Hierarchy of Evidence: An Appraisal Tool for Weighting the Evidence in Healthcare Design Research Based on Internal Validity.

Abstract

This inquiry aims to develop an appraisal tool to offer greater granularity in weighing evidence in the field of healthcare design research. In evidence-based design (EBD), the goal is to implement interventions that result in a meaningful and optimal effect based on current best evidence. Although multiple appraisal tools (many adopted from medical disciplines) have been instrumental in evaluating studies in the field of healthcare design research to identify the best evidence, they do not necessarily consider the unique contexts of healthcare design research, and methodologies appropriate to the field. Five basic types of studies are ranked based on the level of confidence that they offer regarding the estimate of an effect: (1) meta-analysis studies, (2) causal studies, (3) correlational studies, (4) descriptive studies, and (5) anecdotal evidence. Causal studies are further divided into four levels based on the interaction of two factors: (1) type of intervention and (2) groups' equivalency and extraneous variable control. An eight-level hierarchy of evidence for healthcare design research is proposed that is expected to improve upon previous hierarchies in three major ways: (a) including research methods that are more relevant to healthcare design research, (b) enhancing evaluation accuracy and reliability by providing a clearer definition of studies based on their key components rather than using study labels alone, and (c) distinguishing different levels of evidence, particularly in causal studies. The proposed appraisal tool is developed specifically for EBD by reflecting on the unique context of healthcare design research and practice.