Thesis (M.S.)--Washington State University, May 2006. / Includes bibliographical references (p. 81-83).

Thesis (Ph. D.)--University of Texas at Austin, 2005. / Supervisor: Keenan A. Pituch. Vita. Includes bibliographical references.

Undergraduate honors paper--Mount Holyoke College, 2007. Dept. of Mathematics and Statistics. / Includes bibliographical references (leaves 81-82).

No description available.

To obtain a better understanding of the pharmacokinetic and/or pharmacodynamic characteristics of an investigated treatment, clinical data is often analysed with nonlinear mixed effects modelling. The developed models can be used to design future clinical trials or to guide individualised drug treatment. Missing data is a frequently encountered problem in analyses of clinical data, and to not venture the predictability of the developed model, it is of great importance that the method chosen to handle the missing data is adequate for its purpose. The overall aim of this thesis was to develop methods for handling missing data in the context of nonlinear mixed effects models and to compare strategies for handling missing data in order to provide guidance for efficient handling and consequences of inappropriate handling of missing data. In accordance with missing data theory, all missing data can be divided into three categories; missing completely at random (MCAR), missing at random (MAR) and missing not at random (MNAR). When data are MCAR, the underlying missing data mechanism does not depend on any observed or unobserved data; when data are MAR, the underlying missing data mechanism depends on observed data but not on unobserved data; when data are MNAR, the underlying missing data mechanism depends on the unobserved data itself. Strategies and methods for handling missing observation data and missing covariate data were evaluated. These evaluations showed that the most frequently used estimation algorithm in nonlinear mixed effects modelling (first-order conditional estimation), resulted in biased parameter estimates independent on missing data mechanism. However, expectation maximization (EM) algorithms (e.g. importance sampling) resulted in unbiased and precise parameter estimates as long as data were MCAR or MAR. When the observation data are MNAR, a proper method for handling the missing data has to be applied to obtain unbiased and precise parameter estimates, independent on estimation algorithm. The evaluation of different methods for handling missing covariate data showed that a correctly implemented multiple imputations method and full maximum likelihood modelling methods resulted in unbiased and precise parameter estimates when covariate data were MCAR or MAR. When the covariate data were MNAR, the only method resulting in unbiased and precise parameter estimates was a full maximum likelihood modelling method where an extra parameter was estimated, correcting for the unknown missing data mechanism's dependence on the missing data. This thesis presents new insight to the dynamics of missing data in nonlinear mixed effects modelling. Strategies for handling different types of missing data have been developed and compared in order to provide guidance for efficient handling and consequences of inappropriate handling of missing data.

In longitudinal studies, outcomes are measured repeatedly over time, but in reality clinical studies are full of missing data points of monotone and non-monotone nature. Often this missingness is related to the unobserved data so that it is non-ignorable. In such context, pattern-mixture model (PMM) is one popular tool to analyze the joint distribution of outcome and missingness patterns. Then the unobserved outcomes are imputed using the distribution of observed outcomes, conditioned on missing patterns. However, the existing methods suffer from model identification issues if data is sparse in specific missing patterns, which is very likely to happen with a small sample size or a large number of repetitions. We extend the existing methods using latent class analysis (LCA) and a shared-parameter PMM. The LCA groups patterns of missingness with similar features and the shared-parameter PMM allows a subset of parameters to be different among latent classes when fitting a model, thus restoring model identifiability. A novel imputation method is also developed using the distribution of observed data conditioned on latent classes. We develop this model for continuous response data and extend it to handle ordinal rating scale data. Our model performs better than existing methods for data with small sample size. The method is applied to two datasets from a phase II clinical trial that studies the quality of life for patients with prostate cancer receiving radiation therapy, and another to study the relationship between the perceived neighborhood condition in adolescence and the drinking habit in adulthood.

Clinical practice should be based on the best available evidence. Ideally such evidence is obtained through rigorously conducted, purpose-designed clinical studies such as randomised controlled trials and prospective cohort studies. However gathering information in this way requires a massive effort, can be prohibitively expensive, is time consuming, and may not always be ethical or practicable. When answers are needed urgently and purpose-designed prospective studies are not feasible, retrospective healthcare data may offer the best evidence there is. But can we rely on analysis with such data to give us meaningful answers? The current thesis studies this question through analysis with repeated psychological symptom screening data that were routinely collected from over 20,000 outpatients who attended selected oncology clinics in Scotland. Linked to patients’ oncology records these data offer a unique opportunity to study the progress of distress symptoms on an unprecedented scale in this population. However, the limitations to such routinely collected observational healthcare data are many. We approach the analysis within a missing data context and develop a Bayesian model in WinBUGS to estimate the posterior predictive distribution for the incomplete longitudinal response and covariate data under both Missing At Random and Missing Not At Random mechanisms and use this model to generate multiply imputed datasets for further frequentist analysis. Additional to the routinely collected screening data we also present a purpose-designed, prospective cohort study of distress symptoms in the same cancer outpatient population. This study collected distress outcome scores from enrolled patients at regular intervals and with very little missing data. Consequently it contained many of the features that were lacking in the routinely collected screening data and provided a useful contrast, offering an insight into how the screening data might have been were it not for the limitations. We evaluate the extent to which it was possible to reproduce the clinical study results with the analysis of the observational screening data. Lastly, using the modelling strategy previously developed we analyse the abundant screening data to estimate the prevalence of depression in a cancer outpatient population and the associations with demographic and clinical characteristics, thereby addressing important clinical research questions that have not been adequately studied elsewhere. The thesis concludes that analysis with observational healthcare data can potentially be advanced considerably with the use of flexible and innovative modelling techniques now made practicable with modern computing power.

UA Open Access Publishing Fund / Background Missing outcome data is a threat to the validity of treatment effect estimates in randomized controlled trials. We aimed to evaluate the extent, handling, and sensitivity analysis of missing data and intention-to-treat (ITT) analysis of randomized controlled trials (RCTs) in top tier medical journals, and compare our findings with previous reviews related to missing data and ITT in RCTs. Methods Review of RCTs published between July and December 2013 in the BMJ, JAMA, Lancet, and New England Journal of Medicine, excluding cluster randomized trials and trials whose primary outcome was survival. Results Of the 77 identified eligible articles, 73 (95%) reported some missing outcome data. The median percentage of participants with a missing outcome was 9% (range 0 – 70%). The most commonly used method to handle missing data in the primary analysis was complete case analysis (33, 45%), while 20 (27%) performed simple imputation, 15 (19%) used model based methods, and 6 (8%) used multiple imputation. 27 (35%) trials with missing data reported a sensitivity analysis. However, most did not alter the assumptions of missing data from the primary analysis. Reports of ITT or modified ITT were found in 52 (85%) trials, with 21 (40%) of them including all randomized participants. A comparison to a review of trials reported in 2001 showed that missing data rates and approaches are similar, but the use of the term ITT has increased, as has the report of sensitivity analysis. Conclusions Missing outcome data continues to be a common problem in RCTs. Definitions of the ITT approach remain inconsistent across trials. A large gap is apparent between statistical methods research related to missing data and use of these methods in application settings, including RCTs in top medical journals.

This research provides improved support for missing data in the relational model and relational database systems. There is a need for a systematic method to represent and interpret missing data values in the relational model. A system that processes missing data needs to enable making reasonable decisions when some data values are unknown. The user must be able to understand query results with respect to these decisions. While a number of approaches have been suggested, none have been completely implemented in a relational database system. This research describes a missing data model that works within the relational model, is implemented in MySQL, and was validated by a user feasibility study.

Examination of the literature reveals a paucity of dedicated research into collisions and groundings involving UK fishing vessels. The aim of this research was to provide answers to fundamental questions regarding the factors that contribute to fishing vessel traffic losses. Data for this study were gathered from a broad range of sources and an eclectic range of techniques employed in their analysis. The recent development of the UK fishing fleet and the pattern of losses from all causes is investigated for the period 1975 to 1994. Fishing vessel collision and grounding losses are then set in relative perspective by comparison with those arising from other causes. Aspects of the macro-environment in which the UK fishing fleet has operated since 1975 are examined and the results interpreted in the form of a comparative regional analysis. The micro-environment prevailing in the fishing fleet is exemplified through combining an array of observations made at sea on board working fishing vessels with questionnaire responses drawn from representative samples of British fishermen in 22 fishing ports around the country. A previously unattempted composite analysis of the circumstances of fishing vessel collision and grounding losses is presented and this allows for a number of conclusions to be drawn. A causal analysis technique is applied to fishing vessel casualties for the first time and leads to the identification of human factors as a more significant contributor to traffic losses than either technical or environmental factors. A novel programme of cross-validated observations of fishing vessel watchkeepers in their working environment was pursued, providing data on how attention is allocated, workload levels at different stages in the fishing cycle and also on the watchkeeper's cognitive state while on duty. The thesis concludes with a wide ranging discussion and recommendations based on the research that could contribute to reducing loss of life and vessels in traffic events, made with due consideration for the physical and fiscal constraints that impinge upon the UK fishing fleet.