When we think of large simple randomised trials we might think of maybe studies done in cardiovascular disease or cancer and testing one or more drug treatments. But the need for reliable evidence from large randomised trials extends across all areas of medicine and trialists have applied themselves in many areas to develop this evidence. One such area is pregnancy and childbirth.
An excellent example of such a randomised trial is the INFANT trial recently published in the Lancet. The trial set out to assess whether automated computerised interpretation of fetal heart rate during childbirth improved major outcomes, including neonatal death and development at age two.
All women in the trial were having continuous electronic fetal monitoring during labour with half randomised to usual care where interpretation of the fetal heart rate was undertaken by the clinical team and the other half randomised to automated computerised interpretation of the fetal heart rate which alerted the clinical team to potential problems. A variety of abnormal fetal heart rate patterns are an indicator of fetal distress caused by asphyxia.
The trial was conducted at 24 sites in the UK and Ireland and recruited more than 47,000 women. The results clearly showed no difference between the groups in any of the main outcomes, including the composite primary outcome of stillbirth, neonatal death up to 28 days, neonatal encephalopathy requiring cooling and admission to the neonatal unit with evidence of poor condition at birth (control 171/23,351 vs intervention 172/23,263; adjusted risk ratio 1.01, 95% Confidence Interval 0.82-1.25). No difference was also seen between the two groups in the number of women delivering by caesarean section (control 5,555/23,351 vs intervention 5,669/23,263; adjusted risk ratio 0.99, 99% Confidence Interval 0.97-1.01). No differences were also seen in any of the developmental outcomes assessed in a subset of 6,700 children at age two.
The INFANT trial clearly demonstrated that computerised decision support software to interpret fetal heart rate does not improve neonatal outcomes compared to fetal heart rate monitoring alone. The trial was unusual because of its large size being seven times larger than any previous trial done and 20 times larger than a previous small trial that showed some indication of benefit. The wider context is that still few medical technologies, including many medical devices and computerised software/apps undergo proper evaluation like this in large simple randomised trials.