Estimating injury mortality in South Africa and identifying urban-rural differences
| dc.contributor.advisor | Myers, Jonathan | |
| dc.contributor.advisor | Bradshaw, Debbie | |
| dc.contributor.advisor | Matzopoulos, Richard | |
| dc.contributor.author | Prinsloo, Megan | |
| dc.date.accessioned | 2019-05-15T07:25:55Z | |
| dc.date.available | 2019-05-15T07:25:55Z | |
| dc.date.issued | 2019 | |
| dc.date.updated | 2019-05-14T11:29:56Z | |
| dc.description.abstract | The overarching aim of this thesis is to utilise national data on injury mortality in South Africa, to conduct advanced statistical analyses to identify urban-rural differences for injury deaths, and to gain insight into the explanatory variables for homicide in metropolitan- and non-metropolitan (metro- and non-metro) areas. The literature review describes the global and national estimates of injury mortality and reports higher rural than urban injury mortality rates for high-income countries. It further discusses a framework for assessing data quality and reviews South Africa’s fatal and non-fatal injury data sources, issues of under-reporting and misclassification of deaths. The risk factors for violence are reviewed, which inform particular hypotheses on the role of age, sex, race, day of week and firearms with regard to homicide. The Injury Mortality Survey (IMS) data, which estimated 52 493 injury deaths nationally in 2009, is utilised for this PhD study. Data quality is assessed using an internationally developed conceptual framework for mortality data. Exploratory and multiple correspondence analysis identified possible associations between metro/nonmetro and other explanatory variables, prior to more sophisticated multinomial logistic regression analysis, which adjusted for age, sex, race and metro/non-metro for each manner of death (homicide, suicide, transport-related and other unintentional injury deaths) to explore particular hypotheses for the differences in the metro/non-metro injury mortality profile. Age-standardised injury mortality rates were calculated to take into account the effects of different age structures for metro- and non-metro populations. Generalized linear models were fitted in relation to particular hypotheses to determine the explanatory variables for homicide deaths in both metro and nonmetro areas. Main findings include a significantly higher likelihood for homicide in metro areas compared to non-metro areas, while transport-related deaths were significantly lower in metro areas. The risk of homicide for Coloureds was higher than Blacks in metro areas, while Blacks, Coloureds and Asians had similar risks of homicide in non- metro areas. Whites had a similar risk and Asians a higher risk of homicide in nonmetro areas compared with metro areas. Firearm use was shown to significantly explain metro/non-metro differences in homicide risks. This study’s most significant knowledge contribution includes the identification of metro/non-metro as a significant predictor of the injury mortality profile in South Africa. The association of metro/non-metro differences in the pattern of homicide for Blacks and Coloureds, also resolved conflicting statements found in the literature regarding race and homicide in South Africa. The results are of considerable significance to national and provincial policy makers. Recommendations are made in relation to the main findings of this study. | |
| dc.identifier.apacitation | Prinsloo, M. (2019). <i>Estimating injury mortality in South Africa and identifying urban-rural differences</i>. (). ,Faculty of Health Sciences ,Department of Public Health and Family Medicine. Retrieved from http://hdl.handle.net/11427/30083 | en_ZA |
| dc.identifier.chicagocitation | Prinsloo, Megan. <i>"Estimating injury mortality in South Africa and identifying urban-rural differences."</i> ., ,Faculty of Health Sciences ,Department of Public Health and Family Medicine, 2019. http://hdl.handle.net/11427/30083 | en_ZA |
| dc.identifier.citation | Prinsloo, M. 2019. Estimating injury mortality in South Africa and identifying urban-rural differences. . ,Faculty of Health Sciences ,Department of Public Health and Family Medicine. http://hdl.handle.net/11427/30083 | en_ZA |
| dc.identifier.ris | TY - Thesis / Dissertation AU - Prinsloo, Megan AB - The overarching aim of this thesis is to utilise national data on injury mortality in South Africa, to conduct advanced statistical analyses to identify urban-rural differences for injury deaths, and to gain insight into the explanatory variables for homicide in metropolitan- and non-metropolitan (metro- and non-metro) areas. The literature review describes the global and national estimates of injury mortality and reports higher rural than urban injury mortality rates for high-income countries. It further discusses a framework for assessing data quality and reviews South Africa’s fatal and non-fatal injury data sources, issues of under-reporting and misclassification of deaths. The risk factors for violence are reviewed, which inform particular hypotheses on the role of age, sex, race, day of week and firearms with regard to homicide. The Injury Mortality Survey (IMS) data, which estimated 52 493 injury deaths nationally in 2009, is utilised for this PhD study. Data quality is assessed using an internationally developed conceptual framework for mortality data. Exploratory and multiple correspondence analysis identified possible associations between metro/nonmetro and other explanatory variables, prior to more sophisticated multinomial logistic regression analysis, which adjusted for age, sex, race and metro/non-metro for each manner of death (homicide, suicide, transport-related and other unintentional injury deaths) to explore particular hypotheses for the differences in the metro/non-metro injury mortality profile. Age-standardised injury mortality rates were calculated to take into account the effects of different age structures for metro- and non-metro populations. Generalized linear models were fitted in relation to particular hypotheses to determine the explanatory variables for homicide deaths in both metro and nonmetro areas. Main findings include a significantly higher likelihood for homicide in metro areas compared to non-metro areas, while transport-related deaths were significantly lower in metro areas. The risk of homicide for Coloureds was higher than Blacks in metro areas, while Blacks, Coloureds and Asians had similar risks of homicide in non- metro areas. Whites had a similar risk and Asians a higher risk of homicide in nonmetro areas compared with metro areas. Firearm use was shown to significantly explain metro/non-metro differences in homicide risks. This study’s most significant knowledge contribution includes the identification of metro/non-metro as a significant predictor of the injury mortality profile in South Africa. The association of metro/non-metro differences in the pattern of homicide for Blacks and Coloureds, also resolved conflicting statements found in the literature regarding race and homicide in South Africa. The results are of considerable significance to national and provincial policy makers. Recommendations are made in relation to the main findings of this study. DA - 2019 DB - OpenUCT DP - University of Cape Town LK - https://open.uct.ac.za PY - 2019 T1 - Estimating injury mortality in South Africa and identifying urban-rural differences TI - Estimating injury mortality in South Africa and identifying urban-rural differences UR - http://hdl.handle.net/11427/30083 ER - | en_ZA |
| dc.identifier.uri | http://hdl.handle.net/11427/30083 | |
| dc.identifier.vancouvercitation | Prinsloo M. Estimating injury mortality in South Africa and identifying urban-rural differences. []. ,Faculty of Health Sciences ,Department of Public Health and Family Medicine, 2019 [cited yyyy month dd]. Available from: http://hdl.handle.net/11427/30083 | en_ZA |
| dc.language.rfc3066 | eng | |
| dc.publisher.department | Department of Public Health and Family Medicine | |
| dc.publisher.faculty | Faculty of Health Sciences | |
| dc.title | Estimating injury mortality in South Africa and identifying urban-rural differences | |
| dc.type | Doctoral Thesis | |
| dc.type.qualificationlevel | Doctoral | |
| dc.type.qualificationname | PhD |