Recent research from the University of Edinburgh reveals that artificial intelligence transcription tools used to automate clinical notes can sometimes miss important aspects of consultations—such as facial expressions, gestures and emotional state—and may lead to patients withholding sensitive disclosures, raising questions across healthcare settings.
Medical scribe technologies have been used to alleviate administrative burdens. However, an evaluation indicates that reliance on automated summaries can reduce memory recall and skill development.
In Plain English: The Clinical Takeaway
- Cognitive Offloading: AI tools take on some of the mental effort of manual documentation, freeing clinicians to spend their efforts having more meaningful conversations with patients or concentrating on complex tasks, but this can reduce memory recall.
- Missed Nuance: Automated summaries tend to prioritise clinical information over patients’ stories and experiences of their illness, as well as expressions, gestures, and emotional states.
- Disclosure Hesitancy: Patients may be hesitant to reveal sensitive information—such as mental health struggles or substance abuse—if they know the consultation is being recorded and processed by AI.
The Operational Impact of AI Scribes
Deploying automated transcription tools has affected operations in some practices. For instance, doctors in the West Midlands said their workload had been significantly cut by using the transcription tool Heidi. At a Dudley medical facility, the software successfully cut the backlog of unmailed patient letters down from a half-year wait to just 14 days.
Despite these administrative gains, the University of Edinburgh study highlights trade-offs. Automated scribes tend to prioritise clinical information. This prioritisation can come at the expense of recording the patients’ stories and experiences of their illness. Furthermore, researchers identified that patients may be hesitant when revealing vulnerable topics—including domestic abuse or mental health struggles—knowing the consultation is being recorded and processed by AI.
Cognitive Offloading and Documentation Fidelity
The traditional practice of writing notes by hand plays a critical role in helping practitioners think through and process what individuals share during appointments. When clinicians rely on AI to synthesize a patient encounter, they engage in what researchers term “cognitive offloading.”
While offloading mental effort can help practitioners concentrate on complex tasks, the Edinburgh report underscores downsides. Investigators documented scenarios where medical staff failed to recognize records they supposedly authored or completely forgot the individuals involved by the time of their follow-up appointments. Over time, this can reduce memory recall and skill development among medical professionals.
| Documentation Method | Primary Advantage | Identified Risk |
|---|---|---|
| Manual Notetaking | Supports clinical reasoning and reflection. | Administrative burden. |
| AI Scribes (e.g., Heidi) | Reduction of administrative backlogs and workload. | Loss of non-verbal context, patient hesitation, and reduced recall. |
Balancing Technology and Clinical Governance
The findings from the University of Edinburgh indicate that while automated documentation assistants offer administrative relief, further investigation is essential to evaluate their performance across distinct medical environments and ensure their framework adequately serves both practitioners and those in their care.
Contraindications & When to Consult a Doctor
Always consult a qualified medical professional for any urgent health concerns or when managing complex chronic conditions.
