Table 1.

Summary of study characteristics: design and population.

Ref. No.	Ref., Year	Design; level of evidencea	Clinical setting	Users of AIb	Stage of clinical education	Stage of clinical use	Total, n (% male)	Age (years), race or ethnicity (%)	Study duration or follow-up
1	Bien, 2018 [23]	Modeling and evaluation; 2bc	Large academic hospital; imaging department	Orthopedic surgeons; general radiologists	Practicing physicians	Implementation	N/Rd (N/R)	N/R (N/R)	N/R
2	Hirsch, 2015 [22]	Evaluation; 4e	Large private hospital; large academic medical center; nephrology and internal medicine departments	Internal medicine physicians; nephrologists	Graduate medical education (internal medicine residents and interns; nephrology fellows)	Implementation	12 (N/R)	N/R (N/R)	~9 months
3	Jordan, 2010 [21]	Evaluation; 4	Large academic hospital; cardiothoracic intensive care department	Intensive care unit nurses	Practicing nurses	Implementation	N/R (N/R)	N/R (N/R)	N/R
4	Sayres, 2019 [20]	Experimental 3-arm observational study; 2b	Large academic hospitals, large health systems, and specialist office; ophthalmology department	Ophthalmologists	Practicing physicians	Implementation	10 (N/R)	N/R (N/R)	N/R

^aAdapted from Oxford Levels of Evidence [19].

^bAI: artificial intelligence.

^dLevel 2b: individual cohort, modeling, or observational studies.

^cN/R: not reported.

^eLevel 4: case series or poor-quality cohort studies.