Abstract
The role of machine learning on clinical documentation for predictive outcomes remains undefined. We aimed to compare three neural networks on inpatient providers' notes to predict mortality in neonatal hypoxic-ischaemic encephalopathy (HIE). Using Children's Hospitals Neonatal Database, non-anomalous neonates with HIE treated with therapeutic hypothermia were identified at a single-centre. Data were linked with the initial seven days of documentation. Exposures were derived using the databases and applying convolutional and two recurrent neural networks. The primary outcome was mortality. The predictive accuracy and performance measures for models were determined. The cohort included 52 eligible infants. Most infants survived (n = 36, 69%) and 23 had severe HIE (44%). Neural networks performed above baseline and differed in their median accuracy for predicting mortality (P = .0001): recurrent models with long short-term memory 69% (25th , 75th percentile 65, 73%) and gated-recurrent model units 65% (62, 69%) and convolutional 72% (64, 96%). Convolutional networks' median specificity was 81% (72, 97%). The neural network models demonstrated fundamental validity in predicting mortality using inpatient provider documentation. Convolutional models had high specificity for (excluding) mortality in neonatal HIE. These findings provide a platform for future model training and ultimately tool development to assist clinicians in patient assessments and risk stratifications. The role of machine learning on clinical documentation for predictive outcomes remains undefined. We aimed to compare three neural networks on inpatient providers' notes to predict mortality in neonatal hypoxic-ischaemic encephalopathy (HIE). Using Children's Hospitals Neonatal Database, non-anomalous neonates with HIE treated with therapeutic hypothermia were identified at a single-centre. Data were linked with the initial seven days of documentation. Exposures were derived using the databases and applying convolutional and two recurrent neural networks. The primary outcome was mortality. The predictive accuracy and performance measures for models were determined. The cohort included 52 eligible infants. Most infants survived (n = 36, 69%) and 23 had severe HIE (44%). Neural networks performed above baseline and differed in their median accuracy for predicting mortality (P = .0001): recurrent models with long short-term memory 69% (25th , 75th percentile 65, 73%) and gated-recurrent model units 65% (62, 69%) and convolutional 72% (64, 96%). Convolutional networks' median specificity was 81% (72, 97%). The neural network models demonstrated fundamental validity in predicting mortality using inpatient provider documentation. Convolutional models had high specificity for (excluding) mortality in neonatal HIE. These findings provide a platform for future model training and ultimately tool development to assist clinicians in patient assessments and risk stratifications.