Regression analysis of longitudinal covariates with censored and longitudinal outcome

2018 ◽  
Author(s):  
◽  
Li Chen
Keyword(s):  
Regression Analysis ◽  
Longitudinal Data ◽  
Linear Models ◽  
Partial Likelihood ◽  
Response Variable ◽  
Additive Hazards ◽  
Time Points ◽  
Longitudinal Outcome ◽  
Longitudinal Covariates ◽  
Weighting Technique

[ACCESS RESTRICTED TO THE UNIVERSITY OF MISSOURI AT AUTHOR'S REQUEST.] Longitudinal data contain repeated measurements of variables on the same experimental subject. It is often of interest to analyze the relationship between these variables. Typically, there is one or several longitudinal covariates and a response variable that can be either longitudinal or time to an event. Regression models can be employed to analyze these relationships. Ideally, longitudinal variables should be continuously monitored and their complete trajectories along the time are observed. Practically, however, this is unrealistic, either economically or methodologically. Often one only obtains so called sparse longitudinal data, where variables are intermittently observed at relatively sparse time points within the period of study. Such sparse longitudinal data give rise to an issue for the analysis of the response of time to an event, where survival analysis is typically implemented, e.g. the Cox model or additive hazards model. In both models, the values of covariates of all subjects at risk are needed in order to calculate the partial likelihood. But in the case of sparse longitudinal data, the availability of these observations may not be satis fied. Moreover, if the response variable is also longitudinal, it is possible that the response and covariates are not observed altogether, or at least not close to each other enough to be considered as observed simultaneously. Although a wealth of studies have been dedicated to longitudinal data analysis, very few of them have seriously considered and rigorously studied the situation aforementioned. In this dissertation, we discuss the regression analysis of longitudinal cavities with censored and longitudinal outcome. To be specific, Chapter 2 targets the additive hazards models with sparse longitudinal covariates, Chapter 3 studies the partially linear models with longitudinal covariates and response observed at mismatched time points, also known as asynchronous longitudinal data, and Chapter 4 explores longitudinal data with more complex structures with linear models. Kernel weighting technique is the key idea to all the stated researches. Estimators are derived based on kernel weighting technique and their asymptotical properties were rigorously examined, along with simulation studies for their fi nite sample performance, and illustrations using real data sets.

Functional clustering methods for longitudinal data with application to electronic health records

2020 ◽  
pp. 096228022096563
Author(s):  
Bret Zeldow ◽  
James Flory ◽  
Alisa Stephens-Shields ◽  
Marsha Raebel ◽  
Jason A Roy
Keyword(s):  
Electronic Health Records ◽  
Longitudinal Data ◽  
Mixed Models ◽  
Dirichlet Process ◽  
Dirichlet Process Prior ◽  
Health Records ◽  
Time Points ◽  
Longitudinal Outcome ◽  
Electronic Health ◽  
Baseline Covariates

We develop a method to estimate subject-level trajectory functions from longitudinal data. The approach can be used for patient phenotyping, feature extraction, or, as in our motivating example, outcome identification, which refers to the process of identifying disease status through patient laboratory tests rather than through diagnosis codes or prescription information. We model the joint distribution of a continuous longitudinal outcome and baseline covariates using an enriched Dirichlet process prior. This joint model decomposes into (local) semiparametric linear mixed models for the outcome given the covariates and simple (local) marginals for the covariates. The nonparametric enriched Dirichlet process prior is placed on the regression and spline coefficients, the error variance, and the parameters governing the predictor space. This leads to clustering of patients based on their outcomes and covariates. We predict the outcome at unobserved time points for subjects with data at other time points as well as for new subjects with only baseline covariates. We find improved prediction over mixed models with Dirichlet process priors when there are a large number of covariates. Our method is demonstrated with electronic health records consisting of initiators of second-generation antipsychotic medications, which are known to increase the risk of diabetes. We use our model to predict laboratory values indicative of diabetes for each individual and assess incidence of suspected diabetes from the predicted dataset.

Risperidone-treated Children and Adolescents with Behavioral Disorders: Do Drug Dose and Patients’ Gender and Age Relate to Drug and Metabolite Plasma Levels?

2017 ◽  
Vol 41 (S1) ◽  
pp. S104-S104
Author(s):  
D. Piacentino ◽  
M. Grözinger ◽  
A. Saria ◽  
F. Scolati ◽  
D. Arcangeli ◽  
...  
Keyword(s):  
Regression Analysis ◽  
Children And Adolescents ◽  
Behavioral Disorders ◽  
Plasma Levels ◽  
Linear Models ◽  
Parent Drug ◽  
Drug Dose ◽  
Cut Elimination ◽  
Younger Patients ◽  
Gender And Age

IntroductionBehavioral disorders, such as conduct disorder, influence choice of treatment and its outcome. Less is known about other variables that may have an influence.Objectives/AimsWe aimed to measure the parent drug and metabolite plasma levels in risperidone-treated children and adolescents with behavioral disorders and investigate the role of drug dose and patients’ gender and age.MethodsWe recruited 115 children/adolescents with DSM-5 behavioral disorders (females = 24; age range: 5–18 years) at the Departments of Psychiatry of the Hospitals of Bolzano, Italy, and Innsbruck, Austria. We measured risperidone and its metabolite 9-hydroxyrisperidone plasma levels and the parent drug-to-metabolite ratio in the plasma of all patients by using LC-MS/MS. A subsample of 15 patients had their risperidone doses measured daily. We compared risperidone and 9-hydroxyrisperidone plasma levels, as well as risperidone/9-hydroxyrisperidone ratio, in males vs. females and in younger (≤ 14 years) vs. older (15–18 years) patients by using Mann-Whitney U test. We fitted linear models for the variables “age” and “daily risperidone dose” by using log-transformation, regression analysis and applying the R2 statistic.ResultsFemales had significantly higher median 9-hydroxyrisperidone plasma levels (P = 0.000). Younger patients had a slightly lower median risperidone/9-hydroxyrisperidone ratio (P = 0.052). At the regression analysis, daily risperidone doses and metabolite, rather than parent drug–plasma levels were correlated (R2 = 0.35).ConclusionsGender is significantly associated with plasma levels, with females being slower metabolizers than males. Concerning age, younger patients seem to be rapid metabolizers, possibly due to a higher activity of CYP2D6. R2 suggests a clear-cut elimination of the metabolite.Disclosure of interestThe authors have not supplied their declaration of competing interest.

scholarly journals Regression Analysis of Additive Hazards Model With Latent Variables

2015 ◽  
Vol 110 (511) ◽  
pp. 1148-1159 ◽  
Author(s):  
Deng Pan ◽  
Haijin He ◽  
Xinyuan Song ◽  
Liuquan Sun
Keyword(s):  
Regression Analysis ◽  
Latent Variables ◽  
Additive Hazards Model ◽  
Additive Hazards ◽  
Hazards Model

Regression analysis of informative current status data with the additive hazards model

2014 ◽  
Vol 21 (2) ◽  
pp. 241-258 ◽  
Author(s):  
Shishun Zhao ◽  
Tao Hu ◽  
Ling Ma ◽  
Peijie Wang ◽  
Jianguo Sun
Keyword(s):  
Regression Analysis ◽  
Current Status Data ◽  
Current Status ◽  
Additive Hazards Model ◽  
Additive Hazards ◽  
Hazards Model ◽  
Status Data

scholarly journals Are We Drawing the Correct Conclusions? Regression Analysis in the Nonprofit Literature

2016 ◽  
Vol 7 (1) ◽  
Author(s):  
Robert Shearer ◽  
Truman Clark
Keyword(s):  
Regression Analysis ◽  
Linear Models ◽  
Type Ii ◽  
Type Ii Errors ◽  
Analytical Tools

Linear models are the most commonly used analytical tools in the nonprofit literature. Academics and practitioners utilize these models to test different hypotheses in support of their research efforts, seeking to find significant results that substantiate their theories. And yet the authors of this article have discovered a surprisingly large number of insignificant results in articles from established nonprofit journals. Insignificant hypotheses and Type II errors surely account for a number of these results, but the authors believe the majority of these results are due to a different cause, one that is detectable and preventable: multicollinearity.Dans les articles sur les organismes sans but lucratif, les modèles linéaires sont les outils analytiques les plus communément utilisés. En effet, académiques et praticiens utilisent tous les deux ces modèles pour évaluer diverses hypothèses relatives à leurs recherches, espérant trouver des résultats significatifs pouvant confirmer leurs théories. Pourtant, les auteurs de cet article ont découvert un nombre surprenant de résultats non significatifs dans des articles de revues établies sur les organismes sans but lucratif. Des hypothèses non significatives et des erreurs du type II expliquent sûrement certains de ces résultats, mais les auteurs croient que la majorité des résultats ont une cause différente qui est détectable et évitable : la multicolinéarité.

Sign in / Sign up

Export Citation Format

Share Document