Actuarial Modelling for Patients with Acute Lymphoblastic Leukaemia (ALL) based on CNS1 Status

Abstract

Introduction: Worldwide, Acute Lymphoblastic Leukaemia (ALL) is the most frequent cancer in children. One of the major clinical challenges is adequate diagnosis and treatment of Central Nervous System (CNS) involvement in this disease. CNS relapse has been a barrier to the successful treatment of ALL for many years. Recent studies have shown encouraging results in the survival of these patients for a long time. However, their long-term survival depends upon the cost of therapy toxicity and financial distress. The primitive aim of the paper is to propose a yearly insurance plan to assist these patients financially during the diagnosis period. Method: Based on the CNS status 110 patients are categorised to estimate their long-term survival. Survival times of CNS1 status and for all the patients cumulatively are estimated by Kaplan-Meier and Cox-PH model in presence of the prognostic factors. The survival estimates are used to estimate the premium cost. The premium cost is estimated using a deterministic model which is advantageous for the patient and serviceable for the insurance provider. Result: Both the methods Kaplan-Meier and Cox-PH gave higher survival estimates for ALL patients cumulatively as compared to CNS1. Survival estimate from Cox-PH is 0.998 and 0.997 of first year of follow-up for patients taken cumulatively and in CNS1 respectively. For the fifth year the survival estimates are 0.802 and 0.783 respectively. The estimated premium cost for a 100 rupees of sum insured is rupees 4.7 for the first year and rupees 26.69 for the fifth year for patients taken cumulatively. Same for CNS1, it is rupees 6.24 and 29.42. Conclusion: Cox-PH model for estimating the survival is recommended since it includes the prognostic factors. The insurance plan suggests to opt for the premium as early as possibly since it costs less and increases later.