IADR Abstract Archives
Challenges for Analysis Of Head And Neck Cancer Mortality Data
Objectives: To analyze mortality data from head and neck cancer (HNC) in Chilean regions, presenting the challenges for it.
Methods: Secondary analysis of mortality records. Ecological study considering 15 regions as the unit of analysis. Crude and age-standardized mortality rates were calculated for every region and every year with available data. Quality of the data was assessed through cause of death codification analysis. Standardization was done by direct method with the Chilean population of 2014 as a standard (census projections). The annual change rate of the standardized rates was calculated to evaluate change over time. For every step of the analysis, challenges were recorded and discussed
Results: Four main challenges were detected in the process. First, quality of the data was assessed when calculating mortality rates by region and year. In three regions over 15% of data was considered garbage code according to the cause of death. Also, many regions had zero cases which can be corrected by grouping regional data by triennia. Second, for standardization a minimum of 20 cases by strata for stability is required, leaving four regions out. Third, when stratifying by sex, only 9 regions for males and 3 for women were comparable. Finally, joinpoint regression for trend analysis is not possible due to the number of years and minimum breaking points required in time.
Conclusions: Despite availability of mortality data for HCN, analysis need to be planned considering the unit of analysis. The differences in the regional contexts of the country, highlight the need of evaluation for this outcome in each of them. However, many challenges arise so strategies to face these issues are required.
Methods: Secondary analysis of mortality records. Ecological study considering 15 regions as the unit of analysis. Crude and age-standardized mortality rates were calculated for every region and every year with available data. Quality of the data was assessed through cause of death codification analysis. Standardization was done by direct method with the Chilean population of 2014 as a standard (census projections). The annual change rate of the standardized rates was calculated to evaluate change over time. For every step of the analysis, challenges were recorded and discussed
Results: Four main challenges were detected in the process. First, quality of the data was assessed when calculating mortality rates by region and year. In three regions over 15% of data was considered garbage code according to the cause of death. Also, many regions had zero cases which can be corrected by grouping regional data by triennia. Second, for standardization a minimum of 20 cases by strata for stability is required, leaving four regions out. Third, when stratifying by sex, only 9 regions for males and 3 for women were comparable. Finally, joinpoint regression for trend analysis is not possible due to the number of years and minimum breaking points required in time.
Conclusions: Despite availability of mortality data for HCN, analysis need to be planned considering the unit of analysis. The differences in the regional contexts of the country, highlight the need of evaluation for this outcome in each of them. However, many challenges arise so strategies to face these issues are required.