A Semi-nonparametric Approach to Joint Modeling of A Primary Binary Outcome and Longitudinal Data Measured at Discrete Informative Times

Statistics in Biosciences - Tập 4 - Trang 213-234 - 2012
Song Yan1, Daowen Zhang1, Wenbin Lu1, James A. Grifo2, Mengling Liu3
1Department of Statistics, North Carolina State University, Raleigh, USA
2NYU Fertility Center, NYU Langone Medical Center, New York, USA
3Division of Biostatistics, School of Medicine, New York University, New York, USA

Tóm tắt

In a study conducted at the New York University Fertility Center, one of the scientific objectives is to investigate the relationship between the final pregnancy outcomes of participants receiving an in vitro fertilization (IVF) treatment and their β-human chorionic gonadotrophin (β-hCG) profiles. A common joint modeling approach to this objective is to use subject-specific normal random effects in a linear mixed model for longitudinal β-hCG data as predictors in a model (e.g., logistic model) for the final pregnancy outcome. Empirical data exploration indicates that the observation times for longitudinal β-hCG data may be informative and the distribution of random effects for longitudinal β-hCG data may not be normally distributed. We propose to introduce a third model in the joint model for the informative β-hCG observation times, and relax the normality distributional assumption of random effects using the semi-nonparametric (SNP) approach of Gallant and Nychka (Econometrica 55:363–390, 1987). An EM algorithm is developed for parameter estimation. Extensive simulation designed to evaluate the proposed method indicates that ignoring either informative observation times or distributional assumption of the random effects would lead to invalid and/or inefficient inference. Applying our new approach to the data reveals some interesting findings the traditional approach failed to discover.

Tài liệu tham khảo

Bjercke S, Tanbo T, Dale PO, Morkrid L, Abyholm T (1999) Human chorionic gonadotrophin concentrations in early pregnancy after in-vitro fertilization. Hum Reprod 14:1642–1646 Chen HY, Little RJA (1999) Proportional hazards regression with missing covariates. J Am Stat Assoc 94:896–908 Chen J, Zhang D, Davidian M (2001) A Monte Carlo EM algorithm for generalized linear mixed models with flexible random effects distribution. Biostatistics 1:1–27 Chung K, Sammel MD, Coutifaris C, Chalian R, Lin K, Castelbaum AJ, Freedman MF, Barnhart KT (2006) Defining the rise of serum hCG in viable pregnancies achieved through use of IVF. Hum Reprod 21:823–828 Confino E, Demir RH, Friberg J, Gleicher N (1986) The predictive value of hCG β subunit levels in pregnancies achieved by in vitro fertilization and embryo transfer: an international collaborative study. Fertil Steril 45:526–531 Dempster AP, Laird NM, Rubin DB (1977) Maximum likelihood from incomplete data via the EM algorithm. J R Stat Soc, Ser B, Stat Methodol 39:1–38 Gallant AR, Nychka DW (1987) Seminonparametric maximum likelihood estimation. Econometrica 55:363–390 Glatstein IZ, Hornstein MD, Kahana MJ, Jackson KV, Friedman AJ (1995) The predictive value of discriminatory human chorionic gonadotropin levels in the diagnosis of implantation outcome in vitro fertilization cycles. Fertil Steril 63:350–356 Henderson R, Diggle P, Dobson A (2000) Joint modelling of longitudinal measurements and event time data. Biostatistics 1:465–480 Hogan J, Laird N (1997) Model-based approaches to analysing incomplete longitudinal and failure time data. Stat Med 16:259–272 Lambers MJ, Weering HGIV, Grunewold MDV, Lambalk CB, Homburg R, Schats R, Hopes PGA (2006) Optimizing hCG cut-off values: A single determination on day 14 or 15 is sufficient for a reliable prediction of pregnancy outcome. Eur J Obstet Gynecol Reprod Biol 127:94–98 Li E, Zhang D, Davidian M (2004) Conditional estimation for generalized linear models when covariates are subject-specific parameters in a mixed model for longitudinal measurements. Biometrics 60:1–7 Louis TA (1982) Finding the observed information matrix when using the EM algorithm. J R Stat Soc, Ser B, Stat Methodol 44:226–233 Meilijson E (1989) A fast improvement to the EM algorithm on its own terms. J R Stat Soc, Ser B, Stat Methodol 51:127–138 Shamonki M, Frattarelli JL, Bergh PA, Scott RT (2009) Logarithmic curves depicting initial level and rise of serum beta human chorionic gonadotropin and live delivery outcomes with in vetro fertilization: an analysis of 6021 pregnancies. Fertil Steril 91:1760–1764 Strandell A, Thorburn J, Hamberger L (1999) Risk factors for ectopic pregnancy in assisted reproduction. Fertil Steril 71:282–286 Song X, Davidian M, Tsiatis AA (2002) A semiparametric likelihood approach to joint modeling of longitudinal and time-to-event data. Biometrics 58:742–753 Tsiatis AA, Davidian M (2001) A semiparametric estimator for the proportional hazards model with longitudinal covariates measured with error. Biometrika 88:447–458 Vonesh EF, Greene T, Schluchter MD (2006) Shared parameter models for the joint analysis of longitudinal data and event times. Stat Med 25:143–163 Wang CY, Wang N, Wang S (2000) Regression analysis when covariates are regression parameters of a random-effects model for observed longitudinal measurements. Biometrics 56:487–495 Wulfsohn MS, Tsiatis AA (1997) A joint model for survival and longitudinal data measured with error. Biometrics 53:330–339 Zeng D, Cai J (2005) Simultaneous modeling of survival and longitudinal data with an application to repeated quality of life measures. Lifetime Data Anal 11:151–174 Zhang D, Davidian M (2001) Linear mixed models with flexible distributions of random effects for longitudinal data. Biometrics 57:795–802