Background:
Electronic Health Records (EHRs) provide the opportunity to identify undiagnosed individuals likely to have a given disease using Machine Learning (ML) techniques, and who could then benefit from more medical screening and case finding, reducing the number needed to screen with convenience and healthcare cost savings. Ensemble Machine Learning Models (EMLs) combining multiple prediction estimates into one, are often said to provide better predictive performances than non-ensemble models. Yet, to our knowledge, no literature review summarises the use and performances of different types of EMLs in the context of medical pre-screening.
Method:
We aimed to conduct a scoping review of the literature reporting the derivation of EMLs for screening of EHRs. We searched EMBASE and MEDLINE databases across all years applying a formal search strategy using terms related to medical screening, EHR and ML. Data were collected, analysed, and reported in accordance with the PRISMA scoping review guideline.
Results:
A total of 3,355 articles were retrieved, of which 145 articles met our inclusion criteria and were included in this study. EMLs were increasingly employed across several medical specialities and often outperformed non-ensemble approaches. EMLs with complex combination strategies and heterogeneous classifiers often outperformed other types of EMLs but were also less used. EML methodologies, processing steps and data sources were often not clearly described.
Conclusions:
Our work highlights the importance of deriving and comparing the performances of different types of EMLs when screening EHRs and underscores the need for more comprehensive reporting of ML methodologies employed in clinical research.