Machine Learning to Predict Interruption in Treatment (Mozambique)
Data.FI, Mozambique (September 2021)
Intervention Description
Data for Implementation team shared their work with predictive model as part of a software solution connected to OpenMRS, the EMR used at ECHO-supported facilities. They will be creating a software plugin to generate patient risk scores through the EMR.
Presenter: Yoni Friedman, Data.FI
Intervention Details:
Demographics Medical history Publicly available data source | |
Model showed strong predictive power achieving 0.65 AUC-PR compared to an underlying IIT rate of 23%. Learning activities are planned to track several ways in which this project can impact outcomes such as increasing intensity of interventions at high risk patients. | |
Machine learning Predictive model connected to OpenMRS. | |
Precision and recall is used to evaluate model performance. | |
ECHO opted to generate risk prediction in raw form. A challenge will be to properly socialize the intrepretation of prediction. Factors to reach scale – Scalling the intervention include installing the software at additional facilities and should not be technically difficult. Preparing for updates to OpenMRS will be part of scaling plan. | |
Contextual factors help to boost predictive accuracy | |
Last updated