Cluster,Subcluster,Statement ID,Statement,Importance,Feasibility
1,1.1,1,Human oversight during implementation in early days,4.25,3.42
1,1.1,16,Proactive preventative care,3.17,3.0
1,1.1,19,Developing a rigorous privacy and quality assurance framework,3.5,3.42
1,1.1,21,Helps with generation of ideas can act as a sounding board,2.83,3.58
1,1.1,29,Data governance policies and regulations to enhance data quality and accountability,2.92,2.58
1,1.1,32,Regulation framework and a good understanding of the medico-legal implications,2.83,2.25
1,1.1,37,Consultation recording and transcription,3.17,4.18
1,1.1,38,Ensuring cybersecurity of AI tools,3.75,3.17
1,1.1,39,Educating end-users about properly navigating AI,3.83,3.58
1,1.1,43,Ensuring accessibility and translation,3.5,3.67
1,1.1,45,Embed research / quality improvement into all areas,2.92,3.25
1,1.1,68,Phased roll-out,3.17,4.08
1,1.1,88,Ensuring interpretability of AI tools,3.42,3.33
1,1.2,3,Concerns with patient acceptability,4.08,3.17
1,1.2,5,False confidence,3.33,2.73
1,1.2,7,AI could perpetuate existing biases,3.33,3.25
1,1.2,10,Concerns with digital literacy and digital divide barriers,3.08,2.75
1,1.2,20,AI could be misused,3.17,2.75
1,1.2,28,Concerns about personalizing care,2.83,3.25
1,1.2,33,"AI could lead to excessive human delegation, assuming that AI will take care of it!",2.58,2.5
1,1.2,35,AI could be of concern regarding its ongoing validation,3.0,3.09
1,1.2,47,Concerns with attrition of skills in healthcare providers,2.58,2.58
1,1.2,49,Loss of human connection/interactions,3.33,2.75
1,1.2,52,AI could engender issues with trust,2.75,2.67
1,1.2,55,AI hallucinations,3.17,2.83
1,1.2,60,"Concerns with standards for quality, accuracy, etc.",3.08,2.92
1,1.2,63,Distress is nuanced and often detected in the unsaid of human communication and in a trusting therapeutic relationship,3.42,2.5
1,1.2,65,Concerns with difficulty to know older data for training and missing new breakthrough information,2.92,2.92
1,1.2,67,Racism bias through research (history of medical research),3.08,2.5
1,1.2,71,Over reliance by physicians and reduced problem solving skills,2.67,2.75
1,1.2,77,Concerns with going too far and can't come back,2.33,2.5
1,1.2,82,AI could make mistakes,3.33,3.0
1,1.2,83,"Concerns with stability of the system (during virus attack, power outage, system interruption)",2.75,2.42
1,1.2,87,Can existing healthcare servers support the use of AI?,2.17,3.08
1,1.2,93,Minority groups being marginalized,3.0,2.92
1,1.3,2,Concerns with confidentiality,3.75,3.25
1,1.3,13,Too much bureaucracy slowing progress,3.0,3.0
1,1.3,14,Concerns with liability,3.33,2.83
1,1.3,26,Concerns with job security,2.42,2.25
1,1.3,36,AI could be of concern to patient privacy,3.08,3.08
1,1.3,40,Concerns with protection of personal health identifiers,3.42,3.0
1,1.3,54,Concerns with responsibility accountability of AI recommendations,3.17,3.0
1,1.3,61,Concerns with power outages and down time procedures,2.17,2.58
1,1.3,66,Concerns with consent of patient,3.25,3.17
1,1.3,74,"Concerns with maintaining up to date information, sources and programming",2.83,3.42
1,1.3,75,Data privacy concerns,3.33,2.83
1,1.3,78,Control of data by corporations,2.75,2.5
1,1.3,81,Concerns with transparency,3.08,2.75
1,1.3,86,Lack of training/knowledge for those using the AI tool in the real world,3.17,3.25
1,1.3,91,Concerns that data is held by private companies,2.92,2.42
1,1.3,95,Data ownership and standards and guidelines on AI developed are not established,2.83,2.92
1,1.3,96,Possibility of data breaches / data leaks,3.0,2.83
1,1.3,99,Cybersecurity concerns,2.75,2.42
1,1.4,9,Ensuring accountability and oversight of AI,4.0,3.58
1,1.4,12,Diversification of AI training data and sources,3.33,3.33
1,1.4,25,Establishing proper consent,3.0,3.08
1,1.4,46,"Data that the AI model is trained on must be good, clean, large and diverse",3.58,2.83
1,1.4,48,Preservation of human touch,3.08,2.83
1,1.4,56,Multidisciplinary collaboration and training in developing AI models,3.17,3.25
1,1.4,73,Ensuring clinical validation of AI tools,3.5,3.42
1,1.4,85,"Establishing guidelines for ""best practices"" for training AI + clinical validation",3.5,3.17
1,1.4,94,Patient/parent/caregiver/provider acceptability,3.92,2.92
2,2.1,6,AI could provide alternate perspectives,3.17,3.17
2,2.1,11,"Better detection (of moles, cancer lumps)",3.58,3.75
2,2.1,15,AI could decrease false positives,3.42,3.42
2,2.1,24,Benefits in early distress screening,3.33,3.25
2,2.1,27,Benefits with patient access to accurate information,3.25,3.55
2,2.1,34,Knowledge repositing for learning and research,2.92,3.5
2,2.1,41,Facilitation of clinical research,3.08,3.42
2,2.1,42,Improve access to services,3.67,3.58
2,2.1,44,AI could help improve accuracy,3.58,3.75
2,2.1,51,Collate large databases,3.25,4.17
2,2.1,58,Noticing patterns and new things humans haven't noticed,3.58,3.5
2,2.1,62,Increased opportunities for innovation,2.92,3.58
2,2.1,97,"Potential for widening differential diagnosis, widening perspective of patient",3.0,3.33
2,2.2,4,Efficiency improved for documentation,3.58,4.17
2,2.2,18,AI could free up time for pt care,3.75,3.67
2,2.2,23,AI allows access vast information very quickly to help diagnosis,3.75,4.0
2,2.2,30,"Savings in time, can integrate different information and improve in treatment especially when there is less number of human specialist",3.83,3.67
2,2.2,31,AI could remove human bias,3.17,2.75
2,2.2,57,"24/7 availability, AI doesn't get tired or brain fog",3.42,4.25
2,2.2,59,Prediction/prognosis/treatment recommendations,3.92,3.58
2,2.2,70,Potential in navigation and directing to self-management resources,3.42,3.67
2,2.2,72,Ability to make text/conversation more digestable for patients (less technical),3.5,3.83
2,2.2,76,Triage diagnostic results for clinicians,3.67,3.67
2,2.2,79,AI could help address the human health resources issue,3.58,3.25
2,2.2,80,Faster assessments to inform decision-making,3.5,3.58
2,2.2,84,Live translation and/or transcription of conversations,3.73,4.08
2,2.2,89,Allow clinicians to focus on tasks requiring their expertise,4.17,4.0
2,2.2,98,Lists possible outcome and risks for the patient,3.83,3.58
2,2.2,100,Automation of routine tasks,4.0,4.08
2,2.3,8,Minimizing bias and increasing performance and accuracy,4.0,3.33
2,2.3,17,Benefits in equitable access,3.33,2.92
2,2.3,53,Pooling large volume of data can improve accuracy of AI system and allow for more timely diagnosis and treatment recommendation,3.58,3.75
2,2.3,64,Personalization of AI support tools for patients,3.58,3.33
2,2.3,69,Reduce admin burden,3.92,4.17
2,2.3,90,"Identifying and focusing on high-priority areas, such as prevention",3.08,3.42
2,2.4,22,Benefits with accessibility & communication (translating languages lay terms),3.25,3.58
2,2.4,50,Increases digestibility/highlighting key points,3.67,4.33
2,2.4,92,Maintaining competency with changing best practices and approved standards,3.17,2.83