TESTING FOR ASSOCIATION OF DISEASES

被引：11

作者：

BOYD, AV

机构：

[1] Department of Mathematics, University of the Witwatersrand, Johannesburg

来源：

JOURNAL OF CHRONIC DISEASES | 1979年 / 32卷 / 9-10期

关键词：

D O I：

10.1016/0021-9681(79)90097-3

中图分类号：

R1 [预防医学、卫生学];

学科分类号：

1004 ; 120402 ;

摘要：

Berkson has shown that differences in selection rates for various diseases can introduce a bias into tests for association between two diseases when hospitalized patients suffering from a third disease are used as a control group. A formula is derived showing how the odds ratio for a test depends on the hospitalization rates for the three diseases and on the prevalence rate for the control group disease. Rules are then formulated showing under what conditions an observed ratio can have arisen from a Berkson-type bias or is likely to be the result of a real (positive or negative) association between two diseases. © 1979.

引用

页码：667 / 672

页数：6

共 9 条

[1] LIMITATIONS OF THE APPLICATION OF FOURFOLD TABLE ANALYSIS TO HOSPITAL DATA [J].

BERKSON, J .

BIOMETRICS BULLETIN, 1946, 2 (03) :47-53

[2]

BERKSON J, 1955, P STAFF M MAYO CLIN, V30, P319

[3]

COLTON T, 1974, STATISTICS MEDICINE

[4]

LILIENFELD AM, 1976, F EPIDEMIOLOGY

[5] THE RISK OF FALLACIOUS CONCLUSIONS FROM AUTOPSY DATA ON THE INCIDENCE OF DISEASES WITH APPLICATIONS TO HEART DISEASE [J].

MAINLAND, D .

AMERICAN HEART JOURNAL, 1953, 45 (05) :644-654

[6]

MAINLAND D, 1963, ELEMENTARY MEDICAL S

[7] Cancer and tuberculosis [J].

Pearl, R .

AMERICAN JOURNAL OF HYGIENE, 1929, 9 (01) :97-159

[8]

Pearl R, 1929, LANCET, V1, P1078

[9] EMPIRICAL DEMONSTRATION OF BERKSON BIAS [J].

ROBERTS, RS ;

SPITZER, WO ;

DELMORE, T ;

SACKETT, DL .

JOURNAL OF CHRONIC DISEASES, 1978, 31 (02) :119-128

← 1 →