Open Availability of Patient Medical Photographs in Google Images Search Results: Cross-Sectional Study of Transgender Research
Bibliographic record
Abstract
BACKGROUND: This paper focuses on the collision of three factors: a growing emphasis on sharing research through open access publication, an increasing awareness of big data and its potential uses, and an engaged public interested in the privacy and confidentiality of their personal health information. One conceptual space where this collision is brought into sharp relief is with the open availability of patient medical photographs from peer-reviewed journal articles in the search results of online image databases such as Google Images. OBJECTIVE: The aim of this study was to assess the availability of patient medical photographs from published journal articles in Google Images search results and the factors impacting this availability. METHODS: We conducted a cross-sectional study using data from an evidence map of research with transgender, gender non-binary, and other gender diverse (trans) participants. For the original evidence map, a comprehensive search of 15 academic databases was developed in collaboration with a health sciences librarian. Initial search results produced 25,230 references after duplicates were removed. Eligibility criteria were established to include empirical research of any design that included trans participants or their personal information and that was published in English in peer-reviewed journals. We identified all articles published between 2008 and 2015 with medical photographs of trans participants. For each reference, images were individually numbered in order to track the total number of medical photographs. We used odds ratios (OR) to assess the association between availability of the clinical photograph on Google Images and the following factors: whether the article was openly available online (open access, Researchgate.net, or Academia.edu), whether the article included genital images, if the photographs were published in color, and whether the photographs were located on the journal article landing page. RESULTS: We identified 94 articles with medical photographs of trans participants, including a total of 605 photographs. Of the 94 publications, 35 (37%) included at least one medical photograph that was found on Google Images. The ability to locate the article freely online contributes to the availability of at least one image from the article on Google Images (OR 2.99, 95% CI 1.20-7.45). CONCLUSIONS: This is the first study to document the existence of medical photographs from peer-reviewed journals appearing in Google Images search results. Almost all of the images we searched for included sensitive photographs of patient genitals, chests, or breasts. Given that it is unlikely that patients consented to sharing their personal health information in these ways, this constitutes a risk to patient privacy. Based on the impact of current practices, revisions to informed consent policies and guidelines are required.
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 machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.013 | 0.057 |
| Meta-epidemiology (narrow) | 0.000 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.008 | 0.009 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.005 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.005 | 0.001 |
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 source (direct Gemma or distilled Codex), 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".