Healthcare providers are not the only ones raising these kinds of concerns around the adoption of AI based technologies. Sectors like banking, aviation, manufacturing etc. where AI is already widely adopted, have also seen similar discourse at some point in time. However, experience from these sectors has shown no reduction in number of jobs with the introduction of AI. It has only seen the requirement for different skill set for jobs over the years.
Healthcare as a sector has huge gaps currently in terms of access to care as well as affordability and quality of care. There are millions of people around the globe who have no access to specialized care. Even where care is available, the cost and quality of care are a matter of concern. These gaps make healthcare an appropriate use case for AI based technologies that
- Improve access by helping diagnostics in the absence of trained personnel like ophthalmologists, radiologists; supporting telemedicine efforts in fields like radiology, infection management
- Empower providers like nurse practitioners, health assistants by providing decision support systems
- Improve cost-effectiveness of systems by improving efficiencies with the help of data backed algorithms
In fields like infection management where sensitivity patterns and trends are changing rapidly, sometimes as frequently as every three months, the relevance of recent data in treatment decisions becomes more imperative. Given the amount of data required to be processed for identifying trends and patterns in infection, it is difficult for the human mind to be able to stay on top of things. Analyzing infection related data and giving actionable insights is thus a relevant use case where AI is the perfect enabler for the infectious disease specialists to take data backed decisions for precision medicine.
It is worth mentioning here that AI can only handle the large database and its analytics but cannot take care of the finer nuances of clinical decision-making. It would thus be safe to conclude that the current AI systems cannot be robust enough to replace the physician in infection management. They can at best be an enabler for an experienced physician allowing him/ her to take data based decisions and even help them cater to multiple healthcare units at one time (given the shortage of the number of ID physicians in countries like US).
To conclude, in a healthcare system perpetually constrained by costs and availability of trained workforce, the possibilities for AI to solve problems and improve efficiencies is huge. What is required is to allay the fears of the stakeholders by positioning AI an enabler rather than a threat or replacement.
