the 2 new files contains only one category of the FHR: baseline segments (with accelerations) only or decelerations only (Fig. 1). 1 min was always taken as the reference time interval. In order to exclude foetal hypoxia and acidosis during the last 30 min, a small pHUA -'window' was chosen (7.290 up to 7.310) using acid-base variables from umbilical arterial (UA) blood measured soon after delivery with RADIOMETER equipment (mainly ABL500) by trained personal.

RESULTS: Overall 14,520 min of the 484 foetuses were analysed by measuring in UA blood (X ± SD):pH=7.262 ± 0.065, pCO2 = 53.7 ± 8.8 mmHg, BEEcf,ox=-3.3 ± 2.5 mmol/l and sO2 = 23.9 ± 12.4%. In the whole sample and in non-acidotic (pHUA: 7.29-7.31) foetuses (N=50) there exist 3 fundamental principles which combine the 3 FHR variables under investigation: (I) MIC is strongly associated (r=0.631, P < 0.0001) with mean FRQ (bpm): in ca. 40% of all foetal heart beats a turning of the vector occurs (Fig. 4). (II) MIC is associated also with OA (r = -0.480, P < 0.0001); this regression is non-linear: Smaller band-widths are associated with increased MIC [OA = 0.0027 × MIC2 - 0.56 × MIC + 71 (see Fig. 5)]. (III) In non-acidotic foetuses lowering of the mean frequency niveau is associated with increased OA (overall: r = -0.349, P< 0.0001); Using baseline segments only: r = -0.283, Nmin=844, P<0.0001. This regression is linear again: OABL = -0.445 × FRQBL + 94.1. Overall a Delta frequency (ΔFRQ) of + 10 bpm leads to a ΔOA of -4.1 bpm. These 3 rules are valid in isolated baseline segments as well as during artificially isolated deceleration patterns.

CONCLUSIONS: FHR is a unit and should be analysed by computer-aided technologies as a unit. MIC, OA and FRQ belong together and their interaction can be described in non-acidotic foetuses by the 3 basic principles given above. Standard FHR tracings remain untouched.