Continuity of Care in the Age of AI: Supporting Safer Handovers in Primary and Community Health and Social Care
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Background Handover failures during care transitions remain a leading cause of avoidable harm across primary care, community health services, and social care settings. General practitioners, district nurses, community mental health workers, and domiciliary carers regularly exchange responsibility for the same individuals across shifts and team boundaries — often with incomplete, inconsistent, or fragmented information. Human-centred artificial intelligence (AI), designed to support rather than replace professional judgement, offers a practical mechanism for reducing information loss at these transitions while respecting the realities of frontline work. Methods A conceptual analysis drawing on a programme of seven empirical and review papers on AI, communication, workforce experience, and patient safety in UK health and social care, supplemented by primary care continuity literature and UK digital health policy. The analysis proposes a four-component framework for AI-supported handovers applicable across multidisciplinary primary and community care settings. Results Four components of a human-centred AI handover framework are identified: change-focused summaries that reduce informational discontinuity at shift start; cross-team communication support that provides shared situational awareness across multidisciplinary teams; a safety prompt layer that enforces consistent completion of safety-critical actions without directing professional judgement; and a workforce wellbeing integration component that acknowledges and supports the carer as a worker with needs. Empirical evidence from a needs assessment of 21 domiciliary care workers confirms the relevance of each component: 65% reported informational gaps at shift start, 52.4% prioritised medication reminders, and 90.5% wanted integrated wellbeing support. Six ethical and governance requirements are specified for safe implementation. Conclusions AI-supported handovers are a population health intervention, not merely a technology upgrade. When designed around the realities of frontline care workers in primary, community, and social care settings — reducing burden, supporting safety, and acknowledging workforce wellbeing — they have the potential to improve continuity for the populations most dependent on multidisciplinary care coordination. Implementation requires attention to equity, governance, and co-design with frontline professionals.