Patient-Level Modeling Approach Using Discrete-Event Simulation: A Cost-Effectiveness Study of Current Treatment...

Decision tree and Markov models have been the most commonly used modeling methods in health economic evaluations. Both methods are known for their limitations. Discrete-event simulation (DES), an event-driven model in continuous time at the patient level, is a relatively new method in health economic evaluations that addresses some limitations of the common modeling techniques. Specifically, with the advent of personalized medicine, conventional methods for value assessment that are based on population-level measures might not be appropriate. The best treatment would depend on patient characteristics and clinical profiles. Value assessment of health interventions can vary substantially and may lead to different health decision making due to patient heterogeneity. As such, modeling at the patient level is an appropriate approach for value assessment of health interventions. The DES model has several advantages, such as flexibility, ability to reflect patient heterogeneity, increased precision, and better characterization of modeling uncertainty, that may be preferred to decision tree and Markov models. In addition, with increasing health care spending and drug prices, it is important to quantify value of available treatment options for women with postmenopausal osteoporosis (PMO). The purpose of this Viewpoints article was to describe and demonstrate an application of a DES model to evaluate the cost-effectiveness of the current treatment guidelines for women with PMO. In particular, the DES model indicated that the optimal treatment at the common willingness-to-pay thresholds between $100,000 per quality-adjusted life-year (QALY) and $150,000 per QALY was denosumab. Analysis of patient heterogeneity in terms of low, medium, high, and very high risk of fractures resulted in a similar conclusion. DISCLOSURES: Funding for this study was received through the PhRMA Foundation Value Assessment Challenge Award. The author has no conflicts of interest to declare.