Table 1 Minimum data items to be extracted from studies reporting model derivation or model validation
Data item | Model derivation studies | External model validation studies |
|---|---|---|
Study author(s) and year of publication | ✓ | ✓ |
Country of study and publication language | ✓ | ✓ |
Source of funding and conflicts of interest | ✓ | ✓ |
Study design | ✓ | ✓ |
Study population | ✓ | ✓ |
Geographical location | ✓ | ✓ |
Number of participants | ✓ | ✓ |
Number of outcome events (asthma diagnosis confirmed) | ✓ | ✓ |
Number and description of candidate predictors | ✓ | |
Definition and method used for measuring the outcome | ✓ | ✓ |
Number of participants with missing data | ✓ | ✓ |
Methods for handling missing values (e.g., imputation, complete case analysis) | ✓ | ✓ |
Methods for internally validating the model (e.g., bootstrapping, split-sample) | ✓ | |
How the model is presented (e.g., full or simplified equation, score chart) | ✓ | |
Measures of model performance (e.g., calibration, discrimination) | ✓ | ✓ |
Method of external validation (e.g., geographical, temporal) | ✓ | |
Study investigators (developers of the original model or not?) | ✓ |
