Commentary

Abstract

In any evidence-based scientific society such as NASS, it is appropriate to revisit some of the basic tenets of research methodology in evaluating clinical research outcomes. It is rare that we can conduct the “perfect” study to unequivocally document a cause–effect relationship [ [1] Gatchel R.J The hierarchy of clinical research studies. Spine J. 2001; 1: 85-87 Abstract Full Text Full Text PDF PubMed Scopus (2) Google Scholar ]. The randomized controlled trial (RCT) is often viewed as an important benchmark to use in considering the reliability and validity of treatment-outcome results. However, in order to infer cause–effect relationships, RCTs can vary greatly in the degree of internal and external validity that make them less than “perfect” in nature. Most careful researchers would question the possibility of designing the “perfect” study that could prove a cause–relationship with 100% certainty. Indeed, scientific inquiry does not assume to “prove” as much as it seeks to achieve greater degrees of certainty. Achieving closer approximations to certainty requires diverse and creative methodological approaches, and there is no one method that works alone. Although RCTs offer the potential for greater controls than do some other research designs, RCTs are not without limitations. Such limitations leave room for other types of designs to contribute important scientific data toward delineating cause–effect relationships.