Loss to follow-up in cohort studies: How much is too much?

被引：549

作者：

Kristman V. ^{[1
,3
,4
]}

Manno M. ^{[2
,3
]}

Côté P. ^{[1
,3
]}

机构：

[1] Samuel Lunenfeld Research Institute, Mount Sinai Hospital

[2] Department of Public Health Sciences, University of Toronto, Toronto, Ont.

[3] Institute for Work and Health, Toronto, Ont. M5G 2E9

来源：

European Journal of Epidemiology | 2004年 / 19卷 / 8期

基金：

加拿大健康研究院;

关键词：

Bias (epidemiology); Cohort studies; Computer simulation; Epidemiologic methods; Follow-up studies; Logistic models;

D O I：

10.1023/B:EJEP.0000036568.02655.f8

中图分类号：

学科分类号：

摘要：

Loss to follow-up is problematic in most cohort studies and often leads to bias. Although guidelines suggest acceptable follow-up rates, the authors are unaware of studies that test the validity of these recommendations. The objective of this study was to determine whether the recommended follow-up thresholds of 60-80% are associated with biased effects in cohort studies. A simulation study was conducted using 1000 computer replications of a cohort of 500 observations. The logistic regression model included a binary exposure and three confounders. Varied correlation structures of the data represented various levels of confounding. Differing levels of loss to follow-up were generated through three mechanisms: missing completely at random (MCAR), missing at random (MAR) and missing not at random (MNAR). The authors found no important bias with levels of loss that varied from 5 to 60% when loss to follow-up was related to MCAR or MAR mechanisms. However, when observations were lost to follow-up based on a MNAR mechanism, the authors found seriously biased estimates of the odds ratios with low levels of loss to follow-up. Loss to follow-up in cohort studies rarely occurs randomly. Therefore, when planning a cohort study, one should assume that loss to follow-up is MNAR and attempt to achieve the maximum follow-up rate possible. © 2004 Kluwer Academic Publishers.

引用

页码：751 / 760

页数：9

共 23 条

[1]

Kleinbaum D.G., Morgenstern H., Kupper L.L., Selection bias in epidemiologic studies, Am J Epidemiol, 113, pp. 452-463, (1981)

[2]

Greenland S., Response and follow-up bias in cohort studies, Am J Epidemiol, 106, pp. 184-187, (1977)

[3]

Schafer J.L., Graham J.W., Missing data: Our view of the state of the art, Psychol Meth, 7, pp. 147-177, (2002)

[4]

Butler C.W., Snyder M., Wood D.E., Curtis J.R., Albert R.K., Benditt J.O., Underestimation of mortality following lung volume reduction surgery resulting from incomplete follow-up, Chest, 119, pp. 1056-1060, (2001)

[5]

Rothman K.J., Greenland S., Modern Epidemiology, (1998)

[6]

Hollen P.J., Gralla R.J., Cox C., Eberly S.W., Kris M.G., A dilemma in analysis: Issues in the serial measurement of quality of life in patients with advanced lung cancer, Lung Cancer, 18, pp. 119-136, (1997)

[7]

Deeg D.J.H., Attrition in longitudinal population studies: Does it affect the generalizability of the findings? An introduction to the series, J Clin Epidemiol, 55, pp. 213-215, (2002)

[8]

Lohr S.L., Nonresponse, Sampling: Design and Analysis, pp. 255-287, (1999)

[9]

Altman D.G., Statistics in medical journals: Some recent trends, Stat Med, 19, pp. 3275-3289, (2000)

[10]

Babbie E.R., Survey Research Methods, (1973)

← 1 2 3 →