Survival for people diagnosed with lung cancer in Aotearoa New Zealand diverged by ethnicity. The study team reported lower observed 5-year survival for Māori than for Europeans: males 11.3% (95% CI 9.4–13.7) versus 16.8% (95% CI 15.2–18.7), and females 17.1% (95% CI 15.0–19.5) versus 24.2% (95% CI 22.5–26.2). The investigators then applied scenario-based standardisation to examine whether those gaps were driven more by differences in stage distribution at diagnosis or by differences in stage-specific survival after diagnosis. Their goal was simple: estimate how Māori survival would change if Māori had the same stage distribution as Europeans, the same stage-specific survival as Europeans, or both.
The authors used national cancer registry data to identify primary lung cancers diagnosed between 2015 and 2024, with mortality follow-up through December 2025. Survival modelling was performed with sex-specific Cox regression. Covariates in those models included age at diagnosis, diagnosis year, ethnicity, Surveillance, Epidemiology, and End Results (SEER) summary stage, and an ethnicity-by-stage interaction term. Survival probabilities were estimated at 1, 3 and 5 years; the abstract reports the 5-year probabilities and their 95% confidence intervals. The source abstract did not report sample sizes, detailed stage frequencies, or whether survival estimates represented overall or net/cause-specific survival.
The scenario analyses produced distinct shifts in estimated survival. When the authors standardised stage distribution for Māori to match that of Europeans, estimated 5-year survival rose modestly: to 13.5% (95% CI 11.4–16.2) for Māori males and to 20.2% (95% CI 17.9–22.9) for Māori females. The study team observed a larger effect from standardising stage-specific survival alone — estimated 5-year survival increased to 17.2% (95% CI 15.4–19.3) for Māori males and 23.4% (95% CI 21.5–25.6) for Māori females. Applying both standardisations produced the highest 5-year estimates reported: 19.5% (95% CI 17.7–21.7) for males and 26.6% (95% CI 24.7–28.8) for females.
These numerical shifts show that both earlier stage at diagnosis and better survival conditional on stage would be required to close the gap fully. The investigators reported that standardising stage-specific survival produced the larger single change in estimated survival, meaning differences in outcomes after diagnosis — not only differences in stage at presentation — contributed substantially to the observed inequity.
The abstract provides the following 5-year survival comparisons and scenario results, with 95% confidence intervals: observed Māori versus European survival (males 11.3% vs 16.8%; females 17.1% vs 24.2%), stage-distribution-standardised Māori (males 13.5%; females 20.2%), stage-specific-survival-standardised Māori (males 17.2%; females 23.4%), and both standardised (males 19.5%; females 26.6%). The study team supplied CIs for each estimate, and these intervals indicate non-trivial uncertainty around the point estimates. The abstract did not give detailed stage-by-stage survival rates, nor did it present subgroup counts or p-values.
Data indicated that addressing only one axis — for example, earlier diagnosis through screening or improved access to staging investigations — would move the survival needle, but not eliminate the disparity. The larger increase after standardising stage-specific survival suggests that differences in treatment access, receipt of curative therapies, timeliness of care, comorbidity management, or other post-diagnostic factors could be major contributors. The abstract does not specify which health-system or clinical factors underlie the stage-specific survival gap; the authors reported the association but did not document causal mechanisms in the abstract.
The abstract lists core methods and the scenario results but leaves several methodological details unreported. The source did not provide sample sizes, the distribution of SEER stages by ethnicity, whether survival was overall or cause-specific, or which additional covariates (beyond age, year, ethnicity, SEER stage and an ethnicity-by-stage interaction) were considered. The abstract also does not describe sensitivity analyses, model diagnostics, or how missing stage data — if present — were handled. Those details were not reported in the abstract and would need to be consulted in the full article to assess internal validity and the robustness of the scenario estimates.
The authors concluded that differences in both stage distribution and stage-specific survival were associated with inequities in lung cancer survival for Māori, with stage-specific survival producing the larger single change in estimated outcomes. The abstract ends with that conclusion; the full text would be required to evaluate causal explanations, potential interventions, and the statistical assumptions underpinning the standardisation scenarios.