Data Extraction

Two independent reviewers screen titles, abstracts, and full texts, with disagreements resolved by a third reviewer. The extraction framework was guided by CHARMS to capture clinical and technical dimensions of prediction-modelling studies. The extraction process requires 90% agreement before the lead reviewer extracts the remaining studies alone. Vaccination-rate adjustment methods will be described because survivors may have higher vaccination rates. During full extraction, two researchers will independently extract information from each included study. Extracted information will cover bibliographic details, study characteristics, and findings about EDIA in faculty development, curriculum content, and teaching strategies. Data extraction will use a predefined tool designed to support consistency and accuracy. The extraction tool will be piloted in pairs on eight studies before full extraction. The pilot phase is intended to refine the extraction tool and resolve ambiguities. Data extraction will use a predefined standardised form piloted on 3 to 5 studies. The lead author will extract data and a co-author will independently check it. Data extraction uses a standardised, pre-pilo…