Assessing the risk of worsening conditions for diabetes mellitus patients in Italy: a non-parametric and semi-parametric approach to time-to-event data

Background: Diabetes is associated with the development of sequelae that further reduce patients' quality of life. A valuable instrument for assessing the severity of sequelae is the disability weight. The evaluation of the risks associated with the potential deterioration of sequelae has direct implications for healthcare. Objectives: The objective of this study is to estimate the risk of progression from no complications to any complications in both type 2 and type 1 diabetes and to examine the progression of risk over time. Methods: Survival analysis methods are applied. The Kaplan-Meier and the Cox proportional hazards models are employed to ascertain the probability of developing complications in the future, given that the subject is currently free of such complications. Patients were obtained from the longitudinal dataset of the Italian Association of Diabetologists, collected between 2005 and 2017. Results: This is the first study in Italy to estimate the risk of transition to complications in people with diabetes. The results indicate that older males with type 2 diabetes living in the central and northern regions of the country are associated with a higher risk profile. Since the patients excluded from the baseline have the same risk characteristics as the patients studied, the results can be generalized. Our findings provide evidence within a large clinical cohort and suggest potential applicability of the proposed approach in other settings beyond the Italian context. Conclusion: Quantifying risk in a way that is easily understood by policymakers and the general public is a valuable tool, as diabetes complications are a significant burden for individuals and society as a whole.