[Clinical research V. Sample size]. / Investigación clínica V. Tamaño de muestra.
Rev Med Inst Mex Seguro Soc
; 49(5): 517-22, 2011.
Article
em Es
| MEDLINE
| ID: mdl-22185853
In clinical research it is impossible and inefficient to study all patients with a specific pathology, so it is necessary to study a sample of them. The estimation of the sample size before starting a study guarantees the stability of the results and allows us to foresee the feasibility of the study depending on the availability of patients and cost. The basic structure of sample size estimation is based on the premise that seeks to demonstrate, among other cases, that the observed difference between two or more maneuvers in the subsequent state is real. Initially, it requires knowing the value of the expected difference (δ) and its data variation (standard deviation). These data are usually obtained from previous studies. Then, other components must be considered: a (alpha), percentage of error in the assertion that the difference between means is real, usually 5 %; and ß, error rate accepting the claim that the no-difference between the means is real, usually ranging from 15 to 20 %. Finally, these values are substituted into the formula or in an electronic program for estimating sample size. While summary and dispersion measures vary with the type of variable according to the outcome, the basic structure is the same.
Buscar no Google
Coleções:
01-internacional
Base de dados:
MEDLINE
Assunto principal:
Projetos de Pesquisa
/
Tamanho da Amostra
Limite:
Humans
Idioma:
Es
Revista:
Rev Med Inst Mex Seguro Soc
Assunto da revista:
MEDICINA
/
MEDICINA SOCIAL
Ano de publicação:
2011
Tipo de documento:
Article
País de publicação:
México