Table 3.

Characteristics of included studies and quality appraisal scores.

Author(s) and year	Nature/form of AI	Participant age	Participant’s profession	Sample size	Study design and method	MMAT score	Barriers	Facilitators
Blanco et al. (2018)26	CDSS	Not applicable (n.a.)	Nurses, physicians, pharmacists, radiology technicians and environmental services workers	34 (interviews); 13 (survey)	Qualitative semi-structured interviews and surveys	5	Sensitive systems induce alert fatigue
Catho et al. (2020)37	CDSS	n.a.	Physicians	29	Qualitative semi-structured interviews	5	Reduction in time spent with patients
Chow et al. (2015)44	CDSS	n.a.	Physicians	11 (focus group discussions); 265 (survey)	Mixed-methods focus groups and survey	4	Junior physicians were more likely to follow the systems recommendation than senior physicians
Tscholl et al. (2018)43	Monitoring System	35–44 years old	Physicians	128 (interviews); 38 (online survey)	Mixed-methods Interviews and survey	5	Lack of precision in the representation of the information	Visibility of information at a glance enables to interpret the patients‘ situation more quickly
Liberati et al. (2017)25	CDSS	n.a.	Physicians, nurses, managers, IT staff	30	Qualitative semi-structured interviews and surveys	5	Lack of understanding of functionalities
Elahi et al. (2020)46	Prognostic model	n.a.	Physicians	25 (questionnaires); 11 (interviews)	Mixed-methods survey and semi-structured interviews	5	Infeasibility of the system if dependent on a strong internet connection	objective assessment of patient risk and support difficult triage decisions, particularly in resource-limited settings
English et al. (2017)28	CDSS	25–61 years old	Pharmacists	25	Quantitative survey	4		Facilitating conditions influence clinical pharmacists’ use of the system
Fan et al. (2020)15	Medical diagnosis support system	Average age 40 years old	Healthcare professionals in the medical imaging department	191	Quantitative survey	4
Grau et al. (2019)27	CDSS	n.a.	Physicians	21	Qualitative semi-structured interviews	5	Sensitive systems induce alert fatigue
Hand et al. 201882	CDSS	n.a.	Physicians, nurses and allied health professionals	39	Quantitative survey	4		17/37 (45.9%) felt it would help improve clinician satisfaction 31/35 (88.6%) indicated that they were willing to always or often use the CDSS for fertility discussions
Hsiao et al. (2013)83	Pain management decision support systems	n.a.	Nurses	101	Quantitative survey	3		perceived ease of use and perceived usefulness account for 64% of the total explained variance in nurse anaesthetists’ acceptance of PM-DSS.
Jauk et al. (2021)32	CDSS	26–42 years old	Physicians and nurses	47 (questionnaires); 15 (expert group)	Mixed-methods interviews & survey	4	14.9% of participants did not believe that the application can be used to detect delirium at an early stage
Kanagasundaram et al. (2016)29	CDSS	n.a.	Physicians	24	Qualitative interviews	5	Alert fatigue System was cited to be an insult to knowledge Workflow interruption
Khong et al. (2015)34	CDSS	Average age junior nurses: 29,8 years old and average age senior nurses 45,5 years old	Nurses	14	Qualitative semi-structured interviews	5	Worry that too much trust in the system might lead to over-reliance and limit the development of clinical skills Participants doubted systems’ accuracy
Kitzmiller et al. (2019)41	Predictive analytics	n.a.	Physicians and nurses	22	Qualitative semi-structured interviews	5	Distal and inconvenient location was perceived to negatively affect routine engagement with the system
Horsfall et al. (2021)22	AI in surgery	31–61 years old or older	Physicians and nurses	100 for quantitative survey, 33 for qualitative	Mixed-methods survey	5		85% of participants strongly or somewhat agreed to the use of AI to enhance real-time alert of hazards or complications
Liang et al. (2019)35	Robots	30–36 years old	Nurses	23	Qualitative Semi-structured interviews	3	Fear of a loss of job	Perceived to be ideal for performing repetitive actions, routine tasks and assisting with precision treatment Robotics could also be a useful tool in multi-language communication with children and family caregivers from foreign countries, improving their understanding of the healthcare situation
Lin et al. (2021)84	AI in precision medicine	21–40 years old	Physicians and nurses	245 nurses and 40 physicians	Quantitative survey	4		The most dominant determinant for acceptance was perceived usefulness of the system
McBride et al. (2019)39	Robots	18 to over 50 years old	Physicians, nurses and support staff	164	Quantitative survey	4	Most participants had concerns about care and handling (p = 0.056) Nursing (52.6%) and medical staff (59.6%) were concerned that robotic-assisted surgery will add significant cost and financial pressure on the facility	Most of the nursing, medical and support staff agreed that theoretical, practical training, educational guides and staff support would facilitate the introduction of new technology in the workplace
Norton et al. (2015)52	CDSS	<39 to more than 50 years old	Physicians and nurses	32	Quantitative Survey	4	Nonsurgeons reported that the tool would make their job easier more so than surgeons
								Good educational training tool for residents
Oh et al. (2016)23	CDSS	n.a.	Physicians and pharmacists	98	Mixed-methods survey	4	Self-reported lower likelihood to change certain behaviours
O’Leary et al. (2014)31	Clinical pathway support system	n.a.	Physicians, nurses and physiotherapists	19	Mixed-methods Interviews and Surveys	4		Over half of the participants felt that clinical pathway support systems could help the reductions of errors
Omar et al. (2017)38	CDSS	n.a.	Physicians	n.a.	Qualitative Semi-structured interviews	1	Some junior nurses preferred to seek advice from senior nurses rather than AI
Esmaeilzadeh et al. (2015)85	CDSS	n.a.	Physicians	335	Quantitative survey	4	Significant relationship between perceived threat to professional autonomy and intention to use CDSS (β = −0.392, p-value = 0.000)
Petitgand et al. (2020)21	CDSS	n.a.	Physicians	20	Qualitative semi-structured Interviews	5	Systems may favour errors
Sandhu et al. (2020)45	Machine learning	n.a.	Physicians and nurses	15	Qualitative Semi-structured Interviews	5	Unfamiliarity with the system resulted in confusion and misunderstanding	Most useful for residents still developing clinical skills or low-resource community settings
Schulte et al. (2020)50	Automatic speech recognition	Mean age of 41.8 ± 9.8 years	Physicians	185	Quantitative Survey	4		Voice recognizer without headset
Stifter et al. (2018)51	CDSS	21–71 years old	Nurses	60	Quantitative Survey	4	Higher acceptability among participants with less than one year of experience than those with 1 or more years of experience
Walter et al. (2020)53	Automated pain recognition	Mean age of 40.31 years ± 11.5	Physicians and nurses	102	Quantitative Survey	5		Pain detection accuracy of > 80%
Yurdaisik and Aksoy (2021)30	AI	n.a.	Physicians, technicians and medical students	204	Quantitative Survey	4	Only 5.3% of participants stated that they will assume the legal responsibility of imaging results	Among the participants, 51.9% think that AI applications will save time for radiologists
Zheng et al. (2021)33	AI in ophtalmology	Less than 25 to older than 45 years old	Physicians and technicians	562	Quantitative Survey	4	56.4% said that in the current ophthalmic AI practice, medical responsibilities are unclear
Aljarboa et al. (2019)18	CDSS	25–51 years old	Physicians	12	Qualitative Semi-structured interviews	5		Alerts direct attention to important issues
Jones et al. (2022)54	CDSS	29–62 years old	Physicians and nurses	33	Qualitative Interviews	5	Sensitive systems induce alert fatigue
Panicker and Sabu (2020)36	Computer-assisted medical diagnosis system	27–58 years old	Physicians and technicians	18	Qualitative Interviews	5	Participants doubted systems’ accuracy
So et al. (2021)42	AI	25 years old to 55 or older	Physicians, nurses, pharmacists, physiotherapists and technicians	96	Quantitative Survey	5		Working experience significantly favoured use of AI
Strohm et al. (2020)86	AI in radiology	n.a.	Physicians	25	Qualitative Semi-structured interviews	5	Unresolved question of legal responsibility for damage occurred due to e.g. false negatives and false positives resulting from an AI-generated diagnosis
Pumplun et al. (2021)49	Machine learning	n.a.	Physicians, professionals in administrative roles	22	Qualitative Interviews	5	Lack of transparency Limited resources Uncertainties in governmental regulations, strict requirements for the protection of sensitive patient data, and existing medical ethics
Prakash and Das (2021)19	CDSS	82% younger than 40 years old	Physicians	n.a.	Mixed-methods interviews and surveys	5	Lack of understanding of functionalities
Zhai et al. (2021)87	AI	18 to more than 50 years old	Physicians and medical students	307	Mixed-methods Survey	5
Aljarboa and Miah (2021)24	CDSS	25–51 years old	Physicians	54	Qualitative interviews	5		Importance of privacy and security factors as confidentiality and privacy of patient data is essential for use
Nydert et al. (2017)20	CDSS	n.a.	Physicians	17	Qualitative interviews	5	Risk of overreliance on the system; double-check of recommended dosage is needed	Greatest benefit within emergency care
Alumran et al. (2020)47	Electronic triage and acuity scale	n.a.	Nurses	71	Quantitative survey	5		The years of nurse’s experience influenced their usage of the E-CTAS. There was a positive correlation between years of experience likelihood to become an E-CTAS user