Borderline personality disorder (BPD) is a complex psychiatric condition whose severity is compounded by the heterogeneous psychological functioning of those who suffer from the disorder. This heterogeneity has made the identification of a unified treatment strategy difficult and attempts to resolve this variation within the disorder by investigating subtypes of BPD have been made. However, the clinical utility of this approach has not been examined. The major object of this research project was to investigate the presence of subtypes of BPD and to examine whether treatment effectiveness varied as a function of subtype. Data from 61, predominantly female, participants with BPD were entered into a cluster analysis. Using variables that are central to cognitive behavioural models of BPD and have been shown to be heterogeneously distributed in previous BPD samples, two subgroups were identified and defined on the basis of whether participants attributed the causes o f negative events as being themselves or other people. Consistent with hypotheses, the subgroup with a tendency to blame others for negative events showed far lower levels of change between admission and discharge than the subgroup who blamed themselves for negative events, on both measures of statistical and clinical significance. Alternate means of identifying participants who optimally responded to the intervention were explored and a cluster analysis identified two groups of participants that were separated on the basis of whether they had shown clinically significant change on a range of variables. It was found that data from admission to the program could successfully predict which participants would belong in the optimal or mixed response groups upon discharge. Taken together, the findings of this research project suggest that not only can theoretically valid subgroups of BPD be identified, but that they have clinical utility in understanding participants' response to intervention. Further, the findi ngs suggest that profiles of clinical change can be identified and predicted. The findings of this research project are discussed with respect to their methodological limitations, suggestions for future research, and their implications for both theory and practice.

Antibiotic resistant bacteria, a growing health crisis, arise due to antibiotic overuse and misuse. Resistant infections endanger the lives of patients and are financially burdensome. Aggregate antimicrobial susceptibility reports, called antibiograms, are critical for tracking antibiotic susceptibility and evaluating the likelihood of the effectiveness of different antibiotics to treat an infection prior to the availability of patient specific susceptibility data. This research leverages the Massachusetts Statewide Antibiogram database, a rich dataset composed of antibiograms for $754$ antibiotic-bacteria pairs collected by the Massachusetts Department of Public Health from $2002$ to $2016$. However, these antibiograms are at least a year old, meaning antibiotics are prescribed based on outdated data which unnecessarily furthers resistance. Our objective is to employ data science techniques on these antibiograms to assist in developing more responsible antibiotic prescription practices. First, we use model selectors with regression-based techniques to forecast the current antimicrobial resistance. Next, we develop an assistant to immediately identify clinically and statistically significant changes in antimicrobial resistance between years once the most recent year of antibiograms are collected. Lastly, we use k-means clustering on resistance trends to detect antibiotic-bacteria pairs with resistance trends for which forecasting will not be effective. These three strategies can be implemented to guide more responsible antibiotic prescription practices and thus reduce unnecessary increases in antibiotic resistance.

Antibiotic resistant bacteria, a growing health crisis, arise due to antibiotic overuse and misuse. Resistant infections endanger the lives of patients and are financially burdensome. Aggregate antimicrobial susceptibility reports, called antibiograms, are critical for tracking antibiotic susceptibility and evaluating the likelihood of the effectiveness of different antibiotics to treat an infection prior to the availability of patient specific susceptibility data. This research leverages the Massachusetts Statewide Antibiogram database, a rich dataset composed of antibiograms for $754$ antibiotic-bacteria pairs collected by the Massachusetts Department of Public Health from $2002$ to $2016$. However, these antibiograms are at least a year old, meaning antibiotics are prescribed based on outdated data which unnecessarily furthers resistance. Our objective is to employ data science techniques on these antibiograms to assist in developing more responsible antibiotic prescription practices. First, we use model selectors with regression-based techniques to forecast the current antimicrobial resistance. Next, we develop an assistant to immediately identify clinically and statistically significant changes in antimicrobial resistance between years once the most recent year of antibiograms are collected. Lastly, we use k-means clustering on resistance trends to detect antibiotic-bacteria pairs with resistance trends for which forecasting will not be effective. These three strategies can be implemented to guide more responsible antibiotic prescription practices and thus reduce unnecessary increases in antibiotic resistance.

No description available.

The inter-subjective and dialogical nature of narrative therapy, as commonly practiced, remains unarticulated. Further, there currently exists no rigorous empirical research investigating the process or outcome of narrative therapy. The research aim, to investigate the process and outcome of narrative therapy, comprised theoretical and empirical objectives. The first objective was to articulate a theoretical synthesis of narrative theory, research and practice. The process of narrative reflexivity was identified as a theoretical construct linking narrative theory with narrative research and practice. The second objective was to substantiate this synthesis empirically by examining narrative therapy processes, specifically narrative reflexivity and the therapeutic alliance, and their relation to therapy outcomes. The third objective was to support the proposed synthesis of theory, research and practice and provide quantitative evidence for the utility of narrative therapy, by evaluating depressive symptom and inter-personal relatedness outcomes through analyses of statistical significance, clinical significance and benchmarking. Founded in theories of self, language and narrative (James, 1890; Bruner, 1986; Gergen, 1991; Hollway, 2006; Vygotsky, 1934/ 1987), narrative therapy was conceptualized as involving dialogical and intra-personal processes. Narrative therapists generally apply a story metaphor and commonly focus on the inter-personal field (White, 2007). This thesis recognised the storied and inter-personal nature of narrative therapy, but proposed this does not represent narrative therapy in its entirety. The notion of story connotes monological processes, inconsistent with the conversations of narrative practice, and neglect of intra-personal dimensions is inconsistent with narrative notions of inter-subjectivity. This thesis proposed an integration of dialogical narrative theory (Cooper, 2003; Hermans & Kempen, 1993; Lysaker & Lysaker, 2006) and narrative research (Angus, Levitt, & Hardtke, 1999) provides a model for understanding narrative therapy (White, 2007) as involving the inter-subjective and dialogical process of narrative reflexivity. During the process of narrative reflexivity, a person engages in dialogue with his or her own self and others as extensions of self, interpreting experience from diverse perspectives in the context of personal aspects, such as beliefs, values and intentions that give meaning to experience, to achieve a rich narrative and a sense of well-being. To support this theoretical synthesis, a process-outcome trial evaluated eight-sessions of narrative therapy for 47 adults with major depressive disorder. Dependent process variables were narrative reflexivity (assessed at Sessions 1 and 8) and therapeutic alliance (assessed at Sessions 1, 3 and 8). Primary dependent outcome variables were depressive symptoms and inter-personal relatedness. Primary analyses assessed therapy outcome at pre-therapy, post-therapy and three-month follow-up and utilized a benchmarking strategy to the evaluate pre-therapy to post-therapy and post-therapy to follow-up gains, effect size and pre-therapy to post-therapy clinical significance. Results indicated that when a sub-sample of clients were categorised into five least-improved and five most-improved groups (according to depressive symptom change), there was a differential change in the percentage of reflexive sequences in the discourse of clients at the end of therapy depending on outcome. Improvement in the quality of the working alliance was associated with improvements in depressive symptoms and inter-personal relatedness, with working alliance improvement from Session 1 to 8 sharing 19% of the variance in depressive symptom improvement and 17% of the variance in inter-personal relatedness improvement from pre-therapy to post-therapy. The clinical trial provided empirical support for the utility of narrative therapy in improving depressive symptoms and inter-personal relatedness from pre-therapy to post-therapy: the magnitude of change indicating large effect sizes (d = 1.10 to 1.36) for depressive symptoms and medium effect sizes (d = .52 to .62) for inter-personal relatedness. Therapy was effective in reducing depressive symptoms in clients with moderate and severe pre-therapy depressive symptom severity. Improvements in depressive symptoms, but not inter-personal relatedness, were maintained three-months following therapy. The reduction in depressive symptoms and the proportion of clients who achieved clinically significant improvement (53%) in depressive symptoms at post-therapy were comparable to improvements from standard psychotherapies, reported in benchmark research. This research has implications for assisting our understanding of narrative approaches, refining strategies that will facilitate recovery from psychological disorder and providing clinicians with a broader evidence base for narrative practice. Despite limitations of a repeated-measures research design, use of a standardised intervention protocol, coupled with outcome evaluation of clinical significance enhanced internal validity. Future research could examine narrative therapy in a larger sample, with different disorders, and with an alternative therapy or control group. Coding a greater number of therapy transcripts for evaluating associations of narrative reflexivity with working alliance and outcomes could enhance understanding of narrative reflexivity. Thesis strengths included a strong theoretical foundation underpinning the research design and arguments, examination of therapy process in the context of outcome, and a parsimonious evaluation of narrative therapy outcomes.

No description available.