Patient and clinician perspectives in the use of machine learning and artificial intelligence in the context of acute neurology.

Background: Clinician perspectives on machine learning and artificial intelligence (ML/AI) vary with discipline. However, fewer studies describe patient perspectives and address medical situations that require rapid decisions with durable consequences for patient outcomes. This study characterized perspectives (qualitatively) and sentiment (quantitatively) regarding the use of ML/AI for patient management in neurological emergencies. Methods: We conducted semi-structured interviews with survivors (or their proxy) of intracranial hemorrhage, and clinicians who care for such patients. Interviews were analyzed qualitatively using thematic analysis, and quantitatively using sentiment analysis to assess attitudes using a transformer-based language model with scores from -1 (most negative) to 1 (most positive). Results: We enrolled 21 participants (14 patients, 1 proxy, and 6 clinicians) and reached thematic saturation. Help with clinical decision-making was cited as an advantage of ML/AI. Participants noted the importance of considering ML/AI as an adjunct to clinical care, not as a replacement for clinicians. Over-reliance on recommendations, potentially leading to diminution of clinician skill, incorrect ML/AI recommendations, potential liability, and bias were cited as challenges. Clinician and patient education were noted as a potential burden. Median sentiment scores ranged from 0.0 (neutral) to 0.3 (positive). Sentiment varied with question type ( Discussion and Conclusion: Patients and clinicians expressed mixed views about ML/AI. Potential benefits related to improved decision-making and concerns focused on bias, liability, and the need for further education. Future work should address how best to incorporate ML/AI into education and obviate potential burdens as ML/AI is integrated into clinical care.