Lamb also warned that privacy should not be sacrificed and that the health service should get a ‘fair deal’ from companies implementing the AI systems.
The report, by the Reform think-tank, suggests that AI could be utilised to target treatment by analysing and predicting the propensity for individuals to be within a risk group of a particular illness.
The technology could also be used for diagnosis – initial data suggests that an AI system could diagnose breast cancer more accurately and 30 times faster than a human. However, the think-tank believes the best operating model is for the AI system to be an ‘enabler’, working alongside humans to deliver a decision.
To-date only a small number of NHS trusts have started using AI, this, according to Mr Lamb, is due to a lack of funds available and the ‘lack of clarity about the strategic direction to take and where to start’.
In order for widespread adoption of AI to take place, the first step is to clean up and digitise patient data, says Lamb. “If data is not inputted accurately or if there is no consistent approach across the system then the AI can play havoc because it’s operating on the basis of wrong information.”
In the report, Reform believes NHS Digital should start by creating a series of training datasets and make them available for third-party companies to produce algorithms for use in the health service.
Lack of trust in NHS data collection and security practises
At the top of everyone’s concerns is patient data privacy, which has been further compounded following the WannaCry ransomware attack last year which exposed the NHS’ vulnerabilities in its creaking data infrastructure.
Unfortunately, this isn’t the first time public confidence in digital records waivered. The Government had to backtrack on its patient data sharing programme and vowed to restore trust after accusations that the scheme was not transparent and that the necessary opt-in’s had not been collected.
On that point, Mr Lamb believes that the NHS needs to use AI in a ‘way that guarantees the safety of data’. The think-tank report also urged for clarity around consent, and the ability to opt-out if data is being used for anything that is not directly related to the patients care.
The report also found that AI could deliver staffing efficiencies by reducing the ‘burden’ of administrative tasks. When quizzed on whether this would lead to job-losses, Mr Lamb believes that with an ageing population there will ‘always be demand for people to work in the NHS’. The use of AI will strip cost out of the running the service, which means the NHS will have more money to reinvest in actually supporting people instead of a system that is on the brink of imploding.