Simulation based cost-benefit analysis of a telemedical system for closed-loop insulin pump therapy of diabetes

INCA (Intelligent Control Assistant for Diabetes) is a project funded by the EU with the objective to improve diabetes therapy by creating a personal closed loop system interacting with telemedical remote control. Cost-benefit analyses of such systems are needed to decide on the introduction of telemedical systems such as the INCA system to routine care.

To identify and apply suitable methods for a cost-benefit analysis from the perspective of the payor for health services (i.e. a health insurance company).

For the cost analysis MOSAIK-M was used, a method and tool that supports health information systems analysis and design. Two MOSAIK-M models were created during the INCA project. Both, the "As is"-model of conventional insulin pump based diabetes care, and the "To be"-model of the INCA system were parameterised with cost values. With both models a period of one year was simulated to determine the yearly costs of diabetes management and treatment for a patient who does not suffer from diabetes related complications yet. The HbA1c-value was chosen as effectiveness parameter for diabetes therapy. To determine the probability of developing complications and their probable duration the Archimedes-Model was used. It was parameterised with selected HbA1c-values anticipating the effect of INCA. The simulation results in form of years of disease within a 30-years time frame were multiplied with corresponding treatment costs from the KoDiM study.

The yearly costs of conventional insulin pump treatment for a 19 year old diabetes type 1 patient with no complications are 5,907 euro (German health care system). Using the INCA system would raise the yearly costs by 7,348 euro. Almost all (98.53%) of the additional costs are generated by the continuous blood glucose measurement device. HbA1c-decreases from 7% (conventional treatment) to 6.5%, 6%, and 5.8 % would produce yearly savings (benefit) concerning the treatment of complications of 100.50 euro, 189.20 euro and 221.82 euro.

The selected approach produces an estimation of a lower bound for cost savings. Further work is needed to improve the approximation and to include indirect and intangible costs. The INCA approach would be cost efficient from the chosen perspective, only if the costs of system operation were notably lowered.