Correlation of short‐term variation derived from novel ambulatory fetal electrocardiography monitor with computerized cardiotocography

B. Liu,B. Thilaganathan,A. Bhide
DOI: https://doi.org/10.1002/uog.26191
2023-03-04
Ultrasound in Obstetrics and Gynecology
Abstract:Objectives The objective of this study is to compare short‐term variation (STV) outputs from a novel self‐applied non‐invasive fetal electrography (NIFECG) with hospital computerized cardiotocography (cCTG) monitors. Technological and algorithmic limitations, as well as mitigation strategies will be evaluated. Methods This prospective cohort study took place in a tertiary London hospital. Women with a singleton pregnancy over 28+0 weeks' gestation attending hospital for cCTG assessment were monitored with both NIFECG and cCTG simultaneously for up to one hour. Post‐processing of NIFECG using various methods of filtering produced NIFECG STV (eSTV) values, which were compared with cCTG STV (cSTV) outputs. Linear correlation, mean bias, precision, and limits of agreement (LoA) were assessed, using different methods of STV computation and mathematical correction. Results 306 concurrent NIFECG and cCTG traces were collected from 285 women. Fully filtered (F‐filtered) eSTV was very strongly correlated with cSTV (R=0.911, p<0.001), but only generated results in 142/306 (46.4%) of one‐hour traces due to the removal of those with lower‐quality signals. Partial filtering generated more eSTV results (98.4%), but with a weak correlation to cSTV (R=0.337, p<0.001). STV difference (eSTV – cSTV) increased with signal loss, where in traces with ≥60% signal loss, the values became highly discrepant. Removal of traces with ≥60% signal loss resulted in a higher correlation with cSTV, whilst generating eSTV results for 65% of traces. Correcting these remaining eSTV values for signal loss using linear regression further improved correlation with cSTV (R=0.839, p<0.001). Conclusions The causes of STV discrepancy mandate the need for signal filtering, exclusion of poor‐quality traces and eSTV correction. With such correction, the data demonstrate the device's ability to produce eSTV values highly correlated to the cCTG cSTV readings. This evidence‐base for NIFECG monitoring and interpretation is a promising step forward in the development of safe and effective home FHR monitoring. This article is protected by copyright. All rights reserved.
radiology, nuclear medicine & medical imaging,obstetrics & gynecology,acoustics
What problem does this paper attempt to address?