National health systems characteristics and lymphoma outcomes: A global analysis

Abstract Background Lymphomas are a heterogeneous group of neoplasms, many of which are potentially curable. While prognosis has markedly improved in recent years due to advancements in treatment, outcomes vary widely across the world. This study aimed to investigate the influence of health system metrics on global outcomes among people diagnosed with non-Hodgkin lymphoma (NHL) and Hodgkin lymphoma (HL). Methods The following publicly available health system metrics were obtained from World Health Organization (WHO), the World Bank, the DIrectory of RAdiotherapy Centres (DIRAC), the United Nations Development Programme (UNDP), and the International Agency for Research on Cancer (IARC): Health spending as a percentage of Gross Domestic Product (GDP 2021), out of pocket expenditure as percentage of current health expenditure (OOP), physicians per 1000 people (2016-2021 estimates), nurses and midwives per 1000 people (2016-2021 estimates), radiotherapy centers per 1000 people (RT), availability of pathology services (binarily defined), availability of bone marrow transplant (BMT) services (binarily defined), universal healthcare coverage index (UHC index, 2021), gender-inequality index (GII 2022), GDP per capita, Human Development Index (HDI 2024). Age-standardized incidence and mortality rates from 185 countries were retrieved from GLOBOCAN 2022, and mortality-to-incidence ratios (MIR) were manually calculated.Univariable linear regressions evaluated the association between each health system metric and NHL and HL MIR. Metrics with Bonferroni-adjusted p0.0045 were included in multivariable models (MVAs), which were iteratively refined by excluding covariates with variance inflation factor (VIF) 10 (an indicator of high multicollinearity) until all covariates had VIF 10. For the final MVA, α = 0.10 defined statistical significance to optimize inclusivity. Complete-case analysis was performed, and subgroup analyses by sex were conducted. Results In the final MVAs, 114 countries were included for NHL versus 109 countries for HL. In the final MVAs for subgroup analysis by sex, 114 countries were included for NHL for males and females while 108 countries were included for males versus 106 countries for females. HDI and UHC were excluded from all final MVAs due to high VIF. For NHL (all ages, both sexes), the final model had an R2 of 0.64. Greater physicians per 1000 people (Coef: -0.02, p=0.02), RT (Coef: -32.6, p=0.005), and GDP per capita (Coef: -1.4x10-6, p=0.06) were significantly associated with lower (improved) MIR. Greater GII (Coef: 0.3, p=0.03) was associated with a higher (worse) MIR. When analyzing NHL by sex, the final model had an R2 of 0.63 for males and an R2 of 0.65 for females. For males and females, greater physicians per 1000 people (Males, Coef: -0.019, p=0.04; Females, Coef: -0.03, p=0.007) and RT (Males, Coef: -32.19, p = 0.004; Females, Coef: -31.5, p=0.008), and were associated with improved MIR. For males and females, greater GII (Males, Coef: 0.2, p=0.03; Females, Coef: 0.3, p=0.02) was associated with worse MIR. For males only, greater GDP per capita (Coef: -1.62x10-6, p=0.02) was associated with improved MIR. For HL (all ages, both sexes), the final model had an R2 of 0.58. Greater RT (Coef: -31.8, p=0.01), and BMT (Coef: -0.06, p=0.04), were associated with improved MIR. Greater GII (Coef: 0.29, p=0.03) was associated with worse MIR. When analyzing HL by sex, the final model had an R2 of 0.55 for males and an R2 of 0.60 for females. For males and females, greater RT (Males, Coef: -30.4, p=0.01; Females, Coef: -25.3, p=0.04 was associated with improved MIR while greater GII (Males, Coef: 0.2, p=0.07; Females, Coef: 0.3, p=0.01) was associated with worse MIR. For males only, greater BMT (Coef: -0.06, p=0.04) was associated with improved MIR. For females only, greater OOP (Coef: -0.002, p=0.01) was associated with improved MIR. Conclusions This comprehensive global analysis of health system metrics suggests that increasing access to radiotherapy and improving gender equity are important for improving both NHL and HL outcomes. Strengthening physician workforce and increasing GDP were important for NHL while increasing access to BMT was important for HL. This work highlights important levers to strengthen cancer health systems around the world to improve lymphoma outcomes.