Implementing artificial intelligence in clinical practice: a mixed-method study of barriers and facilitators

Schouten, Bo; Schinkel, Michiel; Boerman, Anneroos W.; Pijkeren, Petra van; Thodé, Maureen; Beneden, Marlou van; Panday, Rishi; Jonge, Robert de; Wiersinga, W. Joost; Nanayakkara, Prabath W.B.

doi:10.21037/jmai-22-71

Cited by 5 publications

(3 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, there remains a scarcity of tools that directly enhance clinical practice by offering decision-making support for diagnostic or treatment procedures, especially in acute care. The notable disparity between the development and implementation of clinical AI tools, especially compared to other sectors, can be attributed to various factors, including an inadequate involvement of endusers (e.g., physicians) during development of clinically relevant algorithms, performance degradation over time and in new settings, insufficient evidence of clinical benefits, and a limited drive to change existing practices among healthcare professionals 4 . In the context of acute medicine, the diverse nature of care and the variability in data collection during the acute phase poses additional challenges in developing AI tools to support the diagnostic workup and treatment 5 .…”

Section: Historical Perspective and Current Landscapementioning

confidence: 99%

“…While the potential of AI in acute care is promising, its translation to clinical practice faces many barriers 4 . Although various sectors, such as travel and banking, have embraced the use of AI, its integration into the healthcare sector has been comparatively slower.…”

Section: The Negative: Challenges and Considerationsmentioning

confidence: 99%

“…Ensuring that AI tools are trained on diverse and representative datasets, accounting for various demographic groups and clinical scenarios, is essential. Transparency and accountability are also integral to ethical AI deployment 4 . Healthcare professionals, patients, and other stakeholders should have access to clear and understandable explanations of AI algorithms and their decision-making processes.…”

Section: The Negative: Challenges and Considerationsmentioning

confidence: 99%

See 2 more Smart Citations

Artificial Intelligence: its Future and Impact on Acute Medicine

Schinkel,

Paranjape,

Bhagirath

et al. 2023

Acute Med

View full text Add to dashboard Cite

This commentary explores the potential impact of artificial intelligence (AI) in acute medicine, considering its possibilities and challenges. With its ability to simulate human intelligence, AI holds the promise for supporting timely decision-making and interventions in acute care. While AI has significantly contributed to improvements in various sectors, its implementation in healthcare remains limited. The development of AI tools tailored to acute medicine can improve clinical decision-making, and AI’s role in streamlining administrative tasks, exemplified by ChatGPT, may offer immediate benefits. However, challenges include uniform data collection, privacy, bias, and preserving the doctor-patient relationship. Collaboration among AI researchers, healthcare professionals, and policymakers is crucial to harness the potential of AI in acute medicine and create a future where advanced technologies synergistically enhance human expertise.

show abstract

Section: Historical Perspective and Current Landscapementioning

confidence: 99%

Section: The Negative: Challenges and Considerationsmentioning

confidence: 99%

Section: The Negative: Challenges and Considerationsmentioning

confidence: 99%

See 1 more Smart Citation

Artificial Intelligence: its Future and Impact on Acute Medicine

Schinkel,

Paranjape,

Bhagirath

et al. 2023

Acute Med

View full text Add to dashboard Cite

show abstract

A framework for implementing machine learning in healthcare based on the concepts of preconditions and postconditions

MacKay

Klement

Vanberkel

et al. 2023

Healthcare Analytics

View full text Add to dashboard Cite

Personalising intravenous to oral antibiotic switch decision making through fair interpretable machine learning

Bolton,

Wilson,

Gilchrist

et al. 2024

Nat Commun

View full text Add to dashboard Cite

Antimicrobial resistance (AMR) and healthcare associated infections pose a significant threat globally. One key prevention strategy is to follow antimicrobial stewardship practices, in particular, to maximise targeted oral therapy and reduce the use of indwelling vascular devices for intravenous (IV) administration. Appreciating when an individual patient can switch from IV to oral antibiotic treatment is often non-trivial and not standardised. To tackle this problem we created a machine learning model to predict when a patient could switch based on routinely collected clinical parameters. 10,362 unique intensive care unit stays were extracted and two informative feature sets identified. Our best model achieved a mean AUROC of 0.80 (SD 0.01) on the hold-out set while not being biased to individuals protected characteristics. Interpretability methodologies were employed to create clinically useful visual explanations. In summary, our model provides individualised, fair, and interpretable predictions for when a patient could switch from IV-to-oral antibiotic treatment. Prospectively evaluation of safety and efficacy is needed before such technology can be applied clinically.

show abstract

Implementing artificial intelligence in clinical practice: a mixed-method study of barriers and facilitators

Cited by 5 publications

References 0 publications

Artificial Intelligence: its Future and Impact on Acute Medicine

Artificial Intelligence: its Future and Impact on Acute Medicine

A framework for implementing machine learning in healthcare based on the concepts of preconditions and postconditions

Personalising intravenous to oral antibiotic switch decision making through fair interpretable machine learning

Contact Info

Product

Resources

About