There are few sectors where the notion of human interaction seems so intrinsic as healthcare. When we picture it in our minds, the image that appears isn’t a machine: It’s a doctor, a nurse and a surgeon. A calm, kind, reassuring face.
Now imagine a healthcare sector of a different kind, where most of our interaction is not with a human, but with a smart computer. In this scenario, our discussions are with an intelligent virtual agent, to which we explain our problems, symptoms, and concerns, and from which we receive our diagnosis and treatment plan. When we give blood, or undertake a medical procedure or even surgery, we are in the hands of a robotic device.
If it sounds like a vision from the future, think again. This scenario is not just on the horizon; it is already here, made possible by the exponential technologies in the sphere of healthcare that have emerged over the last few years. At the centre of this unprecedented change in one of our most pivotal and beloved sectors are new solutions that incorporate varying levels of artificial intelligence and robotics that have left the lab and are now being put into practice across the world.
True, they can hardly be called commonplace yet. But make no mistake, they will be.
However, the inevitability of change does not equate to us being stripped of the ability to influence how this change takes place. It may be natural for healthcare to be embracing the digital transformation, given the impact it has already had on other industries. It may be an obvious move to take full advantage of AI and robotics to keep us healthy. But that doesn’t draw a line under the conversation about precisely how these new technologies will be implemented.
Of course, the philosophical debate about technology replacing humans and its social impact is already swirling and will continue to do so. But that is not the only debate in town when we are talking about a technological transformation in healthcare. In many ways, it is not even the most important one. Instead, we need to focus on whether the technology is ready for use in the first place.
The questions we asked ourselves at a recent conference centred on whether there is a way we can be sure these solutions are safe; whether they are more effective than, or even equally effective as, the traditional human approach to healthcare; and how we can adapt existing evidence-based studies to keep pace with innovation.
Digital health solutions are being focused on for one reason: they have proven benefits. There are many examples that demonstrate the capacity of AI-based diagnosis to be several times faster than human intervention. They make advanced knowledge available in standardised form across the world, including in developing countries that would otherwise not have access to this insight. Studies have shown that the vast database of medical records can improve testing and diagnosis as compared to existing methods.
But the exciting and transformational nature of technology does not remove the need for a balanced assessment of how it will be used and whether it is completely safe. ‘New’ does not mean ‘perfect’. And we still face the risk that many of the early results and studies mentioned just now may be limited to their specific scenarios. And for all the positive feedback, not all of them have ensured they are fully compliant with the processes and demands of full clinical trials.
Making matters more complicated is the fact that AI is often based on algorithms designed to evolve over time – such as machine learning scenarios. We can have continuous trials of these mathematical rules every time a new generation of them emerges. But are we really prepared to do that?
If technology is to change the face of healthcare in a way that makes the sector more efficient, effective, and streamlined, enabling it to help more people faster, simply rushing it into operation is not the answer.
AI should help users and doctors make better decisions, rather than making them itself; the improvement of clinical effectiveness through solutions that have completed full clinical trials should be just as much of a priority as regulatory and ethical considerations; and more work needs to be done to ensure that solutions can be adopted across different regions.
There must be recognition of the fact that individual populations’ demographic, social, and genetic profiles will impact how intelligent solutions carry out their analysis, and for this to be taken into account in the development and deployment of these solutions. In healthcare, there is no such thing as one-size-fits-all technology.
Digital health is already here, and pretty soon it will be everywhere. It is undeniable that healthcare, and those it serves, can benefit immensely from it. But this is one sector where, if a technology is ‘ahead of its time’, it is reason for caution rather than excitement. The stakes are too high for it to be any different.
José Mussi is senior director for digital health and innovation at PwC