Evaluation of the accuracy of a low-cost external tocodynamometer in a pilot study in Malawi

Uterine contraction monitoring during labor has been linked to improved maternal outcomes. However, performing this monitoring can be challenging for financial and logistical reasons in low resource settings. This proof-of-concept study aimed to compare the accuracy of a low cost external tocodynamometer we developed to that of a commercially available external tocodynamometer. In total, 60 patients with anticipated vaginal deliveries at a hospital in Blantyre, Malawi were enrolled. Both the research device and the commercial device were secured to the patients, and traces were recorded simultaneously from each device. Trace pairs were split into 10 minute segments, and contraction locations were independently annotated on a selection of 75 contemporaneous trace pairs from 38 out of 60 patients by two expert clinicians. In total, 484 contractions were marked on the research device, and 465 contractions were marked on the commercially available device, 312 of which were marked consistently on both devices. The average consistency of marked contractions on the same device between the two observers was 0.89 for the research device compared with 0.84 for the commercially available device. The average consistency between the two devices using a 10-patient rolling average increased from 0.50 at the beginning of the study to 0.64 at the end. The annotated traces from the two devices suggested the same clinical management 72% of the time. The research device displayed reasonable agreement with the commercially available tocodynamometer in detecting contractions. The increase in the measured consistency over the course of the study suggests that improving usability of the device to ensure better positioning on the patient could result in improved performance. Further studies assessing the accuracy and usability of the device are needed.