Research in clinical medicine often suffers from poor methodological quality and is sometimes based on fraudulent data, which may bias the conclusions of systematic reviews. This dissertation develops the idea that not all research malpractices and misconduct threaten the scientific knowledge, nor may they all bias the conclusions of systematic reviews to the same extent, but rather that the strict procedures applied during proper systematic reviewing actually should make systematic reviews robust to some common errors and misconduct. Moreover, well-performed systematic reviews could become part of the solution to the problem by providing clear research agendas and by providing important methodological information that may be used for future research. Furthermore, systematic reviewers may correct the published literature after having identified errors during critical appraisal of articles, and they may even play the role of whistle-blowers by identifying new, yet unrecognised, cases of misconducts.