1
|
Chen X, Yu B, Zhang Y, Wang X, Huang D, Gong S, Hu W. A machine learning model based on emergency clinical data predicting 3-day in-hospital mortality for stroke and trauma patients. Front Neurol 2025; 16:1512297. [PMID: 40183016 PMCID: PMC11966482 DOI: 10.3389/fneur.2025.1512297] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/16/2024] [Accepted: 03/05/2025] [Indexed: 04/05/2025] Open
Abstract
Background Accurately predicting the short-term in-hospital mortality risk for patients with stroke and TBI (Traumatic Brain Injury) is crucial for improving the quality of emergency medical care. Method This study analyzed data from 2,125 emergency admission patients with stroke and traumatic brain injury at two Grade a hospitals in China from January 2021 to March 2024. LASSO regression was used for feature selection, and the predictive performance of logistic regression was compared with six machine learning algorithms. A 70:30 ratio was applied for cross-validation, and confidence intervals were calculated using the bootstrap method. Temporal validation was performed on the best-performing model. SHAP values were employed to assess variable importance. Results The random forest algorithm excelled in predicting in-hospital 3-day mortality, achieving an AUC of 0.978 (95% CI: 0.966-0.986). Time series validation demonstrated the model's strong generalization capability, with an AUC of 0.975 (95% CI: 0.963-0.986). Key predictive factors in the final model included metabolic syndrome, NEWS2 score, Glasgow Coma Scale (GCS), whether surgery was performed, bowel movement status, potassium level (K), aspartate transaminase (AST) level, and temporal factors. SHAP value analysis further confirmed the significant contributions of these variables to the predictive outcomes. The random forest model developed in this study demonstrates good accuracy in predicting short-term in-hospital mortality rates for stroke and traumatic brain injury patients. The model integrates emergency scores, clinical signs, and key biochemical indicators, providing a comprehensive perspective for risk assessment. This approach, which incorporates emergency data, holds promise for assisting decision-making in clinical practice, thereby improving patient outcomes.
Collapse
Affiliation(s)
- Xu Chen
- Shangrao People's Hospital, Shangrao, China
| | - Bin Yu
- Shangrao People's Hospital, Shangrao, China
| | | | - Xin Wang
- Huaian Hospital of Huaian City, Huai'an, China
| | | | | | - Wei Hu
- School of Nursing, Jinzhou Medical University, Jinzhou, China
| |
Collapse
|
2
|
O'Reilly GM, Afroz A, Curtis K, Mitra B, Kim Y, Solly E, Ryder C, Hunter K, Hendrie DV, Rushworth N, Tee J, Fitzgerald MC. The determinants for death in hospital following moderate to severe traumatic brain injury in Australia. Emerg Med Australas 2025; 37:e14562. [PMID: 39844697 PMCID: PMC11755221 DOI: 10.1111/1742-6723.14562] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/01/2024] [Revised: 12/30/2024] [Accepted: 01/06/2025] [Indexed: 01/24/2025]
Abstract
OBJECTIVES To establish the determinants of death in hospital for patients with moderate to severe traumatic brain injury (TBI) in Australia. DESIGN, SETTING, PARTICIPANTS Retrospective analysis of Australia New Zealand Trauma Registry (ANZTR) data. Cases were included if they presented to a participating hospital between 1 July 2015 and 30 June 2020 and had an Abbreviated Injury Severity (AIS) score - head greater than 2. MAIN OUTCOME MEASURES Death in hospital. RESULTS There were 16 350 patients. Their mean age was 51 years and 71% were male. After adjusting for measures of injury severity, there was an increased odds of in-hospital death for patients whose injury occurred outside daylight hours or first mode of transport was road ambulance, who were not transferred from another hospital, had an endotracheal tube placed prior to definitive hospital arrival or received their definitive hospital care outside Victoria. CONCLUSION Among people presenting to a major trauma hospital in Australia following moderate to severe TBI, there were multiple factors independently associated with death in hospital. The potentially modifiable determinants of in-hospital death included out-of-hours access to emergency care, mode of transfer from the scene of the injury, prior facility care and pre-definitive hospital endotracheal intubation.
Collapse
Affiliation(s)
- Gerard M O'Reilly
- National Trauma Research Institute, Alfred HealthMelbourneVictoriaAustralia
- Emergency and Trauma Centre, Alfred HealthMelbourneVictoriaAustralia
- School of Public Health and Preventive Medicine, Monash UniversityMelbourneVictoriaAustralia
| | - Afsana Afroz
- National Trauma Research Institute, Alfred HealthMelbourneVictoriaAustralia
| | - Kate Curtis
- Susan Wakil School of Nursing and Midwifery, Faculty of Medicine and Health, University of SydneySydneyNew South WalesAustralia
- Emergency DepartmentWollongong Hospital, Illawarra Shoalhaven Local Health DistrictWollongongNew South WalesAustralia
- The George Institute for Global Health, UNSWSydneyNew South WalesAustralia
| | - Biswadev Mitra
- Emergency and Trauma Centre, Alfred HealthMelbourneVictoriaAustralia
- School of Public Health and Preventive Medicine, Monash UniversityMelbourneVictoriaAustralia
| | - Yesul Kim
- National Trauma Research Institute, Alfred HealthMelbourneVictoriaAustralia
- School of Translational Medicine, Monash UniversityMelbourneVictoriaAustralia
| | - Emma Solly
- National Trauma Research Institute, Alfred HealthMelbourneVictoriaAustralia
| | - Courtney Ryder
- The George Institute for Global Health, UNSWSydneyNew South WalesAustralia
- College of Medicine and Public Health, Flinders UniversityAdelaideSouth AustraliaAustralia
| | - Kate Hunter
- The George Institute for Global Health, UNSWSydneyNew South WalesAustralia
| | - Delia V Hendrie
- School of Population Health, Curtin UniversityPerthWestern AustraliaAustralia
| | | | - Jin Tee
- National Trauma Research Institute, Alfred HealthMelbourneVictoriaAustralia
- School of Translational Medicine, Monash UniversityMelbourneVictoriaAustralia
- Neurosurgery, Alfred HealthMelbourneVictoriaAustralia
| | - Mark C Fitzgerald
- National Trauma Research Institute, Alfred HealthMelbourneVictoriaAustralia
- School of Translational Medicine, Monash UniversityMelbourneVictoriaAustralia
- Trauma Service, Alfred HealthMelbourneVictoriaAustralia
| |
Collapse
|
3
|
McIver R, Erdogan M, Parker R, Evans A, Green R, Gomez D, Johnston T. Effect of trauma quality improvement initiatives on outcomes and costs at community hospitals: A scoping review. Injury 2024; 55:111492. [PMID: 38531721 DOI: 10.1016/j.injury.2024.111492] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 12/31/2023] [Revised: 03/05/2024] [Accepted: 03/06/2024] [Indexed: 03/28/2024]
Abstract
BACKGROUND Due to complex geography and resource constraints, trauma patients are often initially transported to community or rural facilities rather than a larger Level I or II trauma center. The objective of this scoping review was to synthesize evidence on interventions that improved the quality of trauma care and/or reduced healthcare costs at non-Level I or II facilities. METHODS A scoping review was performed to identify studies implementing a Quality Improvement (QI) initiative at a non-major trauma center (i.e., non-Level I or II trauma center [or equivalent]). We searched 3 electronic databases (MEDLINE, Embase, CINAHL) and the grey literature (relevant networks, organizations/associations). Methodological quality was evaluated using NIH and JBI study quality assessment tools. Studies were included if they evaluated the effect of implementing a trauma care QI initiative on one or more of the following: 1) trauma outcomes (mortality, morbidity); 2) system outcomes (e.g., length of stay [LOS], transfer times, provider factors); 3) provider knowledge or perception; or 4) healthcare costs. Pediatric trauma, pre-hospital and tele-trauma specific studies were excluded. RESULTS Of 1046 data sources screened, 36 were included for full review (29 journal articles, 7 abstracts/posters without full text). Educational initiatives including the Rural Trauma Team Development Course and the Advanced Trauma Life Support course were the most common QI interventions investigated. Study outcomes included process metrics such as transfer time to tertiary care and hospital LOS, along with measures of provider perception and knowledge. Improvement in mortality was reported in a single study evaluating the impact of establishing a dedicated trauma service at a community hospital. CONCLUSIONS Our review captured a broad spectrum of trauma QI projects implemented at non-major trauma centers. Educational interventions did result in process outcome improvements and high rates of self-reported improvements in trauma care. Given the heterogeneous capabilities of community and rural hospitals, there is no panacea for trauma QI at these facilities. Future research should focus on patient outcomes like mortality and morbidity, and locally relevant initiatives.
Collapse
Affiliation(s)
- Reba McIver
- Dalhousie University, School of Medicine, Halifax, NS, Canada.
| | - Mete Erdogan
- Nova Scotia Health Trauma Program, Halifax, NS, Canada
| | - Robin Parker
- Dalhousie University Libraries, Halifax, NS, Canada
| | - Allyson Evans
- Dalhousie University, School of Medicine, Halifax, NS, Canada
| | - Robert Green
- Nova Scotia Health Trauma Program, Halifax, NS, Canada; Dalhousie University, Faculty of Medicine, Department of Emergency Medicine, Halifax, NS, Canada; Dalhousie University, Faculty of Medicine, Department of Critical Care, Halifax, NS, Canada
| | - David Gomez
- Division of General Surgery, St. Michael's Hospital, Unity Health Toronto, Toronto, ON, Canada
| | - Tyler Johnston
- Dalhousie University, Faculty of Medicine, Department of Emergency Medicine, Halifax, NS, Canada
| |
Collapse
|
4
|
Bulte CSE, Mansvelder FJ, Loer SA, Bloemers FW, Den Hartog D, Van Lieshout EMM, Hoogerwerf N, van der Naalt J, Absalom AR, Peerdeman SM, Giannakopoulos GF, Schwarte LA, Schober P, Bossers SM. Effect of Daytime versus Nighttime on Prehospital Care and Outcomes after Severe Traumatic Brain Injury. J Clin Med 2024; 13:2249. [PMID: 38673522 PMCID: PMC11051010 DOI: 10.3390/jcm13082249] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/15/2024] [Revised: 03/28/2024] [Accepted: 04/10/2024] [Indexed: 04/28/2024] Open
Abstract
Background/Objectives: Severe traumatic brain injury (TBI) is a frequent cause of morbidity and mortality worldwide. In the Netherlands, suspected TBI is a criterion for the dispatch of the physician-staffed helicopter emergency medical services (HEMS) which are operational 24 h per day. It is unknown if patient outcome is influenced by the time of day during which the incident occurs. Therefore, we investigated the association between the time of day of the prehospital treatment of severe TBI and 30-day mortality. Methods: A retrospective analysis of prospectively collected data from the BRAIN-PROTECT study was performed. Patients with severe TBI treated by one of the four Dutch helicopter emergency medical services were included and followed up to one year. The association between prehospital treatment during day- versus nighttime, according to the universal daylight period, and 30-day mortality was analyzed with multivariable logistic regression. A planned subgroup analysis was performed in patients with TBI with or without any other injury. Results: A total of 1794 patients were included in the analysis, of which 1142 (63.7%) were categorized as daytime and 652 (36.3%) as nighttime. Univariable analysis showed a lower 30-day mortality in patients with severe TBI treated during nighttime (OR 0.74, 95% CI 0.60-0.91, p = 0.004); this association was no longer present in the multivariable model (OR 0.82, 95% CI 0.59-1.16, p = 0.262). In a subgroup analysis, no association was found between mortality rates and the time of prehospital treatment in patients with combined injuries (TBI and any other injury). Patients with isolated TBI had a lower mortality rate when treated during nighttime than when treated during daytime (OR 0.51, 95% CI 0.34-0.76, p = 0.001). Within the whole cohort, daytime versus nighttime treatments were not associated with differences in functional outcome defined by the Glasgow Outcome Scale. Conclusions: In the overall study population, no difference was found in 30-day mortality between patients with severe TBI treated during day or night in the multivariable model. Patients with isolated severe TBI had lower mortality rates at 30 days when treated at nighttime.
Collapse
Affiliation(s)
- Carolien S. E. Bulte
- Department of Anesthesiology, Amsterdam University Medical Center, Vrije Universiteit Amsterdam, 1081 HV Amsterdam, The Netherlands; (F.J.M.); (S.A.L.); (L.A.S.); (P.S.); (S.M.B.)
- Helicopter Emergency Medical Service Lifeliner 1, 1045 AR Amsterdam, The Netherlands;
| | - Floor J. Mansvelder
- Department of Anesthesiology, Amsterdam University Medical Center, Vrije Universiteit Amsterdam, 1081 HV Amsterdam, The Netherlands; (F.J.M.); (S.A.L.); (L.A.S.); (P.S.); (S.M.B.)
| | - Stephan A. Loer
- Department of Anesthesiology, Amsterdam University Medical Center, Vrije Universiteit Amsterdam, 1081 HV Amsterdam, The Netherlands; (F.J.M.); (S.A.L.); (L.A.S.); (P.S.); (S.M.B.)
| | - Frank W. Bloemers
- Department of Surgery, Amsterdam University Medical Center, Vrije Universiteit Amsterdam, 1081 HV Amsterdam, The Netherlands;
| | - Dennis Den Hartog
- Trauma Research Unit Department of Surgery, Erasmus MC, University Medical Center Rotterdam, 3015 GD Rotterdam, The Netherlands; (D.D.H.); (E.M.M.V.L.)
| | - Esther M. M. Van Lieshout
- Trauma Research Unit Department of Surgery, Erasmus MC, University Medical Center Rotterdam, 3015 GD Rotterdam, The Netherlands; (D.D.H.); (E.M.M.V.L.)
| | - Nico Hoogerwerf
- Department of Anesthesiology, Radboud University Medical Center, 6525 GA Nijmegen, The Netherlands;
- Helicopter Emergency Medical Service Lifeliner 3, 5408 SM Volkel, The Netherlands
| | - Joukje van der Naalt
- Department of Neurology, University Medical Center Groningen, University of Groningen, 9713 GZ Groningen, The Netherlands;
| | - Anthony R. Absalom
- Department of Anesthesiology, University Medical Center Groningen, University of Groningen, 9713 GZ Groningen, The Netherlands;
| | - Saskia M. Peerdeman
- Department of Neurosurgery, Amsterdam University Medical Center, Vrije Universiteit Amsterdam, 1081 HV Amsterdam, The Netherlands
| | - Georgios F. Giannakopoulos
- Helicopter Emergency Medical Service Lifeliner 1, 1045 AR Amsterdam, The Netherlands;
- Department of Surgery, Amsterdam University Medical Center, Vrije Universiteit Amsterdam, 1081 HV Amsterdam, The Netherlands;
| | - Lothar A. Schwarte
- Department of Anesthesiology, Amsterdam University Medical Center, Vrije Universiteit Amsterdam, 1081 HV Amsterdam, The Netherlands; (F.J.M.); (S.A.L.); (L.A.S.); (P.S.); (S.M.B.)
- Helicopter Emergency Medical Service Lifeliner 1, 1045 AR Amsterdam, The Netherlands;
| | - Patrick Schober
- Department of Anesthesiology, Amsterdam University Medical Center, Vrije Universiteit Amsterdam, 1081 HV Amsterdam, The Netherlands; (F.J.M.); (S.A.L.); (L.A.S.); (P.S.); (S.M.B.)
- Helicopter Emergency Medical Service Lifeliner 1, 1045 AR Amsterdam, The Netherlands;
| | - Sebastiaan M. Bossers
- Department of Anesthesiology, Amsterdam University Medical Center, Vrije Universiteit Amsterdam, 1081 HV Amsterdam, The Netherlands; (F.J.M.); (S.A.L.); (L.A.S.); (P.S.); (S.M.B.)
| |
Collapse
|
5
|
Jang KM, Jang JS. Weekend Admission and Mortality in Patients With Traumatic Brain Injury: A Meta-analysis. Korean J Neurotrauma 2023; 19:422-433. [PMID: 38222828 PMCID: PMC10782108 DOI: 10.13004/kjnt.2023.19.e61] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/20/2023] [Revised: 12/04/2023] [Accepted: 12/04/2023] [Indexed: 01/16/2024] Open
Abstract
Objective Previous studies have reported the presence of a "weekend effect" with respect to mortality in serious emergency admissions, including cases of traumatic brain injury (TBI). However, the relationship between weekend hospitalization and TBI mortality has not been fully established. This study aimed to conduct a systematic review of available evidence and investigate differences in mortality among TBI patients between weekday and weekend admissions. Methods Electronic databases including PubMed, Cochrane Library, and Embase were used to obtain relevant articles. Mortality, as the primary outcome of interest, encompassed in-hospital or 30-day mortality. Mortality rates were compared between the 2 groups, weekend and weekday admissions. Additionally, meta-regression analysis was performed on potential confounders to verify and provide comparative results. Results A total of 7 studies involving 522,942 TBI patients were eligible for inclusion in the synthesis of the systematic review. Of these patients, 71.6% were admitted during weekdays, whereas 28.4% were hospitalized on weekends. The overall integrated mortality was 11.0% (57,286/522,942), with a mortality rate of 10.8% in the weekday group and 11.3% in the weekend group. Pooled analysis revealed no significant difference in mortality between the weekday and weekend groups (risk ratio, 0.99; 95% confidence interval, 0.90-1.09; p=0.78). Furthermore, the meta-regression analysis for sensitivity assessment showed no modifying effect on mortality (p=0.79). Conclusion This study found no difference in mortality rates between weekday and weekend admissions among TBI patients. Additional sensitivity analyses also demonstrated no significant increase in the risk of mortality in the weekend group.
Collapse
Affiliation(s)
- Kyoung Min Jang
- Department of Neurosurgery, Chung-Ang University Gwangmyeong Hospital, Chung-Ang University College of Medicine, Gwangmyeong, Korea
| | - Ju Sung Jang
- Department of Neurosurgery, Chung-Ang University Hospital, Chung-Ang University College of Medicine, Seoul, Korea
| |
Collapse
|