A longitudinal study of computerised cardiotocography in early fetal growth restriction

Abstract

Objectives To explore if in early fetal growth restriction (FGR) the longitudinal pattern of short-term fetal heart rate (FHR) variation (STV) can be used for identifying imminent fetal distress and if abnormalities of FHR registration associate with two-year infant outcome. Methods The original TRUFFLE study assessed if in early FGR the use of ductus venosus Doppler pulsatility index (DVPI), in combination with a safety-net of very low STV and / or recurrent decelerations, could improve two-year infant survival without neurological impairment in comparison to computerised cardiotocography (cCTG) with STV calculation only. For this secondary analysis we selected women, who delivered before 32 weeks, and who had consecutive STV data for more than 3 days before delivery, and known infant two-year outcome data. Women who received corticosteroids within 3 days of delivery were excluded. Individual regression line algorithms of all STV values except the last one were calculated. Life table analysis and Cox regression analysis were used to calculate the day by day risk for a low STV or very low STV and / or FHR decelerations (DVPI group safety-net) and to assess which parameters were associated to this risk. Furthermore, it was assessed if STV pattern, lowest STV value or recurrent FHR decelerations were associated with two-year infant outcome. Results One hundred and fourty-nine women matched the inclusion criteria. Using the individual STV regression lines prediction of a last STV below the cCTG-group cut-off had a sensitivity of 0.42 and specificity of 0.91. For each day after inclusion the median risk for a low STV(cCTG criteria) was 4% (Interquartile range (IQR) 2% to 7%) and for a very low STV and / or recurrent decelerations (DVPI safety-net criteria) 5% (IQR 4 to 7%). Measures of STV pattern, fetal Doppler (arterial or venous), birthweight MoM or gestational age did not improve daily risk prediction usefully. There was no association of STV regression coefficients, a last low STV or /and recurrent decelerations with short or long term infant outcomes. Conclusion The TRUFFLE study showed that a strategy of DVPI monitoring with a safety-net delivery indication of very low STV and / or recurrent decelerations could increase infant survival without neurological impairment at two years. This post-hoc analysis demonstrates that in early FGR the day by day risk of an abnormal cCTG as defined by the DVPI protocol safety-net criteria is 5%, and that prediction of this is not possible. This supports the rationale for cCTG monitoring more often than daily in these high-risk fetuses. Low STV and/or recurrent decelerations were not associated with adverse infant outcome and it appears safe to delay intervention until such abnormalities occur, as long as DVPI is in the normal range.