Table 2.
Summary of study characteristics: clinical competency and performance assessment.
Ref. No. | Ref., Year | Professional education domains of competence | Description (implied or explicit) of competency | User-AIa interface training and description | Performance assessment |
1 | Bien, 2018 [23] |
|
Implied in methods; improve image interpretation | Training N/Rb; interface not described | Metric N/Pc; evaluate if AI assistance improves expert performance in reading MRId images |
2 | Hirsch, 2015 [22] |
|
Implied in methods; improve summarization of longitudinal patient record and information processing in preparation for new patients | Training N/R; authenticated user queries the database for a patient and is provided with a visual summary of content containing all visit, note, and problem information | Questionnaire; evaluate time and efficiency in information processing for patient care |
3 | Jordan, 2010 [21] |
|
Implied in methods; improve handovers in peri-operative patient care by reducing communication and informational errors | Training N/R; patient summarization and visualization tool are used as an overlay to the existing electronic patient record | Questionnaire; evaluate if AI-based tool performs better than physicians to provide clinical information and patient status in ICUe handovers |
4 | Sayres, 2019 [20] |
|
Implied in methods; improve reader sensitivity and increase specificity of fundal images | Readers were provided training and similar instructions for use; interface not described | Metric N/P; evaluate if AI assistance increases severity grades in model predictions by assessing sensitivity and specificity of reader |
aAI: artificial intelligence.
bN/R: not reported.
cN/P: not provided.
dMRI: magnetic resonance imaging.
eICU: intensive care unit.