Department of Epidemiology, University of Groningen, University Medical Center of Groningen, Groningen, The Netherlands.
Value Health. 2013 Jun;16(4):529-35. doi: 10.1016/j.jval.2013.02.008. Epub 2013 May 6.
Early estimates of the commercial headroom available to a new medical device can assist producers of health technology in making appropriate product investment decisions. The purpose of this study was to illustrate how this quantity can be captured probabilistically by combining probability elicitation with early health economic modeling. The technology considered was a novel point-of-care testing device in heart failure disease management.
First, we developed a continuous-time Markov model to represent the patients' disease progression under the current care setting. Next, we identified the model parameters that are likely to change after the introduction of the new device and interviewed three cardiologists to capture the probability distributions of these parameters. Finally, we obtained the probability distribution of the commercial headroom available per measurement by propagating the uncertainty in the model inputs to uncertainty in modeled outcomes.
For a willingness-to-pay value of €10,000 per life-year, the median headroom available per measurement was €1.64 (interquartile range €0.05-€3.16) when the measurement frequency was assumed to be daily. In the subsequently conducted sensitivity analysis, this median value increased to a maximum of €57.70 for different combinations of the willingness-to-pay threshold and the measurement frequency.
Probability elicitation can successfully be combined with early health economic modeling to obtain the probability distribution of the headroom available to a new medical technology. Subsequently feeding this distribution into a product investment evaluation method enables stakeholders to make more informed decisions regarding to which markets a currently available product prototype should be targeted.
新医疗器械的商业空间早期估计可以帮助医疗技术生产者做出适当的产品投资决策。本研究的目的是说明如何通过将概率推断与早期健康经济建模相结合来以概率方式捕获该数量。所考虑的技术是心力衰竭疾病管理中的新型即时检测设备。
首先，我们开发了一个连续时间马尔可夫模型来表示当前护理环境下患者的疾病进展。接下来，我们确定了在引入新设备后可能会改变的模型参数，并采访了三位心脏病专家以捕获这些参数的概率分布。最后，我们通过将模型输入中的不确定性传播到模型结果中的不确定性，获得了每次测量可用的商业空间的概率分布。
对于每例生命年 10,000 欧元的支付意愿，当假设测量频率为每日时，每次测量可用的中位数为 1.64 欧元（四分位间距 0.05-3.16 欧元）。在随后进行的敏感性分析中，对于支付意愿阈值和测量频率的不同组合，此中位数最高可达 57.70 欧元。
概率推断可以成功地与早期健康经济建模相结合，以获得新医疗技术可用空间的概率分布。随后将该分布输入产品投资评估方法，使利益相关者能够就当前可用产品原型应针对哪些市场做出更明智的决策。
