A Comparison of Mortality-to-Incidence Ratio with Survival Analyses in Assessing Racial Breast Cancer Disparities Across South Carolina Counties

Abstract

The mortality-to-incidence rate ratio (MIR) provides a population-based measure of survival which accounts for incidence. The use of MIR as a surveillance tool has shown that South Carolina (SC) exhibits more extreme racial differences in cancer incidence, mortality and MIR than other states or the nation. We assessed the effectiveness of MIR as a proxy for 5-year survival time (5YST) among breast cancer (BrCa) patients in South Carolina. Methods: The 5YST was computed from data on BrCA cases which were obtained retrospectively from the SC Central Cancer Registry from 2002 to 2010. The MIR was computed from Cancer incidence and mortality data which were obtained from the SC Community Access Network (SCAN). The underlying data for SCAN were generated from the SC Central Cancer Registry and SC DHEC Vital Records and used to construct MIRs. ArcGIS 10.2 was utilized to map BrCA MIRs by race for 46 counties within SC. Seven categories of MIR were derived using the national MIR for BrCA as reference. 5YST was computed for all BrCA cases in each county utilizing SAS software and this was mapped with MIR per county. Exploratory and geographically weighted regression analyses were conducted in ArcGIS to determine the relationship between MIR and MST. Results: A total of 2155 breast cancer patients (nWhites = 1557/72%; nBlacks = 598/28%) were reported in the study period. A visual inspection of the MIR maps by race showed that Blacks were in the highest MIR category while the MIR by 5YST map showed that higher MIR was likely associated with lower 5YST. By contrast, the MIRs for Whites were more evenly represented over the seven categories. Overall, the 5YST was 92.8% among blacks and 95.6% among whites. Assessment of MIR with MST in ArcGIS utilizing exploratory ArcGIS regression showed that there was statistically significant Global Moran's I p value indicative of clustering. Conclusions: The MIR proved useful for identifying disparities in BrCA incidence and mortality among Black and White women in SC. Cancer surveillance programs may use the MIR to monitor disparities across racial/ethnic groups and geographic regions going forward. MIRs have the potential to serve as an indicator of the long-term success of cancer surveillance programs.

Improving Quality Indicator of Melanoma Management – Change of Melanoma Mortality-to-Incidence Rate Ratio Based on a Hungarian Nationwide Retrospective Study

“…Mortalityto-incidence ratios (MIR) have been used for more than 40 years, first as a measure of completeness of cancer registries (13,14). Where quality of data is good, MIR can be a valid proxy indicator for survival estimates (calculated as 1-MIR) (14)(15)(16), and has already been used in multiple cancer types (17)(18)(19)(20)(21), including melanoma (22)(23)(24). Despite this, the use of MIR for survival estimation is mathematically questionable as the calculation does not use the same population for incidence and mortality (13).…”

“…MIR is a good tool for international comparisons, as it is easily calculable with access to good quality data and using standardized rates which allows cross-country comparison (25). Studies on various types of cancer in OECD countries also discussed the potential of MIR as a measure for evaluation the success of cancer surveillance and screening (16,17,26). Studies generally found reverse correlation between healthrelated expenditure and MIR or development and MIR in multiple cancer types (18,20,23).…”