Development of machine-learning algorithms for 90-day and one-year mortality prediction in the elderly with femoral neck fractures based on the HEALTH and FAITH trials

H. Dijkstra; J. H.F. Oosterhoff; A. van de Kuit; F. F.A. Ijpma; J. H. Schwab; R. W. Poolman; S. Sprague; S. Bzovsky; M. Bhandari; M. Swiontkowski; E. H. Schemitsch; J. N. Doornberg; L. A.M. Hendrickx

doi:10.1302/2633-1462.43.BJO-2022-0162.R1

Development of machine-learning algorithms for 90-day and one-year mortality prediction in the elderly with femoral neck fractures based on the HEALTH and FAITH trials

H. Dijkstra, J. H.F. Oosterhoff, A. van de Kuit, F. F.A. Ijpma, J. H. Schwab, R. W. Poolman, S. Sprague, S. Bzovsky, M. Bhandari, M. Swiontkowski, E. H. Schemitsch, J. N. Doornberg, L. A.M. Hendrickx

Research output: Contribution to journal › Article › peer-review

2 Scopus citations

Abstract

Aims To develop prediction models using machine-learning (ML) algorithms for 90-day and one-year mortality prediction in femoral neck fracture (FNF) patients aged 50 years or older based on the Hip fracture Evaluation with Alternatives of Total Hip arthroplasty versus Hemi-arthroplasty (HEALTH) and Fixation using Alternative Implants for the Treatment of Hip fractures (FAITH) trials. Methods This study included 2,388 patients from the HEALTH and FAITH trials, with 90-day and one-year mortality proportions of 3.0% (71/2,388) and 6.4% (153/2,388), respectively. The mean age was 75.9 years (SD 10.8) and 65.9% of patients (1,574/2,388) were female. The algorithms included patient and injury characteristics. Six algorithms were developed, in-ternally validated and evaluated across discrimination (c-statistic; discriminative ability between those with risk of mortality and those without), calibration (observed outcome compared to the predicted probability), and the Brier score (composite of discrimination and calibration). Results The developed algorithms distinguished between patients at high and low risk for 90-day and one-year mortality. The penalized logistic regression algorithm had the best performance metrics for both 90-day (c-statistic 0.80, calibration slope 0.95, calibration intercept-0.06, and Brier score 0.039) and one-year (c-statistic 0.76, calibration slope 0.86, calibration intercept-0.20, and Brier score 0.074) mortality prediction in the hold-out set. Conclusion Using high-quality data, the ML-based prediction models accurately predicted 90-day and one-year mortality in patients aged 50 years or older with a FNF. The final models must be externally validated to assess generalizability to other populations, and prospectively evaluated in the process of shared decision-making.

Original language	English (US)
Pages (from-to)	168-181
Number of pages	14
Journal	Bone and Joint Open
Volume	4
Issue number	3
DOIs	https://doi.org/10.1302/2633-1462.43.BJO-2022-0162.R1
State	Published - Mar 2023
Externally published	Yes

Bibliographical note

Funding Information:
The authors disclose receipt of the following financial or material support for the research, authorship, and/or publication of this article: The HEALTH trial was supported by research grants from the Canadian Institutes of Health Research (CIHR) (MCT-90168), National Institutes of Health (NIH) (1UM1AR063386-01), ZorgOnder-zoek Nederland-medische wetensehappen (ZonMw) (17088.2503), Sophies Minde Foundation for Orthopaedic Research, McMaster Surgical Associates, and Stryker Orthopaedics. The FAITH trial was supported by research grants from the Canadian Institutes of Health Research (MOP-106630 and MCT-87771), National Institutes of Health (1R01AR055267-01A1), Stichting NutsOhra (SNO-T-0602-43), the Nether-lands Organisation for Health Research and Development (80-82310-97-11032), and Physicians’ Services Incorporated. The FAITH trial was also supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases of the National Institutes of Health (award number R01AR055267-01A1). Research reported in this publication was also supported by the County Durham and Tees Valley Comprehensive Local Research Network, which operates as part of the National Institute for Health Research Comprehensive Clinical Research Network in England. Each author certifies that there are no funding or commercial associations (consultancies, stock ownership, equity interest, patent/licensing arrangements, etc.) that might pose a conflict of interest in connection with the submitted article related to the author or any immediate family members. The FAITH trial was also supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases of the National Institutes of Health under Award Number R01AR055267-01A1. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Research reported in this publication was also supported by The County Durham & Tees Valley Comprehensive Local Research Network which operates as part of the National Institute for Health Research Comprehensive Clinical Research Network in England. The funding sources had no role in design or conduct of the study; the collection, management, analysis, or interpretation of the data; or the preparation, review, or approval of the manuscript.

Publisher Copyright:
© 2023 Author(s) et al.

Keywords

Artificial intelligence
Hip fracture
Machine learning
Prediction models
Shared decision-making

PubMed: MeSH publication types

Journal Article

Access

10.1302/2633-1462.43.BJO-2022-0162.R1

OpenUrl availability

Full text

Cite this

Dijkstra, H., Oosterhoff, J. H. F., van de Kuit, A., Ijpma, F. F. A., Schwab, J. H., Poolman, R. W., Sprague, S., Bzovsky, S., Bhandari, M., Swiontkowski, M., Schemitsch, E. H., Doornberg, J. N., & Hendrickx, L. A. M. (2023). Development of machine-learning algorithms for 90-day and one-year mortality prediction in the elderly with femoral neck fractures based on the HEALTH and FAITH trials. Bone and Joint Open, 4(3), 168-181. https://doi.org/10.1302/2633-1462.43.BJO-2022-0162.R1

Dijkstra, H, Oosterhoff, JHF, van de Kuit, A, Ijpma, FFA, Schwab, JH, Poolman, RW, Sprague, S, Bzovsky, S, Bhandari, M, Swiontkowski, M, Schemitsch, EH, Doornberg, JN & Hendrickx, LAM 2023, 'Development of machine-learning algorithms for 90-day and one-year mortality prediction in the elderly with femoral neck fractures based on the HEALTH and FAITH trials', Bone and Joint Open, vol. 4, no. 3, pp. 168-181. https://doi.org/10.1302/2633-1462.43.BJO-2022-0162.R1

@article{bbe606b81c764412a30bff95949d0b63,

title = "Development of machine-learning algorithms for 90-day and one-year mortality prediction in the elderly with femoral neck fractures based on the HEALTH and FAITH trials",

abstract = "Aims To develop prediction models using machine-learning (ML) algorithms for 90-day and one-year mortality prediction in femoral neck fracture (FNF) patients aged 50 years or older based on the Hip fracture Evaluation with Alternatives of Total Hip arthroplasty versus Hemi-arthroplasty (HEALTH) and Fixation using Alternative Implants for the Treatment of Hip fractures (FAITH) trials. Methods This study included 2,388 patients from the HEALTH and FAITH trials, with 90-day and one-year mortality proportions of 3.0% (71/2,388) and 6.4% (153/2,388), respectively. The mean age was 75.9 years (SD 10.8) and 65.9% of patients (1,574/2,388) were female. The algorithms included patient and injury characteristics. Six algorithms were developed, in-ternally validated and evaluated across discrimination (c-statistic; discriminative ability between those with risk of mortality and those without), calibration (observed outcome compared to the predicted probability), and the Brier score (composite of discrimination and calibration). Results The developed algorithms distinguished between patients at high and low risk for 90-day and one-year mortality. The penalized logistic regression algorithm had the best performance metrics for both 90-day (c-statistic 0.80, calibration slope 0.95, calibration intercept-0.06, and Brier score 0.039) and one-year (c-statistic 0.76, calibration slope 0.86, calibration intercept-0.20, and Brier score 0.074) mortality prediction in the hold-out set. Conclusion Using high-quality data, the ML-based prediction models accurately predicted 90-day and one-year mortality in patients aged 50 years or older with a FNF. The final models must be externally validated to assess generalizability to other populations, and prospectively evaluated in the process of shared decision-making.",

keywords = "Artificial intelligence, Hip fracture, Machine learning, Prediction models, Shared decision-making",

author = "H. Dijkstra and Oosterhoff, {J. H.F.} and {van de Kuit}, A. and Ijpma, {F. F.A.} and Schwab, {J. H.} and Poolman, {R. W.} and S. Sprague and S. Bzovsky and M. Bhandari and M. Swiontkowski and Schemitsch, {E. H.} and Doornberg, {J. N.} and Hendrickx, {L. A.M.}",

note = "Funding Information: The authors disclose receipt of the following financial or material support for the research, authorship, and/or publication of this article: The HEALTH trial was supported by research grants from the Canadian Institutes of Health Research (CIHR) (MCT-90168), National Institutes of Health (NIH) (1UM1AR063386-01), ZorgOnder-zoek Nederland-medische wetensehappen (ZonMw) (17088.2503), Sophies Minde Foundation for Orthopaedic Research, McMaster Surgical Associates, and Stryker Orthopaedics. The FAITH trial was supported by research grants from the Canadian Institutes of Health Research (MOP-106630 and MCT-87771), National Institutes of Health (1R01AR055267-01A1), Stichting NutsOhra (SNO-T-0602-43), the Nether-lands Organisation for Health Research and Development (80-82310-97-11032), and Physicians{\textquoteright} Services Incorporated. The FAITH trial was also supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases of the National Institutes of Health (award number R01AR055267-01A1). Research reported in this publication was also supported by the County Durham and Tees Valley Comprehensive Local Research Network, which operates as part of the National Institute for Health Research Comprehensive Clinical Research Network in England. Each author certifies that there are no funding or commercial associations (consultancies, stock ownership, equity interest, patent/licensing arrangements, etc.) that might pose a conflict of interest in connection with the submitted article related to the author or any immediate family members. The FAITH trial was also supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases of the National Institutes of Health under Award Number R01AR055267-01A1. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Research reported in this publication was also supported by The County Durham & Tees Valley Comprehensive Local Research Network which operates as part of the National Institute for Health Research Comprehensive Clinical Research Network in England. The funding sources had no role in design or conduct of the study; the collection, management, analysis, or interpretation of the data; or the preparation, review, or approval of the manuscript. Publisher Copyright: {\textcopyright} 2023 Author(s) et al.",

year = "2023",

month = mar,

doi = "10.1302/2633-1462.43.BJO-2022-0162.R1",

language = "English (US)",

volume = "4",

pages = "168--181",

journal = "Bone and Joint Open",

issn = "2633-1462",

publisher = "British Editorial Society of Bone and Joint Surgery",

number = "3",

}

TY - JOUR

T1 - Development of machine-learning algorithms for 90-day and one-year mortality prediction in the elderly with femoral neck fractures based on the HEALTH and FAITH trials

AU - Dijkstra, H.

AU - Oosterhoff, J. H.F.

AU - van de Kuit, A.

AU - Ijpma, F. F.A.

AU - Schwab, J. H.

AU - Poolman, R. W.

AU - Sprague, S.

AU - Bzovsky, S.

AU - Bhandari, M.

AU - Swiontkowski, M.

AU - Schemitsch, E. H.

AU - Doornberg, J. N.

AU - Hendrickx, L. A.M.

N1 - Funding Information: The authors disclose receipt of the following financial or material support for the research, authorship, and/or publication of this article: The HEALTH trial was supported by research grants from the Canadian Institutes of Health Research (CIHR) (MCT-90168), National Institutes of Health (NIH) (1UM1AR063386-01), ZorgOnder-zoek Nederland-medische wetensehappen (ZonMw) (17088.2503), Sophies Minde Foundation for Orthopaedic Research, McMaster Surgical Associates, and Stryker Orthopaedics. The FAITH trial was supported by research grants from the Canadian Institutes of Health Research (MOP-106630 and MCT-87771), National Institutes of Health (1R01AR055267-01A1), Stichting NutsOhra (SNO-T-0602-43), the Nether-lands Organisation for Health Research and Development (80-82310-97-11032), and Physicians’ Services Incorporated. The FAITH trial was also supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases of the National Institutes of Health (award number R01AR055267-01A1). Research reported in this publication was also supported by the County Durham and Tees Valley Comprehensive Local Research Network, which operates as part of the National Institute for Health Research Comprehensive Clinical Research Network in England. Each author certifies that there are no funding or commercial associations (consultancies, stock ownership, equity interest, patent/licensing arrangements, etc.) that might pose a conflict of interest in connection with the submitted article related to the author or any immediate family members. The FAITH trial was also supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases of the National Institutes of Health under Award Number R01AR055267-01A1. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Research reported in this publication was also supported by The County Durham & Tees Valley Comprehensive Local Research Network which operates as part of the National Institute for Health Research Comprehensive Clinical Research Network in England. The funding sources had no role in design or conduct of the study; the collection, management, analysis, or interpretation of the data; or the preparation, review, or approval of the manuscript. Publisher Copyright: © 2023 Author(s) et al.

PY - 2023/3

Y1 - 2023/3

N2 - Aims To develop prediction models using machine-learning (ML) algorithms for 90-day and one-year mortality prediction in femoral neck fracture (FNF) patients aged 50 years or older based on the Hip fracture Evaluation with Alternatives of Total Hip arthroplasty versus Hemi-arthroplasty (HEALTH) and Fixation using Alternative Implants for the Treatment of Hip fractures (FAITH) trials. Methods This study included 2,388 patients from the HEALTH and FAITH trials, with 90-day and one-year mortality proportions of 3.0% (71/2,388) and 6.4% (153/2,388), respectively. The mean age was 75.9 years (SD 10.8) and 65.9% of patients (1,574/2,388) were female. The algorithms included patient and injury characteristics. Six algorithms were developed, in-ternally validated and evaluated across discrimination (c-statistic; discriminative ability between those with risk of mortality and those without), calibration (observed outcome compared to the predicted probability), and the Brier score (composite of discrimination and calibration). Results The developed algorithms distinguished between patients at high and low risk for 90-day and one-year mortality. The penalized logistic regression algorithm had the best performance metrics for both 90-day (c-statistic 0.80, calibration slope 0.95, calibration intercept-0.06, and Brier score 0.039) and one-year (c-statistic 0.76, calibration slope 0.86, calibration intercept-0.20, and Brier score 0.074) mortality prediction in the hold-out set. Conclusion Using high-quality data, the ML-based prediction models accurately predicted 90-day and one-year mortality in patients aged 50 years or older with a FNF. The final models must be externally validated to assess generalizability to other populations, and prospectively evaluated in the process of shared decision-making.

AB - Aims To develop prediction models using machine-learning (ML) algorithms for 90-day and one-year mortality prediction in femoral neck fracture (FNF) patients aged 50 years or older based on the Hip fracture Evaluation with Alternatives of Total Hip arthroplasty versus Hemi-arthroplasty (HEALTH) and Fixation using Alternative Implants for the Treatment of Hip fractures (FAITH) trials. Methods This study included 2,388 patients from the HEALTH and FAITH trials, with 90-day and one-year mortality proportions of 3.0% (71/2,388) and 6.4% (153/2,388), respectively. The mean age was 75.9 years (SD 10.8) and 65.9% of patients (1,574/2,388) were female. The algorithms included patient and injury characteristics. Six algorithms were developed, in-ternally validated and evaluated across discrimination (c-statistic; discriminative ability between those with risk of mortality and those without), calibration (observed outcome compared to the predicted probability), and the Brier score (composite of discrimination and calibration). Results The developed algorithms distinguished between patients at high and low risk for 90-day and one-year mortality. The penalized logistic regression algorithm had the best performance metrics for both 90-day (c-statistic 0.80, calibration slope 0.95, calibration intercept-0.06, and Brier score 0.039) and one-year (c-statistic 0.76, calibration slope 0.86, calibration intercept-0.20, and Brier score 0.074) mortality prediction in the hold-out set. Conclusion Using high-quality data, the ML-based prediction models accurately predicted 90-day and one-year mortality in patients aged 50 years or older with a FNF. The final models must be externally validated to assess generalizability to other populations, and prospectively evaluated in the process of shared decision-making.

KW - Artificial intelligence

KW - Hip fracture

KW - Machine learning

KW - Prediction models

KW - Shared decision-making

UR - http://www.scopus.com/inward/record.url?scp=85151926037&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85151926037&partnerID=8YFLogxK

U2 - 10.1302/2633-1462.43.BJO-2022-0162.R1

DO - 10.1302/2633-1462.43.BJO-2022-0162.R1

M3 - Article

C2 - 37051847

AN - SCOPUS:85151926037

SN - 2633-1462

VL - 4

SP - 168

EP - 181

JO - Bone and Joint Open

JF - Bone and Joint Open

IS - 3

ER -

Development of machine-learning algorithms for 90-day and one-year mortality prediction in the elderly with femoral neck fractures based on the HEALTH and FAITH trials

Abstract

Bibliographical note

Keywords

PubMed: MeSH publication types

Access

OpenUrl availability

Other files and links

Fingerprint

Cite this