Global cancer mortality: The effect of health care system resources on survival from cancer.

Abstract

1576 Background: Per capita Gross Domestic Product (GDP) has been shown to predict improved cancer survival among European countries. Experts have proposed increased spending on radiation facilities as a cost-effective tool for battling cancer globally. We aimed to investigate the relationship between cancer survival and quantifiable measures of national healthcare systems (World Health Organization (WHO) healthcare system ranking, GDP, total healthcare expenditure (THE), physician density (PD), access to radiation oncology (measured by per capita total external beam devices (TEBD)), THE per GDP and TEBD per THE). Methods: The most recent cancer incidence and mortality data (years 2005-2008) for 27 different cancers were obtained from the Globocan cancer registry for the 85 countries with reliable non-estimated data. Mortality per incidence for a given year was used as an inverse estimate of survival. Countries were split in half according to high (GDP>$15,000) or low-income (GDP<$15,000), and multivariate linear regression was run to compare the aforementioned markers with survival. The model controlled for rates of HIV, smoking, obesity, rural residence and ethanol consumption. Results: In low income countries, only WHO ranking was a significant correlate of survival, with a 21 point improvement in rank corresponding to a 1% increase in overall cancer survival. In high-income countries, GDP, TEBD, THE and THE per GDP were all significant correlates of survival. THE and THE per GDP were the strongest correlates, with a $400 increase in THE or a 0.8% increase in THE per GDP each corresponding to a 1% increase in overall cancer survival. WHO ranking, PD and TEBD per THE correlated poorly with survival in high income countries. Conclusions: These results highlight different approaches to predicting cancer survival in low vs. high income countries. In the former, only WHO ranking correlated significantly, surpassing all specific measures of healthcare resources. In the latter, healthcare spending was the most significant predictor of survival in high income countries. For high income countries, though TEBD (but not TEBD per THE) was correlated to cancer mortality, it was less significant than overall healthcare spending.