A risk perspective on technology and professional work

In the FT on Monday June 19, an article titled ‘Hunt sharpens focus on AI benefits’ includes a quote from our PM Rishi Sunak: ‘The more AI progresses the greater the opportunities to grow our economy’.  The rejoinder phrase continues ‘… but with such potential to transform our future, we owe it to our children and grandchildren to ensure AI progresses safely and responsibly’. It is precisely this duality around benefits and potential harm associated with contiguous and overlapping risks that is of critical importance to unpack as we understand the changing risk landscape and the implications for technology and professional work.

At CDI our ongoing digital health research highlights how risk professionals are digitally transforming their work as they navigate  ongoing risks related to COVID and AI. The term risk professionals distinguishes those professionals whose work inherently entails risk in providing abstract, specialised knowledge to exercise professional judgment. For example, specialist doctors are always aware of risk and doing no harm to patients in carrying out medical diagnosis using decision support systems in performing their work. 

Our research, which started just prior to COVID, looked at the implementation of AI algorithms in ophthalmology services. The eye hospital we collaborated with had recently gained special recognition as developing one of the first real-world AI implementations in healthcare. This referred to the use of machine learning algorithms to diagnose diabetic retinopathy from retinal fundus photos. These algorithms boasted similar levels of accuracy in diagnosis to that of a consultant. At the time, there were telemedicine pilots using AI to imagine the future transformation of professional work and reshaping of the future ecosystem around eyecare. While these telemedicine pilots looked very promising, it was widely agreed by the digital champions that possible benefits for clinicians as risk professionals in using telemedicine were between 5-10 years away. The traditional face-to-face examinations using eye slit lamps were preferred by many ophthalmologists and telemedicine was not just unnecessary but involved taking unnecessary harm!

Our research sharpened with the ongoing risks related to COVID and AI to examine the following question:  How are risk professionals digitally transforming in navigating ongoing risks related to COVID and AI?  Rather dramatically, during COVID, video consults went from 0 to 10k in just a few months? Indeed, it was common the world over.  How can we account for this unexpected phenomenon of technology use by risk professionals in risk environments?  How might we go beyond recognizing that it was a necessity for contactless care which became the mother of invention in using telemedicine for digitally transforming eye care?  

One vantage point might be paying close attention to the changing risk profile or the riskscape that emerged with COVID. Literally overnight, telemedicine, which had been deemed far too risky and unnecessary for a largely face-to-face procedure using a slit lamp became an urgent necessity. Taking a risk perspective, our findings showed how the rapid uptake of videoconsults required a lot of riskwork to support and shape the traditional forms of professional work. This riskwork took different forms in navigating potential risk in the adoption of technology as part of professional work. First, care was taken to revise the service as being a triage instead of a diagnosis. Second, in the absence of a physical eye examination more reliance was placed on case history which is not traditionally viewed as being necessary with the presence of a physical eye exam. Third, there was an increased use of what are called red flags to manage the risk in the video consult procedure. With these changes to work, we found that responsibility for handling risk in part shifted to the patient at home. 

Our findings urge us to explore the changing riskscape and risk interactions which may reorder risks as being harmful and/or of benefit. The telemedicine pilots, once only an imagined long future, became a present reality as the greater risk posed by COVID reordered the perceived harm of telemedicine and indeed allowed for its benefit to be realized instead in short order. The impact on the work of the professionals led to some ‘Road to Damascus’ conversions on the value of telemedicine with revisions as to the risk profile of telemedicine. In other words, the duality of risk associated with the technology allowed for an unexpected shifting from harm to benefit in navigating the dynamic riskscape.