Proctor : August 2019
38 PROCTOR | August 2019 Would you trust AI with a scalpel? And how should it be regulated? BY IVY SHI, THE LEGAL FORECAST Robotic surgery originated in a National Aeronautics and Space Administration (NASA) research centre in the ’80s.1 The earliest purpose of a surgical robot was to perform minimally invasive surgery (MIS) using a ‘master and slave approach’ – surgical robots were created as an extension of the human surgeon’s arm, a tool for precision during robot-assisted surgeries. By creating medical tools that could perform what mundane hands could not, robotic surgery was developed to improve the performance of and results from surgical practice, including optimised emergency responses, imaging technology, accuracy, and telesurgery. In 2007, the SAGES-MIRA Robotic Surgery Consensus Group, defined robotic surgery as: “A surgical procedure or technology that adds a computer technology enhanced device to the interaction between a surgeon and a patient during a surgical operation and assumes some degree of control heretofore completely reserved for the surgeon.” 2 Current robotic surgeons Rapid developments in medical robotics have seen the creation of numerous robots in healthcare, including the MAKO system, which is an orthopaedic robotic arm currently residing at Queensland’s St Vincent’s Private Hospital,3 and the CyberKnife, a radiosurgery robot located at Sir Charles Gairdner Hospital in Western Australia. 4 Both of these devices have a relatively high level of autonomy, operating either with close monitoring or minimal execution from the human surgeon.5 The two medical robots show not only the invincible power of technology, but also the slow transitioning from the traditional ‘master and slave approach’ into individually capable devices. In the future, it is expected that the inevitable development and enhancement of medical robots will lead to greater autonomy and decision-making through the power of machine learning and artificial intelligence.6 Legal implications While hospitals may see the benefit in medical efficiency in respect of costs, timing and access to healthcare, there may be issues with identifying medical liability. Currently, if medical fatalities related to device malfunctions were to occur, surgical robots would be governed and regulated under the Therapeutic Goods Act 19897 as medical device malfunction. This may not be applicable when robotic surgeons join the ranks of the medical profession, as their high level of autonomy will not only classify them as medical devices, but also capable of practising medicine.8 So, how should we regulate robotic surgeons? When fully autonomous robotic surgeons arrive in our hospitals, regulations surrounding medical malpractice will need to change. As human surgeons will not be directly operating on patients in these situations, it will become much more difficult to determine if medical responsibility lies with manufacturers (product liability)9 or with human surgeons (medical negligence). 10 The future will inevitably see robotic malpractice in the operating room, when device malfunctions create calculation and decision-making errors. These situations will require the courts to decide the level of responsibility of the involved parties, determined by the level of involvement each had in the surgical procedure.11 The current regulation of surgical robots is not substantial enough for the future scenarios of malpractice outlined above. Manufacturers cannot take over the human surgeon’s role of medical responsibility alone, as there are multiple parties to bear a shared sense of responsibility, including: • robot designers • engineers • programmers • manufacturers • investors • sellers • users. These parties are involved with the design, creation and execution of the surgical robot. However none can be chosen as the ultimate source of responsibility, because each plays a role unique to their expertise.12 If an autonomous robot fails to perform the procedure correctly, would it be solely the manufacturer’s fault for the device error, or would the supervising surgeon be responsible for not preventing the incident? Perhaps the patient’s situation was unexpected, and the robotic surgeon was not programmed to correctly act upon the scenario? In these situations, where full responsibility is not delegated to a sole party, robotics ethics are challenged, and court cases would be difficult and tedious to solve. Where to from here? Fortunately for robotic surgeons, we have already seen satisfactory regulation for autonomous devices. The self-driving vehicle has gone through legislation amendments in some jurisdictions, including the appointment of an automatic driving system entity13 as the legal entity who determines the vehicle’s safety on the road, and is held responsible for potential vehicle malfunction and error. This methodology could potentially be applied to future autonomous robotic surgeons.