Robotics and AI in Hospitals
With robotics and AI playing an ever-increasing part in the way healthcare is managed and conducted, Greg McEwen, partner at insurance law specialist BLM, explores the potential for the technology and how the ‘friend or foe?’ equation will affect medical claims.
The idea of ‘robotic surgery’ is not a new one. The first robot assistant for surgery, the Arthrobot, was developed in the early 1980s and first used in 1984, in an orthopaedic procedure. Since then, robots have been developed and deployed in increasing numbers, to performs all types of surgery from bowel and bladder procedures to eye operations and neurosurgery. One of the most well known systems is the da Vinci Surgical System, approved since 2000 by the US FDA. The system accepts a variety of different surgical instruments, allowing it to be used in different surgical settings.
It’s not just the US however that has led the charge in the field of robotic surgery. The PROBOT, developed at Imperial College London, was used in a world-first to perform robotic prostate surgery at Guy’s & St Thomas’s Hospital in 1992. As AI has progressed, so too have we in its adoption into ever more complicated procedures – you need only look at the first ophthalmic surgery carried out by a miniature robot within the eye, which took place just a few months ago, here in the UK.
However, with the greater ubiquity of robots have come greater concerns and reports of errors. Researchers from the University of Illinois, Michigan Institute of Technology and Rush Medical Center published a manuscript in 2015 entitled Adverse Events in Robotic Surgery: A Retrospective Study of 14 Years of FDA Data, utilising data from the Manufacturer and User Facility Device Experience (MAUDE). The study noted that, from 2007 to 2013, 1.74 million robotic surgical procedures were performed in the US – the majority of which were urological or gynaecological.
In the thirteen years preceding, the data recorded 8,061 ‘device malfunctions’, 1,391 patient injuries and 144 patient deaths. Adverse incidents included electrical arcing, sparking or charring of instruments and the falling of broken or burnt pieces into the patient’s body. Such incidents were said to have contributed to 119 injuries and one patient death.
Clearly, operations utilising robotics are not without their risk. However, it should be noted that incidents relating to broken and/or retained instrumentation are by no means exclusive to robotic surgery, and herein lies one of the difficulties in interpreting the data. It cannot tell us whether a complication is solely or partly attributable to the use of a robot, whether it is patient related, or whether it represents a complication of the surgery itself. As a result, there remains a debate amongst medical professionals over whether the perceived advantages of robotic surgery outweigh the costs, both financial and otherwise. Some studies have suggested that surgical outcomes for robotic procedures are as good as the non-robotic alternative, but is “as good as” good enough?
In fact, there have in recent years been reports of a decline in the sale of surgical robots. Amongst the disadvantages sometimes cited are longer set-up times as well as the time that may be required during surgery to change instruments. Surprisingly perhaps, some surgical procedures may therefore take longer to perform with robot assistance, with the knock on effect of longer periods under anaesthesia. On a practical level, robots also lack the sensation or ‘feedback’ that experienced surgical hands may rely on to apply just the right level of traction or force, and can lead to inadvertent trauma and resultant injury. There have been a number of lawsuits in the US arising from robotic surgery, some involving fatalities.
Assuming however that robotic surgery is here to stay, particularly with private investment growing in markets such as China and Brazil, what does the future hold? Could we see the rise of remote surgery, where the operator and patient are not even in the same place? This is not the stuff of science fiction. The technology already exists, developed for situations where a patient was inaccessible to a surgeon, and was being considered as far back as the 1970s by NASA in the context of space travel.
In 2001, the first transatlantic surgery was performed by surgeons in New York, on a patient in France. In 2006, a robot was used remotely to sew up a cut on an inhabitant of the Aquarius underwater base, 70 feet under the sea. A faster, more robust internet has made the technology more viable, but obvious pitfalls remain such as the potential for a sudden power failure (at either the patient or the surgeon end of proceedings), or a loss of communication.
There are ethical considerations too. Medical tourism already exists, with some patients choosing to travel abroad for surgery for reason of cost, expertise or availability. Should remote surgery become more widespread, the patient would not even need to travel to their surgeon. Treatments that are not available (or not approved) in one country could become available without the patient even needing to find their passport. Obvious questions that arise here include:
- In the event of an adverse outcome, which legal jurisdiction would apply – that of the surgeon or that of the patient?
- How would the duty to ensure the patient is properly consented be discharged, where doctor and patient never meet?
- How do you ensure the patient receives appropriate post-operative follow up? What happens in the event of a serious complication?
These are all issues that the legal, medical and insurance industries are going to have to iron out between them, and we are likely to see more and more ‘precedent’ cases in the coming years. AI clearly has the potential to revolutionise the way we work and play, but it’s fair to say there will be some teething issues along the way, which will inevitably come at some human cost.This is a guest article written by Greg McEwen, Partner, BLM.