Comparing women's unprompted and prompted knowledge of breast and cervical cancer risk factors and symptoms in Sub- Saharan Africa

Breast and cervical cancer are leading causes of female cancer morbidity and mortality in Sub-Saharan Africa (SSA). Despite the high burden of disease, women’s knowledge of evidence-based risk factors and symptoms remains low. To adequately address the apparent knowledge deficits, the underlying knowledge in communities needs to be measured, so as to identify important gaps and contextually address them. To date, cancer knowledge in SSA has been measured using either prompted or unprompted question formats, yielding varying knowledge scores. However, there has been little exploration on the impact of using either question format for assessing disease awareness. This study sought to measure the differences in breast and cervical cancer risk factors and symptoms knowledge reported through prompted and unprompted questions, in South Africa (SA) and Uganda (UG). This was a descriptive cross-sectional study drawing on data collected during validation of an interviewer administered questionnaire (African Woman’s Awareness of Cancer - AWACAN) measuring breast and cervical cancer awareness in SSA. The sample included 139 women recruited from public sector primary health care facilities in two urban districts, Gulu (UG) and Cape Town (SA). Descriptive statistics were used to summarize participant’s socio-demographic characteristics and knowledge about breast and cervical cancer. Composite knowledge scores were calculated by adding up the number of correct responses per individual. The Wilcoxon Singed Rank test was used to compare differences between unprompted and prompted knowledge scores. Regression analyses were used to measure the relationship between unprompted and prompted knowledge. The median age of study participants was 42 years. The majority of women had not completed secondary education (57%) and were unemployed (64%). Unprompted knowledge was considerably lower than prompted knowledge for all breast and cervical cancer risk factors and symptoms. Median scores for unprompted knowledge of breast cancer risk factors (0) and symptoms (1) were significantly lower than for prompted at 6 and 14 respectively. Similarly, the median scores for unprompted knowledge of cervical cancer risk factors (0) and symptoms (1) were lower than prompted knowledge at 6 and 9 respectively. The difference between prompted and unprompted knowledge was least for classical breast and cervical cancer symptoms. For instance, the well-known breast cancer symptom ‘lump in the breast’ was recalled by 57% and 96% with unprompted and prompted questioning respectively. Unprompted questioning identified additional risk lay beliefs such as, ‘itching of the breast’. Combined use of unprompted and prompted questions provides more insight on breast and cervical cancer knowledge patterns in SSA. The low unprompted knowledge scores reported here demonstrate the need for health education interventions to improve knowledge of established breast and cervical cancer risk factors and whilst addressing any predominant lay beliefs about the disease in SSA. This dissertation is divided into three parts. Part A consists of the study protocol outlining the rational for undertaking this study as well as the proposed research methodology. Part B is the literature review that gives a summary of existing literature on the use of prompted and unprompted questions in measuring cancer knowledge thereby providing context for this study. Part C is a journal ready manuscript presenting the results and discussion of study findings