Problems
•
Huge variations in which endpoints are used
and how they are defined
And variable starting points (not always
defined)
Imprecision/heterogeneity makes it difficult if
not impossible to compare results among
studies