Consistency of variety of machine learning and statistical models in predicting clinical risks of individual patients: longitudinal cohort study using cardiovascular disease as exemplar
Open accessThe DOI matches its Crossref record; 22 of the 22 references it cites were found in a registry or an index.
Abstract
Objective: To assess the consistency of machine learning and statistical techniques in predicting individual level and population level risks of cardiovascular disease and the effects of censoring on risk predictions. Design: Longitudinal cohort study from 1 January 1998 to 31 December 2018. Setting and participants: 3.6 million patients from the Clinical Practice Research Datalink registered at 391 general practices in England with linked hospital admission and mortality records. Main outcome measures: Model performance including discrimination, calibration, and consistency of individual risk prediction for the same patients among models with comparable model performance. 19 different prediction techniques were applied, including 12 families of machine learning models (grid searched for best models), three Cox proportional hazards models (local fitted, QRISK3, and Framingham), three parametric survival models, and one logistic model. Results: The various models had similar population level performance (C statistics of about 0.87 and similar calibration). However, the predictions for individual risks of cardiovascular disease varied widely between and within different types of machine learning and statistical models, especially in patients with higher risks. A patient with a risk of 9.5-10.5% predicted by QRISK3 had a risk of 2.9-9.2% in a random forest and 2.4-7.2% in a neural network. The differences in predicted risks between QRISK3 and a neural network ranged between –23.2% and 0.1% (95% range). Models that ignored censoring (that is, assumed censored patients to be event free) substantially underestimated risk of cardiovascular disease. Of the 223 815 patients with a cardiovascular disease risk above 7.5% with QRISK3, 57.8% would be reclassified below 7.5% when using another model. Conclusions: A variety of models predicted risks for the same patients very differently despite similar model performances. The logistic models and commonly used machine learning models should not be directly applied to the prediction of long term risks without considering censoring. Survival models that consider censoring and that are explainable, such as QRISK3, are preferable. The level of consistency within and between models should be routinely assessed before they are used for clinical decision making.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Arrieta, A. B., Díaz-Rodríguez, N., Del Ser, J., Bennetot, A., Tabik, S., Barbado, A., García, S., Gil-López, S., Molina, D., Benjamins, R., et al. (2019) · arXiv · doi:10.48550/arxiv.1910.10045 · Cited by 11,321 · cited 1×
“51 This contrasts with the concept of the “black box” in machine learning, whereby predictions cannot be explained.”
BackgroundBarredo Arrieta, A., Díaz-Rodríguez, N., Del Ser, J., Bennetot, A., Tabik, S., Barbado, A., Garcia, S., Gil-Lopez, S., Molina, D., Benjamins, R., et al. (2020) · Information Fusion · doi:10.1016/j.inffus.2019.12.012 · Cited by 10,423
Gulshan, V., Peng, L., Coram, M., Stumpe, M. C., Wu, D., Narayanaswamy, A., Venugopalan, S., Widner, K., Madams, T., Cuadros, J., et al. (2016) · JAMA · doi:10.1001/jama.2016.17216 · Cited by 7,697
Measurement in Medicine: The Analysis of Method Comparison Studies
Altman, D. G., & Bland, J. M. (1983) · The Statistician · doi:10.2307/2987937 · Cited by 4,300
Clinical Prediction Models: A Practical Approach to Development, Validation, and Updating
Steyerberg, E. W. (2019) · Statistics for Biology and Health · doi:10.1007/978-3-030-16399-0 · Cited by 3,281
Ishwaran, H., Kogalur, U. B., Blackstone, E. H., & Lauer, M. S. (2008) · The Annals of Applied Statistics · doi:10.1214/08-aoas169 · Cited by 2,714
Data Resource Profile: Clinical Practice Research Datalink (CPRD)
Herrett, E., Gallagher, A. M., Bhaskaran, K., Forbes, H., Mathur, R., van Staa, T., & Smeeth, L. (2015) · International Journal of Epidemiology · doi:10.1093/ije/dyv098 · Cited by 2,667
Christodoulou, E., Ma, J., Collins, G. S., Steyerberg, E. W., Verbakel, J. Y., & Van Calster, B. (2019) · Journal of Clinical Epidemiology · doi:10.1016/j.jclinepi.2019.02.004 · Cited by 2,119 · cited 3×
“Changing the number of predictors did not result in lower levels of inconsistencies in cardiovascular disease risk predictions with more predictors included in the models (eTable 11 and eFigure 10),”
BackgroundInfluentialAn updated coronary risk profile. A statement for health professionals.
Anderson, K. M., Wilson, P. W., Odell, P. M., & Kannel, W. B. (1991) · Circulation · doi:10.1161/01.cir.83.1.356 · Cited by 1,834
Hippisley-Cox, J., Coupland, C., & Brindle, P. (2017) · BMJ · doi:10.1136/bmj.j2099 · Cited by 1,703
Cited by
Hu, J., Xu, J., Li, M., Jiang, Z., Mao, J., Feng, L., Miao, K., Li, H., Chen, J., Bai, Z., et al. (2024) · eClinicalMedicine · doi:10.1016/j.eclinm.2023.102409 · Cited by 213
“40 The SHAP method was employed to assist feature selection.”
MethodSelf-citationLiu, T., Krentz, A., Lu, L., & Curcin, V. (2025) · European Heart Journal - Digital Health · doi:10.1093/ehjdh/ztae080 · Cited by 177
Cai, Y., Cai, Y.-Q., Tang, L.-Y., Wang, Y.-H., Gong, M., Jing, T.-C., Li, H.-J., Li-Ling, J., Hu, W., Yin, Z., et al. (2024) · BMC Medicine · doi:10.1186/s12916-024-03273-7 · Cited by 122
Clift, A. K., Dodwell, D., Lord, S., Petrou, S., Brady, M., Collins, G. S., & Hippisley-Cox, J. (2023) · BMJ · doi:10.1136/bmj-2022-073800 · Cited by 97
Young, R. A., Martin, C. M., Sturmberg, J. P., Hall, S., Bazemore, A., Kakadiaris, I. A., & Lin, S. (2024) · The Journal of the American Board of Family Medicine · doi:10.3122/jabfm.2023.230219r1 · Cited by 17
“42 An explanation for this discrepancy is that ML focuses on the strength of the correlation between variables rather than the direction of causality, 26 and ML may add little value to predictions of future events compared with traditional methods.”
Metadata
- Authors
- Li, Y., Sperrin, M., Ashcroft, D. M., & van Staa, T. P.
- Venue
- BMJ
- Publisher
- BMJ
- Published
- 2020-11-04
- Volume
- 371
- Pages
- m3919
- ISSN
- 1756-1833 (ISSN-L)
- OpenAlex
- W3096943439
- Cited by
- 135
- References
- 22
- FWCI
- 13.12
- Type
- journal article
- Language
- en
- Open access
- hybrid
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
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.