Patient safety concerns have emerged regarding AI scribes that transcribe conversations between doctors and patients, according to an alert from Healthwatch England. The technology, which the NHS is rapidly deploying across the health system, has documented instances of significant errors in medical transcription that could compromise patient safety.
In one notable case, an AI scribe incorrectly documented that a patient had demyelination, a serious neurological condition associated with multiple sclerosis, when the correct finding was “null demyelination.” The patient, who works as an NHS health professional, described the experience as traumatising and noted that correction only occurred after she questioned the AI tool’s record. In another instance, an AI scribe confused a prescribed medication with a similarly named drug, an error the patient herself identified. A third case involved an AI-generated summary that omitted critical instructions for a patient to obtain a repeat prescription for migraine medication from their general practitioner.
Healthwatch England has documented multiple instances where patients rather than healthcare professionals identified errors in AI-generated transcripts. The organization emphasized that if such inaccuracies remain undetected, they may persist in patients’ permanent medical records and influence their ongoing care. Approximately 27 different AI scribes are currently in use by GPs and hospital doctors throughout England. The Medicines and Healthcare products Regulatory Agency has determined that AI scribes do not require classification as medical devices, meaning no England-wide safety oversight or effectiveness verification is mandated.
The government’s 10-year health plan for the NHS anticipates that AI scribes will reduce administrative burdens on healthcare staff and increase time available for patient care. However, healthcare professionals and patient advocates have raised concerns about implementation. Research cited in the British Journal of General Practice suggests that doctors may be expected to see additional patients daily despite already working within some of the world’s shortest consultation timeframes. Research from Uppsala University found that GP surveys indicate errors are more likely when consultations involve multiple participants, patients with complex medical histories, or those whose first language is not English.
The Patients Association cautioned that trust in the technology depends on clear communication and genuine partnership with patients, which they characterized as currently lacking. Healthwatch called for clarity regarding patient reporting mechanisms and error correction procedures, noting that while healthcare has never been error-free, urgent attention to these processes is necessary.
Article Attribution | Read More at Article Source
Article summary produced by Claude AI