Reporting of Fairness Metrics in Clinical Risk Prediction Models Used for Precision Health: Scoping Review
Open accessPublished version of a preprint
Abstract
BACKGROUND: Clinical risk prediction models integrated into digitized health care informatics systems hold promise for personalized primary prevention and care, a core goal of precision health. Fairness metrics are important tools for evaluating potential disparities across sensitive features, such as sex and race or ethnicity, in the field of prediction modeling. However, fairness metric usage in clinical risk prediction models remains infrequent, sporadic, and rarely empirically evaluated. OBJECTIVE: We seek to assess the uptake of fairness metrics in clinical risk prediction modeling through an empirical evaluation of popular prediction models for 2 diseases, 1 chronic and 1 infectious disease. METHODS: We conducted a scoping literature review in November 2023 of recent high-impact publications on clinical risk prediction models for cardiovascular disease (CVD) and COVID-19 using Google Scholar. RESULTS: Our review resulted in a shortlist of 23 CVD-focused articles and 22 COVID-19 pandemic-focused articles. No articles evaluated fairness metrics. Of the CVD-focused articles, 26% used a sex-stratified model, and of those with race or ethnicity data, 92% had study populations that were more than 50% from 1 race or ethnicity. Of the COVID-19 models, 9% used a sex-stratified model, and of those that included race or ethnicity data, 50% had study populations that were more than 50% from 1 race or ethnicity. No articles for either disease stratified their models by race or ethnicity. CONCLUSIONS: Our review shows that the use of fairness metrics for evaluating differences across sensitive features is rare, despite their ability to identify inequality and flag potential gaps in prevention and care. We also find that training data remain largely racially and ethnically homogeneous, demonstrating an urgent need for diversifying study cohorts and data collection. We propose an implementation framework to initiate change, calling for better connections between theory and practice when it comes to the adoption of fairness metrics for clinical risk prediction. We hypothesize that this integration will lead to a more equitable prediction world.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Treating 'Asian populations' as a homogeneous entity-or extrapolating models developed in one Asian subgroup to all Asian individuals-can introduce systematic prediction bias.
“Treating “Asian populations” as a homogeneous entity—or extrapolating models developed in one Asian subgroup to all Asian individuals—can introduce systematic prediction bias.”
From AI-driven cardiovascular risk prediction in patients with diabetes: bridging algorithmic innovation to equitable clinical application (2026) · cites this paper as [16]
The Discussion (page unavailable) says that aggregated model results can mask inequalities and that limited data make it difficult to assess models separately across racial or ethnic groups. This supports the general concern about grouping populations, but the paper does not address treating Asian populations as homogeneous or extrapolating a model from one Asian subgroup to all Asian individuals. (LLM, conf=0.83)
Experimental — accuracy not yet measuredAssuming homogeneity across Asian populations or generalizing models developed from a single subgroup to the broader Asian population risks introducing systematic predictive bias and undermining equitable clinical utility.
“Assuming homogeneity across Asian populations or generalizing models developed from a single subgroup to the broader Asian population risks introducing systematic predictive bias and undermining equitable clinical utility.”
From AI-driven cardiovascular risk prediction in patients with diabetes: bridging algorithmic innovation to equitable clinical application (2026) · cites this paper as [16]
The Discussion (page not provided) says model-development data were often dominated by one racial or ethnic group, leaving insufficient data to assess fairness separately across groups. The Introduction also warns that bias can affect prediction accuracy and equity. The paper does not specifically address differences among Asian populations or generalizing a model from one Asian subgroup to all Asian populations. (LLM, conf=0.83)
Experimental — accuracy not yet measuredA recent scoping review found that fairness metrics are rarely reported in clinical risk prediction studies.
“This is especially important because a recent scoping review found that fairness metrics are rarely reported in clinical risk prediction studies, while a recent cardiovascular AI perspective emphasized that bias can arise throughout problem formulation, dataset construction, model development, validation, and implementation (16, 17).”
From AI-driven cardiovascular risk prediction in patients with diabetes: bridging algorithmic innovation to equitable clinical application (2026) · cites this paper as [16]
Even within the same geographic region, cardiovascular risk characteristics vary markedly across ethnic and healthcare subgroups.
“Taken together, the evidence points to a central conclusion: even within the same geographic region, cardiovascular risk characteristics vary markedly across ethnic and healthcare subgroups.”
From AI-driven cardiovascular risk prediction in patients with diabetes: bridging algorithmic innovation to equitable clinical application (2026) · cites this paper as [16]
The Results section on CVD describes which ethnic, chronic-condition, and geographic subgroups the reviewed prediction studies included; the Discussion addresses limited ethnic and geographic diversity in their data. Neither passage reports that cardiovascular risk characteristics vary markedly between subgroups within the same geographic region. A search for within-region risk differences found no supporting passage in our copy. (LLM, conf=0.76)
Experimental — accuracy not yet measuredSubstantial heterogeneity exists even within Asian populations.
“Furthermore, substantial heterogeneity exists even within Asian populations, adding another layer of complexity to risk prediction.”
From AI-driven cardiovascular risk prediction in patients with diabetes: bridging algorithmic innovation to equitable clinical application (2026) · cites this paper as [16]
The Discussion describes limited diversity across racial and ethnic groups in the reviewed datasets, but does not address heterogeneity within Asian populations. Searches for Asian populations and Asian ethnic subgroups found no passage supporting that specific claim; the indexed copy provides no page numbers. (LLM, conf=0.74)
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Cites
A Survey on Bias and Fairness in Machine Learning
Mehrabi, N., Morstatter, F., Saxena, N., Lerman, K., & Galstyan, A. (2022) · ACM Computing Surveys · doi:10.1145/3457607 · Cited by 5,812 · see citing works
Collins, G. S., Reitsma, J. B., Altman, D. G., & Moons, K. G. M. (2015) · Annals of Internal Medicine · doi:10.7326/m14-0697 · Cited by 3,267 · see citing works
Fair prediction with disparate impact: a study of bias in recidivism prediction instruments
Chouldechova, A. (2017) · Big Data · doi:10.1089/big.2016.0047 · Cited by 2,159 · see citing works
Hidden in Plain Sight — Reconsidering the Use of Race Correction in Clinical Algorithms
Vyas, D. A., Eisenstein, L. G., & Jones, D. S. (2020) · New England Journal of Medicine · doi:10.1056/nejmms2004740 · Cited by 1,731 · see citing works
Fairness definitions explained
Verma, S., & Rubin, J. (2018) · Proceedings of the International Workshop on Software Fairness · doi:10.1145/3194770.3194776 · Cited by 1,124 · see citing works
Khan, S. S., Coresh, J., Pencina, M. J., Ndumele, C. E., Rangaswami, J., Chow, S. L., Palaniappan, L. P., Sperling, L. S., Virani, S. S., Ho, J. E., et al. (2023) · Circulation · doi:10.1161/cir.0000000000001191 · Cited by 797 · see citing works
Algorithmic Fairness: Choices, Assumptions, and Definitions
Mitchell, S., Potash, E., Barocas, S., D'Amour, A., & Lum, K. (2021) · Annual Review of Statistics and Its Application · doi:10.1146/annurev-statistics-042720-125902 · Cited by 687 · see citing works
Sex and gender analysis improves science and engineering
Tannenbaum, C., Ellis, R. P., Eyssel, F., Zou, J., & Schiebinger, L. L. (2019) · Nature · doi:10.1038/s41586-019-1657-6 · Cited by 587 · see citing works
Collins, G. S., Reitsma, J. B., Altman, D. G., & Moons, K. G. M. (2015) · Annals of Internal Medicine · doi:10.7326/l15-5093-2 · Cited by 493 · see citing works
Sex as a Biological Variable: A 5-Year Progress Report and Call to Action
Arnegard, M. E., Whitten, L. A., Hunter, C., & Clayton, J. A. (2020) · Journal of Women s Health · doi:10.1089/jwh.2019.8247 · Cited by 355 · see citing works
Hippisley‐Cox, J., Coupland, C. A. C., Mehta, N., Keogh, R. H., Diaz‐Ordaz, K., Khunti, K. K., Lyons, R. A., Kee, F., Sheikh, A., Rahman, S., et al. (2021) · BMJ · doi:10.1136/bmj.n2244 · Cited by 308 · see citing works
Fletcher, R. R., Nakeshimana, A., & Olubeko, O. (2021) · Frontiers in Artificial Intelligence · doi:10.3389/frai.2020.561802 · Cited by 305 · cited 2× · see citing works
“For further discussion of the subtle differences between these concepts, we refer the reader to the nuanced comparisons in [12,13].”
A clarification of the nuances in the fairness metrics landscape
Castelnovo, A., Crupi, R., Greco, G., Regoli, D., Penco, I., & Cosentini, A. (2021) · Scientific Reports · doi:10.1038/s41598-022-07939-1 · Cited by 283 · see citing works
Paulus, J. K., & Kent, D. M. (2020) · npj Digital Medicine · doi:10.1038/s41746-020-0304-9 · Cited by 274 · see citing works
The Critical Role of Racial/Ethnic Data Disaggregation for Health Equity
Kauh, T. J., Read, J. G., & Scheitler, A. J. (2021) · Population Research and Policy Review · doi:10.1007/s11113-020-09631-6 · Cited by 229 · see citing works
Use of race, ethnicity, and ancestry data in health research
Lu, C., Ahmed, R., Lamri, A., & Anand, S. S. (2022) · PLOS Global Public Health · doi:10.1371/journal.pgph.0001060 · Cited by 162 · see citing works
Lee, T. C., Shah, N. U., Haack, A., & Baxter, S. L. (2020) · Informatics · doi:10.3390/informatics7030025 · Cited by 133 · see citing works
Sharma, V., Ali, I., van der Veer, S. N., Martin, G. P., Ainsworth, J. D., & Augustine, T. (2021) · BMJ Health & Care Informatics · doi:10.1136/bmjhci-2020-100253 · Cited by 110 · see citing works
In Pursuit of Interpretable, Fair and Accurate Machine Learning for Criminal Recidivism Prediction
Wang, C., Han, B., Patel, B., & Rudin, C. (2023) · Journal of Quantitative Criminology · doi:10.1007/s10940-022-09545-w · Cited by 101 · see citing works
Foryciarz, A., Pfohl, S. R., Patel, B., & Shah, N. H. (2022) · BMJ Health & Care Informatics · doi:10.1136/bmjhci-2021-100460 · Cited by 36 · see citing works
Race and Ethnicity: A Part of the Equation for Personalized Clinical Decision Making?
Paulus, J. K., & Kent, D. M. (2017) · Circulation: Cardiovascular Quality and Outcomes · doi:10.1161/circoutcomes.117.003823 · Cited by 35 · see citing works
Cardiometabolic risk estimation using exposome data and machine learning
Atehortúa, A., Gkontra, P., Camacho, M., Díaz, O., Bulgheroni, M., Simonetti, V., Chadeau‐Hyam, M., Felix, J. F., Sebert, S. P., & Lekadir, K. (2023) · International Journal of Medical Informatics · doi:10.1016/j.ijmedinf.2023.105209 · Cited by 30 · see citing works
A transformer-based deep learning approach for fairly predicting post-liver transplant risk factors
Li, C., Jiang, X., & Zhang, K. (2024) · Journal of Biomedical Informatics · doi:10.1016/j.jbi.2023.104545 · Cited by 25 · see citing works
Prediction performance and fairness heterogeneity in cardiovascular risk models
Kartoun, U., Khurshid, S. S., Kwon, B. C., Patel, A. P., Batra, P., Philippakis, A., Khera, A. V., Ellinor, P. T., Lubitz, S. A., & Ng, K. (2022) · Scientific Reports · doi:10.1038/s41598-022-16615-3 · Cited by 22 · see citing works
Liu, Y., Joly, R., Turchioe, M. R., Benda, N. C., Hermann, A. D., Beecy, A., Pathak, J., & Zhang, Y. (2024) · Journal of the American Medical Informatics Association · doi:10.1093/jamia/ocae056 · Cited by 22 · see citing works
Can medical algorithms be fair? Three ethical quandaries and one dilemma
Bærøe, K., Gundersen, T., Henden, E., & Rommetveit, K. (2022) · BMJ Health & Care Informatics · doi:10.1136/bmjhci-2021-100445 · Cited by 22 · see citing works
Nabi, R., & Shpitser, I. (2018) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v32i1.11553 · Cited by 20 · see citing works
AI as an Ecosystem — Ensuring Generative AI Is Safe and Effective
Coiera, E. W., & Navarro, D. F. (2024) · NEJM AI · doi:10.1056/aip2400611 · Cited by 13 · see citing works
Lin, Y. C., Mallia, D., Clark-Sevilla, A. O., Catto, A., Leshchenko, A., Yan, Q., Haas, D. M., Wapner, R. J., Pe’er, I., Raja, A., et al. (2024) · BMC Pregnancy and Childbirth · doi:10.1186/s12884-024-06988-w · Cited by 13 · see citing works
Self-citationIs calibration a fairness requirement?
Loi, M., & Heitz, C. (2022) · ACM Conference on Fairness, Accountability, and Transparency (FAccT) · doi:10.1145/3531146.3533245 · Cited by 10 · see citing works
An intersectional framework for counterfactual fairness in risk prediction
Wastvedt, S., Huling, J. D., & Wolfson, J. A. (2023) · Biostatistics · doi:10.1093/biostatistics/kxad021 · Cited by 9 · see citing works
Operationalising fairness in medical algorithms
Parbhoo, S., Gichoya, J. W., Celi, L. A., & de la Hoz, M. Á. A. (2022) · BMJ Health & Care Informatics · doi:10.1136/bmjhci-2022-100617 · Cited by 8 · see citing works
Ethnic Classifications in Algorithmic Fairness: Concepts, Measures and Implications in Practice
Jaime, S., & Kern, C. A. (2024) · ACM Conference on Fairness, Accountability, and Transparency (FAccT) · doi:10.1145/3630106.3658902 · Cited by 8 · see citing works
Topaz, M., Davoudi, A., Evans, L., Sridharan, S., Song, J., Chae, S., Barrón, Y., Hobensack, M., Scharp, D., Cato, K., et al. (2024) · Journal of the American Medical Directors Association · doi:10.1016/j.jamda.2024.105417 · Cited by 8 · see citing works
Toward Realizing the Promise of AI in Precision Health Across the Spectrum of Care
Wiens, J., Spector‐Bagdady, K., & Mukherjee, B. (2024) · Annual Review of Genomics and Human Genetics · doi:10.1146/annurev-genom-010323-010230 · Cited by 7 · see citing works
Self-citationHow do you question what you read?
Godfrey, J. (2018) · doi:10.1057/978-1-352-00298-0_4 · Cited by 5 · see citing works
Mitigating Clinical Algorithmic Discrimination
Basu, A. (2024) · JAMA Health Forum · doi:10.1001/jamahealthforum.2024.4190 · Cited by 4 · see citing works
Self-citationModel-Agnostic Bias Measurement in Link Prediction
Schwertmann, L., Ravi, M. P. K., & de Melo, G. (2023) · Findings of the Association for Computational Linguistics: EACL · doi:10.18653/v1/2023.findings-eacl.121 · Cited by 3 · see citing works
Mitigating Discriminatory Biases in Success Prediction Models for Venture Capitals
Te, Y.-F., Wieland, M., Frey, M., & Gräbner, H. (2023) · IEEE Swiss Conference on Data Science (SDS) · doi:10.1109/sds57534.2023.00011 · Cited by 2 · see citing works
Flagged first, then by citations
Research works printed but not linked: 2
- [15]Adebayo JA Massachusetts institute of technology FairML: ToolBox for diagnosing bias in predictive modeling 2016 2025-02-20 http://hdl.handle.net/1721.1/108212
- [21]Hardt M Price E Srebro N Equality of Opportunity in Supervised Learning arXiv 2016 2023-10-07 https://arxiv.org/abs/1610.02413 · arXiv
Cited by
Sorted by citations
Ten Core Concepts for Ensuring Data Equity in Public Health
wang, Y., Boyd, A. E., Rountree, L., Ren, Y., Nyhan, K., Nagar, R., Higginbottom, J., Ranney, M. L., Parikh, H., & Mukherjee, B. (2026) · JAMA Health Forum · doi:10.1001/jamahealthforum.2025.6031 · Cited by 5 · see citing works
Self-citationJiang, Z., Chen, H., Wu, Y., Qin, Y., Pei, C., Zeng, D., Sheng, B., & Wong, T. Y. (2026) · Intelligent Medicine · doi:10.1016/j.imed.2026.01.001 · Cited by 5 · see citing works
AI-Driven Transformation: Methodologies, Governance, and Strategic Imperatives
León, M. (2025) · WSEAS TRANSACTIONS ON BUSINESS AND ECONOMICS · doi:10.37394/23207.2025.22.219 · Cited by 3 · see citing works
Li, H., Xu, Z., Cen, Y., & Liu, X. (2026) · Frontiers in Medicine · doi:10.3389/fmed.2026.1831220
“Where feasible, model comparison should therefore extend beyond AUC to include calibration, subgroup error distribution, and decision-curve-oriented evaluation of clinical usefulness (9, 10, 16, 17).”
Bioethics of Kidney Transplantation in the Older Adult Population
Long, J. J., Diwan, T. S., Donnelly, C. B., Liyanage, L. N., Long, M. J., & Kimberly, L. L. (2025) · Current Transplantation Reports · doi:10.1007/s40472-025-00497-1
FairCareNLP: An AI-driven patient review analyzer for healthcare
Esfahani, S. M. P. M., Seeman, D., Dharma, C., Noaeen, M., Guha, S., & Abad, Z. S. H. (2026) · PLoS ONE · doi:10.1371/journal.pone.0337676
(2025) · International Journal of Chemical Engineering and Materials · doi:10.37394/232031.2025.4.11
Development and Validation of the Intensive Documentation Index for ICU Mortality PredictionPreprint
Collier, A. M., & Shalhout, S. Z. (2026) · medRxiv · doi:10.64898/2026.02.10.26345827
AI-Driven Multi-Disease Risk Prediction Using Routine Laboratory Data in Indian Healthcare Settings
Srinivasan, N., S, H. A. A., & Harish Ragavendar, S. (2025) · International Conference on Advancement in Futuristic Technologies (ICAFT) · doi:10.1109/icaft66710.2025.11453178
Hung, M. J., Strickler, I., Jensen, A., Gunnell, B., Parry, A., Dyal, S., & Lipsky, M. S. (2026) · Journal of Imaging Informatics in Medicine · doi:10.1007/s10278-026-01974-4
Joshi, U., & Lanzas, C. (2026) · Antimicrobial Stewardship & Healthcare Epidemiology · doi:10.1017/ash.2026.10747
Lu, L., J, J., Ma, Y., Pang, Y., & Cheng, P. (2026) · Healthcare · doi:10.3390/healthcare14152249
Адылова, Ф. Т., Тригулова, Р. Х., Раxимович, Д. Р., & Хайдарова, Г. А. (2026) · Кардиология в Беларуси · doi:10.34883/pi.2026.18.5.010
- Authors
- Rountree, L., Lin, Y.-T., Liu, C., Salvatore, M., Admon, A., Nallamothu, B., Singh, K., Basu, A., Bu, F., & Mukherjee, B.
- Affiliations
- University of Michigan (US)
- Michigan State University (US)
- University of Pennsylvania (US)
- University of California San Diego (US)
- University of Washington (US)
- Yale University (US)
- Corresponding author
- Mukherjee, Bhramar
- ORCID
- Venue
- Online Journal of Public Health Informatics
- Publisher
- JMIR Publications Inc.
- Published
- 2025-03-19
- Volume
- 17
- Pages
- e66598
- ISSN
- 1947-2579 (ISSN-L)
- PMC
- PMC11966066
- OpenAlex
- W4407634575
- Cited by
- 13
- References
- 38
- Citations by year
- 2026: 9 · 2025: 4
- FWCI
- 2.01
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- cc-by
- Locations
- Online Journal of Public Health Informatics (publishedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Funders
- National Heart, Lung, and Blood Institute (K08HL155407)
- National Cancer Institute (UG3CA267907, UH3CA267907)
- Topics
- Social Sciences › Safety Research › Ethics and Social Impacts of AI
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Physical Sciences › Computer Science › Artificial Intelligence › Machine Learning in Healthcare
- Keywords
- fairness metrics, clinical prediction model, algorithmic fairness, precision health, cardiovascular disease, COVID-19 pandemic, racial and ethnic disparities, training data representativeness, health equity, implementation framework
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
