Confounding in publications of observational intervention studies.

Confounding in publications of observational intervention studies.