GPs Voice Top 10 AI Concerns

A review of queries received by the Medical Protection Society (MPS) last year has highlighted the top concerns of GPs about the growing use of artificial intelligence (AI) in medical practice. 

The MPS told Medscape News UK that all AI-themed queries came from GPs. No questions were raised by secondary care doctors.

Key Concerns Identified

The assessment revealed 10 main areas of concern for GPs:

• Patient safety
• Data protection
• Patient consent
• Liability
• Indemnity
• Transcribing tools
• Generation of fit notes
• Clinical prompts
• Processing of laboratory results
• Generative AI

GPs Wary of AI Risks

Medicolegal advice on these themes has been published on the MPS website. 

“We know members are keen to explore and adopt AI tools, which may enhance patient care and help to facilitate more efficient working,” Dr Ben White, MPS deputy medical director, said in a press release. 

However, calls to the medicolegal advice line showed members were wary of potential risks.

Particular concerns included liability and defence society cover, safety for patients, data protection, and patient consent. Such concerns echo a survey by Medscape last year which found that 1 in 5 respondents were apprehensive about using AI in clinical practice. 

More specific advice was sought on the medicolegal implications of using AI software for transcribing patient consultations, generating fit-to-fly letters or fit notes, using clinical prompts, and processing laboratory results.

“AI is of course fast-evolving, and we will continue to revisit the guidance,” White said. 

The MPS Foundation has contributed to a white paper on AI in healthcare, aimed at ensuring AI tools are developed and integrated to be usable, useful, and safe for both patients and clinicians. 

“Bringing about greater confidence in AI among clinicians is vital if the potential benefits are to be unlocked,” White said. 

Doctors Support AI but Remain Concerned

Nell Thornton, improvement fellow at the Health Foundation, said the MPS findings echoed the foundation’s own research. 

“If AI is to help fulfil its promise to improve patient care, further clarification is needed on issues like regulation and professional liability,” she told Medscape News UK. 

Doctors are keen to ensure that the human dimension of care is protected in an increasingly digitised system, she added. “Addressing these issues will be crucial for ensuring the development and adoption of AI is responsible and works for everyone.”

Professor Kamila Hawthorne, chair of the Royal College of General Practitioners, told Medscape News UK that the college was “always open to introducing new technologies that can improve the experience of patients — and GPs are often at the forefront of this.” However, AI use is “not without potential risks” and its implementation “must be closely regulated to guarantee patient safety and the security of their data.” 

“Technology will always need to work alongside and complement the work of doctors and other healthcare professionals, and it can never be seen as a replacement for the expertise of a qualified medical professional,” Hawthorne emphasised. “Clearly there is big potential for the use of AI in general practice, but it’s vital that it is carefully implemented and closely regulated in the interest of patient safety.”

Consultation Summaries Show Promise and Peril

Dr Rosie Shire, from the Doctors’ Association UK GP committee, told Medscape News UK, “As more AI tools become available, the possibilities of how they could assist GPs grow, but so do the concerns, as highlighted by the MPS.”

The idea of AI software providing a summary of a GP consultation was “very appealing,” she said. It could enable GPs to focus fully on the patient rather than taking notes, which could improve communication and patient satisfaction, and save time on recall and typing up notes. 

“However, it’s vital software is accurate and reliable, able to distinguish who is speaking, and understands accents and dialects,” Shire stressed. “Otherwise it could lead to increased workload, as GPs spend additional time checking the AI has performed correctly.”

AI Can’t Replace ‘Gut Feeling’

There is potential for inaccuracies being introduced into patient records if clinicians become overreliant on AI records of consultations, Shire said.

In time-pressured clinics, GPs might not read through AI summaries.

“It’s important clinicians are given time to review and confirm any AI-generated outcome to ensure accuracy. We also need to remember that AI can’t replace that gut feeling you occasionally get as a GP when you see a patient you know and feel that something just isn’t right,” she added.

Liability Questions Remain

Shire said that a standard policy on AI use in general practice was needed. Clarity is also required over who would be responsible if AI software malfunctioned and led to patient harm. 

“Would it be the practice partners, as the data owners, the GP who saw the patient, or the AI software provider?”

White said the MPS would not normally provide indemnity for issues relating to AI software failure. While the society is not currently aware of any binding case law, “we would expect the designers or producers of the AI software to be liable for any issues relating to the failure of the AI software.”

Where a practice has purchased an AI system, the individual GP using it would remain potentially liable for any harm that results from its use. 

“Doctors remain responsible for the decisions they make whilst using such technology,” he said. “In order to mitigate the risk of a complaint or claim, doctors should assure themselves that any system they are using is fit for purpose and appropriate for the patients that they are seeing.”

Dr Sheena Meredith is an established medical writer, editor, and consultant in healthcare communications, with extensive experience writing for medical professionals and the general public. She is qualified in medicine and in law and medical ethics.