Background
Cancer outcomes among people who experience incarceration are poorly documented, in part because individual-level linkage between correctional systems and cancer registries is rarely available. Yet incarceration may act as a marker of life-course accumulation of preventable exposures and barriers to prevention and timely care, with cancers often occurring years after release. Responding to recent calls to strengthen evidence on cancer inequalities in this population, we leveraged a unique data platform in Geneva.
Objectives
To quantify cancer incidence and cancer mortality among adults with incarceration experience in Geneva compared with the general population, and to identify cancer sites contributing most to excess burden to inform prevention and care strategies.
Methods
We built a population-based retrospective cohort of adults (≥18 years) with at least one incarceration episode in Geneva (1990-2022). We performed individual-level deterministic record linkage within the Geneva Cancer Registry secure environment using personal identifiers (sex, full name, date of birth, and a national personal identification number when available). Invasive cancers, excluding non-melanoma skin cancer, were considered. We calculated standardized incidence ratios (SIRs) and standardized mortality ratios (SMRs) by comparing observed numbers of cancers/deaths with expected numbers derived from sex-, age-, and calendar-period–specific Geneva population rates. Incidence follow-up ended on 31-12-2022 (114,909 person-years) and mortality follow-up on 31-12-2023 (119,960 person-years). Confidence intervals assumed a Poisson distribution.
Results
The linkage identified 8,922 individuals with incarceration experience (8,069 men; 853 women) eligible for follow-up. We observed 648 malignant cancers versus 471.8 expected (SIR 1.37, 95% CI 1.27–1.48), higher in women (SIR 1.70, 95% CI 1.36-2.09) than men (SIR 1.34, 95% CI 1.23-1.45). In the total cohort (sexes combined), excess incidence clustered in head and neck cancers (SIR 3.08, 95% CI 2.42-3.88), larynx (SIR 3.86, 95% CI 2.42-5.85), lung (SIR 2.83, 95% CI 2.38-3.34), and liver (SIR 2.75, 95% CI 2.00-3.68). Conversely, in the total cohort, melanoma and prostate cancer were diagnosed less often than expected (SIR 0.30, 95% CI 0.15–0.54 and 0.69, 95% CI 0.54–0.87 respectively). We observed 251 cancer deaths versus 130.6 expected (SMR 1.92, 95% CI 1.69–2.18), higher in women (SMR 2.06, 95% CI 1.32–3.05) than men (SMR 1.91, 95% CI 1.67–2.17). In the total cohort (sexes combined), mortality excess was particularly marked for larynx (SMR 4.35, 95% CI 1.41-10.15), head and neck cancers (SMR 4.11, 95% CI 2.81-5.80), lung (SMR 2.46, 95% CI 1.95-3.07), and liver (SMR 2.46, 95% CI 1.59–3.64). 
Conclusions/Implications
Using a unique, deterministic, individual-level linkage between prison records, population office and population-based cancer registry data, we show substantially elevated cancer incidence and nearly doubled cancer mortality among people with incarceration experience. Taken together, the site-specific pattern supports detention and re-entry as potential points of contact to deliver primary prevention measures, targeting modifiable exposures, and access to primary care and timely diagnostic pathways. Routine linkage provides an actionable surveillance platform to guide and evaluate equity-oriented cancer prevention and control policies.