A systematic review of networks for prognostic prediction of health outcomes and diagnostic prediction of health conditions within Electronic Health Records

Zoe Hancox, Allan Pang, Philip G. Conaghan, Sarah R. Kingsbury, Andrew Clegg, Samuel D. Relton

Abstract: Highlights•Systematic review of graph-based models in clinical prediction tasks.•Only three out of 27 eligible papers had a low PROBAST risk of bias.•Few clinical prediction models provide calibration curves or appropriate metrics.•Graphs may improve healthcare prediction over other machine learning methods.•Suggestions are provided to guide future research to reduce bias in this domain.