The “DeepSeek effect” and the adoption–integration gap of generative artificial intelligence in clinical practice: a national online convenience cross-sectional survey of academic critical care physicians in China - Scorecard - MDSpire
Advertisement
The “DeepSeek effect” and the adoption–integration gap of generative artificial intelligence in clinical practice: a national online convenience cross-sectional survey of academic critical care physicians in China
Clinical Scorecard: The Impact of the 'DeepSeek Effect' on the Integration of Generative Artificial Intelligence in Clinical Settings: A National Online Cross-Sectional Survey of Academic Critical Care Physicians in China
At a Glance
Category
Detail
Condition
Generative Artificial Intelligence in Clinical Settings
Key Mechanisms
Integration of GAI in critical care medicine for improved diagnosis and treatment.
Target Population
Academic critical care physicians in China.
Care Setting
Tertiary hospitals
Key Highlights
Self-reported GAI utilization increased from 64.7% to 94.1% post-DeepSeek.
Formal GAI training participation remained low at around 13.2%.
Structured training linked to improved self-reported professional competence (AOR = 22.2).
Physicians prioritized critical integration skills over basic technical proficiency.
Guideline-Based Recommendations
Diagnosis
Utilize GAI for precise diagnoses in critical care settings.
Management
Develop structured training frameworks for effective GAI integration.
Monitoring & Follow-up
Assess the impact of GAI on clinical decision-making and patient outcomes.
Risks
Address ethical concerns regarding dependence on GAI and patient data confidentiality.
Patient & Prescribing Data
Patients in critical care settings managed by academic physicians.
GAI can enhance personalized treatment strategies in critical care.
Clinical Best Practices
Encourage critical evaluation of GAI outputs among physicians.
Implement institutional governance for GAI use in clinical practice.
Heart rate monitoring and atrial fibrillation detection had the strongest supporting evidence, but investigators found limited evidence for broader outpatient self-monitoring applications.
