Something about this just reminds me of when I did a literature review in my anatomy class to address the question: "Is running bad for your knees?"
I had to decide which of two sets of peer-reviewed publications that contradict each other was least guilty using the data to support the conclusion rather than letting the data speak for itself and making an honest conclusion.
Compared to PhDs, MDs hate designing an experiment and would rather just extrapolate a different conclusion from the same longitudinal study by cherry-picking a different set of variables. The only articles I bother reading from the NEJM anymore are case studies because they're the only publications that consist of mostly-original information.
replies(6):