在存活分析中，使用誤判模型 (misspecified model)配適存活資料時，常會影響存活模型中的參數估計值，而且與正確模型配適存活資料相比較的話，甚至會導致其共變數對風險函數的顯著性產生不同的結果，所以本篇藉由統計模擬，分別在時間獨立及時間相依的共變數存活模型下，使用Cox模型、AFT模型及擴充風險模型配適存活資料，我們所感興趣的是Cox模型與AFT模型的相互誤用是否會對共變數的迴歸參數產生影響，而我們的方法是利用擴充風險模型來做Cox模型及AFT模型誤判的研究，而主要的原因在於擴充風險模型為一個較廣義的模型，它同時包含了Cox模型及AFT模型的特性。在實例分析的部份，我們討論台灣愛滋病病患資料 (Taiwan AIDS data)以及Miller (1981)所分析的史丹佛心臟移植資料 (Stanford heart transplant data)。在這兩筆資料中，我們分別使用Cox模型及AFT模型時，觀察其共變數對於風險函數的影響程度。

In the survival analysis, when we use the misspecified model to fit the survival data, it often affects the parameter estimates of the survival model, and this even leads to produce different results of significance from covariates on the hazard function, if we compare with the result of using the correct model to fit the survival data. Therefore, through statistical simulation under the time-independent covariate of survival model and the time-dependent covariate of survival model, we use Cox model, AFT model, and extended hazard model to fit the survival data, and we have interest in the effects of misspecification of regression parameters between Cox and AFT model. Our approach is to use extended hazard model to discuss the effect of misspecification between Cox and AFT model mainly, because the extended hazard model is a more generalized model, which includes the features of Cox model and AFT model. In the real data analysis, we analyze Stanford heart transplant data from Miller (1981), and the Taiwan AIDS data. We use Cox model and AFT model to analyzed these data, and to observe the degree of influence from the covariate on the hazard function.

[1] Andersen, P. K. and Borgan, Q. and Gill, R. D. and Keiding, N. (1993). Statistical Models Based On Counting Processes. Springer: New York.
[2] Andersen, P. K. and Gill, R. D. (1982). Cox''s regression model for counting processes: a large sample study. The Annals of Statistics 10: 1100-1120.
[3] Bretagnolle, J. and Huber-Carol, C. (1985). Sous-estimation des contrastes due à I’oubli de variables pertinentes dans le modèle de Cox pour les durées de survie avec censure. C.R. Acad. Sc Paris, t. 300, Série I 11:359-362.
[4] Buckley, J. and James, I. (1979). Linear regression with censored data. Biometrika 66: 429-436.
[5] Chen, Y. Q. and Jewell, N. P. (2001). On a general class of semiparametric hazards regression models. Biometrika 88: 687-702.
[6] Ciampi, A. and Etezadi-Amoli, J. (1985). A general model for testing the proportional hazards and the accelerated failure time hypothesis in the analysis of censored survival data with covariates. Communication in Statistics-Theory and Methods 14: 651-667.
[7] Cox, D. R. (1975). Partial likelihood. Biometrika 62: 269-276.
[8] Crowley, J. and Hu, M. (1976). Covariance analysis of heart transplant survival data. Journal of the American Statistical Association 72: 27-36.
[9] Etezadi-Amoli, J. and Ciampi, A. (1987). Extended hazard regression for censored survival data with covariates: A spline approximation for the baseline hazard function. Biometrics 43: 181-192.
[10] Gehan, E. A. (1965). A generalized Wilcoxon test for comparing arbitrary single censored samples. Biometrika 52: 203-223.
[11] Hutton, J. L. and Monaghan, P. F. (2002). Choice of parametric accelerated life and proportional hazards models for survival data: Asymptotic results. Lifetime Data Analysis 8: 375-393.
[12] Jones, M. C. (1990). The performance of kernel density functions in kernel distribution function estimation. Statistics & Probability 9: 129-132.
[13] Jones, M. C. and Sheather, S. J. (1991). Using non-stochastic terms to advantage in kernel-based estimation of integrated squared density derivatives. Statistics & Probability 11: 511-514.
[14] Latouche, A. and Boisson, V. and Chevret, S. and Porcher, R. (2007). Misspecified regression model for the subdistribution hazard of a competing risk. Statistics in Medicine 26: 965-974.
[15] Lin, D. Y. and Ying, Z. (1995). Semiparametric inference for the accelerated life model with time-dependent covariates. Journal of Statistical Planning and Inference 44: 47-63.
[16] Lin, D. Y. and Ying, Z. (1997). Additive regression models for survival data. In Proceedings of the First Seattle Symposium in Biostatistics: Survival Analysis. Springer: New York.
[17] Lu, W. (2010). Efficient estimation for an accelerated failure time model with a cure fraction. Statistica Sinica 20: 661-674.
[18] Miller, R. G. (1976). Least squares regression with censored data. Biometrika 63: 449-464.
[19] Miller, R. G. (1981). Survival Analysis. New York: Wiley.
[20] Nelder, J. A. and Mead, R. (1965). A simplex method for function minimization. The Computer Journal 7: 308-313.
[21] O’Neill, T. J. (1986). Inconsistency of the misspecified proportional hazards model. Statistics & Probability 4: 219-222.
[22] Solomon, P. J. (1984). Effect of misspecification of regression models in the analysis of survival data. Biometrika 71: 291-298.
[23] Struthers, C. A. and Kalbfleisch, J. D. (1986). Misspecified proportional hazard models. Biometrika 73: 363-369.
[24] Tseng, Y. K. and Shu, K. N. (2011). Efficient estimation for a semiparametric extended hazards model. Communications in Statistics-Simulation and Computation 40: 258-273.
[25] Tsiatis, A. A. (1981). A large sample study of Cox''s regression model. The Annals of Statistics 9: 93-108.
[26] Zeng, D. and Lin, D. Y. (2007). Efficient estimation for the accelerated failure time model. Journal of the American Statistical Association 102: 1387-1396.