What Is AI?

Starr et al. cited (https://doi.org/10.1188/23.CJON.595-601) AI’s broad definition as the use of “machines that mimic human intelligence and human cognitive functions like problem-solving and learning.” They explained that it has several branches and subsets:

• Shared science: Engages disciplines that include computer science, mathematics, psychology, linguistics, philosophy, biology, and neuroscience.
• Artificial intelligence: See the broad definition cited above; an example in health care is patient deterioration prediction models.
• Machine learning: AI that is trained and retrained by data and human experts to use dataset patterns to reach a conclusion.
• Deep learning: Machine learning that engages several layers of neural networks to operate. Examples include large language models like ChatGPT and Google Bard.

“AI has been increasingly interwoven throughout daily life,” Starr et al. said (https://doi.org/10.1188/23.CJON.595-601). They gave examples such as digital voice assistants (e.g., Apple’s Siri, Amazon’s Alexa), personalized recommendation algorithms (e.g., Amazon, Google), predictive text in emails, chatbots, and self-driving vehicles.

How Is AI Used in Oncology?

Most widely used in cancer screenings, AI is assisting radiologists to analyze radiographic images to detect, diagnose, and monitor cancer as well as dermatologists in identifying early-stage melanoma lesions, Starr et al. reported (https://doi.org/10.1188/23.CJON.595-601). “An emerging strategy for personalized early cancer detection care is pan-cancer screening, which uses AI to analyze whole blood samples and identify circulating tumor DNA. This method may be cost-effective because of the potential to identify rare cancers sooner,” they added.

In clinical practice, AI can help identify patients who are eligible for clinical trials and “enhance clinician decision-making support tools by analyzing electronic health record data to predict future acute events and prompt clinicians to act,” Starr et al. said (https://doi.org/10.1188/23.CJON.595-601).

What Does AI Mean for Oncology Nursing Practice?

From clinical decision support models to best practice alerts, mobile health and sensor-based technologies like remote telemetry monitoring, voice assistants, and workflow optimization models like iQueue, nurses are using AI tools every day.

“Combined with evidence-based clinical nursing knowledge, AI’s efficiency and power has the potential to further elevate nursing practice and improve patient outcomes,” Starr et al. said (https://doi.org/10.1188/23.CJON.595-601). They added that “the application of AI may reduce labor-intensive nursing workloads and provide technology to better support nurses who have been affected by the nursing shortage.”

In clinical practice, Starr et al. explained (https://doi.org/10.1188/23.CJON.595-601) that to produce accurate and reliable predictions and information, AI requires accurate data and documentation. Nurses must accurately document and enter data in electronic health records and other dataset tools that AI systems use for learning.

Nurses must also be aware of AI’s ethical considerations, Starr et al. said (https://doi.org/10.1188/23.CJON.595-601), including the potential for bias and the absence of full regulation. “Because AI learns from training data generated by humans with implicit and explicit biases, unintentional bias can occur, exacerbating social inequities,” they explained. “Creating a diverse development team can help address this bias.” They encouraged nurses to use their unique understanding of patients to ensure that the AI training data represent the tool’s patient population.

How Can Oncology Nurses Contribute to AI?

By identifying clinical issues suited for AI, recommending documentation for training data, and providing user feedback, nurses have a role in all stages of AI development. Nurses’ feedback about AI’s results can be used to retrain AI models and improve their future accuracy, Starr et al. said (https://doi.org/10.1188/23.CJON.595-601). Nurses can also advocate for keeping patients front and center when considering AI tools to improve outcomes and workflows and for federal, state, and local regulations for responsible AI development and use.

All of that requires a foundational understanding and awareness of AI, and nurses have a responsibility to their profession and  patients to engage in continued learning about AI in health care. Starr et al. provided a list of AI resources and continuing education websites for nurses in their article (https://doi.org/10.1188/23.CJON.595-601). Nurses can also listen to conversations with two experts in nursing and cancer AI on the Oncology Nursing Podcast:

• Episode 284: How AI Is Influencing Cancer Care and Oncology Nursing (https://www.ons.org/podcasts/episode-284-how-ai-influencing-cancer-care-and-oncology-nursing)
• Episode 281: Nursing’s Role in AI in Health Care (https://www.ons.org/podcasts/episode-281-nursings-role-ai-health-care)

Ultimately, AI is not a solution for every problem, but it may be used more frequently as technology advances and continue to affect nursing practice, Starr et al. concluded (https://doi.org/10.1188/23.CJON.595-601). “Applying nursing clinical knowledge and critical thinking skills throughout the AI life cycle can elevate nursing practice and aid in the development of products that will meaningfully improve patient outcomes.”