Kedar S. Mate

Drawing on the experience of a national patient safety advocate who used a consumer AI tool to obtain a correct Bell’s palsy diagnosis after being dismissed by an emergency physician, this article describes and presents survey and AI usage evidence that some patients are embracing AI as a means of navigating a health system that too often fails them. Against a backdrop of eroding institutional trust and diagnostic errors affecting an estimated 12 million Americans annually, patients are turning to AI for diagnostic reasoning, medication safety, and second opinions. The authors term this a “trust inversion”, contrasting some patients’ AI adoption with many clinicians’ continued hesitation. They call on clinicians and healthcare leaders to invest in AI literacy and study the outcomes of encounters in which both patient and physician arrive AI-enabled.

Health systems are rapidly adopting artificial intelligence (AI), with 2 in 3 physicians now using it for tasks like chart summaries and diagnoses—a 78% increase since 2023.1 Yet most clinical AI tools face limited US Food and Drug Administration (FDA) oversight, leading to inadequate evaluation.2 Over half of 1000 FDA-authorised AI devices reported no performance metrics and nearly half documented no clinical studies.3 Unlike pharmaceuticals, which require randomised trials and rigorous surveillance, AI tools are being deployed without comparable standards.

Trust is fundamental to the effective functioning of healthcare systems, influencing access, utilisation and adherence to evidence-based practices. While quality improvement (QI) processes are widely recognised for addressing technical challenges, their role in fostering trust and relationships within health systems remains underexplored. This study examines the relationship dynamics in QI teams and how trust-building frameworks align with adaptive processes in healthcare settings. We conducted a qualitative study involving 30 healthcare workers from six African countries, recruited through the Africa Consortium for Quality Improvement in Frontline Healthcare. Data were collected through semistructured interviews, transcribed and analysed using both inductive and deductive methods. Deductive analysis was guided by a published trust-building framework, while insights from a large language model were incorporated in addition to a traditional analysis to provide an unbiased perspective. Results identified three theoretically described dimensions of trust-building within QI teams: common goals, self-interest and gratitude/indebtedness. Common goals fostered teamwork, multidisciplinary collaboration and effective communication, while self-interest motivated personal and professional growth. Gratitude and recognition reinforced team cohesion and sustained motivation. Participants highlighted the importance of trust in achieving project success, noting that robust relationships within teams correlated with improved outcomes. The study underscores the dual nature of QI processes, which simultaneously address technical improvements and adaptive challenges, including trust and relationship-building. Trust-building, framed as an iterative process of aligning common goals, recognising contributions and addressing individual interests, complements technical QI methodologies like Plan-Do-Study-Act cycles. These findings support expanding QI frameworks to emphasise relational dynamics, contributing to more sustainable and impactful healthcare improvements. Further research should continue to explore the adaptive dimensions of QI, integrating recent research on culturally relevant frameworks prioritising kindness in healthcare systems, to enhance trust and collaboration within healthcare systems.

Coronavirus disease 2019 laid bare the gaps in our health systems. Isolation and discoordination of both individuals and systems, inequities at local and global scales, and false choices between our prosperity and our health, all exacerbated the crisis. To build a better “normal” and not just a new one, health care should employ the approach of targeted universalism, which demonstrates that we can often get to universally held societal objectives by using targeted strategies that help provide an advantage to those that have been systematically disadvantaged. The goal is universal, but achieving it requires multiple strategies that target the needs of various groups to help them share in the universal goal. This approach is perhaps most easily understood, and most urgently needed, in the context of improving health equity. Using targeted strategies to permanently remake our health systems will honor the lives of those we lost far too early.