Evaluation of published clinical studies for reproducibility, comparability and adherence to evidence-based methods

To evaluate the "Materials and Methods" of long-term clinical studies in relation to documentation, reproducibility and comparability with and without employing the systematic methods of evidence-based medicine.

The "Materials and Methods" sections in 45 clinical long-term published studies of direct posterior resin-based composite restorations were evaluated for their use of systematic methods of evidence-based medicine. The search was limited to the years 1988-1997, using the key words "clinical study/evaluation/results/report, long-term, in vivo, posterior, Class I/II, composite, restoration". Special attention was directed to comparisons of the underlying documentation, descriptions of the operative techniques used, and their reproducibility. In addition, an evidence-based search was carried out using the Internet PubMed interface for MEDLINE, using identical synonyms, to identify studies with high levels of quality of evidence. Documentation, reproducibility, and comparability of "Materials and Methods" were also evaluated.

Results revealed how difficult it is to interpret results based on tenuous premises, subjective standards, and inadequate study designs. Only one article could be identified when the search was limited to "humans" and "randomized clinical trials". None of the articles, even when fulfilling the highest quality of evidence, showed sufficient or satisfactory quality of reproducibility in their descriptions in Materials and Methods.