aNo artificial intelligence story mentioned harms only or appeared to deemphasize harms.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Identify all potential conflicts of interest that might be relevant to your comment.
Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.
Err on the side of full disclosure.
If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.
Not all submitted comments are published. Please see our commenting policy for details.
O’Keeffe M, Barratt A, Fabbri A, et al. Global Media Coverage of the Benefits and Harms of Early Detection Tests. JAMA Intern Med. 2021;181(6):865–867. doi:10.1001/jamainternmed.2021.0261
Innovations in technologies for early detection of diseases, such as breast cancer, dementia, and atrial fibrillation, are gaining increasing attention. The media is a key avenue through which tests are promoted to asymptomatic individuals, and it could have an important role in encouraging realistic expectations of the benefits and harms of early detection, including unnecessary diagnoses.1 Evidence suggests that medical media coverage tends to overplay benefits, downplay harms, and ignore conflicts of interest,2 but there are few data on coverage of early detection tests.
We performed a cross-sectional study to examine global media coverage of the benefits and harms of early detection tests for asymptomatic individuals. We also examined conflicts of interest among commentators in stories, as well as media disclosure of conflicts. We studied 5 early detection tests: (1) blood-based liquid biopsy tests for cancer(s), (2) 3-dimensional mammography for breast cancer, (3) Apple Watch Series 4 electrocardiogram for atrial fibrillation, (4) blood biomarker tests for dementia, and (5) artificial intelligence technology for dementia. Our published protocol3 provides detail regarding selection of these tests. The University of Sydney Human Research Ethics Committee reviewed the application and stated that it was exempt from review because it was applied to publicly available information in the media.
We searched for all English-language stories from 2016 to 2019 in LexisNexis, ProQuest, and Google News. We included all nonfictional story types from newspapers, blogs, magazines, broadcast and podcast transcripts, and wire feeds/services if they mentioned or implied a test benefit or harm. Pairs of independent reviewers screened stories for inclusion and coded stories for coverage of health benefits (eg, early treatment, saves lives) and/or harms (eg, false positives, overdiagnosis) and inclusion of commentator views with or without disclosure of conflicts of interest. Analyses were performed using Stata, version 16.0 (StataCorp LLC).
We included 1173 stories: liquid biopsy (n = 124), 3-dimensional mammography (n = 579), Apple Watch (n = 273), blood biomarker tests (n = 128), and artificial intelligence (n = 69). Overall, 97% (95% CI, 96%-98%) reported on the benefits, 37% (95% CI, 34%-40%) reported any harms, and only 34% (95% CI, 31%-36%) reported on both benefits and harms; 63% (95% CI, 60%-66%) of stories reported on benefits only, while only 3% (95% CI, 2%-4%) reported on harms only (Figure 1). There was variation between tests in the percentage reporting any harms (from 10 of 69 [14%] for artificial intelligence to 157 of 273 [58%] for Apple Watch and 72 of 124 [58%] for liquid biopsy) and both benefit and harm (from 10 of 69 [14%] for artificial intelligence to 67 of 124 [54%] for liquid biopsy). Harms were mentioned but deemphasized in just more than one-quarter (317 of 1173 [27%]) of stories.
Overdiagnosis was only mentioned in 57 of 432 (13%) of stories that mentioned any harms—5% of stories overall. Overall, 445 of 1173 (38%) of stories quantified a benefit. However, only 62 (14%) of these used absolute numbers. Overall, 127 of 432 (29%) of stories quantified a harm. Almost half (58 of 127 [46%]) of these provided absolute numbers. More than half (55% [95% CI, 52%-57%]) of all stories included the views of commentators with conflicts of interest, but these conflicts were only disclosed in 12% (95% CI, 10%-14%) of these stories (Figure 2).
In this study, we examined how innovative early detection tests are covered in the media. The findings are important because of the potential of such tests to harm healthy people, in contrast to tests promoted for investigation of symptoms. Yet coverage emphasized benefits far more than harms, and the risk of overdiagnosis received little coverage. Our findings align with other medical media coverage studies.2,4,5 Coding benefits and harms involves subjective judgments, although we piloted an explicit coding scheme and had pairs of independent coders to minimize bias. We did not examine social media coverage. Higher-quality reporting by journalists could encourage more healthy skepticism toward health options6 and curb overdiagnosis.1 Strategies to improve media reporting so that professionals, patients, and the public receive more balanced information about early detection tests are urgently needed.
Accepted for Publication: December 2, 2020.
Published Online: April 5, 2021. doi:10.1001/jamainternmed.2021.0261
Corresponding Author: Mary O’Keeffe, PhD, Institute for Musculoskeletal Health, Sydney Local Health District and University of Sydney, Level 10N, King George V Bldg, Royal Prince Alfred Hospital (C39), PO Box M179, Missenden Rd, Camperdown, NSW 2050, Australia (firstname.lastname@example.org).
Author Contributions: Dr O’Keeffe had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: O’Keeffe, Barratt, Moynihan.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: O’Keeffe, Barratt.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: O’Keeffe.
Administrative, technical, or material support: O’Keeffe, Fabbri, Zadro, Ferreira.
Supervision: O’Keeffe, Barratt, Moynihan.
Conflict of Interest Disclosures: Dr O’Keeffe reported receiving grants from the European Commission (Marie Skłodowska-Curie Postdoctoral Fellowship) outside the submitted work. Dr Barratt reported receiving grants from the National Health and Medical Research Council of Australia during the conduct of the study and serving as a scientific steering committee member and receiving travel and accommodation reimbursement from Preventing Overdiagnosis international conferences outside the submitted work. Dr Fabbri was supported as a postdoctoral fellow by the National Health and Medical Research Council of Australia, project grant No. 1122332, during the conduct of the study. Drs Zadro and Moynihan are supported by research fellowships funded by the National Health and Medical Research Council of Australia. Dr Moynihan reported receiving grants from the National Health and Medical Research Council during the conduct of the study and serving on the scientific steering committee of the Preventing Overdiagnosis conference outside the submitted work. No other disclosures were reported.
Additional Contributions: We thank Steven Woloshin, MD, MS (Center for Medicine in the Media, Dartmouth Institute for Health Policy and Clinical Practice, Lebanon, New Hampshire, and the Lisa Schwartz Foundation for Truth in Medicine, Norwich, Vermont), Chris G. Maher, PhD (Institute for Musculoskeletal Health, Sydney Local Health District and The University of Sydney, Sydney, Australia), and Mark Jones, PhD (Institute for Evidence-Based Healthcare, Bond University, Gold Coast, Queensland, Australia), for providing feedback on an earlier draft of the manuscript. They did not receive compensation for these contributions.