Socio-economic inequalities in cancer survival: how do they translate into Number of Life-Years Lost?

Patients living in more socioeconomically deprived areas have worse cancer outcomes than those in less deprived areas in the UK and elsewhere. Despite successive NHS policies since 2000 aimed at improving cancer survival through early diagnosis, optimized treatment, and better resources, socio-economic inequalities in cancer outcomes have persisted. Awareness, comorbidities, and tumor stage explain only part of these inequalities, with additional contributions from lower screening uptake and differences in patient management. Traditional measures (survival or mortality probabilities) do not fully capture the societal burden, motivating the use of crude probability of cancer death (CPr) and the Number of Life-Years Lost (NLYL). This study quantifies the population burden of socio-economic inequalities in cancer survival in England using CPr and NLYL, to identify targets for improvement and inform public health policy and resource allocation.

The paper summarizes extensive prior evidence showing persistent socio-economic disparities in cancer survival in England and internationally, despite national initiatives (e.g., NHS Cancer Plan, Cancer Reform Strategy, NAEDI). Previous research indicates that variations in cancer awareness, comorbidities, and tumor stage account for only part of the inequality; lower uptake of screening and differences in treatment pathways and patient management contribute importantly. Inequalities are also seen in screening participation and vaccine coverage. Prior work has largely quantified inequalities via survival or mortality probabilities; alternative, more communicable metrics like crude probability of death from cancer and NLYL can better reflect societal burden. The study builds on methodological literature in relative survival and competing risks to estimate CPr and NLYL without cause-of-death information.

Design and data source: Population-based cohort study using the National Cancer Registry of England. Inclusion criteria: all patients aged 15–99 years diagnosed with a primary, invasive, malignant neoplasm (ICD-O behavior code 3), across 23 common cancer sites, from 1 January 2010 to 31 December 2014; follow-up through 31 December 2015. Tumor sites were coded per ICD-10. Population coverage: approximately 1.2 million patients, representing 92.3% of all incident cancers in England during the period. Exposure: Socio-economic deprivation measured at area level using the income domain of the English Indices of Deprivation, grouped into five levels (1 least deprived to 5 most deprived) based on area of residence at diagnosis. Outcomes: Crude probability of death from cancer (CPr) and Number of Life-Years Lost (NLYL) due to cancer within 3 years since diagnosis. Analytic approach: Non-parametric estimation within a relative survival framework to separate death due to cancer from other causes without requiring cause-of-death information. Estimates were stratified by sex, age group (15–44, 45–54, 55–64, 65+), and deprivation level. The 3-year time horizon ensured all patients had adequate follow-up within the cohort. Results were presented overall and by cancer prognosis categories (good, moderate, poor) derived from CPr ranges. Confidence intervals are reported for selected comparisons. Life tables appropriate to the population were used to support relative survival estimation. The approach allows quantifying both the number and the proportion of life-years lost attributable to each cancer across deprivation strata.

- Cohort profile: Over 1.2 million patients with one of 23 cancers (92.3% of incident cancers) were included (2010–2014). Deprivation distribution: ~20–21% in levels 1–4; 17% in level 5 (most deprived). Cancer distribution varied by deprivation: colon, prostate, and female breast more common in less deprived; lung, cervical, stomach, liver, and oesophageal more common in more deprived; pancreatic similar across groups. - Prognosis categories within 3 years: Poor-prognosis cancers (brain, lung, pancreatic, liver, oesophagus, stomach) had CPr 0.75–1.0 and NLYL 1.75–2.3 years. Good-prognosis cancers (Hodgkin lymphoma, thyroid, melanoma, female breast, prostate, testis, cervix, uterus) had CPr <0.25 and NLYL <0.5 years. Intermediate cancers (colon, rectum, kidney, bladder, larynx [men], ovary, leukaemia, myeloma, NHL) had CPr 0.25–0.50 and NLYL 0.5–1.2 years. - Age gradient: NLYL within 3 years increased with age across sexes and cancers. - Socio-economic inequalities (selected estimates within 3 years): • Poor-prognosis cancers: inequalities largest in patients <45 years. Under 45, most deprived males with pancreatic cancer lost 1.81 years (95% CI: 1.56–2.07) vs 1.38 years (95% CI: 1.05–1.71) in least deprived (≈0.43-year gap). Under 45, most deprived females with lung cancer lost 1.49 years (95% CI: 1.34–1.63) vs 0.95 years (95% CI: 0.77–1.16) in least deprived (0.54-year gap). In older (65+) brain cancer patients (especially males), NLYL approached 2.5 years with minimal deprivation gap. • Moderate/good-prognosis cancers: for many sites (colon, rectum, kidney, female leukaemia, male myeloma, NHL, testis, female breast, ovary, uterus), deprivation gaps generally widened with age. For thyroid cancer, the deprivation difference peaked at 65+. Gaps narrowed with age for female bladder and male laryngeal cancers. • Notable inequalities: Hodgkin lymphoma ages 55–64 showed ≈0.4 additional years lost in most vs least deprived (both sexes). Female bladder cancer <45: most deprived lost 1.26 years (95% CI: 0.89–1.65) vs 0.63 (95% CI: 0.16–1.15) in least deprived (0.63-year gap). In males, large gaps were also seen for laryngeal cancer <45, and for thyroid and testicular cancers at 65+. • Small or reversed gaps: minimal deprivation differences for melanoma (both sexes), and relatively small for prostate, cervical, and thyroid (women). Reversal observed for ovarian cancer <45 and Hodgkin lymphoma in women 65+. - Aggregate burden: Most deprived patients lost 1.5 times more NLYL within 3 years than least deprived (0.98 vs 0.67 years; results referenced). Poor-prognosis cancers accounted for the largest share of total LYL across ages and deprivation. - Proportion of total LYL: Lung cancer was the dominant contributor, especially among the most deprived: contribution ranged from ~13% (young females) to >40% (65+ both sexes) of total LYL, despite lung comprising ~21% of incident cases in the most deprived. In the least deprived, lung’s contribution remained below ~30% of LYL (65+ males; lung ~10% of cases). For young females (15–44), breast (both deprivation groups) and cervical (most deprived) were the largest NLYL contributors. - Overall interpretation: More deprived patients, particularly younger adults with lethal cancers, systematically lose more life-years than less deprived patients.

The study demonstrates that socio-economic inequalities in England translate into substantial additional life-years lost due to cancer, especially among the most deprived and in younger adults with poor-prognosis, often smoking-related cancers (pancreatic, lung, oesophageal). For many moderate/good-prognosis cancers, deprivation gaps in NLYL widen with age, whereas for poor-prognosis cancers the extremely low survival at older ages limits observable differences. In good-prognosis cancers among younger patients, a ceiling effect may constrain further gains in the least deprived, yielding smaller gaps. Mechanisms likely include differences in receipt of potentially curative treatments (e.g., lung cancer surgery declines with age and deprivation, even after accounting for comorbidity), diagnostic pathways and promptness (e.g., gender-related diagnostic delays for bladder and colon cancers), and barriers in navigating complex care (e.g., laryngeal cancer). Lung cancer is the single largest driver of NLYL across deprivation groups, supporting targeted lung cancer screening in high-risk, deprived populations. Cervical cancer’s contribution among young deprived women underscores the need to improve screening and HPV vaccination uptake. From a health policy perspective, systematically incorporating an equity lens into early diagnosis and treatment initiatives, and aligning resource allocation with area deprivation, are essential to reduce these inequalities. The COVID-19 pandemic has likely exacerbated existing disparities, emphasizing the urgency of interventions that prevent further widening.

This study quantifies the societal burden of socio-economic inequalities in cancer outcomes using crude probability of death and life-years lost within 3 years of diagnosis across 23 cancers in England. It shows that more deprived patients, particularly younger adults with poor-prognosis cancers, systematically lose more life-years than the less deprived, and that for many moderate/good-prognosis cancers, inequalities widen with age. Policy implications include prioritizing equity in early detection and treatment strategies, targeted screening (notably for lung cancer) in deprived high-risk groups, and aligning resource allocation with deprivation to improve access and outcomes. Future work should further elucidate mechanisms driving inequalities within a universal health system and evaluate the impact of equity-focused interventions, including post-pandemic recovery measures.

The primary limitation noted is the time-bounded nature of the NLYL metric to 3 years post-diagnosis; while facilitating robust estimation without extrapolation, it does not capture longer-term life-years lost or the full social and economic costs beyond 3 years. The study uses area-level deprivation measures, and while relative survival methods obviate the need for cause-of-death data, residual confounding by unmeasured clinical factors (e.g., stage, treatment pathways) cannot be fully assessed within this framework as presented.