A multilevel investigation of inequalities in clinical and psychosocial outcomes for women after breast cancer

Background In Australia, breast cancer is the most common cancer affecting Australian women. Inequalities in clinical and psychosocial outcomes have existed for some time, affecting particularly women from rural areas and from areas of disadvantage. We have a limited understanding of how individual and area-level factors are related to each other, and their associations with survival and other clinical and psychosocial outcomes. Methods/Design This study will examine associations between breast cancer recurrence, survival and psychosocial outcomes (e.g. distress, unmet supportive care needs, quality of life). The study will use an innovative multilevel approach using area-level factors simultaneously with detailed individual-level factors to assess the relative importance of remoteness, socioeconomic and demographic factors, diagnostic and treatment pathways and processes, and supportive care utilization to clinical and psychosocial outcomes. The study will use telephone and self-administered questionnaires to collect individual-level data from approximately 3, 300 women ascertained from the Queensland Cancer Registry diagnosed with invasive breast cancer residing in 478 Statistical Local Areas Queensland in 2011 and 2012. Area-level data will be sourced from the Australian Bureau of Statistics census data. Geo-coding and spatial technology will be used to calculate road travel distances from patients' residence to diagnostic and treatment centres. Data analysis will include a combination of standard empirical procedures and multilevel modelling. Discussion The study will address the critical question of: what are the individual- or area-level factors associated with inequalities in outcomes from breast cancer? The findings will provide health care providers and policy makers with targeted information to improve the management of women with breast cancer, and inform the development of strategies to improve psychosocial care for women with breast cancer.

A comparison of methods for classifying clinical samples based on proteomics data : a case study for statistical and machine learning approaches

The discovery of protein variation is an important strategy in disease diagnosis within the biological sciences. The current benchmark for elucidating information from multiple biological variables is the so called “omics” disciplines of the biological sciences. Such variability is uncovered by implementation of multivariable data mining techniques which come under two primary categories, machine learning strategies and statistical based approaches. Typically proteomic studies can produce hundreds or thousands of variables, p, per observation, n, depending on the analytical platform or method employed to generate the data. Many classification methods are limited by an n≪p constraint, and as such, require pre-treatment to reduce the dimensionality prior to classification. Recently machine learning techniques have gained popularity in the field for their ability to successfully classify unknown samples. One limitation of such methods is the lack of a functional model allowing meaningful interpretation of results in terms of the features used for classification. This is a problem that might be solved using a statistical model-based approach where not only is the importance of the individual protein explicit, they are combined into a readily interpretable classification rule without relying on a black box approach. Here we incorporate statistical dimension reduction techniques Partial Least Squares (PLS) and Principal Components Analysis (PCA) followed by both statistical and machine learning classification methods, and compared them to a popular machine learning technique, Support Vector Machines (SVM). Both PLS and SVM demonstrate strong utility for proteomic classification problems.