Improving reporting standards for phenotyping algorithm in biomedical research: 5 fundamental dimensions.

Here, we propose five independent fundamental dimensions of phenotyping algorithms-complexity, performance, efficiency, implementability, and maintenance-through which researchers can describe, measure, and deploy any algorithms efficiently and effectively. These dimensions must be considered in the context of explicit use cases and transparent methods to ensure that they do not reflect unexpected biases or exacerbate inequities.