Long-term survival for myeloid neoplasms and national health expenditure: A EUROCARE-6 retrospective, population-based study

Background: EUROCARE, a European population-based cancer project, has shown geographical survival disparities for myeloid neoplasms (MN) suggesting healthcare inequalities. Total National Health Expenditure (TNHE) might influence population-based MN survival. We studied the long-term survival of 15 MN and the TNHE impact on prognosis between European countries. Methods: In this observational, retrospective, population-based study, we analysed the EUROCARE-6 dataset (comprising data from 27 European countries) of patients aged 15 or older diagnosed with a MN between 2001 and 2013, and followed up to December 31, 2014. Countries were classified according to their mean TNHE quartile in 2001–13. For each MN, 5-year and 10-year age-standardised relative survival (ASRS) was calculated using the period approach. Generalised linear models indicated the effects of age at diagnosis, sex, and TNHE on the relative excess risk of death (RER). Findings: We identified 267 968 MN cases [median follow-up time: 13 years (IQR 13–14)]. The 10-year ASRS was highest for myeloproliferative neoplasms (MPN) (61.2 % [95 % CI 60.3–62.0 %]), in particular chronic myeloid leukaemia (52.5 % [95 % CI 51.1–53.9 %]), polycythemia vera (70.9 % [95 % CI 69.0–72.8 %]), essential thrombocythemia (75.2 % [95 % CI 73.3–77.1 %]), and lowest for acute myeloid leukaemia (AML) (15.6 % [95 % CI 15.2–16.1 %]), except for acute promyelocytic leukaemia (52.2 % [95 % CI 46.3–58.7 %]). For AML, MPN, myelodysplastic syndrome, 5-year and 10-year ASRS was higher in countries in the highest TNHE quartile (high expenditure) than countries in the lowest quartile (low expenditure). Ten-year RERs for almost all MN increased with decreasing health expenditure, albeit not uniformly. Interpretation: TNHE is associated with geographical inequalities in MN prognosis. Policy decisions on allocating economic resources are needed to reduce these differences.