Artificial Intelligence, Cancer, Shared Decision Making
Conditions
Keywords
Large Language Models, Shared Decision Making, Oncology, Patient Acceptance
Brief summary
This study examines how cancer patients, the general public, and healthcare professionals use and perceive large language models (such as ChatGPT) for health-related shared decision-making in oncology. A cross-sectional survey was conducted among 7,151 participants across 30 countries using a questionnaire developed and validated through a two-round Delphi process involving 44 experts. The study assessed current patterns of large language model use for health information, barriers to adoption including concerns about reliability and privacy, future expectations regarding these tools in shared decision-making, and demographic predictors of adoption. Participants were recruited through the Prolific platform between March and May 2025, with stratified sampling across three groups: cancer patients diagnosed within the past five years, general population members from the United States and United Kingdom, and licensed healthcare professionals with active patient contact.
Detailed description
Shared decision-making is a collaborative process in which clinicians support patients in reaching treatment decisions. Despite its importance in oncology, structured shared decision-making remains uncommon in routine clinical practice. Large language models offer a new way for patients to access and understand medical information, yet little is known about how key stakeholders perceive and use these tools for health decisions. This observational study used a sequential mixed-methods design combining Delphi consensus methodology with cross-sectional survey deployment. A 44-expert panel across eight domains (clinical artificial intelligence, technical development, oncology, psychology, epidemiology, patient advocacy, ethics, and legal expertise) developed and validated the assessment instrument through two Delphi rounds, achieving consensus on 89 items. The final instrument contained 52 quantitative items and 8 qualitative prompts, distinguishing between general and healthcare-specific large language model use. The study recruited three cohorts: 2,316 cancer patients with self-reported diagnosis within five years, 2,000 general population members from the United States and United Kingdom, and 2,835 licensed healthcare professionals. Quality control included attention checks, completion time monitoring, consistency validation, and verification procedures, resulting in exclusion of 694 responses (8.8%) from an initial 7,845. Primary analyses included chi-squared testing and ANOVA with Bonferroni correction, multivariable logistic regression with hierarchical model building to identify adoption predictors, and user segmentation through cross-tabulation combined with k-means clustering. The study was approved by the institutional review board of the Technical University of Munich (TUM2024-89-S-SB).
Interventions
None listed
Sponsors
Technical University of Munich
Study design
Observational model
OTHER
Time perspective
CROSS_SECTIONAL
Eligibility
Sex/Gender
ALL
Age
18 Years to No maximum
Healthy volunteers
Yes
Inclusion criteria
* Age 18 years or older * English language proficiency * Regular internet access * Registered on the Prolific research platform * For cancer patient cohort: self-reported cancer diagnosis within the past five years * For healthcare professional cohort: licensed healthcare practitioner with active patient contact * For general population cohort: resident of the United States or United Kingdom
Exclusion criteria
* Failure on embedded attention check questions (4 checks) * Survey completion time less than 5 minutes or greater than 60 minutes * Straight-line responding pattern detected by consistency validation algorithms * Failure of cohort verification procedures
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Healthcare-specific large language model usage rate | At time of survey completion (single assessment, March-May 2025) | Proportion of participants reporting use of large language models specifically for health-related information, measured on a 5-point Likert frequency scale and dichotomised as use versus non-use. |
| Future belief in large language model improvement of shared decision-making | At time of survey completion (single assessment, March-May 2025) | Proportion of participants believing that large language models will improve the quality of shared decision-making in oncology, assessed via Likert-scale response. |
| Barriers to large language model adoption | At time of survey completion (single assessment, March-May 2025) | Prevalence of concerns regarding large language model use for health decisions, including reliability concerns, privacy concerns, and preference for human interaction, each assessed as binary (present or absent). |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Independent predictors of large language model adoption | At time of survey completion (single assessment, March-May 2025) | Odds ratios from hierarchical multivariable logistic regression identifying demographic and health-related predictors of healthcare-specific large language model acceptance, including age, sex, ethnicity, education, digital literacy, and confidence in understanding health information. |
| User segmentation | At time of survey completion (single assessment, March-May 2025) | Distribution of participants across data-driven user segments derived from cross-tabulation of current usage with perceived benefit, refined through k-means clustering: potential adopters, believing users, resistant non-users, and sceptical users. |
| Healthcare professional recommendation patterns | At time of survey completion (single assessment, March-May 2025) | Proportion of healthcare professionals who recommend large language models to patients for health information, compared with their personal use rate. |
Countries
Germany
Outcome results
None listed