O2178: Feasibility and Efficacy of Telemedicine for hypothermia evaluation in infants with suspected Hypoxic ischemic encephalopathy

Background: Hypoxic ischemic encephalopathy (HIE) is a major cause of neurologic disability in newborn infants. Immediate therapeutic hypothermia is crucial but often delayed by transfers to equipped facilities. Transport teams can initiate hypothermia, but this requires precise neurological assessment. Hypothermia evaluation using telemedicine (TM) could expedite this process. However, there is little evidence on the efficacy of telemedicine for HIE assessments compared to the standard of care in-person exam (IP). In this study, we aim to compare telemedicine evaluation to the gold standard in-person neurological exam in newborns ≥ 36 weeks suspected of being affected by HIE.

Methods: Based on study criteria, we enrolled 85 infants at two academic health centers, Parkland Health System and Texas Health Presbyterian Hospital, from March 2022 to December 2023. All infants had neurological assessments using sarnat exams by in-person and telemedicine physicians. TM exams were performed using a Tytocare Pro device and iPad, and hotspots established connectivity. The data included neurologic exam scores for the Sarnat exam for both IP and TM, the HIE stage, and decisions for therapeutic hypothermia. We further analyzed the scores of components of the sarnat exam, final diagnosis, and telemedicine metrics such as completion time and technology issues. Cohen's Kappa statistic (κ) was selected to measure the interrater agreement between IP and TM exams with the following interpretation for the values: ≤ 0 - no agreement, 0.01–0.20 as none to slight, 0.21–0.40 as fair, 0.41– 0.60 as moderate, 0.61–0.80 as substantial, 0.81–1.00 almost perfect agreement.

Results: 170 neurological assessments (IP and TM) were performed on 85 infants eligible for HIE evaluation. From a technology standpoint, 92.9% of telemedicine exams were completed within 0-15 minutes. 78.8% of TM encounters had no technology issues, and 10.6% had unclear video. The results show almost perfect agreement on signs of moderate or severe HIE (κ=0.84) and the decision for hypothermia intervention(κ =0.84). We report a fair agreement (κ=0.21-0.40) for individual categories of the neurologic exam and a moderate agreement (κ=0.59) on severity gradings (normal, mild, moderate, and severe). Table 1 shows the measure of agreement using Cohen’s kappa between IP and TM for various exam components and therapeutic intervention decisions. Figure 1 highlights the concurrence in treatment decisions across IP and TM exams (Yes – 34.1 vs. 32.9%, No – 65.9% vs. 67.1%).

Conclusion: This is the first study to compare in-person and telemedicine neurologic assessment of infants needing hypothermia evaluation. The study findings show that telemedicine evaluation for infants with suspected HIE is feasible and is non-inferior to in-person assessments for making decisions on therapeutic interventions in infants with suspected HIE. These findings have translational benefits in using telemedicine for remote neurological evaluation and initiating early therapeutic interventions to improve outcomes.