1 | Appropriateness of statistical methods used to derive the function. |
 | • Representativeness of the algorithm’s derivation sample, optimal statistical power and methods, and clarity of end point predicted by the function. |
2 | Performance of the function: internal and external validity. |
 | • Discrimination, calibration, and sensitivity of the algorithm(s) in the derivation and external datasets. |
3 | Usability of the algorithm. |
 | • Impact of an algorithm’s format on its use and uptake in clinical settings. |
4 | Inclusion of appropriate risk factors. |
 | • Inclusion of major risk factors known to be prevalent in the target population. |
5 | Measurable health gains associated with the use of the algorithm(s). |
 | • Tangible clinical benefits associated with use of the algorithm(s). |