1996 Remington Lecture: Modeling multivariate longitudinal data that are incomplete

被引:16
作者
Espeland, MA [1 ]
Craven, TE [1 ]
Miller, ME [1 ]
D'Agostino, R [1 ]
机构
[1] Wake Forest Univ, Sch Med, Biostat Sect, Winston Salem, NC 27157 USA
关键词
missing data; mixed models; carotid artery atherosclerosis; clinical trials; maximum likelihood;
D O I
10.1016/S1047-2797(98)00069-6
中图分类号
R1 [预防医学、卫生学];
学科分类号
1004 ; 120402 ;
摘要
PURPOSE: We describe the impact that missing data may have on model selection for longitudinal multivariate data. METHODS: Maximum likelihood was used to fit several models to ultrasonographic measurements from the Asymptomatic Carotid Artery Progression Study (ACAPS). Graphical techniques were used to examine evidence concerning the underlying missing data mechanisms associated with each model. RESULTS: Using statistical methodology that addressed missing data substantially increased the statisti cal efficiency of our analysis of ultrasonographic data. Only complex models that included segment-specific parameterizations for longitudinal correlations appeared to allow missing data to be assumed to occur at random. CONCLUSION: Ignoring the nature of missing data in conducting statistical analyses can have serious consequences when missingness is not rare. It may be necessary to fit models of high dimension with maximum likelihood techniques to address missing data appropriately, however these approaches may improve statistical efficiency. Ann Epidemiol 1999;9:196-205. (C) 1999 Elsevier Science Inc. All rights reserved.
引用
收藏
页码:196 / 205
页数:10
相关论文
共 25 条
[1]   EFFECT OF CHOLESTEROL-LOWERING MEDICATIONS ON PROGRESSION OF MILD ATHEROSCLEROTIC LESIONS OF THE CAROTID ARTERIES AND ON THE RISK OF STROKE [J].
ADAMS, HP ;
BYINGTON, RP ;
HOEN, H ;
DEMPSEY, R ;
FURBERG, CD .
CEREBROVASCULAR DISEASES, 1995, 5 (03) :171-177
[2]  
[Anonymous], BMDP STATISTICAL SOF
[3]  
D'Agostino RB, 1998, STAT MED, V17, P2265, DOI 10.1002/(SICI)1097-0258(19981015)17:19<2265::AID-SIM918>3.0.CO
[4]  
2-B
[5]  
DIGGLE P, 1994, APPL STAT, V50, P1003
[6]   Reliability of longitudinal ultrasonographic measurements of carotid intimal-medial thicknesses [J].
Espeland, MA ;
Craven, TE ;
Riley, WA ;
Corson, J ;
Romont, A ;
Furberg, CD .
STROKE, 1996, 27 (03) :480-485
[7]   ANALYSIS STRATEGIES FOR SERIAL MULTIVARIATE ULTRASONOGRAPHIC DATA THAT ARE INCOMPLETE [J].
ESPELAND, MA ;
BYINGTON, RP ;
HIRE, D ;
DAVIS, VG ;
HARTWELL, T ;
PROBSTFIELD, J .
STATISTICS IN MEDICINE, 1992, 11 (08) :1041-1056
[8]  
Fried Linda P., 1991, Annals of Epidemiology, V1, P263
[9]   EFFECT OF LOVASTATIN ON EARLY CAROTID ATHEROSCLEROSIS AND CARDIOVASCULAR EVENTS [J].
FURBERG, CD ;
ADAMS, HP ;
APPLEGATE, WB ;
BYINGTON, RP ;
ESPELAND, MA ;
HARTWELL, T ;
HUNNINGHAKE, DB ;
LEFKOWITZ, DS ;
PROBSTFIELD, J ;
RILEY, WA ;
YOUNG, B .
CIRCULATION, 1994, 90 (04) :1679-1687
[10]   LESSONS LEARNED FROM CLINICAL-TRIALS WITH ULTRASOUND END-POINTS [J].
FURBERG, CD ;
BYINGTON, RP ;
CRAVEN, TE .
JOURNAL OF INTERNAL MEDICINE, 1994, 236 (05) :575-580