Includes abstract. / Includes bibliographical references. / Breast cancer is the leading cancer among women in South Africa (SA). Studies conducted in developing countries have shown that the majority of women present with advanced stage breast cancer at diagnosis. There is a gap in terms of recent data on the profile of breast cancer patients in SA. The purpose of the study was to obtain recent data with regards to the sociodemographic, clinical and risk factor profile of breast cancer in patients who presented at a Breast Clinic linked to a tertiary public hospital in the Western Cape in order to underpin the development of strategies for earlier detection and diagnosis of breast cancer. A cross-sectional descriptive medical record review was conducted. The studypopulation included all newly diagnosed patients with histological or cytological confirmedbreast cancer who presented at the Breast Clinic during the period 01 January 2009 to 31December 2010. All patients with a previous diagnosis of breast cancer were excluded. Data onthe socio-demographic, clinical and risk factor profile of breast cancer patients were collectedusing a standardised data capture sheet. Data was entered using Epidata version 3.1 and analysedusing Stata Statistical package version 12.After calculation of initial descriptive analysis for the whole sample, male subjects wereexcluded and further analysis was restricted to 585 female subjects. Stage at presentation wascategorised as “early stage” (stage 0, I, IIA, IIB) and “late stage” (stage IIIA, IIIB, IIIC, IV). Crude associations of potential predictors with stage at presentation were tested using Wilcoxon rank-sum tests for medians and Chi-square tests and Fischer Exact tests for proportions. Logistic regression was used to create a model with stage at presentation as dependent variable. Age and racial group were introduced in the model as possible confounders. Based on literature findings other variables present in the dataset were considered as potential predictors of stage at presentation (namely place of residence, employment status, medical aid status, family history of breast cancer, menopausal status, parity, having ever smoked or used alcohol, clinical signs of breast cancer as well as duration of symptoms) and introduced in the model if their bivariate association with the outcome (adjusted for age and race) was statistically significant. A significance level of p <; 0.15 was used. The only variable showing a significant association according to this criterion was the ordinal variable duration of symptoms. The final logistic regression model, therefore, included stage at presentation as the dependent variable and age, racial group and symptom duration as predictors.
| Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:uct/oai:localhost:11427/6033 |
| Date | January 2013 |
| Creators | Daries, Valdiela |
| Contributors | Moodley, Jennifer |
| Publisher | University of Cape Town, Faculty of Health Sciences, Department of Public Health and Family Medicine |
| Source Sets | South African National ETD Portal |
| Language | English |
| Detected Language | English |
| Type | Master Thesis, Masters, MPH |
| Format | application/pdf |
Page generated in 0.011 seconds