As part of the EPC Methods Guide, we intend that this paper will guide EPCs when selecting studies for inclusion in an SR. Guidance is intended to reduce inconsistencies and risk of bias. Unfortunately, because there are no available studies to guide us how best to reduce this variation, what follows is based on fundamental principles of SRs and the experience of the EPC program.
Inconsistencies and bias can certainly occur during the development of key questions, which define the scope of the review and details the population(s), intervention(s), comparator(s), outcome(s), timing, and setting (PICOTS), and sometimes even the study designs or study characteristics of interest. The methods used by the EPC program at this earlier stage are discussed elsewhere. Likewise, we recognize that bias can also be introduced during the searching stage, or in how reviewers handle assessment of reporting biases, and guidance on these methods are provided elsewhere., This paper focuses on what to do with the literature once it is identified. We first describe the types of bias then stratify the guidance on addressing these biases into sections: Setting Inclusion Criteria to Avoid Bias in Selecting Studies, Study Selection Process, and Using Gray Literature to Assess and Reduce Bias.