Time filter

Source Type

West Haven, CT, United States

Zhou B.,Yale University | Zhou B.,Cooperative Studies Program Coordinating Center | Fine J.,University of North Carolina at Chapel Hill | Latouche A.,French National Conservatory of Arts and Crafts | Labopin M.,Paris University
Biostatistics | Year: 2012

A population average regression model is proposed to assess the marginal effects of covariates on the cumulative incidence function when there is dependence across individuals within a cluster in the competing risks setting. This method extends the Fine-Gray proportional hazards model for the subdistribution to situations, where individuals within a cluster may be correlated due to unobserved shared factors. Estimators of the regression parameters in the marginal model are developed under an independence working assumption where the correlation across individuals within a cluster is completely unspecified. The estimators are consistent and asymptotically normal, and variance estimation may be achieved without specifying the form of the dependence across individuals. A simulation study evidences that the inferential procedures perform well with realistic sample sizes. The practical utility of the methods is illustrated with data from the European Bone Marrow Transplant Registry. The Author 2011. Published by Oxford University Press. All rights reserved. Source

Azad N.,Hines Veterans Administration Hospital | Agrawal L.,Hines Veterans Administration Hospital | Emanuele N.V.,Hines Veterans Administration Hospital | Klein R.,University of Wisconsin - Madison | And 2 more authors.
Diabetologia | Year: 2014

Aims/hypothesis: The aim of this study was to test the hypothesis that intensive glycaemic control (INT) and higher plasma C-peptide levels in patients with poorly controlled diabetes would be associated with better eye outcomes. Methods: The incidence and progression of diabetic retinopathy (DR) was assessed by grading seven-field stereoscopic fundus photographs at baseline and 5 years later in 858 of 1,791 participants in the Veterans Affairs Diabetes Trial (VADT). Results: After adjustment for all covariates, risk of progression (but not incidence) of DR increased by 30% for each 1% increase in baseline HbA 1c (OR 1.3; 95% CI 1.123, 1.503; p∈=∈0.0004). Neither assignment to INT nor age was independently associated with DR in the entire cohort. However, INT showed a biphasic interaction with age. The incidence of DR was decreased in INT participants ≤55 years of age (OR 0.49; 95% CI 0.24, 1.0) but increased in those ≥70 years old (OR 2.88; 95% CI 1.0, 8.24) (p∈=∈0.0043). The incidence of DR was reduced by 67.2% with each 1 pmol/ml increment in baseline C-peptide (OR 0.328; 95% CI 0.155, 0.7; p∈=∈0.0037). Baseline C-peptide was also an independent inverse risk factor for the progression of DR, with a reduction of 47% with each 1 pmol/ml increase in C-peptide (OR 0.53; 95% CI 0.305, 0.921; p∈=∈0.0244). Conclusions/interpretation: Poor glucose control at baseline was associated with an increased risk of progression of DR. INT was associated with a decreased incidence of DR in younger patients but with an increased risk of DR in older patients. Higher C-peptide at baseline was associated with reduced incidence and progression of DR. © 2014 Springer-Verlag Berlin Heidelberg (outside the USA). Source

Zhou B.,Yale University | Zhou B.,Cooperative Studies Program Coordinating Center | Fine J.,University of North Carolina at Chapel Hill | Laird G.,Bristol Myers Squibb
Statistics in Medicine | Year: 2013

This paper concerns using modified weighted Schoenfeld residuals to test the proportionality of subdistribution hazards for the Fine-Gray model, similar to the tests proposed by Grambsch and Therneau for independently censored data. We develop a score test for the time-varying coefficients based on the modified Schoenfeld residuals derived assuming a certain form of non-proportionality. The methods perform well in simulations and a real data analysis of breast cancer data, where the treatment effect exhibits non-proportional hazards. © 2013 John Wiley & Sons, Ltd. Source

Acharjee S.,Einstein Medical Center Philadelphia | Boden W.E.,Albany Medical College | Hartigan P.M.,Cooperative Studies Program Coordinating Center | Teo K.K.,McMaster University | And 8 more authors.
Journal of the American College of Cardiology | Year: 2013

Objectives This study sought to assess the independent effect of high-density lipoprotein-cholesterol (HDL-C) level on cardiovascular risk in patients with stable ischemic heart disease (SIHD) who were receiving optimal medical therapy (OMT). Background Although low HDL-C level is a powerful and independent predictor of cardiovascular risk, recent data suggest that this may not apply when low-density lipoprotein-cholesterol (LDL-C) is reduced to optimal levels using intensive statin therapy. Methods We performed a post-hoc analysis in 2,193 men and women with SIHD from the COURAGE trial. The primary outcome measure was the composite of death from any cause or nonfatal myocardial infarction (MI). The independent association between HDL-C levels measured after 6 months on OMT and the rate of cardiovascular events after 4 years was assessed. Similar analyses were performed separately in subjects with LDL-C levels below 70 mg/dl (1.8 mmol/l). Results In the overall population, the rate of death/MI was 33% lower in the highest HDL-C quartile as compared with the lowest quartile, with quartile of HDL-C being a significant, independent predictor of death/MI (p = 0.05), but with no interaction for LDL-C category (p = 0.40). Among subjects with LDL-C levels <70 mg/dl, those in the highest quintile of HDL-C had a 65% relative risk reduction in death or MI as compared with the lowest quintile, with HDL-C quintile demonstrating a significant, inverse predictive effect (p = 0.02). Conclusions In this post-hoc analysis, patients with SIHD continued to experience incremental cardiovascular risk associated with low HDL-C levels despite OMT during long-term follow-up. This relationship persisted and appeared more prominent even when LDL-C was reduced to optimal levels with intensive dyslipidemic therapy. (Clinical Outcomes Utilizing Revascularization and Aggressive Drug Evaluation; NCT00007657). © 2013 by the American College of Cardiology Foundation. Source

Li X.,Cooperative Studies Program Coordinating Center | Hedeker D.,University of Illinois at Chicago
Statistics in Medicine | Year: 2012

In studies using ecological momentary assessment (EMA), or other intensive longitudinal data collection methods, interest frequently centers on changes in the variances, both within-subjects and between-subjects. For this, Hedeker et al. (Biometrics 2008; 64: 627-634) developed an extended two-level mixed-effects model that treats observations as being nested within subjects and allows covariates to influence both the within-subjects and between-subjects variance, beyond their influence on means. However, in EMA studies, subjects often provide many responses within and across days. To account for the possible systematic day-to-day variation, we developed a more flexible three-level mixed-effects location scale model that treats observations within days within subjects, and allows covariates to influence the variance at the subject, day, and observation level (over and above their usual effects on means) using a log-linear representation throughout. We provide details of a maximum likelihood solution and demonstrate how SAS PROC NLMIXED can be used to achieve maximum likelihood estimates in an alternative parameterization of our proposed three-level model. The accuracy of this approach using NLMIXED was verified by a series of simulation studies. Data from an adolescent mood study using EMA were analyzed to demonstrate this approach. The analyses clearly show the benefit of the proposed three-level model over the existing two-level approach. The proposed model has useful applications in many studies with three-level structures where interest centers on the joint modeling of the mean and variance structure. © 2012 John Wiley & Sons, Ltd. Source

Discover hidden collaborations