Data recorder helps to measure robotic surgery skills

Register now

Keck Medicine of USC is the first healthcare system in the world to use data from a recording tool—similar in concept to a flight recorder on an airplane—to objectively measure surgeons’ proficiency in robotic-assisted prostate cancer surgery.

The recorder—called the dVLogger—is made by Intuitive Surgical; it attaches to the vendor’s da Vinci Surgical System, a robotic surgery platform approved by the Food and Drug Administration, and records both anonymized video and movement data in order to capture where the instruments are and how the surgeon is moving them.

According to Andrew Hung, MD, assistant professor of clinical urology at the Keck School of Medicine of USC, subjective opinion from surgeons is currently the “gold standard” for them to be credentialed by hospitals to use the robotic system. However, he contends that the problem is surgeons are evaluated by their peers for a handful of procedures but the evaluations are not ongoing, and sometimes evaluators don’t agree on what constitutes proficiency.

“However, we can measure objective data—it’s not based on a person’s opinion,” says Hung, an expert in robotic, laparoscopic and traditional open surgery for diseases of the adrenal, kidney, ureter, bladder and prostate. “The dVLogger is recording the system’s data of the robot, and it’s deciphering that system’s data to make these performance metrics. Those are perhaps best quantified by a computer, which is essentially what the dVLogger is. It’s a bean counter. It will count everything that the surgeon is doing. We call it motion tracking or kinematics. Everything that the surgeon’s hand and feet are doing is all being recorded.”

Also See: Loyola develops system for better reporting robotic surgeries

Keck Medicine of USC has been using the dVLogger since August 2016 during robotic-assisted radical prostatectomy procedures—the most common treatment for prostate cancer that can take between three and four hours for the operation—providing the ability to differentiate between novice and expert surgeons.

“Robotic surgery has been widely adopted by urologic surgeons, but methods of assessing proficiency vary widely between institutions,” notes Hung, who is lead author of a dVLogger pilot study appearing in the January 2018 edition of The Journal of Urology. “To have a surgeon review footage of a case for that duration (three to four hours), is not really practical. This study is a big deal because for the first time we’re able to objectively capture surgeon performance during live surgery.”

In the study, data were recorded during robot-assisted radical prostatectomy cases for both novice and expert surgeons to test the dVLogger’s ability to measure proficiency. Four basic prostate surgery steps were analyzed in the study—bladder mobilization, seminal vesicle dissection, anterior vesicourethral anastomosis and right pelvic lymphadenectomy—with results showing that expert surgeons completed operative steps faster with less instrument travel distance, less aggregate instrument idle time, shorter camera path length and more frequent camera movements.

“Objective metrics revealed experts to be more efficient and directed during preselected steps of robot-assisted radical prostatectomy,” concludes the study. “These findings lay the foundation for developing standardized metrics for surgeon training and assessment.”

The dVLogger “should be treated as a research tool,” observes Hung, who notes that the data recorder is not yet commercially available but is now being evaluated by other healthcare systems besides Keck Medicine of USC.

“At the moment, this is strictly a research-based project,” he adds. “One day we will be able to provide meaningful feedback to surgeons—how that data is packaged, that is the next challenge. We are able to capture tremendous amounts of data, but how do you make it meaningful to the surgeons such that they can improve their approach to the operation and do a better job next time? That remains to be spelled out. We’re hoping that in the future these become automated performance metrics.”

For reprint and licensing requests for this article, click here.