Ohidul Siddiqui - Academia.edu (original) (raw)

Papers by Ohidul Siddiqui

Http Dx Doi Org 10 1080 10543409808835259, Mar 29, 2007

The last-observation-carried-forward imputation method is commonly used for imputting data missin... more The last-observation-carried-forward imputation method is commonly used for imputting data missing due to dropouts in longitudinal clinical trials. The method assumes that outcome remains constant at the last observed value after dropout, which is unlikely in many clinical trials. Recently, random-effects regression models have become popular for analysis of longitudinal clinical trial data with dropouts. However, inference obtained from random-effects regression models is valid when the missing-at-random dropout process is present. The random-effects pattern-mixture model, on the other hand, provides an approach that is valid under more general missingness mechanisms. In this article we describe the use of random-effects pattern-mixture models under different patterns for dropouts. First, subjects are divided into groups depending on their missing-data patterns, and then model parameters are estimated for each pattern. Finally, overall estimates are obtained by averaging over the missing-data patterns and corresponding standard errors are obtained using the delta method. A typical longitudinal clinical trial data set is used to illustrate and compare the above methods of data analyses in the presence of missing data due to dropouts.

Addict Behav, 1997

The present study tested whether smokeless tobacco (ST) fits a unidimensional model of drug invol... more The present study tested whether smokeless tobacco (ST) fits a unidimensional model of drug involvement and tried to locate ST along the dimension that underlies drug use. The latent trait analysis was employed to quantify drug involvement in a sample of high school students. Analyses showed that although ST use fits the unidimensional model of drug involvement, the place of ST use along the continuum of drug involvement is not stable and differs by gender and ethnicity. Particularly for males, ST use is likely to be preceded by soft drug use and followed by hard drug use; for females, however, ST use is closely associated with hard drug use. The analyses also revealed that the fit of the unidimensional model and the location of ST use along the dimension vary with different ethnic groups.

This study examines the predictors of inconsistent responses from adolescents to questions about ... more This study examines the predictors of inconsistent responses from adolescents to questions about whether they ever used alcohol, cigarettes, and marijuana. Male adolescents had significantly higher rates of inconsistent responses than female adolescents. Black and Hispanic adolescents had significantly higher rates of inconsistent responses regarding ever using alcohol and cigarettes (only for Black) than White adolescents. The subjects' living status and academic achievements were significant predictors of inconsistent responses regarding ever using marijuana. Thus, these results are consistent with the notion that inconsistent responses may bias the estimation of the prevalence of ever using drugs in multivariate analyses.

American journal of preventive medicine

This study describes the patterns and predictors of smokeless tobacco (ST) use in a large sample ... more This study describes the patterns and predictors of smokeless tobacco (ST) use in a large sample of urban public school students in Los Angeles and San Diego. The use of ST is more common among men than women and among Caucasians than African Americans, Hispanics, and others. Approximately 20% of the male respondents and 5% of the female respondents reported use of ST at least once, and 10.1% of male students and 3.1% of female students who had never tried ST by seventh grade started to use it by eighth grade. Among Caucasians, about 30% of boys reported trying ST at least once and 13.7% of those who had never used ST by seventh grade initiated experimentation by eighth grade. These data are used to examine the family, peer, and intrapersonal predictors of ST onset. The family risk factors for ST onset include living with a single parent, parent-child conflicts, and parental alcohol use. The peer risk factors for ST use include peer drug use and activities with friends, such as part...

Journal of biopharmaceutical statistics, 2009

Statistics in Medicine, 2010

In a clinical trial, if there are three or more co-primary endpoints, the type II error could inc... more In a clinical trial, if there are three or more co-primary endpoints, the type II error could increase depending on the correlation among the endpoints and their treatment effect sizes. To keep the type II error under control one may have to consider larger sample sizes. However, in cases where treatment effect size of at least one of the endpoints is likely to be small, the required sample size estimates can exceed reasonable bounds. Patel (1991) proposed an approach that adjusts the significance level for testing each primary endpoint based on the idea of restricting the null space. In Chuang-Stein et al. (2007), the upward adjustment to the significance levels is based on controlling an average type I error rate. In the scenario that statistical significance of each individual hypothesis is not required, we introduce a compromise testing approach in which the significance level for a co-primary endpoint is adjusted upward only if the treatment shows high significance in one (or more than one) of the remaining co-primary endpoints. The adjustment depends on the correlation among the endpoints: larger adjustment is needed for cases of smaller correlation. The method is applicable for the scenario where the null space is restricted. Our testing approach controls maximum joint false positive rate over the restricted null space.

Substance Use & Misuse, 1999

Statistical Methods in Medical Research, 2000

Random-effects regression modelling is proposed for analysis of correlated grouped-time survival ... more Random-effects regression modelling is proposed for analysis of correlated grouped-time survival data. Two analysis approaches are considered. The first treats survival time as an ordinal outcome, which is either right-censored or not. The second approach treats survival time as a set of dichotomous indicators of whether the event occurred for time periods up to the period of the event or censor. For either approach both proportional hazards and proportional odds versions of the random-effects model are developed, while partial proportional hazards and odds generalizations are described for the latter approach. For estimation, a full-information maximum marginal likelihood solution is implemented using numerical quadrature to integrate over the distribution of multiple random effects. The quadrature solution allows some flexibility in the choice of distributions for the random effects; both normal and rectangular distributions are considered in this article. An analysis of a dataset where students are clustered within schools is used to illustrate features of random-effects analysis of clustered grouped-time survival data.

Preventive Medicine, 1999

Background. In school-based smoking prevention reof smoking. search, it is still debatable whethe... more Background. In school-based smoking prevention reof smoking. search, it is still debatable whether parents or peers are most influential to maintained smoking among adolescents. As a result, this study examines the effects of tion of risk factors for smoking onset or initiation among strong for black, Hispanic, and Asian adolescents.

Preventive Medicine, 1996

or the escalation of smoking over time. This design is Background. In longitudinal smoking preven... more or the escalation of smoking over time. This design is Background. In longitudinal smoking prevention also useful for studying developmental sequences: for studies, a difficulty in evaluating treatment effects is example, when smoking cigarettes is followed by mariunderstanding whether bias is associated with those juana use and later by other illegal drug use. 2 One who do not complete the study. This study presents important characteristic of the longitudinal design is the significant predictors of attrition and suggests how the ability to compare the same individual at different to reduce attrition bias in evaluating program effects. time points, that is, to conduct within-individual analy-Methods. Survival analysis methods were used to asses of change. Information about within-individual sess factors associated with attrition at different time change over time is useful to draw valid conclusions points of the study.

Random-e®ects regression modeling is proposed for analysis of correlated grouped-time survival da... more Random-e®ects regression modeling is proposed for analysis of correlated grouped-time survival data. Two analysis approaches are considered. The¯rst treats survival time as an ordinal outcome, which is either right-censored or not. The second approach treats survival time as a set of dichotomous indicators of whether the event occurred for time periods up to the period of the event or censor. For either approach both proportional hazards and proportional odds versions of the random-e®ects model are developed, while partial proportional hazards and odds generalizations are described for the latter approach. For estimation, a full-information maximum marginal likelihood (MML) solution is implemented using numerical quadrature to integrate over the distribution of multiple random e®ects. The quadrature solution allows some°exibility in the choice of distributions for the random e®ects; both normal and rectangular distributions are considered in this article. An analysis of a dataset where students are clustered within schools is used to illustrate features of random-e®ects analysis of clustered grouped-time survival data. Short Running Title: Random-e®ects survival model

The Journal of Heart and Lung Transplantation, 2010

Journal of Biopharmaceutical Statistics, 1999

Assessment of quality of life (QOL) in clinical trials becomes a challenging task from the viewpo... more Assessment of quality of life (QOL) in clinical trials becomes a challenging task from the viewpoint of clinical biostatistics. The responses of the items for measuring QOL indices usually vary widely from patient to patient and from time to time. Measurement errors might be present in the responses of the items, and they might be correlated. Hence, in analyzing QOL data, the usual assumption that there are no measurement errors in responses is too liberal. Because the QOL indices are likely to be correlated, separate analysis of each index might not be efficient from the point of view of statistical methodology. We apply linear structural equation modeling (LISREL) in assessing the QOL data obtained from a clinical trial. A basic premise of the LISREL approach is that the abstract concepts (latent constructs) that are not directly measurable can be studied. LISREL is a statistical procedure for conceiving and testing structural hypotheses that cannot be tested adequately with other statistical procedures. It allows us to specify relations between unobserved and observed variables while controlling for measurement errors and correlations among both the measurement errors and the latent constructs.

Journal of Biopharmaceutical Statistics, 2009

In recent years, the use of the last observation carried forward (LOCF) approach in imputing miss... more In recent years, the use of the last observation carried forward (LOCF) approach in imputing missing data in clinical trials has been greatly criticized, and several likelihood-based modeling approaches are proposed to analyze such incomplete data. One of the proposed likelihood-based methods is the Mixed-Effect Model Repeated Measure (MMRM) model. To compare the performance of LOCF and MMRM approaches in analyzing incomplete data, two extensive simulation studies are conducted, and the empirical bias and Type I error rates associated with estimators and tests of treatment effects under three missing data paradigms are evaluated. The simulation studies demonstrate that LOCF analysis can lead to substantial biases in estimators of treatment effects and can greatly inflate Type I error rates of the statistical tests, whereas MMRM analysis on the available data leads to estimators with comparatively small bias, and controls Type I error rates at a nominal level in the presence of missing completely at random (MCAR) or missing at random (MAR) and some possibility of missing not at random (MNAR) data. In a sensitivity analysis of 48 clinical trial datasets obtained from 25 New Drug Applications (NDA) submissions of neurological and psychiatric drug products, MMRM analysis appears to be a superior approach in controlling Type I error rates and minimizing biases, as compared to LOCF ANCOVA analysis. In the exploratory analyses of the datasets, no clear evidence of the presence of MNAR missingness is found.

Journal of Biopharmaceutical Statistics, 2000

Statistical analysis based on multiple imputation (MI) of missing data when analyzing data with m... more Statistical analysis based on multiple imputation (MI) of missing data when analyzing data with missing observations is gaining popularity among statisticians because of availability of computing softwares; it might be tempting to use MI whenever data is missing. An important assumption behind MI is the "ignorability of missingness." In this paper, we demonstrate the use of MI in conjunction with random effects models and several other methods that are devised to handle nonignorable missingness (informative dropouts). We then compare the results to assess sensitivity to underlying assumptions. Our focus is primarily to estimate and compare rates of change (of a primary variable). The application dataset has a high dropout rate and has features to suggest informativeness of the dropout process. The estimates obtained under random effects modeling with multiple imputation were found to differ substantially from those obtained by methods devised to handle informative dropouts.

Journal of Biopharmaceutical Statistics, 1998

Journal of Applied …, Jan 1, 1995

American Journal of Epidemiology, 1996

Most school-based smoking prevention studies employ designs in which schools or classrooms are as... more Most school-based smoking prevention studies employ designs in which schools or classrooms are assigned to different treatment conditions while observations are made on individual students. This design requires that the treatment effect be assessed against the between-school variance. However, the betweenschool variance is usually larger than the variance that would be obtained if students were individually randomized to different conditions. Consequently, the power of the test for a treatment effect is reduced, and it becomes difficult to detect important treatment effects. To assess the potential loss of power or to calculate appropriate sample sizes, investigators need good estimates of the intraclass correlations for the variables of interest. The authors calculated intraclass correlations for some common outcome variables in a school-based smoking prevention study, using a three-level model-i.e., students nested within classrooms and classrooms nested within schools. The authors present the intraclass correlation estimates for the entire data set, as well as separately by sex and ethnicity. They also illustrate the use of these estimates in the planning of future studies. Am J Epidemiol 1996; 144:425-33.