Article type
Year
Abstract
Background: In spite of meta-analysis remaining the cornerstone of evidence synthesis, numerous alternative methods of statistical and non-statistical synthesis have recently emerged. Their proliferation has been driven by needs that the 'exemplar systematic review’ meets incompletely, such as the need to address broad research questions and heterogeneous study types or to incorporate multiple perspectives and levels of analysis. Healthcare quality and patient safety are prime examples of areas where such demands arise. The field of emerging research synthesis methods is poorly mapped but lively. Four review publications identify over 40 such methods among themselves. These methods have developed in relative isolation. Their terminology is highly varied. Some of their constitutive practices are at odds with traditional systematic review ones. Many of them simply extend single study designs and lack the level of abstraction, articulation and detail required for wide usage. There are thus substantial barriers to their uptake. Yet those methods try to solve important research synthesis concerns. Are we giving them sufficient attention?
Objectives: This presentation will begin with an overview of the field of emerging health research synthesis methods. It will then address the higher-order dimensions along which they tend to be compared and the lines along which they translate desiderata for systematicity, transparency and robustness into specific procedures. Finally, it will propose further criteria for evaluating the capacity of a synthesis method to instantiate such desiderata.
Methods and data sources: A case study of the contents and structure of a heterogeneous pool of research papers on cancer. Ideas from the philosophy and social studies of science.
Conclusions: The current edition of the Cochrane Handbook offers limited guidance to non-standard synthesis methods. A debate is needed as to whether extending the Handbook in this direction is required or whether there are positive reasons for taking aconservative attitude.
Objectives: This presentation will begin with an overview of the field of emerging health research synthesis methods. It will then address the higher-order dimensions along which they tend to be compared and the lines along which they translate desiderata for systematicity, transparency and robustness into specific procedures. Finally, it will propose further criteria for evaluating the capacity of a synthesis method to instantiate such desiderata.
Methods and data sources: A case study of the contents and structure of a heterogeneous pool of research papers on cancer. Ideas from the philosophy and social studies of science.
Conclusions: The current edition of the Cochrane Handbook offers limited guidance to non-standard synthesis methods. A debate is needed as to whether extending the Handbook in this direction is required or whether there are positive reasons for taking aconservative attitude.