Format

Send to

Choose Destination
Rev Port Cardiol. 2011 Feb;30(2):235-42.

Bias in clinical studies.

[Article in English, Portuguese]

Author information

1
Centro de Estudos de Medicina Baseada na EvidĂȘncia Faculdade de Medicina da Universidade de Lisboa, Lisboa, Portugal. ave@fm.ul.pt

Abstract

Clinical studies are designed fundamentally to obtain data to support practice, whether diagnostic, therapeutic or prognostic. No matter how well a study was designed, implemented or analyzed, it is never possible to guarantee the truth of its results with absolute certainty. This is due to two factors that can change the study's results: random error and systematic error (also called bias). The first appears without warning and cannot be modulated, while the second--defined as the systematic introduction of error into a study--can be removed (if not completely, at least partially). In this paper we present the most common types of bias found in clinical studies, whether therapeutic (clinical trials), or with other experimental or observational designs.

PMID:
21553616
[Indexed for MEDLINE]
Free full text

Supplemental Content

Full text links

Icon for Sociedade Portuguesa de Cardiologia / Portuguese Society of Cardiology
Loading ...
Support Center