Transitioning Toward Evidence-Based Research in the Health Sciences for the XXI Century

Abstract

This article discusses some of the misconceptions of evidence-based research in the health sciences. It proposes that since not all treatments in medicine and dentistry can be evidence-based, clinical applications of the evidence-based process should become a specialty. The case is particularly evident in dentistry. Therefore dentistry is taken in this article as a model for discussion. We propose that to approach dentistry from the viewpoint of the patient-oriented evidence that matters (POEM) is perfectly acceptable so far as we also engage in the process of research evaluation and appraisal in dentistry (READ). We distinguish between dentistry based on the evidence, and evidence-based dentistry. We argue that when invoking an evidence-based approach to dentistry or medicine, it is not sufficient to establish the 'levels of evidence', but rather that all evidence-based clinical intervention must undergo the stringent process of evidence-based research so that clinical practice guidelines be revised based on the best available evidence.

Keywords: clinical practice guidelines; evidence-based research; levels of evidence.

Figure 1.

The process of evidence-based research, which corresponds from the validation of the data in published reports. The process involves the systematic and critical evaluation of the research reliability and validity of each published report pertaining to the PICO question. This is achieved by means of specialized evaluation instruments, which yield quantitative estimates of the strengths and the weakness of each report with respect to research design, methodology and statistical analysis. These estimates are assessed by means of acceptable sampling analysis, and when appropriate, by meta-analysis. The outcome of the systematic review process is the generation of the consensus statement, which leads to revisions of the clinical practice guidelines based on the best available research evidence. [Note: an early version of this figure was published in (10); we present this modification here because it bests represents the READ process].