Development and validation of a machine learning-based model to assess probability of systemic inflammatory response syndrome in patients with severe multiple traumas.
Decision making
Machine learning
Polytrauma
SIRS
Journal
BMC medical informatics and decision making
ISSN: 1472-6947
Titre abrégé: BMC Med Inform Decis Mak
Pays: England
ID NLM: 101088682
Informations de publication
Date de publication:
27 Aug 2024
27 Aug 2024
Historique:
received:
09
05
2024
accepted:
20
08
2024
medline:
28
8
2024
pubmed:
28
8
2024
entrez:
27
8
2024
Statut:
epublish
Résumé
Systemic inflammatory response syndrome (SIRS) is a predictor of serious infectious complications, organ failure, and death in patients with severe polytrauma and is one of the reasons for delaying early total surgical treatment. To determine the risk of SIRS within 24 h after hospitalization, we developed six machine learning models. Using retrospective data about the patient, the nature of the injury, the results of general and standard biochemical blood tests, and coagulation tests, six models were developed: decision tree, random forest, logistic regression, support vector and gradient boosting classifiers, logistic regressor, and neural network. The effectiveness of the models was assessed through internal and external validation. Among the 439 selected patients with severe polytrauma in 230 (52.4%), SIRS was diagnosed within the first 24 h of hospitalization. The SIRS group was more strongly associated with class II bleeding (39.5% vs. 60.5%; OR 1.81 [95% CI: 1.23-2.65]; P = 0.0023), long-term vasopressor use (68.4% vs. 31.6%; OR 5.51 [95% CI: 2.37-5.23]; P < 0.0001), risk of acute coagulopathy (67.8% vs. 32.2%; OR 2.4 [95% CI: 1.55-3.77]; P < 0.0001), and greater risk of pneumonia (59.5% vs. 40.5%; OR 1.74 [95% CI: 1.19-2.54]; P = 0.0042), longer ICU length of stay (5 ± 6.3 vs. 2.7 ± 4.3 days; P < 0.0001) and mortality rate (64.5% vs. 35.5%; OR 10.87 [95% CI: 6.3-19.89]; P = 0.0391). Of all the models, the random forest classifier showed the best predictive ability in the internal (AUROC 0.89; 95% CI: 0.83-0.96) and external validation (AUROC 0.83; 95% CI: 0.75-0.91) datasets. The developed model made it possible to accurately predict the risk of developing SIRS in the early period after injury, allowing clinical specialists to predict patient management tactics and calculate medication and staffing needs for the patient. Level 3. The study was retrospectively registered in the ClinicalTrials.gov database of the National Library of Medicine (NCT06323096).
Sections du résumé
BACKGROUND
BACKGROUND
Systemic inflammatory response syndrome (SIRS) is a predictor of serious infectious complications, organ failure, and death in patients with severe polytrauma and is one of the reasons for delaying early total surgical treatment. To determine the risk of SIRS within 24 h after hospitalization, we developed six machine learning models.
MATERIALS AND METHODS
METHODS
Using retrospective data about the patient, the nature of the injury, the results of general and standard biochemical blood tests, and coagulation tests, six models were developed: decision tree, random forest, logistic regression, support vector and gradient boosting classifiers, logistic regressor, and neural network. The effectiveness of the models was assessed through internal and external validation.
RESULTS
RESULTS
Among the 439 selected patients with severe polytrauma in 230 (52.4%), SIRS was diagnosed within the first 24 h of hospitalization. The SIRS group was more strongly associated with class II bleeding (39.5% vs. 60.5%; OR 1.81 [95% CI: 1.23-2.65]; P = 0.0023), long-term vasopressor use (68.4% vs. 31.6%; OR 5.51 [95% CI: 2.37-5.23]; P < 0.0001), risk of acute coagulopathy (67.8% vs. 32.2%; OR 2.4 [95% CI: 1.55-3.77]; P < 0.0001), and greater risk of pneumonia (59.5% vs. 40.5%; OR 1.74 [95% CI: 1.19-2.54]; P = 0.0042), longer ICU length of stay (5 ± 6.3 vs. 2.7 ± 4.3 days; P < 0.0001) and mortality rate (64.5% vs. 35.5%; OR 10.87 [95% CI: 6.3-19.89]; P = 0.0391). Of all the models, the random forest classifier showed the best predictive ability in the internal (AUROC 0.89; 95% CI: 0.83-0.96) and external validation (AUROC 0.83; 95% CI: 0.75-0.91) datasets.
CONCLUSIONS
CONCLUSIONS
The developed model made it possible to accurately predict the risk of developing SIRS in the early period after injury, allowing clinical specialists to predict patient management tactics and calculate medication and staffing needs for the patient.
LEVEL OF EVIDENCE
METHODS
Level 3.
TRIAL REGISTRATION
BACKGROUND
The study was retrospectively registered in the ClinicalTrials.gov database of the National Library of Medicine (NCT06323096).
Identifiants
pubmed: 39192291
doi: 10.1186/s12911-024-02640-x
pii: 10.1186/s12911-024-02640-x
doi:
Banques de données
ClinicalTrials.gov
['NCT06323096']
Types de publication
Journal Article
Validation Study
Langues
eng
Sous-ensembles de citation
IM
Pagination
235Subventions
Organisme : Ministry of Science and Higher Education of the Republic of Kazakhstan
ID : AP13067824
Informations de copyright
© 2024. The Author(s).
Références
Gebhard F, Huber-Lang M. Polytrauma - Pathophysiology and management principles. Langenbeck’s Arch Surg. 2008;393:825–31. https://doi.org/10.1007/s00423-008-0334-2 .
doi: 10.1007/s00423-008-0334-2
Nicola R. Early total care versus damage control: current concepts in the Orthopedic Care of Polytrauma patients. ISRN Orthop. 2013;2013:1–9. https://doi.org/10.1155/2013/329452 .
doi: 10.1155/2013/329452
Halvachizadeh S, Baradaran L, Cinelli P, et al. How to detect a polytrauma patient at risk of complications: a validation and database analysis of four published scales. PLoS ONE. 2020;15:1–16. https://doi.org/10.1371/journal.pone.0228082 .
doi: 10.1371/journal.pone.0228082
Pape H-C, Tscherne H. Early definitive fracture fixation with Polytrauma: advantages Versus Systemic/Pulmonary consequences. Multiple organ failure. New York, NY: Springer New York; 2000. pp. 279–90.
doi: 10.1007/978-1-4612-1222-5_29
Pfeifer R, Teuben M, Andruszkow H, et al. Mortality patterns in patients with multiple trauma: a systematic review of autopsy studies. PLoS ONE. 2016;11:1–9. https://doi.org/10.1371/journal.pone.0148844 .
doi: 10.1371/journal.pone.0148844
Fröhlich M, Lefering R, Probst C, et al. Epidemiology and risk factors of multiple-organ failure after multiple trauma: an analysis of 31,154 patients from the TraumaRegister DGU. J Trauma Acute Care Surg. 2014;76:921–7. https://doi.org/10.1097/TA.0000000000000199 .
doi: 10.1097/TA.0000000000000199
pubmed: 24662853
Cimbanassi S, O’Toole R, Maegele M et al. (2020) Orthopedic injuries in patients with multiple injuries: Results of the 11th trauma update international consensus conference Milan, December 11, 2017.
Rau CS, Wu SC, Kuo PJ, et al. Polytrauma defined by the new berlin definition: a validation test based on propensity-score matching approach. Int J Environ Res Public Health. 2017;14:4–13. https://doi.org/10.3390/ijerph14091045 .
doi: 10.3390/ijerph14091045
Carlino W. Damage control resuscitation from major haemorrhage in polytrauma. Eur J Orthop Surg Traumatol. 2014;24:137–41. https://doi.org/10.1007/s00590-013-1172-7 .
doi: 10.1007/s00590-013-1172-7
pubmed: 23412314
Schwing L, Faulkner TD, Bucaro P, et al. Trauma Team activation: accuracy of Triage when Minutes Count: a synthesis of literature and performance improvement process. J Trauma Nurs. 2019;26:208–14. https://doi.org/10.1097/JTN.0000000000000450 .
doi: 10.1097/JTN.0000000000000450
pubmed: 31283750
Cole E, Gillespie S, Vulliamy P, et al. Multiple organ dysfunction after trauma. Br J Surg. 2020;107:402–12. https://doi.org/10.1002/bjs.11361 .
doi: 10.1002/bjs.11361
pubmed: 31691956
Baek JH, Kim MS, Lee JC, Lee JH. Systemic inflammation response syndrome score predicts the mortality in multiple trauma patients. Korean J Thorac Cardiovasc Surg. 2014;47:523–8. https://doi.org/10.5090/kjtcs.2014.47.6.523 .
doi: 10.5090/kjtcs.2014.47.6.523
pubmed: 25551073
pmcid: 4279832
Bochicchio GV, Napolitano LM, Joshi M, et al. Systemic inflammatory response syndrome score at admission independently predicts infection in blunt trauma patients. J Trauma - Inj Infect Crit Care. 2001;50:817–20. https://doi.org/10.1097/00005373-200105000-00007 .
doi: 10.1097/00005373-200105000-00007
Kuhne CA, Ruchholtz S, Kaiser GM, et al. Mortality in severely injured elderly trauma patients–when does age become a risk factor? World J Surg. 2005;29:1476–82. https://doi.org/10.1007/S00268-005-7796-Y .
doi: 10.1007/S00268-005-7796-Y
pubmed: 16228923
Kocuvan S, Brilej D, Stropnik D, et al. Evaluation of major trauma in elderly patients - a single trauma center analysis. Wien Klin Wochenschr. 2016;128:535–42. https://doi.org/10.1007/S00508-016-1140-4 .
doi: 10.1007/S00508-016-1140-4
pubmed: 27896468
Mörs K, Wagner N, Sturm R, et al. Enhanced pro-inflammatory response and higher mortality rates in geriatric trauma patients. Eur J Trauma Emerg Surg. 2021;47:1065–72. https://doi.org/10.1007/s00068-019-01284-1 .
doi: 10.1007/s00068-019-01284-1
pubmed: 31875239
Vourc’h M, Roquilly A, Asehnoune K. Trauma-induced damage-associated molecular patterns-mediated remote organ injury and immunosuppression in the acutely ill patient. Front Immunol. 2018;9. https://doi.org/10.3389/fimmu.2018.01330 .
Scherer J, Coimbra R, Mariani D, et al. Standards of fracture care in polytrauma: results of a Europe-wide survey by the ESTES polytrauma section. Eur J Trauma Emerg Surg. 2022. https://doi.org/10.1007/s00068-022-02126-3 .
doi: 10.1007/s00068-022-02126-3
pubmed: 36227354
pmcid: 11249422
Stonko DP, Guillamondegui OD, Fischer PE, Dennis BM. Artificial intelligence in trauma systems. Surg (United States). 2021;169:1295–9. https://doi.org/10.1016/j.surg.2020.07.038 .
doi: 10.1016/j.surg.2020.07.038
Choi A, Choi SY, Chung K, et al. Development of a machine learning-based clinical decision support system to predict clinical deterioration in patients visiting the emergency department. Sci Rep. 2023;13:1–10. https://doi.org/10.1038/s41598-023-35617-3 .
doi: 10.1038/s41598-023-35617-3
Ehrlich H, McKenney M, Elkbuli A. The niche of artificial intelligence in trauma and emergency medicine. Am J Emerg Med. 2021;45:669–70. https://doi.org/10.1016/j.ajem.2020.10.050 .
doi: 10.1016/j.ajem.2020.10.050
pubmed: 33129644
Pape HC, Lefering R, Butcher N, et al. The definition of polytrauma revisited: an international consensus process and proposal of the New Berlin definition. J Trauma Acute Care Surg. 2014;77:780–6. https://doi.org/10.1097/TA.0000000000000453 .
doi: 10.1097/TA.0000000000000453
pubmed: 25494433
Osler T, Baker SP, Long W. A modification of the Injury Severity score that both improves accuracy and simplifies Scoring. J Trauma Inj Infect Crit Care. 1997;43:922–6. https://doi.org/10.1097/00005373-199712000-00009 .
doi: 10.1097/00005373-199712000-00009
Chakraborty RK, Burns B. (2023) Systemic Inflammatory Response Syndrome.
Gennarelli TA, Wodzin E. others (2008) Abbreviated injury scale 2005: update 2008. Russ Reeder 200:2008.
Shi X, Cui Y, Wang S, et al. Development and validation of a web-based artificial intelligence prediction model to assess massive intraoperative blood loss for metastatic spinal disease using machine learning techniques. Spine J. 2024;24:146–60. https://doi.org/10.1016/j.spinee.2023.09.001 .
doi: 10.1016/j.spinee.2023.09.001
pubmed: 37704048
NeSmith EG, Weinrich SP, Andrews JO, et al. Systemic inflammatory response syndrome score and race as predictors of length of stay in the intensive care unit. Am J Crit Care. 2009;18:339–46. https://doi.org/10.4037/ajcc2009267 .
doi: 10.4037/ajcc2009267
pubmed: 19556412
Butcher NE, Balogh ZJ. The practicality of including the systemic inflammatory response syndrome in the definition of polytrauma: experience of a level one trauma centre. Injury. 2013;44:12–7. https://doi.org/10.1016/j.injury.2012.04.019 .
doi: 10.1016/j.injury.2012.04.019
pubmed: 22607995
Schefzik R, Hahn B, Schneider-Lindner V. Dissecting contributions of individual systemic inflammatory response syndrome criteria from a prospective algorithm to the prediction and diagnosis of sepsis in a polytrauma cohort. Front Med. 2023;10. https://doi.org/10.3389/fmed.2023.1227031 .
Mica L, Niggli C, Bak P, et al. Development of a Visual Analytics Tool for Polytrauma patients: Proof of Concept for a New Assessment Tool using a multiple layer Sankey Diagram in a single-center database. World J Surg. 2020;44:764–72. https://doi.org/10.1007/s00268-019-05267-6 .
doi: 10.1007/s00268-019-05267-6
pubmed: 31712843
Fachet M, Mushunuri RV, Bergmann CB, et al. Utilizing predictive machine-learning modelling unveils feature-based risk assessment system for hyperinflammatory patterns and infectious outcomes in polytrauma. Front Immunol. 2023;14:1–15. https://doi.org/10.3389/fimmu.2023.1281674 .
doi: 10.3389/fimmu.2023.1281674
Li X, Lu Y, Chen C, et al. Development and validation of a patient-specific model to predict postoperative SIRS in older patients: a two-center study. Front Public Heal. 2023;11:1–10. https://doi.org/10.3389/fpubh.2023.1145013 .
doi: 10.3389/fpubh.2023.1145013
Thompson HJ, Tkacs NC, Saatman KE, et al. Hyperthermia following traumatic brain injury: a critical evaluation. Neurobiol Dis. 2003;12:163–73. https://doi.org/10.1016/S0969-9961(02)00030-X .
doi: 10.1016/S0969-9961(02)00030-X
pubmed: 12742737
Keane RW, Hadad R, Scott XO, et al. Neural–cardiac Inflammasome Axis after Traumatic Brain Injury. Pharmaceuticals. 2023;16:1382. https://doi.org/10.3390/ph16101382 .
doi: 10.3390/ph16101382
pubmed: 37895853
pmcid: 10610322
Ma RN, He YX, Bai FP, et al. Machine Learning Model for Predicting Acute Respiratory failure in individuals with moderate-to-severe traumatic brain Injury. Front Med. 2021;8. https://doi.org/10.3389/fmed.2021.793230 .
Kerr N, de Rivero Vaccari JP, Dietrich WD, Keane RW. Neural-respiratory inflammasome axis in traumatic brain injury. Exp Neurol. 2020;323:113080. https://doi.org/10.1016/j.expneurol.2019.113080 .
doi: 10.1016/j.expneurol.2019.113080
pubmed: 31626746