Faculty Research

Take a close look the next time you see a hospital advertisement on your commute: there’s a good chance that ad will mention robotic surgery. First approved a bit over two decades ago, this new technology has dramatically changed how patients get treated for cancer and many other conditions. Now, it’s a common feature on highway billboards across the country. In the past, surgeons would hold and manipulate instruments by hand. With robotics, the device controls the instruments and a surgeon manipulates them remotely.    

As health economists, my collaborators Danea Horn (Stanford), Annetta Zhou (RAND Corporation), and I were intrigued by the rise of surgical robotics. We wondered if hospitals focused so heavily on their surgical robots in advertising because these devices appealed to patients and helped grow market shares. We also worried that these developments could even be bad for patients: robotic surgery is expensive, and the evidence it improves patient outcomes is lacking.    

With funding from the National Institute on Aging, we embarked on an academic study to trace out the effects of robotic surgery. We focused on a setting where robotics had an outsized impact, prostate cancer treatment. We leveraged nearly two decades of data on older adults covered by the Original Medicare program, including their hospitalizations and diagnoses, as well as data from the American Hospital Association covering all U.S. hospitals. Between 2001 and 2015, about half of hospitals that treat cancer installed a surgical robot. The main hospital treatment for prostate cancer is prostatectomy, and over the same period, robot-assisted prostatectomy went from 0% to almost four-fifths of prostatectomies.    

We then studied whether hospitals that adopted robots attracted more prostate cancer patients. Comparing hospitals that installed a robot to those that didn’t, we found that robotic surgery raised patient volume by 49% to 82% (depending on the statistical model). This phenomenon could reflect that patients value having access to robotic surgery and prefer to receive treatment at hospitals that offer this approach. It also might be the result of patients’ physicians believing robotics would improve treatment – or surgeons might prefer to use the robot over older modalities of treatment.    

That result also raises the question of where these patients came from. Did adopting hospitals engage in “business stealing”, taking patients who otherwise would have been treated at other facilities? Or did they expand the market for surgery, in which case the patients they attracted would have otherwise received medical, not surgical, treatment?    

To tell these two stories apart, we zoomed out to study entire regional markets, not individual hospitals. Patients tend to stay in the regional market they live in when they go to the hospital. If there were only business stealing, the number of patients hospitalized for prostate cancer across the entire regional market wouldn’t change as hospitals in that market adopted robots. Yet we estimated that if all the hospitals in a market adopted a robot, prostate cancer hospitalizations would rise by 34%. An effect of that size means that robotics led to business stealing and market expansion.    

Some of the patterns we uncover in the study are encouraging. First, the patients drawn into care by robotics are relatively young and healthy. Policymakers have tried to discourage older and sicker patients from receiving excessive care that is less likely to benefit them, so we interpreted this finding as good news. Second, there were signs that as hospitals adopted, patients became more likely to receive care at bigger hospitals and teaching hospitals; these results weren’t statistically significant, but they all pointed in the direction of patients moving to the types of hospitals that tend to provide better care. On the other hand, because of data limitations we were able to say less about whether this new technology actually improved patient outcomes. We also weren’t able to evaluate its effects on the costs of care.  

Regardless, our results show that new technology can rapidly change how and where patients receive care. Going forward, our team will continue this research agenda and study the effects of surgical robotics on other cancers and patient outcomes.  

– Adam Sacarny, June 2022