Machine learning algorithms performed no better than regression models for prognostication in traumatic brain injury
Open accessThe DOI matches its Crossref record; 22 of the 22 references it cites were found in a registry or an index.
Abstract
OBJECTIVE: We aimed to explore the added value of common machine learning (ML) algorithms for prediction of outcome for moderate and severe traumatic brain injury. STUDY DESIGN AND SETTING: We performed logistic regression (LR), lasso regression, and ridge regression with key baseline predictors in the IMPACT-II database (15 studies, n = 11,022). ML algorithms included support vector machines, random forests, gradient boosting machines, and artificial neural networks and were trained using the same predictors. To assess generalizability of predictions, we performed internal, internal-external, and external validation on the recent CENTER-TBI study (patients with Glasgow Coma Scale <13, n = 1,554). Both calibration (calibration slope/intercept) and discrimination (area under the curve) was quantified. RESULTS: In the IMPACT-II database, 3,332/11,022 (30%) died and 5,233(48%) had unfavorable outcome (Glasgow Outcome Scale less than 4). In the CENTER-TBI study, 348/1,554(29%) died and 651(54%) had unfavorable outcome. Discrimination and calibration varied widely between the studies and less so between the studied algorithms. The mean area under the curve was 0.82 for mortality and 0.77 for unfavorable outcomes in the CENTER-TBI study. CONCLUSION: ML algorithms may not outperform traditional regression approaches in a low-dimensional setting for outcome prediction after moderate or severe traumatic brain injury. Similar to regression-based prediction models, ML algorithms should be rigorously validated to ensure applicability to new populations.
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
Breiman, L. (2001) · Machine Learning · doi:10.1023/a:1010933404324 · Cited by 133,538
DeLong, E. R., DeLong, D. M., & Clarke-Pearson, D. L. (1988) · Biometrics · doi:10.2307/2531595 · Cited by 22,649
Assessing the Performance of Prediction Models: A Framework for Traditional and Novel Measures
Steyerberg, E. W., Vickers, A. J., Cook, N. R., Gerds, T., Gonen, M., Obuchowski, N., Pencina, M. J., & Kattan, M. W. (2010) · Epidemiology · doi:10.1097/ede.0b013e3181c30fb2 · Cited by 5,170
Self-citationGradient boosting machines, a tutorial
Natekin, A., & Knoll, A. (2013) · Frontiers in Neurorobotics · doi:10.3389/fnbot.2013.00021 · Cited by 3,887 · cited 1×
“Frequently used methods are support vector machines (SVMs) [6], deep neural networks (NNs) [7], random forests (RFs) [8], and gradient boosting machine (GBM) [9].”
MethodClinical 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
Self-citationScalable and accurate deep learning with electronic health records
Rajkomar, A., Oren, E., Chen, K., Dai, A. M., Hajaj, N., Hardt, M., Liu, P. J., Liu, X., Marcus, J., Sun, M., et al. (2018) · npj Digital Medicine · doi:10.1038/s41746-018-0029-1 · Cited by 2,648 · cited 1×
“This potentially limits the performance of ML techniques because it has been suggested that flexible ML techniques perform better than traditional regression techniques when a large number of predictors are being considered, that is, high-dimensional data [50,51].”
BackgroundTraumatic brain injury: integrated approaches to improve prevention, clinical care, and research
Maas, A. I. R., Menon, D. K., Adelson, P. D., Andelic, N., Bell, M. J., Belli, A., Bragge, P., Brazinova, A., Büki, A., Chesnut, R. M., et al. (2017) · The Lancet Neurology · doi:10.1016/s1474-4422(17)30371-x · Cited by 2,609 · cited 1×
“Traumatic brain injury (TBI) is a common disease, with a significant societal burden [1]: TBI is estimated to be responsible for around 300 hospital admissions and 12 deaths per 100,000 persons per year in Europe [2].”
BackgroundSelf-citationBig Data and Machine Learning in Health Care
Beam, A. L., & Kohane, I. S. (2018) · JAMA · doi:10.1001/jama.2017.18391 · Cited by 2,506 · cited 1×
“This potentially limits the performance of ML techniques because it has been suggested that flexible ML techniques perform better than traditional regression techniques when a large number of predictors are being considered, that is, high-dimensional data [50,51].”
BackgroundChristodoulou, 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 2×
“A recent systematic review compared flexible ML methods to traditional statistical techniques in relatively small data sets (median sample size was 1,250) and did not find incremental value [14].”
MethodSelf-citationFlexible Imputation of Missing Data
van Buuren, S. (2012) · doi:10.1201/b11826 · Cited by 2,022
Towards better clinical prediction models: seven steps for development and an ABCD for validation
Steyerberg, E. W., & Vergouwe, Y. (2014) · European Heart Journal · doi:10.1093/eurheartj/ehu207 · Cited by 1,928
Self-citationSteyerberg, E. W. (2009) · Statistics for Biology and Health · doi:10.1007/978-0-387-77244-8 · Cited by 1,491
Self-citationBig data and machine learning algorithms for health-care delivery
Ngiam, K. Y., & Khor, I. W. (2019) · The Lancet Oncology · doi:10.1016/s1470-2045(19)30149-4 · Cited by 1,424
Prediction models need appropriate internal, internal-external, and external validation
Steyerberg, E. W., & Harrell, F. E. (2016) · Journal of Clinical Epidemiology · doi:10.1016/j.jclinepi.2015.04.005 · Cited by 1,304 · cited 1×
“This procedure has been referred to as ‘Internal-external cross-validation’ [33,34].”
MethodSelf-citationRisk prediction models: II. External validation, model updating, and impact assessment
Moons, K. G. M., Kengne, A. P., Grobbee, D. E., Royston, P., Vergouwe, Y., Altman, D. G., & Woodward, M. (2012) · Heart · doi:10.1136/heartjnl-2011-301247 · Cited by 1,195
Classification of traumatic brain injury for targeted therapies
Saatman, K. E., Duhaime, A.-C., Bullock, R., Maas, A. I. R., Valadka, A., & Manley, G. T. (2008) · Journal of Neurotrauma · doi:10.1089/neu.2008.0586 · Cited by 1,141 · cited 1×
“TBI is a heterogeneous disease in terms of phenotype and prognosis [3].”
BackgroundA calibration hierarchy for risk models was defined: from utopia to empirical data
Van Calster, B., Nieboer, D., Vergouwe, Y., De Cock, B., Pencina, M. J., & Steyerberg, E. W. (2016) · Journal of Clinical Epidemiology · doi:10.1016/j.jclinepi.2015.12.005 · Cited by 910 · cited 2×
“Poor calibration of prediction models may lead to harmful decision-making when applying these models [22 e 24].”
MethodSelf-citationModern modelling techniques are data hungry: a simulation study for predicting dichotomous endpoints
van der Ploeg, T., Austin, P. C., & Steyerberg, E. W. (2014) · BMC Medical Research Methodology · doi:10.1186/1471-2288-14-137 · Cited by 809 · cited 2×
“This was perhaps to be expected because modern ML methods are known to be data hungry compared with classical statistical techniques [13,46].”
BackgroundSelf-citationTwo Further Applications of a Model for Binary Regression
Cox, D. R. (1958) · Biometrika · doi:10.2307/2333203 · Cited by 557
Two further applications of a model for binary regression
COX, D. R. (1958) · Biometrika · doi:10.1093/biomet/45.3-4.562 · Cited by 556 · cited 1×
“First, calibration was examined graphically and quantified using a calibration slope and the calibration intercept: the calibration test proposed by Cox [35].”
MethodIMPACT Database of Traumatic Brain Injury: Design And Description
Marmarou, A., Lu, J., Butcher, I., McHugh, G. S., Mushkudiani, N. A., Murray, G. D., Steyerberg, E. W., & Maas, A. I. R. (2007) · Journal of Neurotrauma · doi:10.1089/neu.2006.0036 · Cited by 236 · cited 1×
“12), which were conducted between 1984 and 2004 [26].”
BackgroundSelf-citationDelahanty, R. J., Kaufman, D., & Jones, S. S. (2018) · Critical Care Medicine · doi:10.1097/ccm.0000000000003011 · Cited by 105 · cited 1×
“Published ML-based prediction algorithms often include a large number of predictors, sometimes with the suggestion to result in high discriminative performance [52,53].”
Background
Cited by
Machine learning in medical applications: a review of state-of-the-art methods
Shehab, M., Abualigah, L., Shambour, Q., Abu-Hashem, M. A., Shambour, M. K. Y., Alsalibi, A. I., & Gandomi, A. H. (2022) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2022.105458 · Cited by 542
The promise of machine learning in predicting treatment outcomes in psychiatry
Chekroud, A. M., Bondar, J., Delgadillo, J., Doherty, G., Wasil, A., Fokkema, M., Cohen, Z., Belgrave, D., DeRubeis, R., Iniesta, R., et al. (2021) · World Psychiatry · doi:10.1002/wps.20882 · Cited by 531
Steps to avoid overuse and misuse of machine learning in clinical research
Volovici, V., Syn, N. L., Ercole, A., Zhao, J. J., & Liu, N. (2022) · Nature Medicine · doi:10.1038/s41591-022-01961-6 · Cited by 110
Self-citationDe Hond, A., Raven, W., Schinkelshoek, L., Gaakeer, M., Ter Avest, E., Sir, O., Lameijer, H., Hessels, R. A., Reijnen, R., De Jonge, E., et al. (2021) · International Journal of Medical Informatics · doi:10.1016/j.ijmedinf.2021.104496 · Cited by 66
“However, recent articles [22,23] state that, on average, the performance of ML was no different from that of logistic regression.”
MethodResultSelf-citationZhang, H., Chen, D., Shao, J., Zou, P., Cui, N., Tang, L., Wang, X., Wang, D., Wu, J., & Ye, Z. (2021) · Risk Management and Healthcare Policy · doi:10.2147/rmhp.s328180 · Cited by 16
Metadata
- Authors
- Gravesteijn, B. Y., Nieboer, D., Ercole, A., Lingsma, H. F., Nelson, D., van Calster, B., Steyerberg, E. W., Åkerlund, C., Amrein, K., Andelic, N., Andreassen, L., Anke, A., Antoni, A., Audibert, G., Azouvi, P., Azzolini, M. L., Bartels, R., Barzó, P., Beauvais, R., Beer, R., Bellander, B.-M., Belli, A., Benali, H., Berardino, M., Beretta, L., Blaabjerg, M., Bragge, P., Brazinova, A., Brinck, V., Brooker, J., Brorsson, C., Buki, A., Bullinger, M., Cabeleira, M., Caccioppola, A., Calappi, E., Calvi, M. R., Cameron, P., Lozano, G. C., Carbonara, M., Chevallard, G., Chieregato, A., Citerio, G., Cnossen, M., Coburn, M., Coles, J., Cooper, D. J., Correia, M., Čović, A., Curry, N., Czeiter, E., Czosnyka, M., Dahyot-Fizelier, C., Dawes, H., De Keyser, V., Degos, V., Della Corte, F., Boogert, H. D., Depreitere, B., Đilvesi, Đ., Dixit, A., Donoghue, E., Guy-Loup Dulière, J. D., Ercole, A., Esser, P., Martin Fabricius, E. E., Feigin, K. F., Frisvold, S., Furmanov, A., Gagliardo, P., Galanaud, D., Gantner, D., Gao, G., George, P., Ghuysen, A., Giga, L., Glocker, B., Golubovic, J., Gomez, P. A., Gratz, J., Gravesteijn, B., Grossi, F., Gruen, R. L., Gupta, D., Haagsma, J. A., Haitsma, I., Helbok, R., Helseth, E., Horton, L., Huijben, J., Hutchinson, P. J., Jacobs, B., Jankowski, S., Ji-yao Jiang, M. J., Jones, K., Karan, M., Kolias, A. G., Kompanje, E., Kondziella, D., Koraropoulos, E., Koskinen, L.-O., Kovács, N., Lagares, A., Lanyon, L., Laureys, S., Lecky, F., Lefering, R., Legrand, V., Lejeune, A., Levi, L., Lightfoot, R., Lingsma, H., Maas, A. I. R., Castaño-León, A. M., Maegele, M., Majdan, M., Manara, A., Manley, G., Martino, C., Maréchal, H., Mattern, J., McMahon, C., Melegh, B., Menon, D., Menovsky, T., Mulazzi, D., Muraleedharan, V., Murray, L., Nair, N., Negru, A., Nelson, D., Newcombe, V., Nieboer, D., Noirhomme, Q., Nyirádi, J., Olubukola, O., Oresic, M., Ortolano, F., Palotie, A., Parizel, P. M., Payen, J.-F., Perera, N., Perlbarg, V., Persona, P., Peul, W., Piippo-Karjalainen, A., Pirinen, M., Ples, H., Polinder, S., Pomposo, I., Posti, J. P., Puybasset, L., Radoi, A., Ragauskas, A., Raj, R., Rambadagalla, M., Real, R., Rhodes, J., Richardson, S., Richter, S., Ripatti, S., Rocka, S., Roe, C., Roise, O., Rosand, J., Rosenfeld, J. V., Rosenlund, C., Rosenthal, G., Rossaint, R., Rossi, S., Rueckert, D., Rusnák, M., Sahuquillo, J., Sakowitz, O., Sanchez-Porras, R., Sandor, J., Schäfer, N., Schmidt, S., Schoechl, H., Schoonman, G., Schou, R. F., Schwendenwein, E., Sewalt, C., Skandsen, T., Smielewski, P., Sorinola, A., Stamatakis, E., Stanworth, S., Kowark, A., Stevens, R., Stewart, W., Steyerberg, E. W., Stocchetti, N., Sundström, N., Synnot, A., Takala, R., Tamás, V., Tamosuitis, T., Taylor, M. S., Ao, B. T., Tenovuo, O., Theadom, A., Thomas, M., Tibboel, D., Timmers, M., Tolias, C., Trapani, T., Tudora, C. M., Vajkoczy, P., Vallance, S., Valeinis, E., Vámos, Z., Van der Steen, G., van der Naalt, J., van Dijck, J. T. J. M., van Essen, T. A., Van Hecke, W., van Heugten, C., Van Praag, D., Vyvere, T. V., Vanhaudenhuyse, A., van Wijk, R. P. J., Vargiolu, A., Vega, E., Velt, K., Verheyden, J., Vespa, P. M., Vik, A., Vilcinis, R., Volovici, V., von Steinbüchel, N., Voormolen, D., Vulekovic, P., Wang, K. K. W., Wiegers, E., Williams, G., Wilson, L., Winzeck, S., Wolf, S., Yang, Z., Ylén, P., Younsi, A., Zeiler, F. A., Zelinkova, V., Ziverte, A., & Zoerle, T.
- Venue
- Journal of Clinical Epidemiology
- Publisher
- Elsevier BV
- Published
- 2020-06
- Volume
- 122
- Pages
- 95-107
- ISSN
- 0895-4356 (ISSN-L)
- OpenAlex
- W3011571399
- Cited by
- 184
- References
- 22
- FWCI
- 10.94
- 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.