Résumé
Background: Clinical trials are increasingly perceived as a therapeutic opportunity for cancer patients. Favoring their concentration in few high-expertise academic centers maximizes quality of data collection but poses an issue of access equality. Analytical tools to quantify trial accessibility are needed to rationalize resources. Materials and methods: We constructed a distance-based accessibility index (dAI) using publicly available data on demographics, cancer incidence and trials. Multiple strategies were applied to mitigate or quantify clear sources of bias: reporting biases by text mining multiple registries; reliability of simple geographical distance by comparison with high-quality travel cost data for Italy; index inflation due to highly heterogeneous cancer incidence by log-transformation. We studied inequalities by Gini index and time trend significance by Mann–Kendall test. We simulated different resource allocation models in representative countries and identified locations where new studies would maximally improve the national index. Results: The dAI approximated well a more realistic but not widely applicable travel cost-based index. Accessibility was unevenly distributed across and within countries (Gini index ∼0.75), with maximal inequalities in high- and upper-middle-income countries (China, United States, Russian Federation). Over time, accessibility increased but less than the total number of trials, most evidently in upper-middle-income countries. Simulations in representative countries (Italy and Serbia) identified ideal locations able to maximally raise the national index. Conclusions: Access to clinical trials is highly uneven across and within countries and is not mitigated by simple increase in the number of trials; a rational algorithmic approach can be used to mitigate inequalities.
langue originale | Anglais |
---|---|
Numéro d'article | 100515 |
journal | ESMO Open |
Volume | 7 |
Numéro de publication | 3 |
Les DOIs | |
état | Publié - 1 juin 2022 |
Modification externe | Oui |