Thesis (Ph. D.)--University of Hong Kong, 2009. / Includes bibliographical references (leaves 138-148). Also available in print.

Indiana University-Purdue University Indianapolis (IUPUI) / The present thesis is an analysis of longitudinal data collected through the years 2011-2013, from a complex of four hospitals located in Indiana, USA. The aim of the analysis was the detection of changes (especially a decline) in the disease related group (DRG) weights (and thus, the case mix index (CMI)), and the determination of the predictors that significantly affect these changes. The document is divided in four major parts. In the first part it is described the statistical theory required for the the analysis, in the second part the reimbursement strategies for the hospitals in the USA, are briefly described and the concept of the DRG and CMI are explained. In the third part the actual analysis is presented while the last part contains a summary of the findings and some conclusions. The correlation between the observations was taken into account by modeling the data using linearmixed models (LMM). Three major factors were studied for their effect on the DRG weight of thehospitals: the changes in the type of cases (i.e. the product lines), the changes in the number of the Surgical cases, and also the changes of the length of stay (LOS). The analysis did not indicate any significant DRG change in any of the hospitals except from the H4. The H4 hospital has a significant decline over time regarding the Cardio-vascular (CV) DRG weights. For the hospitals H1, H2 and H3 the only decline observed in the product lines was that for the Medical-Surgical DRG. Finally, no significant change was observed for the LOS, or the number of Surgical cases. In addition to the three predictors studied, changes in the coding system, the documentation etc. may also affect the DRG and CMI. However, these changes are not possible to be detected through this analysis, since no available information was given in the present data.

No description available.

published_or_final_version / Dentistry / Master / Master of Philosophy

published_or_final_version / Dentistry / Master / Master of Dental Surgery

published_or_final_version / Statistics and Actuarial Science / Doctoral / Doctor of Philosophy

Seven Vermont school districts participated in a five year professional development program sponsored jointly by the National Science Foundation and the United States Department of Education from 2002-2007. Using a robust mixed methods evaluation, teachers and students demonstrate pronounced organizational and academic growth. Analysis of data from focus groups held with teachers over the course of the period from fall 2004-spring 2006 provides strong supporting evidence for the growth. The purpose of this dissertation is to reanalyze the focus group data to document institutional and longitudinal change at the first person level. With focus groups as the unit of analysis, themes rising from the anonymous participants‟ I statements form the substance for this review. By revisiting an extensive pre-existing data set with a different method of analysis, this work expands on what is known about how teachers process change on the ground level. The findings reveal how complex individual feelings about one‟s experiences serve to describe degrees of institutional as well as personal change. New thematic coding confirms the original findings of the program evaluation. More importantly, the findings provide new details and understandings about organizational change and growth previously unobserved in the aggregate reports. By way of a methodological contribution, the research findings suggest and demonstrate an alternative approach to the analysis of focus group data in the aggregate.

Introduction: The process by which children's bones grow has not been fully charcterised. The current dogma is that girls fill in their medullary canal area by forming bone at the endosteum. It has been argued that the sex difference in how bone strength is conferred -- favouring boys -- may contribute to the relative protection that aging men have over aging women with respect to fracture incidence and the prevalence of osteoporosis. Primary Objectives: 1)To compare bone surface changes at the periosteal and endosteal surface of the tibial midshaft in boys and girls. 2)To compare how bone density at the tibial midshaft is accrued in boys and girls. 3) To compare sex differences in bone strength accrual. Methods: Design and Participants: Participants were obtained from a 20-month randomized, controlled school-based physical activity intervention. As we found no difference in the effect of the intervention on pQCT bone outcome variables, both groups were combined for our current study. A total of 183 participants (93 boys, 89 girls) received a pQCT scan at baseline. Results: Sex-specific comparisons of the pQCT bone outcome variables showed significantly greater rates of change (slope) for boys for the total area (ToA), cortical area (CoA), medullary canal area (MedA) and strength-strain index (SSI) measures, p<0.001. No significant differences were observed for CoD, p=0.904. The magnitude of these differences is 60.8% for ToA, 55.7% for CoA, 75.6% for MedA, 1.3% for CoD, and 54.7% for SSI. Examination of differences between the sexes (intercept) revealed significant differences with greater gains observed for boys for all measures p<0.001 except for CoD where girls exhibited greater gains p<0.001. Conclusion: Girls showed a similar pattern of cortical bone growth at the tibial midshaft- periosteal apposition dominated over endosteal resorption. Boys' increased changes and pattern of growth were of a greater magnitude at both surfaces compared to girls. This resulted in a greater increase in strength as measured by SSI in boys which can partly be explained by their larger size. Girls exhibited greater increases in CoD; however, no significant difference in the change in CoD was observed between the two.

For longitudinal data where the response and time-dependent predictors within each individual are measured at distinct time points, traditional longitudinal models such as generalized linear mixed effects models or marginal models cannot be directly applied. Instead, some preprocessing such as smoothing is required to temporally align the response and predictors. In Chapter 2, we propose a binning method, which results in equally spaced bins of time for both the response and predictor(s). Hence, after incorporating binning, traditional models can be applied. The proposed binning approach was applied on a longitudinal hemodialysis study to look for possible contemporaneous and lagged effects between occurrences of a health event (i.e., infection) and levels of a protein marker of inflammation (i.e., C-reactive protein). Both Poisson mixed effects models and zero-inflated Poisson (ZIP) mixed effects models were applied to the subsequent binned data, and some important biological findings about contemporaneous and lagged associations were uncovered. In addition, a simulation study was conducted to investigate various properties of the binning approach. In Chapter 3, asymptotic properties have been derived for the fixed effects association parameter estimates following binning, under different data scenarios. In addition, we propose some leave-one-subject-out cross-validation algorithms for bin size selection. In Chapter 4, in order to identify levels of a predictor that might be indicative of recently occurred event(s), we propose a generalized mixed effects regression tree (GMRTree) based method which estimates the tree by standard tree method such as CART and estimates the random effects by a generalized linear mixed effects model. One of the main steps in this method was to use a linearization technique to change the longitudinal count response into a continuous surrogate response. Simulations have shown that the GMRTree method can effectively detect the underlying tree structure in an applicable longitudinal dataset, and has better predictive performance than either a standard tree approach without random effects or a generalized linear mixed effects model, assuming the underlying model indeed has a tree structure. We have also applied this method to two longitudinal datasets, one from the aforementioned hemodialysis study and the other from an epilepsy study.

Modeling covariance structure is important for efficient estimation in longitudinal data models. Modified Cholesky decomposition (Pourahmadi, 1999) is used as an unconstrained reparameterization of the covariance matrix. The resulting new parameters have transparent statistical interpretations and are easily modeled using covariates. However, this approach is not directly applicable when the longitudinal data are unbalanced, because a Cholesky factorization for observed data that is coherent across all subjects usually does not exist. We overcome this difficulty by treating the problem as a missing data problem and employing a generalized EM algorithm to compute the ML estimators. We study the covariance matrices in both fixed-effects models and mixed-effects models for unbalanced longitudinal data. We illustrate our method by reanalyzing Kenwards (1987) cattle data and conducting simulation studies.