Hostname: page-component-cd9895bd7-q99xh Total loading time: 0 Render date: 2024-12-23T13:34:28.407Z Has data issue: false hasContentIssue false

Distorsioni («bias») in epidemiologia analitica

Published online by Cambridge University Press:  11 October 2011

Walter A. Rocca*
Affiliation:
Studio Multicentrico Italiano sulla Demenza, Centro SMID, Firenze
*
Indirizzo per la corrispondenza: Dr. W. A. Rocca, Centro SMID, Via il Prato 58, 50123 Firenze. Fax (+39) 055-230.2914

Abstract

Riassunto

Questo articolo descrive i più comuni tipi di distorsione («bias») che si possono incontrare in studi di epidemiologia analitica. Le distorsioni vengono presentate in relazione al disegno degli studi di coorte o caso-controllo. Per questa ragione, nella prima parte dell'articolo, vengono brevemente illustrati i concetti elementari del disegno e la terminologia degli studi di coorte e caso-controllo. Vengono distinti due gruppi principali di distorsioni: le distorsioni df selezione (o di campionamento) e le distorsioni di misura (o di raccolta deirinformazione). Negli studi di coorte, la principale distorsione di selezione è quella dei non partecipanti allo studio; la principale distorsione di misura è quella del sospetto diagnostico. Negli studi caso-controllo, le principali distorsioni di selezione sono: la distorsione prevalenza-incidenza, la distorsione del ricovero ospedaliero e la distorsione dei non partecipanti; le principali distorsioni di misura sono: la distorsione del ricordo, la distorsione deH'informazione familiare e la distorsione del sospetto di esposizione. Alcune di queste distorsioni possono essere prevenute o minimizzate mediante appropriate strategic di disegno dello studio.

Parole chiave

distorsioni, epidemiologia, metodi, studio di coorte, studio caso-controllo.

Summary

This article describes the most common types of bias encountered in analytic epidemiologic studies. Bias is presented in relation to the design of cohort and case-control studies. Therefore, in the firstpart of the article, the basic design concepts and the terminology of cohort and case-control studies are briefly illustrated. Two major groups of bias are described: selection (or sampling) bias and measurement (or data collection) bias. In cohort studies, the most important selection bias is the non-respondent bias; the most important measurement bias is the diagnostic suspicion bias. In case-control studies, the most important selection biases are the incidence-prevalence bias, the admission rate (Berksonian) bias, and the non-respondent bias; the most important measurement biases are the recall bias, the family information bias, and theexposure suspicion bias. Some of these biases may be prevented or minimized by appropriate design strategies.

Type
Articoli
Copyright
Copyright © Cambridge University Press 1992

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

BIBLIOGRAFIA

Breslow, N. E. & Day, N. E. (1980). Statistical Methods in Cancer Research: Volume I - The Analysis of Case-control Studies. International Agency for Research on Cancer: Lyon.Google ScholarPubMed
Breslow, N. E. & Day, N. E. (1987). Statistical Methods in Cancer Research: Volume II - The Design and Analysis of Cohort Studies. International Agency for Research on Cancer: Lyon.Google ScholarPubMed
Chandra, V., Kokmen, E., Schoenberg, B. S. & Beard, M. (1989). Head Trauma with loss of consciousness as a risk factor for Alzheimer's disease. Neurology 39, 15761578.CrossRefGoogle ScholarPubMed
Feinstein, A. R. & Horwitz, R. (1982). Double standards, scientific methods, and epidemiologic research. New England Journal of Medicine 307, 16111617.CrossRefGoogle ScholarPubMed
Horwitz, O. & Lysgaard-Hansen, B. (1975). Medical observations and bias. American Journal of Epidemiology 101, 391399.CrossRefGoogle ScholarPubMed
Last, J. M. (1988). A Dictionary of Epidemiology. Oxford University Press: New York.Google Scholar
Lilienfeld, A. M. & Lilienfeld, D. E. (1980). Foundations of Epidemiology. Oxford University Press: New York.Google Scholar
Mantel, N. & Haenszel, W. (1959) Statistical aspects of the analysis of data from retrospective studies of disease. Journal of the National Cancer Institute 22,719748.Google ScholarPubMed
Meinert, C. L. (1986). Clinical Trials: Design, Conduct, and Analysis. Oxford University Press: New York.CrossRefGoogle Scholar
Rocca, W. A. & Amaducci, L. (1988). The familial aggregation of Alzheimer's disease: an epidemiological review. Psychiatric Developments 6, 2336.Google ScholarPubMed
Rocca, W. A. & Amaducci, L. (1991). Epidemiology of Alzheimer's disease. In Neuroepidemiology: A Tribute to Bruce Schoenberg (ed. Anderson, D. W.), pp. 5596. CRC Press: Boca Raton.Google Scholar
Rocca, W. A., Sharbrough, F. W., Hauser, W. A., Annegers, J. F. & Schoenberg, B. S. (1987). Risk factors for generalized tonicclonic seizures: a populazion-based case-control study in Rochester, Minnesota. Neurology 37, 13151322.CrossRefGoogle ScholarPubMed
Rothman, K. J. (1986). Modern Epidemiology. Little, Brown & Company: Boston.Google Scholar
Sackett, D. L. (1979). Bias in analytic epidemiology. Journal of Chronic Diseases 32, 5163.CrossRefGoogle Scholar
Schlesselman, J. J. (1982). Case-control Studies: Design, Conduct, Analysis. Oxford University Press: New York.Google Scholar
Schoenberg, B. S. (1978). Analytic, experimental, and theoretical epidemiology. In Advances in Neurology, Vol 19 (ed. Schoenberg, B. S.), pp. 4354. Raven Press: New York.Google Scholar
Spizer, W. O. (1979). The case-control study: consensus and controversy. Journal of Chronic Disease 32, 144.Google Scholar