Cancer disparities: Projection, COVID-19, and scenario-based diagnosis delay impact

There has been limited research on how disparities in cancer mortality may evolve in the future, although relevant socio-economic and regional disparities in cancer risk are well-documented. We studied future trends in breast cancer (BC) and lung cancer (LC) mortality up to 2036 across affluent and deprived communities in nine regions of England, motivated by the distinct socio-economic patterns and burden of these cancer types. We used cancer death registrations from the Office for National Statistics on population and deaths in nine regions of England by underlying cause of death from 2001 to 2018, stratified by sex, 5-year age group, and income deprivation. We applied a gender- and cause-specific Bayesian hierarchical model to obtain robust estimates of cancer mortality by age group, gender, deprivation quintile, and region, up to 2036. In these models, we also used a data-driven proxy for age-at-diagnosis as an additional risk factor, and non-smoker prevalence rates as a proxy for smoking. We found that if pre-COVID conditions and trends remained the same, socio-economic disparities in LC would persist during our projection period. LC mortality rates for women in 2036 were found to be around 60% lower in the least deprived areas of London, as compared to the most deprived in the same region, with the disparities being even higher in northern regions and among men. Using data from the period 2011-2018, our model estimated 2% fewer LC deaths than those registered during the pandemic years (2020-2022) across England (and 4% fewer for men). Scenarios linked to delays in LC diagnosis led to stark differences in future excess mortality - significantly higher excesses in the northern regions compared to the southern regions, and in the most deprived areas compared to the least deprived areas. Additionally, our findings show that if pre-COVID conditions and trends remained unchanged, BC mortality would continue to decline up to 2036, with comparable rates in the regions of England. During the pandemic years, BC deaths were estimated to decline by 1% across England compared to the pre-pandemic trends (2001-2018). However, our analysis shows 10% to 13% increase in BC deaths for women aged 80+ in the same years. Cancer disparities are predicted to persist in the future unless targeted interventions are implemented. Our results underscore the notable impact of delays in cancer diagnosis on cancer mortality and related inequalities. Future research that models different causes of death while adjusting model outputs for competing risk factors might be beneficial. Further models with individual-level socio-economic risk factors would also be useful.