Systematic Pyramid Reviews – A new method of evidence synthesis from different research approaches for routine health care decision making and practice

Background: Science-based clinical decision-making and practice in everyday healthcare requires synthesized evidence from different research approaches. The usual focus on RCTs (randomized controlled trials) as the methodological gold standard and the associated neglect of observa-tional studies and qualitative studies does not meet these requirements. The aim of the project presented here was to develop a type of review with which study results from all relevant re-search approaches can be synthesized and the significance of this overall state of evidence on the effects of therapeutic interventions can be critically appraised.Methods: Over a period of almost 20 years, a four-sided evidence hierarchy (research pyramid) was developed on the basis of general epistemological and methodological principles, which is based on two fundamental distinctions: experimental and observational as well as quantitative and qualitative research approaches. From this the type of a Pyramid Review was developed. Results: A Pyramid Review is based on the usual steps for systematic reviews - but without re-striction to specific research approaches or study designs. The specific steps of a Pyramid Review are the following: (A) Assignment of the individual studies and reviews to the corresponding research approach on the research pyramid; (B) Preliminary appraisal of the level of evidence; (C) Critical appraisal of the quality of the conduct of the study, and downgrading of the level of evidence as indicated; (D) Critical appraisal of the quality of results, and upgrading of the level of evidence if appropriate; (E) Clustering the individual studies according to research approach and conducting partial reviews; (F) Critical synthesis of the overall state of evidence from the partial reviews. Conclusion: A Pyramid Review for clinical practice and research currently seems to be the most meaningful form of evidence synthesis for supporting routine health care. It comprehensively considers evidence from different research approaches and evaluates it according to strict re-spective methodological standards. It also makes it clear which research approaches may be overrepresented and which are underrepresented in relation to a particular clinical question.