The economic burden of multiple myeloma. Definition of a model for forecasting patients’ costs

Abstract

Background: The aim of this study was to evaluate healthcare costs in a single-centre population of patients with multiple myeloma (MM), in an attempt to develop a model for forecasting costs. Methods: A cohort of 387 MM patients, diagnosed at Policlinico San Matteo (Pavia, Italy), between 2002 and 2014, was analysed grouping patients into those eligible (n=223) or not eligible (n=164) for transplantation. After descriptive statistics, the benchmark model-Ordinary Least Squares-and different variations of the Generalized Linear Model were adopted. Results: The average total cost per patient was around €28,500 for patients not eligible for transplantation and around €87,000 for the eligible ones. The difference in marginal costs for transplant-eligible patients was probably due to higher costs for hospitalisation and the costs of the transplant procedure itself. The analysis highlighted four determinants useful for building a model to forecast expenditure: age, bortezomib use, lenalidomide use, and number of lines of therapies. The two most important determinants of expenditure were use of the novel agents and the total number of lines of therapy, which reflects a higher number of doses and a greater need for accesses to hospital. Conclusion: In conclusion, using a Generalized Linear Model, we identified four determinants in our cohort which were useful for building a model to predict expenditure for MM patients. Although the analysis was performed in a particular setting in a single hospital, the model could be applied to any scenario of patients..