Written November 15, 2022 for MHST 602: Transforming Healthcare Through Informatics
Despite coordinated government efforts to address gaps in access to health care between urban and rural populations, inequities still exist in today’s health care system. Health service access is a social determinant of health and unfortunately, rural residents continue to face greater challenges with access to care than their urban counterparts, leading to poorer health outcomes (Wilson et al., 2020; World Health Organization, 2022). For cancer patients, this also means less chance of long-term survival (Dobson et al., 2020). In this paper, data from the Cancer Care Ontario Cancer Survival Profile database will be examined to analyze five-year relative survival ratios (RSR) between urban, urban-rural and rural public health units (PHUs).
Definition and Scope of the Problem
In Canada, 28.2% of all deaths are cancer-related (Canadian Cancer Society, 2022). In Ontario, early detection and more effective treatment regimens improved the five-year RSR for all cancers to 64% between 2015-17, up from 55% in the early 1990s (Canadian Cancer Society, 2022; Cancer Care Ontario, 2016). When determining the effectiveness of cancer treatment, survival statistics for the first five years post-diagnosis are of critical importance; patients are most likely to access healthcare services during this period (Cancer Care Ontario, 2020). This paper examines the disparity in cancer survival rates between urban and rural PHUs, where access to care differs. Our question therefore is what is the size of any continued disparity in five-year cancer RSRs between rural or urban PHUs in Ontario?
Dataset and Source
Databases compiling clinical and administrative health data are invaluable tools in guiding research and health system quality improvement (Verma, et al., 2021). Cancer Care Ontario is Ontario’s primary cancer advisor, providing health care stakeholders with the data, statistics and tools to prevent cancer, identify new treatment options and evolve high-quality patient care. Their Cancer Profiles are sourced from Cancer Registry records, drawing from provincial pathology reports, data from hospitals and Ontario’s 14 regional cancer centers, discharge and admission data, and cause-of-death data (Cancer Care Ontario, n.d; Ontario Health, 2021). Data is excluded if the patient’s age is unknown, the person was diagnosed based on autopsy only, the place of residence was unknown or the date of diagnosis and death was the same. The raw data used to complete the analysis is found in Appendix A.
A percentage is used to express the five-year RSR where five years after diagnosis, the survival of those with cancer is divided by the general population’s expected survival during the same period; in other words, as the RSR approaches 100%, the closer the cancer survival rate is to those without cancer (Ontario Health, 2021). RSRs are standardized rates published by the International Cancer Survival Standards (see Appendix B) and are based on the first primary cancer per person (Ontario Health, 2021). Expected survival proportion for the general population to compare this number “were derived using the Ederer II approach from Ontario life tables produced by Statistics Canada” (Ontario Health, 2021, p. 8).
Data Structure
The cancer survival dataset was retrieved online from the Cancer Care Ontario website as a downloadable Excel spreadsheet. The dataset includes survival statistics for both LHIN areas and PHUs, including the five-year RSR, standard error, lower 95% and upper 95% confidence interval. The data can be filtered by primary cancer site (nominal data), sex, age range, and region for a more fulsome analysis. In this paper, the five-year RSR was considered for all cancer sites, ages and genders across PHUs between 2014 and 2018. During analysis, the PHUs were divided into groups representing urban, urban-rural and rural areas; this was achieved using the rurality ranking discussed in Ahmed’s 2019 position paper on how to measure Ontario’s rural-urban divide (see Appendix C).
Results
The overall five-year RSR was calculated for the ten urban PHUs, 14 urban-rural PHUs and 11 rural PHUs, and measures of central tendency were used to compare the variability in survival rates (see Appendix D). Urban PHUs saw a five-year RSR of 67.8%; cancer patients in urban areas are 67.8% as likely to live five or more years past their diagnosis compared to people without cancer. Despite efforts by government agencies to address gaps in access to care in rural areas, the five-year RSR across the rural PHUs averaged out to 63.5%. The rural-urban PHUs saw the lowest comparative survival rate at 62.5%, a full 5% lower than the urban areas.
The table in Appendix E shows the normal distribution by primary population density of each PHU region and interprets the standard error and confidence interval measures. There is less variation within the urban-rural regions, although the mean RSR was lower. One explanation for this may be that in urban-rural regions, access to care is a greater problem. It may be assumed that those in rural regions of urban-rural PHUs have equitable access to care with urban location, but the assumption is incorrect. Further analysis would be required to confirm or reject this hypothesis. There is also greater average variability in five-year RSRs within rural and mixed PHU regions with a mean SE or 1.34 and 1.1 respectively, compared to urban regions with a mean 0.58 standard error.
Evaluation of the Data’s Usefulness
The five-year RSR in the database is combined with the standard error, an interval measure of multiple sampling error; this determines how representative the data is of the population, calculated by the standard deviation / square root of the total number of samples. The dataset used is consistent with those published by the Public Health Agency of Canada and follows all six dimensions of quality data outlined by Statistics Canada (2021). The data is highly relevant to the posited question, and the results would be useful for health system future planning. The source notes are thorough; the data is accurate and most up-to-date. CCO is a highly reputable organization in Ontario; their published data is peer reviewed and checked for accuracy and relevance. It would be more useful if there were multiple year ranges, as trending information cannot be inferred. The interpretability of the dataset is improved for consumers who understand standard error and confidence interval measures; they can accept or reject the raw data by reviewing the range of accuracy of the complex calculations.
Conclusion
The Cancer Care Ontario Survival Profile database has been shown to be useful in revealing a disparity between five-year RSRs in urban, mixed or rural PHUs in Ontario between 2014 and 2018. The results have demonstrated the five-year relative survival ratios to be higher for urban as opposed to rural areas, thereby further emphasizing the association between poorer cancer survival and rural residence. Although greater attention towards addressing these inequities in access and survival has already been a focus for Ontario health care stakeholders, it should remain a priority in the future.
References
Ahmed, W. (2019, November 13). Measuring Ontario’s urban-rural divide. Ontario 360. https://on360.ca/policy-papers/measuring-ontarios-urban-rural-divide/
Caldwell, W., Kraehling, P., Kaptur, S., & Huff, J. (2015). Healthy rural communities tool kit-A guide for rural municipalities. https://www.ruralhealthycommunities.ca/toolkit
Canadian Cancer Society. (2022). Cancer statistics at a glance. https://cancer.ca/en/research/cancer-statistics/cancer-statistics-at-a-glance
Cancer Care Ontario. (n. d.). Ontario cancer registry. https://www.cancercareontario.ca/en/cancer-care-ontario/programs/data-research/ontario-cancer-registry
Cancer Care Ontario. (2016). Survival improving for the 4 most common cancers in Ontario. https://www.cancercareontario.ca/en/cancer-facts/survival-improving-common-cancers
Cancer Care Ontario. (2020). Ontario cancer statistics 2020. https://www.cancercareontario.ca/en/statistical-reports/ontario-cancer-statistics-2020
Cancer Care Ontario. (2021). Ontario cancer profiles: Cancer survival (2014-2018 by PHU) [Data set]. https://profiles.cancercare.on.ca/Survival/atlas.html
Cooke, C. R., & Iwashyna, T. J. (2013). Using existing data to address important clinical questions in critical care. Critical Care Medicine, 41(3), 886–896. https://doi.org/10.1097/ccm.0b013e31827bfc3c
Dobson, C., Rubin, G., Murchie, P., Macdonald, S., & Sharp, L. (2020). Reconceptualizing rural cancer inequalities. International Journal of Environmental Research and Public Health, 17(4), 1455. https://doi.org/10.3390/ijerph17041455
Ontario Health. (2021). Ontario cancer profiles: Data sources and notes (Updated September 2021) [Source Notes]. https://www.cancercareontario.ca/sites/ccocancercare/files/assets/OCPSourceNotes_0_0.pdf
Statistics Canada. (2021). Data quality in six dimensions. [Video]. Retrieved November 9, 2022, from https://www.statcan.gc.ca/en/wtc/data-literacy/catalogue/892000062020001
Verma, A. A., Pasricha, S. V., Jung, H. Y., Kushnir, V., Mak, D. Y. F., Koppula, R., Guo, Y., Kwan, J. L., Lapointe-Shaw, L., Rawal, S., Tang, T., Weinerman, A., & Razak, F. (2021). Assessing the quality of clinical and administrative data extracted from hospitals: the The General Medicine Inpatient Initiative (GEMINI) experience. Journal of the American Medical Informatics Association: JAMIA, 28(3), 578–587. https://doi.org/10.1093/jamia/ocaa225
Wilson, C. R., Rourke, J., Oandasan, I. F., & Bosco, C. (2020). Progress made on access to rural health care in Canada. Canadian Family Physician, 66(1), 31–36. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7012120/
World Health Organization. (2022). Social determinants of health. https://www.who.int/health-topics/social-determinants-of-health#tab=tab_1
Comments