Online Program

Diffusion of Robotic Surgical Systems and Its Impact on Treatment Pattern of Localized Prostate Cancer: An Instrumental Variable Analysis

*Chan Shen, University of Texas, MD Anderson Cancer Center 
Ya-Chen Tina Shih, University of Chicago 

Keywords: Endogeneity, instrumental variables, panel data, technology diffusion

The average number of robotic surgical systems per state increased from 2 in 2002 to 26.3 in 2010. With the rapid adoption of robotic-assisted surgery in US, it is important to understand its impact on treatment patterns for prostate cancer. Using state-level data from the National Cancer Database, we applied regression analyses with state fixed effect to examine the impact of the number of robots on the rate of surgery, radiation therapy, and active surveillance. We considered the possibility that the number of robotic systems was not exogenous, and used instrumental variable (IV) method to deal with this potential endogeneity problem. The exclusion restrictions included the rate of hysterectomy and the total state population in year (t-1). Findings suggest that the effect of the number of robotic systems on the rate of surgery and radiation therapy will be under-estimated when endogeneity is neglected.