Virtual Health Assistants in Preventive Cancer Care Communication: Systematic Review
Bibliographic record
Abstract
Background: Virtual health assistants (VHAs), interactive digital programs that emulate human communication, are being increasingly used in health care to improve patient education and care and to reduce the burden on health care providers. VHAs have the potential to promote cancer equity through facilitating patient engagement, providing round-the-clock access to information, and reducing language barriers. However, it is unclear to what extent audience-centeredness is being considered in the development of cancer-related applications. Objective: This systematic review identifies and synthesizes strategies used to make VHA-based cancer prevention and screening interventions audience-centered. Methods: Following PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) 2020 guidelines, we searched 4 databases (PubMed, Embase, Web of Science, and EBSCOhost) for peer-reviewed studies on VHA interventions promoting cancer screening (January 2022). Included studies focused on adult populations in primary care settings, with interventions emphasizing interactivity and immediacy (key VHA features). Excluded studies were on cancer treatment, noninteractive decision aids, or technical VHA development. Screening, data extraction, and quality assessment (Mixed Methods Appraisal Tool) were performed independently by multiple reviewers. Thematic synthesis was used to analyze audience-centered strategies. Results: Of 1055 records screened, 17 studies met inclusion criteria. Most (n=11) targeted colorectal cancer, with others addressing prostate, breast, cervical, or lung cancer. A total of 16 studies were US-based; 1 study focused on Uganda. Key strategies for audience-centered design included: (1) Demographic Concordance: Race or gender alignment between VHA and users (eg, African American participants interacting with Black-coded avatars); (2) User Feedback: Iterative testing via interviews, think-aloud protocols, or pilot studies to refine interventions; (3) Preintervention Needs Assessment: Identifying cultural, linguistic, or literacy barriers (eg, myths about screening in Ugandan communities); (4) Theoretical Frameworks: The Health Belief Model (most common), the Modality, Agency, Interactivity, and Navigability (MAIN) model, or tailored messaging theories guided design; (5) Information Customization: Culturally adapted content (eg, Spanish-language interfaces, narratives addressing racial disparities); and (6) Feature Customization: Adjusting VHA appearance (eg, animations and fonts) based on user preferences. Notably, 7/17 studies focused on racially minoritized groups (eg, African Americans, Hispanic farmworkers), addressing systemic barriers like mistrust in health care. However, gaps persisted in intersectional tailoring (eg, rurality and income) and non-English languages (only 2/17 studies). Recruitment methods influenced diversity; community-based strategies yielded more representative samples than solely internet-based recruitment approaches. Conclusions: The systematic review identified the audience-centered development practices currently being used for VHA-based interventions in preventive cancer care. The majority of the studies included processes to diversify and segment the intended audience, focused on medically underrepresented population groups, and implemented strategies to be culturally sensitive to the population of interest. However, opportunities remain to address multidimensional inequities (eg, rural access and low literacy). Future interventions should integrate intersectional frameworks, expand language diversity, and measure social presence to enhance engagement. This review provides a roadmap for developing equity-focused eHealth tools in cancer prevention.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.004 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".