Medical Studies Frequently Not Confirmed

According to a study published in the July 13, 2005 Journal of the American Medical Association (JAMA) by Dr. John Ioannidis, almost a third of the medical studies included in his sample, were eventually either contradicted by subsequent studies (16%) or else required significant modification. (One media report summarizing the study appears at: http://www.newsday.com/news/health/ny-hsdrug4348592jul19,0,2629446.story?coll=ny-health-headlines)
This is only surprising in the face of the certainty with which the media and parts of the medical establishment, totally embrace each new study as it appears. Perhaps the tentativenss, and revisability of medical research argues for allowing patients more choice in their treatment?