Novel Uterine Contraction Monitoring to Enable Remote, Self-administered Non-stress Testing.

ABSTRACT Background The serial fetal monitoring recommended for women with high-risk pregnancies places a substantial burden on the patient, often disproportionately affecting underprivileged and rural populations. A telehealth solution that can empower pregnant women to obtain recommended fetal surveillance from the comfort of their own home has the potential to promote health equity and improve outcomes. We have previously validated a novel, wireless pregnancy monitor that can remotely capture fetal and maternal heart rates. However, such a device must also detect uterine contractions if it is to be used to robustly conduct remote non-stress tests. Objective(s) To describe and validate a novel algorithm that utilizes biopotential and acoustic signals to non-invasively detect uterine contractions via a wireless pregnancy monitor. Study Design A prospective, open-label, two-center study evaluated simultaneous detection of uterine contractions by the wireless pregnancy monitor and an intrauterine pressure catheter in women carrying singleton pregnancies ≥32+0 weeks gestation who were in the first stage of labor (NCT03889405). The study consisted of a training phase and a validation phase. Simultaneous recordings from each device were passively acquired for 30 to 60 minutes. In a subset of the monitoring sessions in the validation phase, tocodynamometry was also deployed. Three maternal-fetal medicine specialists, blinded to the data source, identified and marked contractions in all modalities. The positive agreement and false positive rates of both the wireless monitor and tocodynamometry were calculated compared to intrauterine pressure catheter. Results A total of 118 participants were included, 40 in the training phase, and 78 in the validation phase (of which 39/78 were monitored simultaneously by all 3 devices) at a mean gestational age of 38.6 weeks. In the training phase, the positive agreement for the wireless monitor was 88.4% (1440/1692 contractions), with a false positive rate of 15.3% (260/1700). In the validation phase, using the refined and finalized algorithm, the positive agreement for the wireless pregnancy monitor was 84.8% (2722/3210), with a false positive rate of 24.8% (897/3619). For the subgroup who were monitored only with the wireless monitor and intrauterine pressure catheter, the positive agreement was 89.0% (1191/1338), with a similar false positive rate of 25.4% (406/1597). For the subgroup monitored by all 3 devices, the positive agreement for the wireless monitor was significantly better than for tocodynamometry (p Conclusion This novel method to non-invasively monitor uterine activity, via a wireless pregnancy monitoring device designed for self-administration at home, was more accurate than the commonly used tocodynamometry and unaffected by body mass index. Together with the previously reported remote fetal heart rate monitoring capabilities, this added ability to detect uterine contractions creates a complete telehealth solution for remote administration of non-stress tests.