To explore the perceptions and acceptability of AI tools for use in primary care.
Patient and public consultations (PPIE) were undertaken with six stakeholders including patient, professional body representatives and healthcare clinicians to understand the requirements, acceptance, and co-design considerations for an AI tool developed for use in primary care. PPIE conversations were recorded and transcribed verbatim. Data were analysed using Braun and Clarke’s inductive thematic analysis framework.
Two main themes were identified. The first theme ‘Clinical decision making’ had three subthemes: ‘Clinical enabler’; ‘Enhanced consultation’; ‘Acceptability’. The second theme ‘Technology’ had five subthemes: ‘Digital systems and Infrastructure’; ‘Cost of AI tool’; ‘Governance and legalities of healthcare’; ‘AI tool usability’; ‘The patient dataset’.
All stakeholders were positive and emphasised the need to introduce AI into clinical practice in primary care. They felt AI could be a clinical enabler, enhancing patient safety in part by managing vast amount of ‘noisy patient data’ to identify and link important information from electronic patient records, facilitating an earlier diagnosis and improved patient outcomes. Moreover, it could help support the generalist/less experienced clinician with clinician reasoning, especially in identifying serious less common conditions, and support non-clinical staff in prioritising patient appointments. However, caution was raised by all that AI should support clinician’s decision making, not replace it. Related was who would be responsible and liable if things went wrong. Additional concerns were raised around connectivity in the NHS, both within and between primary and secondary care, an aging NHS infrastructure, and cybersecurity. Finally, the cost of AI and who pays, tech literacy, AI hallucinations, quality of data and impact on health inequalities were raised as further considerations.
The NHS and primary care are facing significant issues, with greater demand leading to stretched systems. We need to think differently to address these issues and AI technology may help to do this. However, despite growing interest, AI in primary healthcare in the UK is not without challenges. These findings lay the groundwork for future research into the development of AI in healthcare.
UK NHS
Perceptions
