Nadarzynski, T. et al. (2024) | Achieving health equity through conversational AI: A roadmap for design and implementation of inclusive chatbots in healthcare | PLOS Digital Health 3(5): e0000492 | https://doi.org/10.1371/journal.pdig.0000492
Artificial intelligence (AI) systems like conversational chatbots are increasingly being used in healthcare settings. However, there are concerns that these AI tools may perpetuate biases and worsen health inequalities if they are not designed and implemented carefully with diverse populations in mind.
This study aimed to develop a roadmap to help ensure that AI chatbots for healthcare are created and deployed in an equitable and inclusive way. The authors reviewed existing guidelines on ethical AI use in healthcare and interviewed 33 stakeholders including community members, healthcare professionals, and AI developers. Based on this, a 10-stage roadmap is outlined covering key activities such as involving underrepresented communities in co-production, conducting research on AI biases, implementing safety protocols, evaluating real-world effectiveness, and planning for chatbot maintenance or termination while minimising disruption to care.
The roadmap emphasises the importance of close collaboration between chatbot creators, healthcare providers, policymakers, and the communities being served. It provides a checklist of equity considerations at each stage of the chatbot lifecycle. Following this roadmap can help mitigate unfair biases, increase acceptability among diverse patient groups, and harness the potential of conversational AI to improve healthcare access and reduce inequalities.
Full paper: Achieving health equity through conversational AI: A roadmap for design and implementation of inclusive chatbots in healthcare