DOI QR코드

DOI QR Code

Comparison of GEE Estimation Methods for Repeated Binary Data with Time-Varying Covariates on Different Missing Mechanisms

시간-종속적 공변량이 포함된 이분형 반복측정자료의 GEE를 이용한 분석에서 결측 체계에 따른 회귀계수 추정방법 비교

  • Park, Boram (Biometric Research Branch, National Cancer Center) ;
  • Jung, Inkyung (Department of Biostatistics, Yonsei University College of Medicine)
  • 박보람 (국립암센터 바이오메트릭연구과) ;
  • 정인경 (연세대학교 의학통계학과)
  • Received : 2012.10.25
  • Accepted : 2013.09.06
  • Published : 2013.10.31

Abstract

When analyzing repeated binary data, the generalized estimating equations(GEE) approach produces consistent estimates for regression parameters even if an incorrect working correlation matrix is used. However, time-varying covariates experience larger changes in coefficients than time-invariant covariates across various working correlation structures for finite samples. In addition, the GEE approach may give biased estimates under missing at random(MAR). Weighted estimating equations and multiple imputation methods have been proposed to reduce biases in parameter estimates under MAR. This article studies if the two methods produce robust estimates across various working correlation structures for longitudinal binary data with time-varying covariates under different missing mechanisms. Through simulation, we observe that time-varying covariates have greater differences in parameter estimates across different working correlation structures than time-invariant covariates. The multiple imputation method produces more robust estimates under any working correlation structure and smaller biases compared to the other two methods.

다시점 자료 연구에서 일반화추정방정식은 가상관행렬을 잘못 가정하더라도 모수의 일치추정량을 도출하므로 많이 이용된다. 하지만, 결측 체계가 완전임의결측이 아닌 경우에는 편의추정량을 제공하고, 시간-종속적 공변량이 포함된 경우에는 가상관행렬에 따라 회귀계수 추정값이 다르게 도출될 수 있는 문제점이 있다. 결측 체계가 임의결측인 경우에 발생하는 문제를 해결하기 위해 가중 방법과 다중대체 방법을 사용하는 것이 제안되었다. 본 논문에서는 시간-종속적 공변량이 포함된 이분형 반복측정자료를 GEE를 이용하여 분석할 때 다양한 결측 체계에서 일반화추정방정식 방법, 가중 방법, 다중대체 방법의 회귀계수 추정에 대한 로버스트성과 정확성을 모의실험을 통하여 비교해 보았다. 세 가지 방법 모두에서 시간-종속적 공변량의 회귀계수가 시간-독립적 공변량의 회귀계수에 비해 가상관행렬에 따라 추정값의 차이가 크게 나타났다. 다른 두 방법에 비해 다중대체 방법이 가상관행렬의 형태에 대해 더 로버스트하고 편의도 작은 추정치를 도출하였다.

Keywords

References

  1. Beunckens, C., Sotto, C. and Molenberghs, G. (2008). A simulation study comparing weighted estimating equations with multiple imputation based estimating equations for longitudinal binary data, Computational Statistics & Data Analysis, 52, 1533-1548. https://doi.org/10.1016/j.csda.2007.04.020
  2. Faught, E., Wilder, B. J., Ramsay, R. E., Reife, R. A., Kramer, L. D., Pledger, G. W. and Karim, R. M. (1996). Topiramate placebo-controlled dose-ranging trial in refractory partial epilepsy using 200-, 400-, and 600-mg daily dosages, Neurology, 46, 1684-1690. https://doi.org/10.1212/WNL.46.6.1684
  3. Fitzmaurice, G. M. (1995). A caveat concerning independence estimation equations with multiple multivariate binary data, Biometrics, 51, 309-317. https://doi.org/10.2307/2533336
  4. Kim, T. H. (2004). Handling data in GEE with missing response, Sungkyunkwan University.
  5. Liang, K. Y. and Zeger, S. L. (1986). Longitudinal data analysis using generalized linear models, Biometrika, 73, 13-22. https://doi.org/10.1093/biomet/73.1.13
  6. Little, R. J. A. and Rubin, D. B. (2002). Statistical Analysis with Missing Data, John Wiley & Sons.
  7. Pepe, M. S. and Anderson, G. (1994). A cautionary note on inference for marginal regression models with longitudinal data and general correlated response data, Communication in Statistics B, 23, 939-951. https://doi.org/10.1080/03610919408813210
  8. Preisser, J. S., Lohman, K. K. and Rathouz, P. J. (2002). Performance of weighted estimating equations for longitudinal binary data with drop-outs missing at random, Statistics in Medicine, 21, 3035-3054. https://doi.org/10.1002/sim.1241
  9. Robins, J. M., Rotnitzky, A. and Zhao, L. P. (1994). Estimation of regression coefficients when some regressors are not always observed, Journal of the American Statistical Association, 189, 846-866.
  10. Robins, J. M., Rotnitzky, A. and Zhao, L. P. (1995). Analysis of semiparametric regression models for repeated outcomes in the presence of missing data, Journal of the American Statistical Association, 90, 106-121. https://doi.org/10.1080/01621459.1995.10476493
  11. Rubin, D. B. (1987). Multiple Imputation for Nonresponse in Surveys, John Wiley & Sons.
  12. Song, J. W. and An, H. (2009). Handling and Analysis of Missing Data, Statistical Training Institute, Seoul.
  13. Troxel, A. B., Lipsitz, S. R. and Brennan, T. A. (1997). Weighted estimating equations with nonignorably missing response data, Biometrics, 53, 857-869. https://doi.org/10.2307/2533548
  14. Wall, M. M., Dai, Y. and Eberly, L. E. (2005). GEE estimation of a misspecified time-varying covariate: An example with the effect of alcoholism treatment on medical utilization, Statistics in Medicine, 24, 925-939. https://doi.org/10.1002/sim.1966