Evidence based health care begins with a clinical question and the search on data bases to retrieve the relevant information, that was the issue of two preceding articles of this series. At present it will be discussed how to critically appraise the medical literature using the clinical epidemiological methodology. Clinical research aims to develop diagnostic and therapeutic procedures measuring association and causality between the exposure and outcome. In this case the exposures are signs, symptoms, laboratorial or image exam, and therapy intervention. It is a mistake to take surrogate end-points instead of clinical outcomes. The main types of clinical study design are diagnostic, prognostic, therapeutic and harm/etiology. Experimental, physiologic and animal studies are useful for the medical undergraduate education, but do not contribute with clinical decisions. The study designs are classified according with the presence of a control group, patient's follow-up, and therapy interventions. The evidence hierarchy was done by the previous characteristics and the presence of systematic bias. Systematic reviews are stronger than the primary observational studies and are on the top when they revised randomized clinical trial. Since 1998 the proportion of evidence based practice guidelines was increasing compared with systematic reviews or other types of practice guidelines, although the former still are in a few numbers. The article critical appraisal must answer the clinical question, and need to have consistent study design and bias under control. In conclusion we ought to offer methodological actualization to interested physicians and put the information already critically assessed on evidence-based practice guidelines.
Critical appraisal; Outcome; Study design; Evidence based medicine; Guidelines; Clinical research
