Is it time we let computers drive our medical care? According to the World Health Organization, there was a global shortage of 7.2 million health providers in 2013, and this figure will to increase to 12.9 million by 2035. Establishing new medical schools, which every country has adopted, isn’t a long term solution. Increasing the number of schools is often accompanied by insufficient human resources to train medical students, too limited patient access and equipment-poor facilities.

On the diagnostic side physicians are already being replaced. There is a spectrum of sensors available that measure a range of bodily functions, including temperature, heart and other muscle function, blood pressure and blood glucose. Pill cameras now give doctors a view of our intestines without any need for invasive scopes. They can produce over 800,000 images in 8 hours, with the camera turning at around 60 degrees every 12 seconds. That’s an awful lot of gastroenterologist work that can be done by a computer.

Humans are not inanimate objects like cars. We are more like the weather; often unpredictable.

On the therapeutic side, there are a range of externally controlled devices already in use. In cardiology, some of these devices such as pacemakers and external mechanical hearts, have been around for nearly a century. In other areas, such as diabetes, there has been some recent successes with artificial pancreas and external insulin pumps, both of which can be remotely controlled.

Not all the driverless changes are related to devices. Machine learning and data mining using computer generated models are becoming drivers of clinical decision making. In cancer, for example, there can be a lot of treatment options with similar outcomes. Algorithms can help physicians make appropriate choices.

All sounding good for our future, but humans are not inanimate objects like cars. Though we might be able to be driven consistently, we are more like the weather; subject to what the programmers might call inexplicable, non logical or random effects.

This individual variability can cloud the interpretation or selection of particular tests or treatments and result in physician confusion, diagnostic dilemmas and payer uncertainty. Many groups are trying to codify and integrate this individual component and minimise the risk of a wrong decision, treatment or appliance in the same way that doctors have done for centuries by combining art and science.

For now, clinicians remain indispensable.