Sage Journals: Discover world-class research

Abstract

With artificial intelligence (AI) on the rise in healthcare, we investigated the acceptance and adoption of AI among cancer survivors who are seeking information. We searched 3 databases—PubMed, Scopus, and CINAHL for original research studies examining cancer survivors’ acceptance and adoption of AI while seeking information. We screened 858 articles, resulting in the inclusion of 12 studies. Cancer survivors’ perceived benefits of AI include diagnostic accuracy, improved healthcare access, and reduced healthcare costs. Concerns included the loss of human interaction, data loss, and privacy issues. Adoption and acceptance of AI-based Chatbots and cancer trial search tools were also demonstrated. Regarding the ethical use of AI, cancer survivors shared a desire for data de-identification, disclosure, and transparency. Cancer survivors accepted AI as an adjunct to physicians. Patients demonstrated acceptance and adoption of both AI-based educational tools and search engines. Developing AI models with better explainability and education can improve cancer survivors’ acceptance and adoption of AI.

Keywords

artificial intelligence cancer patient experience trust technology patient satisfaction clinician–patient relationship

Introduction

Though artificial intelligence (AI) use has been explored in healthcare since the 1970s, it has seen a rapid increase over the past decade.¹ AI offers opportunities to improve prognosis, reduce costs, and improve healthcare for clinicians, allied health personnel, and patients. AI is a multidisciplinary field that consists of subfields such as machine learning (ML), deep learning, natural language processing, and more.² As the current healthcare system moves toward patient-centric care, patients’ perspectives and involvement, including cancer survivors, become more critical in AI tool development and integration in healthcare.^3,4

“An individual is considered a cancer survivor from the time of diagnosis through the balance of life,” based on the National Cancer Institute's definition.⁵ Cancer survivorship brings multiple challenges, including lack of ongoing, personalized care and living in constant fear and uncertainty. These fears are heightened among those who are in a linguistic, ethnic, and racial minority group.^6,7

AI applications have demonstrated positive outcomes for mitigating these challenges for cancer survivors with AI-driven digital health technologies, including increased motivation of patient-reported outcomes, reduced fatigue and pain levels, improved quality of life (QOL), and improved physical function.⁸ Further, an AI-based platform, ASCAPE (Artificial Intelligence Supporting Cancer Patients across Europe) has utilized big data and ML to predict and support the health status of patients with breast and prostate cancer.⁹

Other key challenges cancer survivors face include difficulty contacting healthcare teams, receiving timely replies, and information overload from online search engines.¹⁰ With its ability to analyze and summarize large amounts of information concisely, generative AI can become an important tool to bridge the knowledge gap, provide support, and answer healthcare-related questions for cancer survivors. AI can also potentially improve cancer survivors’ clinical care along with screening and rediagnosis of cancer after the initial cure.

The degree of AI acceptance and adoption among cancer survivors is unknown. As with any emerging technology, AI's impact on cancer survivors’ care and experience will depend on their acceptance and adoption of AI tools, including those used by their healthcare team. Technology acceptance is the first step toward its adoption and reflects the attitude toward technology.¹¹ Research examining cancer survivors’ perceptions and adoption of AI tools is limited. While the technology acceptance model suggests that attitudes predict intentions, which influences actual technology adoption, this study focused on attitudes and intentions rather than the explicit behavioral changes made in the later stages of adoption.¹²

The objective of this study is to evaluate the acceptance and adoption of (AI) among cancer survivors seeking information. We hypothesize that there will be variations in the acceptance and adoption of AI among cancer survivors, which will be influenced by factors such as age and digital literacy. We also hypothesized that cancer survivors’ acceptance and adoption of AI tools would improve patient-provider communication and QOL factors.

Methods

We created search terms to explore all existing literature to further our understanding of cancer survivors’ acceptance and adoption of AI. Search terms were derived from extant literature on each topic and are included in the appendix. We systematically searched 3 databases—PubMed, Scopus, and CINAHL Plus with Full Text (EBSCOhost)—on June 17, 2024, for original literature on this topic without language or geography limitations. These databases were chosen due to the availability of research relevant to cancer survivorship and AI acceptability and usability. The complete search strategy for all databases is included in Figure 1. Given that AI is a relatively new tool to the public, we chose to search for articles published within the past 10 years. A single reviewer performed title and abstract screening, followed by a detailed full-text review of articles based on predefined eligibility criteria.

Inclusion criteria were studies investigating individuals 18 and 99 years of age; studies investigating cancer survivors; studies investigating cancer survivors’ acceptance of AI or adoption of AI for seeking information; and original research studies.

Figure 1.

Patient acceptance of artificial intelligence (AI)/adoption of AI for seeking information—cancer survivors.

Our exclusion criteria were wrong patient population: children (under the age of 18), pregnant women, nonviable neonates or neonates of uncertain viability, prisoners, others involuntarily detained or incarcerated (including those helped in treatment centers as a condition of their parole); wrong setting: studies with only provider's acceptance and adoption of AI; wrong intervention and wrong patient population: studies that did not focus on patients with cancer; wrong study design: studies published as conference proceedings, systematic reviews, and scoping reviews; and wrong outcomes: Studies assessed AI tools but did not assess acceptance, adoption, usability, perceptions, and/or engagement.

We included a few studies with both provider and cancer survivor participation due to the limited literature on cancer survivors’ perspectives only.

Following the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines, we conducted a review that culminated in the selection of 12 articles deemed most relevant to our study objectives.¹³ The PRISMA framework is widely recognized for its rigorous and reliable standard of guidelines for conducting empirical literature reviews.

For the analysis, we employed a hybrid approach that combined inductive and deductive reasoning. Inductive reasoning allowed us to identify recurring patterns and themes that emerged organically from the data. These included factors such as trust in AI, perceived usability of AI tools, and the role of personalized AI in enhancing patient engagement. Simultaneously, deductive reasoning was applied to test pre-established concepts and frameworks. This approach helped us situate our findings within broader theoretical paradigms, providing a structured lens through which to interpret the data. This process also enabled a comprehensive synthesis of findings, revealing both commonalities across the studies and unique perspectives influenced by demographic and digital differences among participants. By combining inductive and deductive reasoning, we bridged empirical observations with theoretical constructs, offering a nuanced understanding of how cancer survivors perceive and engage with AI technologies in their healthcare experiences. This approach provided a balanced and in-depth analysis that captured the complexity of patient interactions with AI.

Results

Our search returned 858 references, and after removing duplicates, we screened 581 references and identified 12 studies for inclusion in this literature review. This was completed using a strict inclusion and exclusion criteria list agreed upon by the authors. The first author applied inductive and deductive reasoning to exclude article abstracts according to the inclusion and exclusion criteria list created. Screening was completed using the systematic review tool Covidence and the automation tools within to aid in abstract and article screening.¹⁴ A PRISMA diagram related to the selection process can be found in Figure 1.¹⁵ Key information about the studies is included in Table 1. Of these 12 included studies, 5 were from the United States, 2 from England, 2 from Germany, and 1 from Canada, Norway, and China. All studies were published from 2018 onwards and included cancer survivors of skin cancer, meningioma, prostate cancer, breast cancer, gastrointestinal cancer, and metastatic cancer. Six studies employed patient questionnaires or surveys, 3 employed interviews, 2 were prospective trials, and 1 held focus groups. Four studies included patients without cancer^16–19 and 2 included providers and caregivers.^19,20

Table 1.

Summary of Key Characteristics, Demographics, and Primary Results of the Studies Included in the Literature Review.

Sl no	Author	Title	Country/duration	Cancer	Methodology	Sample size	Demography/prior digital experience	Results
1	Lim et al ¹²	Perceptions of the use of artificial intelligence in the diagnosis of skin cancer: an outpatient survey	England /03/2018 - 08/2018	Skin cancer	Cross-section questionnaire (Quantitative study)	603	18 to 100 years with majority between 51 and 70 years 25% of respondents had used some form of digital technology	1. Confirmation of AI-assisted diagnosis along with discussion was important for the majority of participants (81%) 2. Use of AI technology to aid cancer diagnosis did not concern about half (47%) of the participants
2	Nelson et al ¹⁵	Patient perspectives on the use of artificial intelligence for skin cancer screening: a qualitative study	USA /05/2019 - 07/2019	Skin cancer	Semistructured interviews. Half of the patients were interviewed about direct to patient tool and half for clinician decision support AI tools (Qualitative study)	48	Mean age—53.3, 54% participants were women 90% of patients reported use of digital services for health	1. Perceived benefits of AI included increased diagnostic accuracy (60%) and healthcare access (60%) 2. Perceived risk of AI included increased patient anxiety (40%), loss of human interaction (38%), loss of patient privacy (29%), patient loss to follow up (29%)
3	Jutzi et al ¹³	Artificial intelligence in skin cancer diagnostics: the patients’ perspective	Germany /11/2019 - 01/2020	Skin cancer (melanoma)	Survey (Quantitative study)	298	34.2% of participants between 31 and 45 years and 41.3% between 46 and 60 years, 75% females 88% of participants had heard about AI	1. AI use in medical approaches was supported by nearly 94% [95% CI = 0.91–0.97] of respondents 2. AI's use as assistance system for physicians was supported by 94% [95% CI = 0.92–0.97] participants 3. Patient concerns about AI included insufficient data protection, impersonality and susceptibility to errors
4	Lysø et al ¹⁴	Men's sociotechnical imaginaries of artificial intelligence for prostate cancer diagnostics—a focus group study	Norway /11/2021–12/2021	Prostate cancer	Focus groups with reflexive thematic analysis (Qualitative study)	48	54 and 85 years, all men Text-based vignette was used in 6 focus groups defining AI to stimulate conversation. 4 focus groups did not need it for discussions on AI	1. AI is expected to improve cancer diagnoses accuracy and support more personalized medicine 2. AI is expected to threaten patient and healthcare professionals’ interpersonal and communicational relationships along with trust
5	Rodler et al ¹⁸	Patients’ trust in artificial intelligence-based decision-making for localized prostate cancer: results from a prospective trial	Germany/not specified	Prostate cancer	Prospective trial with survey (Quantitative study)	466	61 and 75 years Majority of participants had experience with digital technology (84.8%), with a mean cumulative affinity for technology being 2.29 (SD = 1.19)	1. Cumulative affinity for technology was positively correlated with trust in AI (correlation coefficient 0.094; p = .04. 2. Patients had more trust in physician capability than AI for responding in an individual's way while communicating a diagnosis (4.51 vs 3.38, p < 0.001) and for explaining information in an understandable way (4.57 vs 3.18, p < .001) 3. AI controlled by a physician had higher patient trust compared to AI not controlled by a physician (4.31 vs 1.75, p < 0.001) 4. Patient preferred AI-assisted physicians (66.74%) compared with physicians alone (29.61%), physicians controlled by AI (2.36%), AI alone (0.64%)
6	Temple et al ¹⁹	Patient views on the implementation of artificial intelligence in radiotherapy	England /11/2021–03/2022	Not specified	Patient questionnaire (Quantitative study)	95	Demographic correlation mentioned No correlation was found between smartphone use and 5 questionnaire factors	1. Patients, overall, had a moderately negative view regarding the use of AI in radiotherapy 2. Patients want significant personal interaction with healthcare professionals during treatment
7	Hildebrand et al ¹⁷	Study of patient and physician attitudes toward automated prognostic models for patients with metastatic cancer	USA /07/2021-02/2022	Metastatic cancer	Semistructured interviews (Qualitative study)	30	50 and 89 years, close to equal male and female patients	1. Participants wanted option for the AI model to be used in patient care 2. Patients would want to be informed regarding use of AI model in their care and trust results of AI model if presented by their physician
8	Yang et al ²⁰	Attitudes of Chinese cancer patients toward the clinical use of artificial intelligence	China/not specified	Multiple cancers	Patient questionnaire (Quantitative study)	402	Mean age—47.86 years with 50.5% of females Only 6% were familiar with AI in medicine	1. Most cancer patients trusted AI for diagnosis and treatment 2. Participants who had heard of AI were more likely to trust it during diagnosis
9	Al-Hill et al ²¹	A randomized trial comparing the effectiveness of pretest genetic counseling using an artificial intelligence automated Chatbot and traditional in-person genetic counseling in women newly diagnosed with breast cancer	USA /07/2021 - 03/2023	Breast cancer	Prospective randomized controlled trial with questionnaire (Quantitative study)	37	48 and 82 years with a median range of 62 years, all females	1. Median patient knowledge score between Chatbot and traditional counseling (11 vs 12, p = .09) or median patient satisfaction score (30 vs 30, p = .19) had no significant difference 2. Patients with breast cancer undergoing pretest genetic counseling using an automated Chatbot and in-person genetic testing had comparable satisfaction and comprehension
10	Visvanathan et al ²²	The ENGAGE study: evaluation of a conversational virtual agent that provides tailored pretest genetic education to cancer patients	USA /04/2021- 02/2022	Breast and prostate cancer	Patient survey (Quantitative study)	51	30 and 90 years with a mean age of 61 years	1. Patients found AI conversational virtual agent secure, easy to use, convenient to use at home 2. Patients found that AI conversational virtual agent provides valuable information 3. Overall, patients who adoption AI conversation tool had high satisfaction with its usability and convenience
11	Jordan et al ²³	Deploying an artificial intelligence-based online search tool to increase patients’ access to and understanding of solid tumor gastrointestinal clinical trials	USA/not specified	Gastrointestinal cancers	Patient survey (Quantitative study)	29	68% between 35 and 54 years of age	1. AI-based online search tool made it easier to find trials, led to better ease of understanding of presented information with more clarity on the next steps 2. New tool led to higher patient satisfaction
12	McCradden et al ¹⁶	Ethical concerns around use of artificial intelligence in health care research from the perspective of patients with meningioma, caregivers and health care providers: a qualitative study	Canada /08/2018-02/2019	Meningioma	Patient interview (Qualitative study)	30	50% patients between age 51 and 70 years	1. Most participants advocated for disclosure and transparency 2. Majority felt it inappropriate to delegate allocation of health resource to a computer 3. Prohibition on selling health data was supported by almost all participants 4. Participants expressed desire for data de-identification and its use within trusted institutions

The table highlights study designs, participant demographics, sample sizes, geographic locations, AI applications, and outcomes related to patient acceptance and adoption of AI in healthcare contexts.

Though a patient's age can impact their views on AI, this was not seen in multiple included studies. No association between age and preference for AI involvement at different stages of diagnosis has been found.¹⁶ Patients’ age and level of education were not positively correlated with trust in AI.²² Similarly, no significant correlation was found between age and gender with strength of views on AI use in radiotherapy.²³

Prior digital experience can also affect cancer survivors’ acceptance and adoption of AI tools. There was variation in prior knowledge or experience with digital technology or AI by study participants in the included studies (Table 1) and a correlation between the cumulative use of technology and trust in AI.²² Additionally, patients who heard of AI were more likely to trust it in making a diagnosis.²⁴ We divided the identified studies into 3 major themes.

Theme 1: Cancer Survivors’ Acceptance of AI

Cancer Diagnosis and Trust in Decision Making

We identified 5 studies to further our understanding of cancer survivors’ perspectives on using AI in cancer diagnosis. Three studies used surveys, 1 used semistructured interviews and 1 employed focus groups. Most participants in these studies were older than 46 years old. Close to half of respondents were not concerned about using AI technology to aid specialists in skin cancer diagnosis.¹⁶ Perceived benefits of AI included increased diagnostic accuracy, healthcare access, and reduced healthcare cost.¹⁹ This study noted patient education as one of AI's strengths. Further, patients expected AI to improve cancer diagnosis accuracy and facilitate more personalized medicine.¹⁸

Patients’ concerns included data protection, data misuse, error susceptibility, and lack of transparency.¹⁷ Perceived risks of AI included increased patient anxiety, loss of human interaction, loss of patient privacy, patient loss to follow up, nefarious use of AI, and human deskilling.¹⁹ While many respondents supported AI use and were willing to learn more, they wanted to retain the human element provided by the involvement of physicians in their care.¹⁶ Patients also expressed concern about the disruption of interpersonal relationships with physicians and loss of physician skills due to dependence on the AI system.^17,18

Another theme was the importance of collaboration between humans and AI.¹⁹ Most respondents supported AI as an adjunct to physicians.¹⁷ Cancer survivors with melanoma were more supportive of using AI applications for early skin cancer detection, with a preference for physicians and AI classifying lesions independently. Most patients felt it essential for a dermatologist to confirm and discuss the cancer diagnosis.¹⁶ These findings were similar to a prospective trial assessing patients’ trust in AI.²² In this study, patients had more confidence in physicians than in AI for responding to diagnoses, communicating individually, and explaining information in understandable ways. Patients also had higher trust in diagnoses made by AI controlled by physicians than AI not controlled by physicians. Moreover, patients overwhelmingly preferred AI-assisted physicians to physicians alone for current clinical scenario treatment. Most cancer survivors support using AI as an adjunct to increase diagnostic accuracy. Together, these studies represent the patients’ desire for AI tools to be used as support rather than to replace physicians.

AI in Therapy, Prognosis, and Clinical Care

We reviewed 3 studies that examined cancer survivors’ perspectives on this theme. In a validated questionnaire, cancer survivors’ positivity or negativity toward AI implementation in radiotherapy suggested a moderately negative patient view of AI use.²³ In questions about “personal interaction,” results indicated a strong patient preference for some level of personal interaction. Patients reported more trust in AI model results if presented by their physicians and expressed a desire to be informed when AI models were used in their care.²⁴ Predictive models can elicit negative responses, as noted in some patients after seeing this study's median survival prediction. Chinese cancer patients’ attitudes toward the clinical use of AI revealed that most patients with cancer trusted AI in diagnosis and treatment. Additionally, patients who were familiar with AI were more likely to trust them being used in diagnosis.²⁴ Patients trusted oncologists more when the opinions of the oncologist and AI diverged.

Theme 2: Cancer Survivors’ Adoption of AI for Seeking Information

AI-Based Automated Chatbots

Significant patient acceptance among newly diagnosed patients with breast cancer was also found when comparing the effectiveness of AI-powered Chatbots for pretest genetic counseling to traditional genetic counseling methods.²⁵ Further, AI-based conversational virtual agents were evaluated to provide tailored pretest genetic education and assessed patients’ responses to adoption, use, and experience with AI applications.²⁶ All participants in this study found AI-based conversational virtual agents secure and easy to use. Patients found them convenient for home use and were very satisfied with their usability and convenience.

AI-Based Online Search Tools

Other assessments include a novel AI-based online search tool to increase patients’ access and understanding of gastrointestinal solid tumor clinical trials.²⁷ This study included a new cancer trial search tool based on natural language processing with restructured trial information to improve search efficiency and automated identification of relevant trials for an individual patient. Patients found it easier and faster to find relevant trials and it easier to understand trial information with clearer next steps. Patients had better overall satisfaction with the novel online search tool than currently available tools for finding cancer trial information.

Theme 3: Cancer Survivors’ Perspective Regarding the Ethical Use of AI

Insights regarding meningioma patients’ ethical concerns for using AI in healthcare research were assessed by providing hypothetical issues.²⁰ Most participants in this study favored disclosure and transparency while supporting nonconsented healthcare data use. Most participants did not favor the commercial selling of healthcare data and delegating healthcare resource utilization to a computer. This study also identified a desire for data de-identification and the use of data within trusted institutions. Providers were more skeptical regarding the use of AI in healthcare research. These findings reiterate earlier studies regarding patients’ concerns for data misuse.

Discussion

The clinical care of cancer survivors involves diagnosis, clinical decision making, treatment, prognosis, and education. AI is expected to play a more significant role in improving these aspects of clinical care. Studies included in this literature review provide a better understanding of cancer survivors’ perspectives regarding accepting and adopting AI, including seeking more information.

Patient acceptance is crucial for the successful integration of technology in healthcare. Cancer survivors are open to using AI to enhance diagnostic accuracy; however, they prefer it as a support tool alongside physicians, valuing the patient-physician relationship. This preference has important design implications for AI healthcare solutions aimed at cancer survivors. Future research should explore the meaningful human interactions necessary for effectively integrating AI into their care. Future research should also explore the lack of correlation between patient age, gender, and education level for AI trust and adoption.^22,23 Prior studies show that AI has significant potential to speed up early diagnosis and improve cancer survivors’ outcomes.²⁸ AI's image-based screening tools, like Google's automated retina disease assessment, are comparable to or better than physician evaluation.^29,30

Cancer survivors showed interest in AI prognostic models but wanted results shared by their physicians, again highlighting higher trust in physicians and healthcare teams.^21,23 Some patients reacted negatively to the predictive model, which revealed emotional elements as a response to technology output.²¹ Patients with poor prognosis may respond negatively to AI-generated prognostic models. This highlights the importance of human guidance in presenting AI data, especially when it's emotionally challenging. Further research is needed to understand how human interactions can enhance trust and acceptance among cancer survivors during the development and implementation of these models. AI and machine learning algorithms have also started to play a role in cancer treatment and predictive models, with their use expected to increase with time.³¹

We also found that cancer survivors’ education and knowledge about their health condition have a significant effect on a patient's treatment, prognosis, and satisfaction.³² With limited resources and time, healthcare professionals often need help to spend adequate time with patients and caregivers for effective patient education.³³ AI has significant potential to fill this gap with AI technologies such as Chatbots. Chatbots are AI-based interactive auditory and written communication tools.²⁵ Chatbot use in healthcare is steadily growing, including in oncology.^34,35 Some of the uses of Chatbots in medicine have included family history intake, risk assessment, pretesting education, and sharing average results.²⁵ Cancer survivors found Chatbots easy to use, convenient, and secure.²⁶ Knowledge and comprehension of cancer survivors between automatic Chatbot-based counseling and in-person genetic counseling were also comparable, supporting their role as an adjunct to the current system in improving the accessibility of these services.²⁵ Studies using Chatbots, like digital genetic conversation Chatbot “ROSA,” with patients without cancer also had positive patient feedback.^34,36 These studies further demonstrate cancer survivors’ willingness to adopt AI-based Chatbots to seek information. One promising area that can be implemented at scale using AI-based Chatbots is general pretrained models (GPT) like ChatGPT and Gemini.^37,38 However, this technology is not designed to provide patient education and has issues with the validity of results.³⁹ Further development efforts focusing on output to support patients, including cancer survivors’ questions and robust validation, can help improve adoption and unleash the complete potential of these tools.

AI can improve cancer survivors’ access to clinical trials by simplifying information. Those in rural areas or smaller hospitals may struggle with guidance from oncologists due to limited access to in-person and virtual care. By quickly analyzing data, AI helps survivors find and understand relevant clinical trials. The AI-based cancer trial search tool can provide more relevant information and better patient satisfaction than current tools.²⁷ AI-based cancer trial search tools can improve clinical trial accessibility of disadvantaged populations.⁴⁰ They can also help clinical trial organizers recruit faster, improve trial completion time, and share results with the broader community.⁴⁰ Further studies should incorporate AI-based search tools for various cancer types and survivor populations to better understand factors influencing the wider adoption of this tool.

As AI use grows in healthcare, it is important to understand cancer survivors’ perspectives about its ethical use. One potential issue in the ethical use of AI is a lack of transparency with black box issues and a poor understanding of some AI systems beyond input and output.⁴¹ When AI is used in clinical care, obtaining informed consent may become more complicated. Improving AI model explainability and educating patients and healthcare personnel can enhance the acceptance of AI tools. For example, Mirada Medical Ltd in Oxford, UK, has implemented targeted educational efforts to boost clinician and patient understanding of AI.²³

Cancer survivors also have valid concerns regarding data protection, misuse, and a lack of transparency.^17,19,20 This perspective is shared by providers and the public as well.²⁰

Concerns about AI misuse highlight the importance of data protection and transparency for cancer survivor tools. Trust between survivors and healthcare teams is vital; without it, survivors may not share their data, limiting AI model training. De-identifying data and using it within trusted institutions may help. Future studies should explore innovative methods to enhance data protection as AI becomes more integrated into our lives.

Limitations

A key limitation of this study is that the abstract review, article inclusion, and subsequent data abstraction included studies were completed by one reviewer due to resource constraints. To mitigate bias, however, a checklist of inclusion and exclusion criteria was formed by 4 authors under the guidance of the Principal Investigator. Uncertainties for study inclusion were resolved through group discussion of mentioned authors. Future research should be done to include 2 reviewers to systematically reduce bias in chosen articles through the PRISMA guidelines. Other limitations include the heterogeneity of patient populations, intervention type, and setting, and outcome due to the small number of articles chosen.

We found that one area that can significantly impact cancer survivors’ care is education using AI driven technologies like Chatbots. Given limited studies incorporating patients’ perspectives in this area, more studies are needed to ensure that cancer survivors’ concerns and expectations are included in Chatbot tool development.

These studies offer insights into cancer survivors’ views on various common cancers, highlighting that experiences and expectations of AI may differ among them. More research is needed to grasp the perspectives of diverse stakeholders. Early consideration of explainability and patient education about AI in clinical care is crucial, particularly for surgical procedures where informed consent is vital.⁴²

A key limitation of many studies is the lack of diversity in patient demographics. Most participants had prior digital experience, which may not reflect the broader population, particularly among socially disadvantaged groups and older cancer survivors. This experience can shape their views on AI. Including a more diverse demographic is crucial for understanding cancer survivors’ perspectives and reducing bias in AI development.

Moreover, many of these studies were done before introducing next-generation AI-based search tools like “ChatGPT” and “Gemini.”^37,38 Cancer survivors’ perspectives may have evolved since the introduction of these tools. Future studies should strive to understand cancer survivors’ acceptance and adoption of these newer AI-based tools.

Recommendations for Improved Cancer Survivor Adoption and Acceptance of AI

Focus on preserving human interaction and patient-physician relationships while developing AI tools to augment diagnosis and decision making.

Ensure adequate data protection and misuse safeguards during the AI tool's development and deployment.

Develop AI tools with clear explainability of their design, functionality, and output.

Develop clear plans for communicating and educating clinical care teams and patients about the deployed AI tool's role in diagnosis, decision making, and information seeking, as well as its limitations.

To reduce bias and increase usability, include training data from diverse patient demographics, educational levels, and digital technology adoption levels.

Conclusion

As AI furthers its integration into the clinical care of cancer survivors, cancer survivors’ perspectives should be further explored and incorporated into their development. Cancer survivors in studies examined in this review were accepting of AI use in clinical care but also have some valid concerns, especially around data misuse and disruption of patient-physician relationships with the use of AI. Cancer survivors have shown interest in adopting AI-based search tools like Chatbots. These tools can also raise awareness and decrease patient concerns about AI integration in clinical care. More studies with diverse patient populations are needed to reduce bias and ensure equitable development of AI-based clinical care tools for cancer survivors.

Supplemental Material

sj-docx-1-jpx-10.1177_23743735251380949 - Supplemental material for Acceptance and Adoption of Artificial Intelligence Among Cancer Survivors Seeking Information: A Literature Review

Supplemental material, sj-docx-1-jpx-10.1177_23743735251380949 for Acceptance and Adoption of Artificial Intelligence Among Cancer Survivors Seeking Information: A Literature Review by Prabal Chourasia, Melissa Coffel, Jamie Lynn Conklin, William A. Wood and Saif Khairat in Journal of Patient Experience

Footnotes

Authors’ Contributions

PC was involved in methodology, formal analysis, investigation, writing original draft, and writing—review & editing; MC in methodology, formal analysis, resources, writing—review & editing, and project administration; JLC in methodology, data curation, writing—review & editing; WAW in writing—review & editing; and SK in conceptualization, methodology, resources, writing—review & editing, supervision, and funding acquisition.

Data Availability

Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.

Declaration of Conflicting Interest

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Statement of Informed Consent

There are no human subjects in this article and informed consent is not applicable.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Research reported in this publication was supported by the National Center For Advancing Translational Sciences of the National Institutes of Health under Award Number RC2TR004380. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Ethical Approval

Ethical approval is not applicable for this article.

Statement of Human and Animal Rights

All procedures in this study were conducted in accordance with the University of North Carolina at Chapel Hill's Institutional Review Board (IRB # 23-1451) approved protocols.

ORCID iDs

Melissa Coffel

Jamie Lynn Conklin

Saif Khairat

Supplemental Material

Supplemental material for this article is available online.

References

AI’s Ascendance in Medicine: A Timeline | Cedars-Sinai. https://www.cedarssinai.org/discoveries/ai-ascendance-in-medicine.html (accessed 8 Jul2024).

brookfieldinstitute.ca/wp-content/uploads/AI_Intro-Policymakers_ONLINE.pdf .

Person-Centered Care | CMS. https://www.cms.gov/priorities/innovation/keyconcepts/person-centered-care (accessed 24 Jun2024).

Lau

AYS

Staccini

. Section editors for the IMIA yearbook section on education and consumer health informatics. Artificial intelligence in health: new opportunities, challenges, and practical implications. Yearb Med Inform. 2019;28(1):174-178. doi:10.1055/s-0039-1677935

Definitions | Division of Cancer Control and Population Sciences (DCCPS) . https://cancercontrol.cancer.gov/ocs/definitions (accessed 19 Jun2024).

Chan

Crawford-Williams

Crichton

, et al. Effectiveness and implementation of models of cancer survivorship care: an overview of systematic reviews. J Cancer Surviv. 2023;17(1):197-221. doi:10.1007/s11764-021-01128-1

Kong

Huang

Gautam

Envisioning possibilities and challenges of AI for personalized cancer care. In Companion Publication of the 2024 Conference on Computer-Supported Cooperative Work and Social Computing. doi: 10.1145/3678884.3681885

Pan

X-R

, et al.

Artificial intelligence empowered digital health technologies in cancer survivorship care: a scoping review

Asia-Pac J Oncol Nurs. 2022;9(12):100127. doi: 10.1016/j.apjon.2022.100127

Ilić

Ivanović

Jakovetić

, et al. ASCAPE - an intelligent approach to support cancer patients. In: Rocha Â, Reis L, Costanzo S, et al., eds. Information systems and technologies. Lecture notes in computer science. Cham: Springer Nature; 2023:779. doi:10.1007/978-3-031-45642-8_27

10.

Cancer Survivorship: Symptoms and Common Problems . https://www.massgeneral.org/cancer-center/patient-and-family-resources/cancersurvivorship/symptoms-and-common-problems (accessed 1 Jul2024).

11.

Granić

. Technology acceptance and adoption in education. In: Zawacki-Richter

Jung

(eds.) Handbook of open, distance and digital education. Springer Nature Singapore; 2023:183-197. doi:10.1007/978-981-19-2080-6_11

12.

Davis

. User acceptance of information technology: system characteristics, user perceptions and behavioral impacts. Int J Man Mach Stud. 1993;38(3):475-487. doi:10.1006/imms.1993.1022

13.

Page

McKenzie

Bossuyt

, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ (Clin Res ed.). 2021;372(71). doi10.1136/bmj.n71

14.

Covidence - Better systematic review management . https://www.covidence.org/ (accessed 26 Jun2024).

15.

Page

McKenzie

Bossuyt

, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. Syst Rev. 2021;10(89). doi:10.1186/s13643- 021-01626-4

16.

Lim

Neal-Smith

Mitchell

, et al. Perceptions of the use of artificial intelligence in the diagnosis of skin cancer: an outpatient survey. Clin Exp Dermatol. 2022;47(3):542-546. doi:10.1111/ced.14969

17.

Jutzi

Krieghoff-Henning

Holland-Letz

, et al. Artificial intelligence in skin cancer diagnostics: the patients’ perspective. Front Med (Lausanne). 2020;7:233. doi:10.3389/fmed.2020.00233

18.

Lysø

Hesjedal

Skolbekken

J-A

, et al. Men’s sociotechnical imaginaries of artificial intelligence for prostate cancer diagnostics—a focus group study. Soc Sci Med. 2024;347:116771. doi:10.1016/j.socscimed.2024.116771

19.

Nelson

Pérez-Chada

Creadore

, et al. Patient perspectives on the use of artificial intelligence for skin cancer screening: a qualitative study. JAMA Dermatol. 2020;156(5):501-512. doi:10.1001/jamadermatol.2019.5014

20.

McCradden

Baba

Saha

, et al. Ethical concerns around use of artificial intelligence in health care research from the perspective of patients with meningioma, caregivers and health care providers: a qualitative study. CMAJ Open. 2020;8(1):E90-E95. doi:10.9778/cmajo.20190151

21.

Hildebrand

Chang

Ewongwoo

, et al. Study of patient and physician attitudes toward automated prognostic models for patients with metastatic cancer. JCO Clin Cancer Inform. 2023;7:e2300023. doi:10.1200/CCI.23.00023

22.

Rodler

Kopliku

Ulrich

, et al. Patients’ trust in artificial intelligence-based decision-making for localized prostate cancer: results from a prospective trial. Eur Urol Focus. 2023;10(4):654-661. doi:

10.1016/j.euf.2023.10.020

23.

Temple

Rowbottom

Simpson

. Patient views on the implementation of artificial intelligence in radiotherapy. Radiography (Lond). 2023;29(Suppl 1):S112-S116. doi:10.1016/j.radi.2023.03.006

24.

Yang

Zeng

Peng

, et al. Attitudes of Chinese cancer patients toward the clinical use of artificial intelligence. Patient Prefer Adherence. 2019;13:1867-1875. doi:10.2147/PPA.S225952

25.

Al-Hilli

Noss

Dickard

, et al. A randomized trial comparing the effectiveness of pre-test genetic counseling using an artificial intelligence automated Chatbot and traditional in-person genetic counseling in women newly diagnosed with breast cancer. Ann Surg Oncol. 2023;30(10):5990-5996. doi:10.1245/s10434-023-13888-4

26.

Visvanathan

Petry

McCullough

, et al. The ENGAGE study: evaluation of a conversational virtual agent that provides tailored pre-test genetic education to cancer patients. J Cancer Surviv. 2023;19(2):623-632. doi:

10.1007/s11764-023-01495-x

27.

Jordan

Jahreiss

Kasi

. Deploying an artificial intelligence-based online search tool to increase patients’ access to and understanding of solid tumor gastrointestinal clinical trials. J Gastrointest Oncol. 2021;12(5):2045-2051. doi:10.21037/jgo-21-175

28.

AI and Cancer - NCI. https://www.cancer.gov/research/infrastructure/artificialintelligence# (accessed 25 Jun2024).

29.

ARDA : Using Artificial Intelligence in Ophthalmology - Google Health. https://health.google/caregivers/arda/ (accessed 5 Jul2024).

30.

Tschandl

Rinner

Apalla

, et al. Human-computer collaboration for skin cancer recognition. Nat Med. 2020;26:1229-1234. doi:10.1038/s41591-020-0942-0

31.

Zhang

Shi

Wang

. Machine learning and AI in cancer prognosis, prediction, and treatment selection: a critical approach. J Multidiscip Healthc. 2023;16:1779-1791. doi:10.2147/JMDH.S410301

32.

Tian

Jia

Cheng

. Relationships between patient knowledge and the severity of side effects, daily nutrient intake, psychological status, and performance status in lung cancer patients. Curr Oncol. 2015;22(4):e254-e258. doi:10.3747/co.22.2366

33.

Paterick

Patel

Tajik

, et al. Improving health outcomes through patient education and partnerships with patients. Proc (Bayl Univ Med Cent). 2017;30(1):112-113. doi:10.1080/08998280.2017.11929552

34.

Heald

Keel

Marquard

, et al. Using Chatbots to screen for heritable cancer syndromes in patients undergoing routine colonoscopy. J Med Genet. 2021;58(12):807-814. doi:10.1136/jmedgenet-2020-107294

35.

Schmidlen

Jones

Campbell-Salome

, et al. Use of a Chatbot to increase uptake of cascade genetic testing. J Genet Couns. 2022;31(5):1219-1230. doi:10.1002/jgc4.1592

36.

Siglen

Vetti

Lunde

ABF

, et al. Ask rosa—the making of a digital genetic conversation tool, a Chatbot, about hereditary breast and ovarian cancer. Patient Educ Couns. 2022;105(6):1488-1494. doi:10.1016/j.pec.2021.09.027

37.

Ling

Tran

Fan

, et al. GEMINI. Proc VLDB Endow. 2014;7(13):1766-1771. doi:10.14778/2733004.2733081

38.

Introducing ChatGPT | OpenAI. https://openai.com/index/chatgpt/ (accessed 3 Jul2024).

39.

Hershenhouse

Mokhtar

Eppler

, et al. Accuracy, readability, and understandability of large language models for prostate cancer information to the public. Prostate Cancer Prostatic Dis. 2025;28(2):394-399. doi:10.1038/s41391-024-00826-y

40.

7 Best AI Tools for Clinical Trials . https://www.ominext.com/en/blog/7-best-ai-toolsfor-clinical-trials (accessed 5 Jul2024).

41.

Castelvecchi

. Can we open the black box of AI? Nature. 2016;538(7623):20-23. doi:10.1038/538020a

42.

Schiff

Borenstein

. How should clinicians communicate with patients about the roles of artificially intelligent team members? AMA J Ethics. 2019;21(2):E138-E145. doi:10.1001/amajethics.2019.138

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.03 MB