The use of the E-value for sensitivity analysis

The practice of evidence-based medicine considers randomized-controlled trials (RCTs) to be the highest quality of research available to clinicians. In a well implemented RCT, randomization ensures that the groups being compared are “equal” in their characteristics by attributing any differences in outcomes to the intervention of interest. However, RCTs are not always feasible. Investigations of population health research may require decades of follow-up time, or pose issues of generalizability when selecting populations more amenable to RCT conditions (e.g., patients at university hospitals) [ [1] Sanson-Fisher R.W. Bonevski B. Green L.W. D’Este C. Limitations of the randomized controlled trial in evaluating population-based health interventions. Am J Prev Med. 2007; 33: 155-161 Abstract Full Text Full Text PDF PubMed Scopus (361) Google Scholar ]. Surgery is another field where RCTs are difficult to implement. Surgical treatment often cannot be blinded, and the randomization of treatments themselves may present ethical dilemmas [ [2] Hoppe D.J. Schemitsch E.H. Morshed S. Tornetta P.I. Bhandari M. Hierarchy of evidence: where observational studies fit in and why we need them. J Bone Joint Surg Am. 2009; 91: 2 Crossref PubMed Scopus (88) Google Scholar ]. In such cases, observational studies are an effective alternative [ [3] D’Agostino R.B. Estimating treatment effects using observational data. JAMA. 2007; 297: 314-316 Crossref PubMed Scopus (275) Google Scholar , [4] Concato J. Shah N. Horwitz R.I. Randomized, controlled trials, observational studies, and the hierarchy of research designs. N Engl J Med. 2000; 342: 1887-1892 Crossref PubMed Scopus (2752) Google Scholar ].