Cross design synthesis : a new strategy for medical effectiveness research : report to Congressional requesters / United States General Accounting Office.

  • United States. General Accounting Office.
Date:
[1992]
Catalogue details

Licence: Public Domain Mark

Credit: Cross design synthesis : a new strategy for medical effectiveness research : report to Congressional requesters / United States General Accounting Office. Source: Wellcome Collection.

    19/128 (page 17)
    Review of Existing Study Designs Early Approaches Chapter 1 Introduction and Review of Study Designs for Evaluating Treatment Effectiveness Through the years, a variety of studies—with different designs—have been used to evaluate treatment effects. Some study designs have been deemed superior to others; each design, however, has characteristic strengths and weaknesses. In this review, we discuss early approaches, randomized studies, meta-analyses, and data base analyses. Considering that the practice of medicine dates back to antiquity, it is only relatively recently (that is, within the past 150 to 200 years) that studies of treatment effectiveness have moved beyond personal observations by individual clinicians. The personal-observation approach relies on the expertise and acumen of the clinician to interpret the observed consequences of treating a specific patient with a particular therapy. The strength of this approach is that it incorporates clinically relevant experiences into the conclusions and thus provides richness to them.® But the weaknesses of personal observation are many and include the possibility that the patient outcomes observed are coincidental to the treatment rather than caused by it. Perhaps the earliest instance of a more objective approach is Lind’s 18th century research on treatments for scurvy conducted on board the Salisbury at sea (see Pocock, 1983, citing Lind, 1753). Lind took 12 very similar cases and assigned two each to six different treatments, finding that the treatment consisting of oranges and lemons produced the best result. P. Louis (1834, 1835) made explicit the limitations of the usual subjective approach and developed the “numerical method” in 19th century France.’ The numerical method emphasized exact observations, exact recordings of treatments (and deviations from intended treatment), and numerical comparisons of patient outcomes. Since that time, a variety of controlled designs have been developed. For example, historical control trials compare outcomes for patients currently receiving a new treatment to historically recorded outcomes for patients who, at earlier times, had received different treatments. Such studies apparently provided the evidence supporting the majority of cancer ®Indeed, case studies are still routinely reported in the New England Journal of Medicine, and many medical decisions are probably still made on the basis of individual physicians “trying out” a new drug or a new diagnostic test to “see if it works” (Sechrest and Figueredo, 1991, p. 7). 7It was Louis’ use of this method that led to a “decline in bleeding as a standard treatment” for many illnesses (Pocock, 1983, p. 15, citing P. Louis, 18365).