The future of healthcare: interview with microbiologist Alberto Rizzo
In the debate on artificial intelligence in healthcare, between enthusiasm and caution, a central question emerges: how can we find the right balance between technological innovation and clinical responsibility? Onelia Onorati discussed this with Alberto Rizzo, Clinical Microbiology, Virology and Bioemergencies “Luigi Sacco” University Hospital, ASST Fatebenefratelli Sacco, who in various articles and interviews has used the metaphor of the “porcupine dilemma” to describe the complexity of human relationships and, today, also the relationship between doctors and technology.
Rizzo is one of the authors of the study Artificial intelligence in clinical microbiology: results from the first National survey by the Italian association of clinical microbiologists, the first national survey promoted by the Italian Association of Clinical Microbiologists to assess the level of awareness and use of artificial intelligence among healthcare professionals. Based on the results of the research, we explore the opportunities, limitations and prospects of AI in everyday clinical practice, with a close look at the crucial issue of training.
Let's discuss some of the results together. What does the research tell us about healthcare professionals' awareness of the adoption of artificial intelligence? Do your colleagues feel confident in using AI?
First of all, let's start from the awareness that those who responded to the research questions are probably people who are particularly interested in AI and motivated to learn more about it. However, the data is promising, because 25% of operators said they use the tools, i.e. 1 in 4, and almost all of them said they were interested in learning more. This highlights not a fear of innovation but an openness to it. It is important to highlight this propensity, which also indicates a willingness to participate in training initiatives. It is indeed of primary importance that healthcare professionals are trained to understand when and within what limits AI can be used, how to consult it and in what contexts.
In your daily experience, how would you like to see AI used in everyday clinical practice?
Healthcare professionals are overwhelmed every day by small “nuisances” that distract them, albeit minimally, from the ultimate goal of their work, which is to provide the best possible service to the population in terms of care. AI tools should relieve us of these repetitive tasks, such as the more bureaucratic aspects of routine work. How many of us, during medical appointments, notice the time spent by the doctor writing documents on the computer! This is where Artificial Intelligence can be configured as a valuable “assistant”, as a support. Many believe that the fruitful use of AI goes hand in hand with balanced human-machine integration.
A thought-provoking question: will we eventually have diagnoses written by AI?
Today, this is not possible, and although I share the enthusiasm for these new tools, we must not get carried away! Some tools still tend to please the user, so we must be very careful. But let's think about the future: if the growth curve in the quality of the tools is confirmed, I am confident that they will be integrated into everyday practice. It may also be possible to request support in formulating diagnoses. I believe that an overly conservative attitude is just as dangerous as uncritical enthusiasm, because falling behind on innovation risks leaving us behind. We must always govern the development of technology and, to do so, we must remain informed.
How can we integrate, as you hope, the training of clinical staff and the supervision of the results proposed by AI?
Let's go back to what we said at the beginning: training. It is essential to have the tools and knowledge to effectively supervise AI outputs. Above all, it is important to know how to read the answers because often, for example in the case of generative AI, they are presented with great logical consistency even though they contain errors. Only a thorough knowledge of the tools can enable effective supervision.
