Tradition. Each study is a riff on past studies. To make things more comparable, you want to change only enough to make your study novel. Therefore, you often u

Tradition. Each study is a riff on past studies. To make things more comparable, you want to change only enough to make your study novel. Therefore, you often use the same analytical methods and report the same indicators as past studies. 5-year survival is just one measure of health. You proposed two others, but none of them are the complete picture. They're all about the same really. Best is actually looking at the distribution of each metric you care about, but then those curves are hard to compare.