Wie wird die Validität wissenschaftlicher Arbeiten im klinischen Alltag bewertet?
Clinicians are trained to apply the methods of evidence-based medicine to assess the validity of publications but may only partially apply it in daily clinical practice. We investigated in a first study which of 11 evidence-based criteria were considered important to assess the validity and in a second study if evidence-based criteria were considered more important than "traditional" criteria. First, we asked 13 medical students, 228 residents and consultants and 31 non-medical staff at Ulm University hospital (total = 272) to rate the importance of 9 evidence-based criteria (e.g. risk profiles similar?) and 2 unrelated validity criteria (e.g. patients recruited in time) on a scale 5 to 1 point (very to not important). Second, we asked 184 physicians, 42 Germans, 16 Italians and 126 Brazilians to rank the importance of 8 evidence-based criteria and 8 "raditional" criteria (e.g. study was supported by a known grant agency). The evidence-based criteria were rated between 4.62 - 3.93 points on a scale of 5 to 1 (very important to not important at all). The two unrelated criteria were rated 3.29 and 3.87. In the second experiment 75 % of German, 64 % of Italian, and 51% of Brazilian physicians filled in the questionnaires. Unexpectedly, the "traditional" criteria were rated less important than evidence-based criteria. Physicians will agree on the validity of scientific publications if most or none of the validity criteria are met but will likely disagree when some of the criteria are not met. The compliance to fill in questionnaires seems to be different in physicians from different cultures. Physicians of all cultures claim to use preferably evidence-based instead of "traditional" criteria to assess the validity of scientific publications. As these results may reflect true or socially desired answers we will continue this challenging project.
Subject HeadingsKriterium [GND]
Evidence-based medicine [MeSH]
Outcome assessment (Health care) [MeSH]