Skip to main content
Download PDF
  • Systematic Review
  • Open access
  • Published:

Clinical prediction models for the management of blunt chest trauma in the emergency department: a systematic review

BMC Emergency Medicine volume 24, Article number: 189 (2024) Cite this article

Abstract

Background

The aim of this systematic review was to investigate how clinical prediction models compare in terms of their methodological development, validation, and predictive capabilities, for patients with blunt chest trauma presenting to the Emergency Department.

Methods

A systematic review was conducted across databases from 1st Jan 2000 until 1st April 2024. Studies were categorised into three types of multivariable prediction research and data extracted regarding methodological issues and the predictive capabilities of each model. Risk of bias and applicability were assessed.

Results

41 studies were included that discussed 22 different models. The most commonly observed study design was a single-centre, retrospective, chart review. The most widely externally validated clinical prediction models with moderate to good discrimination were the Thoracic Trauma Severity Score and the STUMBL Score.

Discussion

This review demonstrates that the predictive ability of some of the existing clinical prediction models is acceptable, but high risk of bias and lack of subsequent external validation limits the extensive application of the models. The Thoracic Trauma Severity Score and STUMBL Score demonstrate better predictive accuracy in both development and external validation studies than the other models, but require recalibration and / or update and evaluation of their clinical and cost effectiveness.

Review registration

PROSPERO database (https://www.crd.york.ac.uk/PROSPERO/display_record.php?RecordID=351638).

Peer Review reports

Introduction

Patients with blunt chest trauma present an ongoing challenge for accurate triage in the Emergency Department (ED). Whilst the majority of patients with blunt chest trauma will have an uncomplicated recovery, clinical presentation at the time of ED assessment is no guarantee that a patient will be of suitable acuity for discharge to home, or for admission to award setting, as up to 10% of patients will decompensate after 48–72 h [1,2,3]. Progressive impaired cough and atelectasis can occur when respiratory excursion is limited by pain due to rib fractures, potentially leading to retained pulmonary secretions and pneumonia. Other complications associated with blunt chest trauma include pneumothorax and haemothorax. Intensive Care Unit (ICU) referral from the ED must be carefully considered and as a result, much has been published over the last 20 years investigating the predictors of poor outcome in this patient cohort [4, 5]. These predictors include patient age, severity of injury, number and location of rib fractures, pre-injury anticoagulants, chronic lung disease and others [4, 6,7,8].

A common aim of such primary prognostic studies is the development of clinical prediction models. The clinical prediction model is intended to estimate the individualised probability or risk that a condition, for example mortality or pulmonary complications, will occur in the future by combining multiple prognostic factors / predictors from an individual [9, 10]. A number of different clinical prediction models have been developed for patients with blunt chest trauma, however there is still no universally accepted model in clinical practice. A recent survey study highlighted that there were 20 different clinical prediction models and pathways used when assessing whether a patient with blunt chest trauma is safe for ED home discharge [11].

There is often conflicting evidence regarding the predictive capabilities of developed clinical prediction models, leading to a growing demand for evidence synthesis of external validation studies that assess model performance in a new patient cohort [10, 12, 13]. This is applicable to the range of clinical prediction models used for the management of patients with blunt chest trauma. The aim of this systematic review therefore was to investigate how clinical prediction models compare in terms of their methodological development, validation, and predictive capabilities, for clinical and healthcare utilisation outcomes for patients with blunt chest trauma presenting to the Emergency Department.

Methods

Search strategy

The CHARMS Checklist was followed for completion of this review. A broad search strategy was employed in order to capture all relevant studies. The search filter was used for PubMed and Embase Databases, the Cochrane Library, and OpenGrey from 1st Jan 2000 until 1st April 2024. The search term combinations were based on Geersing et al. (2012) [12] and used Medical Subject Heading terms, text words and word variants for blunt chest trauma. These were combined with relevant terms for both outcomes and clinical prediction model development and validation methods. An additional file shows the search strategy [see Additional file 1]. The reference lists of all relevant studies were hand-searched in order to identify any evidence missed in the electronic search. The Annals of Emergency Medicine, Emergency Medicine Journal, Injury and the Journal of Trauma and Acute Care Surgery were hand-searched for relevant studies. Searches were international and no search limitations were used.

Study selection

Studies were included that focussed on patients aged ≥ 16 presenting to the Emergency Department with blunt chest trauma (defined as a blunt chest injury resulting in chest wall contusion or rib fractures, with or without underlying lung injury). Prognostic multivariable prediction studies were included where the aim of the study was to predict an outcome using two or more independent variables, in order to develop a multivariable (at least two variables) weighted clinical prediction model for any outcome following blunt chest trauma. Based on the ‘Critical Appraisal and Data Extraction for Systematic Reviews of Prediction Modelling Studies: CHARMS guidance [13], studies were categorised into three types of multivariable prediction research; 1) model development studies without external validation. 2) model development studies with external validation in independent data, and 3) external validation studies without or with model updating.

Studies were excluded which included patients presenting with: (a) Penetrating trauma only, (b) Multi-trauma only and no reference to chest trauma, (c) Severe intra-thoracic injuries only (e.g. bronchial, cardiac, oesophageal, aortic or diaphragmatic rupture) and no chest wall trauma, (d) Children aged < 16 years. Other exclusion criteria included, studies that investigated a single predictor (such as single prognostic marker studies), studies that investigated only causality between one or more variables and an outcome, and studies that do not contribute to patient care. For multiple publications from the same dataset, only the most relevant study to this reviews aims was included. Studies for which only an abstract was available were also excluded.

Data extraction

A two-step process was used to reduce potential selection bias. Two researchers (CB and EB) analysed each title and abstract independently and then met to discuss any discrepancies. The full paper of selected studies was analysed by the reviewers. Data were extracted relating to both the reporting of and use of methods known to influence the quality of multivariable prediction studies. A data extraction form based on CHARMS Checklist was used to record relevant information, shown in additional file 2 [see Additional file 2]. Study authors were contacted for any missing data and response time set at six weeks. Included studies were grouped according to the clinical prediction model under investigation for the analysis.

Data were extracted regarding the methodological issues that are considered to be important in prediction research, focussed broadly on the reporting of the domains outlined in the CHARMS Checklist. Data regarding the predictive capabilities of each model were also extracted where available, for the following outcomes; (a) clinical outcomes such as mortality and any pulmonary complications, and (b) healthcare utilisation outcomes such as length of stay, need for mechanical ventilation or ICU admission.

Quality assessment

Risk of bias and applicability were assessed using the “Prediction model Risk Of Bias ASsessment Tool” (PROBAST) [14] where: “Risk of bias refers to the extent that flaws in the design, conduct, and analysis of the primary prediction modelling study lead to biased, often overly optimistic, estimates of predictive performance measures such as model calibration, discrimination, or (re)classification (usually due to over-fitted models). Applicability refers to the extent to which the primary study matches the review question, and thus is applicable for the intended use of the reviewed prediction model(s) in the target population” (Moons et al., 2014). PROBAST includes 20 signalling questions across four domains (participants, predictors, outcome, and analysis) which were scored low, high or unclear. For each included study, an overall final score for judgement of risk of bias and applicability was allocated. This process was completed independently by two reviewers (CB and EB), with a third reviewer (EC) used to resolve any discrepancies. An additional file shows the PROBAST Score in more detail [see Additional file 3].

Data synthesis and analysis

Narrative synthesis of included study results was conducted, grouped according to clinical prediction models. Model performance was evaluated through assessment of model discrimination, a measure of how well the model can separate those who do and those who do not have the disease of interest, and calibration, a measure of how well predicted probabilities agree with the actual observed risk. The discrimination ‘C-statistic’ (balance between negative and positive predictive value) was defined as low (below 0⋅70), moderate (0⋅70–0⋅79) or good (at least 0⋅80). Where available in the studies, the correlation between observed and expected (calibration) outcome, as measured by the Hosmer–Lemeshow (H-L) test, was presented using a p > 0⋅050 to indicate a good model fit [13].

Results

Study selection

The initial search strategy identified 9495 citations. Following screening titles and abstracts, we identified 174 potentially relevant studies and following full-text review, a total of 41 studies met the inclusion criteria. No additional citations were identified through the grey literature or reference list searches. Figure 1 outlines the flow diagram of study selection.

Fig. 1
figure 1

PRISMA flow diagram

Full size image

Study characteristics

The 41 studies were categorised as; 12 model development studies without external validation, three model development studies with external validation in independent data, and 26 external validation studies without or with model updating. The most commonly observed study design was a single-centre, retrospective, chart review. A total of 22 different clinical prediction models were studied and therefore included in this review. Study design, clinical prediction model, study population (including diversity data where possible, such as age, sex, frailty and ethnicity), total sample size, outcomes and results of the included studies are outlined in Table 1.

Table 1 Characteristics of included studies
Full size table

Quality assessment

The quality of the included studies in this review was variable. Risk of bias was high across most of the included studies for the analysis. Selection of predictors was commonly based on univariable analysis result, handling of missing data was inadequately described and the model performance measures, in particular the model’s calibration, was infrequently reported. The studies scored mostly low risk of bias in terms of the predictors included. Risk of bias for participants was variable across the studies as some used a trauma registry for their participant data. In terms of applicability, some studies scored high risk for participants, as they included paediatric patients, which this review was not investigating. The full PROBAST results are outlined in Table 2; Fig. 2.

Table 2 Risk of bias and applicability of included studies: PROBAST results
Full size table
Fig. 2
figure 2

Risk of bias and applicability of included studies: PROBAST results

Full size image

Figure 2 demonstrates the overall judgment of the included studies.

Clinical prediction models

Thoracic trauma severity score (TTSS)

The TTSS was originally developed and externally validated by Pape et al. (2000) to predict the risk of thoracic-trauma related complications in patients with blunt polytrauma, admitted to ICU [15]. Based on high risk of bias results, the c-index demonstrated good discrimination, as demonstrated by a value of 0.924 for the development set and 0.916 for the validation set, although 95% confidence intervals were not reported. Since 2002, there have been ten external validation studies [16,17,18,19,20,21,22,23,24,25] of high risk of bias, that have reported various cut off values on the TTSS, with moderate to good level c-indices ranging between 0.723 and 0.848. Model calibration was not reported in any of the included studies.

STUMBL score

The STUMBL Score was original developed and externally validated by Battle et al. (2014) to predict risk of pulmonary complications in patients with isolated blunt chest wall trauma presenting to the ED [26]. Based on low risk of bias results, the final model demonstrated good discrimination with a reported c-index of 0.96 (95% CI: 0.93 to 0.98). The model showed good calibration when evaluated with the Hosmer Lemeshow test (9.22, P = 0.32). Since development, there have been four external validation studies [27,28,29,30] completed of variable risk of bias, that have reported various cut off values on the STUMBL Score, with moderate to good level c-indices ranging between 0.61 and 0.90 (95% CI 0.88–0.93).

Rib fracture score (RFS)

The RFS was originally developed by Easter et al. (2001), as a protocol for the management of pain, respiratory care and mobility in patients with multiple rib fractures [31]. The score allocated to the patient (based on number of fractures, number of sides and the patient’s age), determines the treatment recommendations, rather than a risk of a particular outcome. The protocol was based on literature, rather than patient data and as a result was at high risk of bias. No predictive capabilities were reported in the original development study. Five external validation studies [21, 25, 27, 32, 33] of high risk of bias, have been completed, demonstrating a low level of discrimination with c-indices ranging from 0.64 to 0.67 for the prediction of a number of clinical and healthcare resource outcomes. Model calibration was not reported in the included studies.

Chest trauma score (CTS)

The CTS was originally developed by Pressley et al. (2012) for patients presenting with rib fractures, using clinical data available at the time of initial evaluation. It predicts the likelihood of mechanical ventilation and prolonged courses of care [34]. The development study did not report predictive capabilities of the score and was considered high risk of bias. Seven external validation studies [21, 27, 32, 34,35,36,37] of high risk of bias have been completed, demonstrating a low to good level of discrimination with c-indices of 0.67 to 0.91. Model calibration was not reported in any of the studies.

RibScore

The RibScore, originally developed by Chapman et al. (2016) for blunt trauma patients with rib fractures, was based on six candidate radiographic variables, identified on CT imaging [38]. They reported c-indices the outcomes pneumonia, respiratory failure and tracheostomy were 0.71, 0.71, and 0.75, respectively in a high risk of bias study. Three high risk of bias external validation studies [21, 32, 39] have been completed in which low and moderate c-indices of 0.62 and 0.79 (95% CI 0.69 to 0.89) were reported. Model calibration was not reported in any of the studies.

Revised intensity battle score (RIBS)

RIBS was developed and later externally validated by Buccholz et al. (2022 and 2024 respectively), in which the authors revised the STUMBL (Battle) Score, for use with patients admitted to ICU [27, 40]. Using the STUMBL Score, RIBS was developed by re-weighting the predictor variables according to their predictive capacity to identify in hospital complications. A good discrimination for the final model was reported in both development and external validation studies (c-indices: 0.86 and 0.73 respectively), although both studies were at high risk of bias [40]. Model calibration was not reported in either study.

Other clinical prediction models

Table 1 outlines 18 other clinical prediction models which were identified, for which only one study (all high risk of bias) per model met the inclusion criteria for this review. A number of new clinical prediction models have been developed (all high risk of bias studies) but not yet validated were included in the review. These included the Tracheostomy in Thoracic Trauma Prediction Score [41] (T3P-Score, c-index for tracheostomy: 0.938, 95% CI: 0.920–0.956), Sequential Clinical Assessment of Respiratory Function [42] (SCARF Score, c-index for pneumonia: 0.86), Rib Injury Guidelines [43] (RIG, c-index not reported), the Lung Organ Failure Score [44] (c-index not reported), and a new scoring system [45] (c-index: 0.82; 95% CI: 0.77–0.88).

Other models developed and validated by the original authors, but yet to be externally validated in further studies included The Rib Fracture Frailty Index [46] (RFFI) (c-index not reported), Quebec Minor Thoracic Injury Decision Rule [47] (c-index: 0.78; 95% CI 0.74–0.82), a single rib fracture nomogram [48] (c-index: 0.70), and the Trauma Induced Pulmonary Event (TIPE Score) (c-index: 0.85) [49].

The chest wall components of the Abbreviated Injury Scale (AIS) and Organ Injury Scale (OIS) were externally validated in a high risk of bias study by Baker et al. (2020) which reported a low level of discrimination for both the OIS (c-index: 0.68; 95% CI: 0.64–0.73) and AIS (c-index: 0.59; 95%CI: 0.55 to 0.63) for patients with rib and sternal fractures presenting to the ED [50].

There were four model development studies that did not meet the inclusion criteria for this review, but subsequent validation studies were included (all high risk of bias). These included the Revised Cardiac Risk Index [51] (RCRI, originally developed to predict 30-day postoperative myocardial infarction, cardiac arrest, or mortality following non-cardiac surgery, c-index not reported), Pain Inspiratory Effort Cough Score [52] (PIC Score, c-index not reported), Revised Trauma Scale [53] (RTS, c-index: 0.76, 95%CI: 0.72–0.79), Lung Ultrasound Score [54] (LUS, c-index not reported), and the ROX Index [55] (which combines respiratory rate and oxygenation values, c-index: 0.88; 95%CI: 0.80–0.94).

Discussion

This systematic review has highlighted that there are numerous clinical prediction models used for the management of patients with blunt chest trauma in various healthcare settings. These models differ widely in terms of their target patient population, included risk factors and outcomes predicted. They also differ in terms of the methods used for both their development and validation. These findings impede comparison between the models and generalisability for the patient with blunt chest wall trauma. These inherent differences also contribute to the lack of consensus in clinical practice, regarding the optimal clinical prediction model for this patient population [56, 57].

This review highlights the difficulties in developing, validating and using a clinical prediction model. Instead of updating existing models and improving their predictive capabilities, most studies have developed and presented a new model. This has resulted in better performance in their population compared with existing models that were developed in another population and validated externally. Furthermore, there were no impact studies retrieved in this review that explored the clinical or cost effectiveness of any of the models. Traditional impact studies are reported to be costly to undertake and as a result, very few exist for any patient condition [57]. It is reasonable therefore to suggest that the ideal model does not yet exist.

Not all studies calculated a c-index to describe the discriminative abilities of the model and only one study reported an H-L analysis for calibration. Other studies may have used alternative measurements, or it must be assumed that they have compared observed with expected results, but did not report the comparison statistic. Overall, discrimination is more straightforward to calculate when compared with calibration, and the latter can be easily improved using updating methods applied to a new patient cohort [13, 57]. Good calibration is necessary however for calculating predictions, independent of the reported c-index [57]. The clinical usefulness of a model can only be determined when both discrimination and calibration are available, and a model’s cut-off value has been defined for reported sensitivity and specificity values [13, 57].

The models developed specifically for the management of patients with blunt chest trauma according to methodological guidance and most widely externally validated demonstrating moderate to good discrimination, were the TTSS [15] and STUMBL Score [26]. These models were developed for use in different healthcare settings and only the STUMBL Score had been assessed for calibration. Neither model has undergone any recalibration or updating or revision, nor have been assessed for clinical or cost effectiveness. The STUMBL Score has been revised by other authors into the RIBS prediction model, for higher acuity patients [40]. There is limited reference to different diverse patient groups in any of the included studies, with exception to the STUMBL Score, which was the only model that was reported to have been specifically externally validated on patients of varying ethnic groups. Health inequalities across ethnic groups are reported in other disease populations [58, 59] but currently it isn’t clear if existing blunt chest trauma clinical prediction models account for diversity-related differences.

This systematic review has a number of limitations. For pragmatic reasons we were only able to hand-search a selection of key journals. The different age groups selected for investigation in each of the included papers will impact not only their own validity, but that of this review. This heterogeneity needs to be considered when interpreting the review findings. A large number of the included studies failed to report confidence intervals for the reported c-indices, resulting in incomplete comparisons between the models. Most of these models had been developed on Causcian populations, and it remains unknown (other than the STUMBL Score New Zealand validation study [30]) whether these models would perform equally well in other ethnic groups. Frailty as a potential candidate predictor was not considered in any of the included model development studies, other than the RFFI study [46]. It is well-recognised that frailty identification has an important role in any clinical decision-making related in older trauma patients [60, 61], therefore this needs further consideration in future studies and existing model updates. Finally, the lead author of this review is also the researcher who developed the STUMBL Score, so there is the potential for interpretive bias.

Conclusions

This systematic review has examined the methodological development, validation, and predictive capabilities of the clinical prediction models, for clinical and healthcare utilisation outcomes for patients with blunt chest trauma presenting to the Emergency Department. The predictive ability of some of the existing clinical prediction models is acceptable, but high risk of bias and lack of subsequent external validation limits the extensive application of the models in the general blunt chest trauma population. The TTSS and STUMBL Score demonstrate better predictive accuracy in both development and external validation studies than the other models, but both potentially still require recalibration and / or update and evaluation of their clinical and cost effectiveness.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Abbreviations

AIS:

Abbreviated Injury Scale

ED:

Emergency Department

CI:

Confidence interval

CTS:

Chest Trauma Score

H-L:

Hosmer-Lemeshow

ICU:

Intensive Care Unit

LUS:

Lung Ultrasound Score

OIS:

Organ Injury Scale

PROBAST:

Prediction model Risk Of Bias ASsessment Tool

PICS:

Pain Injury Cough Score

RFFI:

Rib Fracture Frailty Index

RFS:

Rib Fracture Score

RIBS:

Revised Intensity Battle Score

RIG:

Rib Injury Guidelines

RTS:

Revised Trauma Scale

SCARF:

Sequential Clinical Assessment of Respiratory Function Score

TIPE:

Trauma Induced Pulmonary Event Score

TTSS:

Thoracic Trauma Severity Score

References

  1. Blecher GE, Mitra B, Cameron PA, Fitzgerald MC. Failed Emergency Department disposition to the ward of patients with thoracic injury. Injury. 2008;39:586–91.

    Article  PubMed  Google Scholar 

  2. Baker E, Battle C. What is the optimal care pathway for patients with blunt chest wall trauma presenting to the ED? Trauma. 2023. https://doi.org/10.1177/14604086221142384.

    Article  Google Scholar 

  3. Plourde M, Émond M, Lavoie A, Guimont C, Le Sage N, Chauny JM, et al. Cohort study on the prevalence and risk factors of late pulmonary complications in adults following a closed minor chest trauma. Can J Emerg Med. 2013;15:337–44.

    Article  Google Scholar 

  4. Battle C, Carter K, Newey L, Giamello JD, Melchio R, Hutchings H. Risk factors that predict mortality in patients with blunt chest wall trauma: an updated systematic review and meta-analysis. Emerg Med J. 2022;14:14.

    Google Scholar 

  5. Christmas AB, Freeman E, Chisolm A, Fischer PE, Sachdev G, Jacobs DG, et al. Trauma intensive care unit ‘bouncebacks’: identifying risk factors for unexpected return admission to the intensive care unit. Am Surg. 2014;80:778–82.

    Article  PubMed  Google Scholar 

  6. Carver TW, Milia DJ, Somberg C, Brasel K, Paul J. Vital capacity helps predict pulmonary complications after rib fractures. J Trauma Acute Care Surg. 2015;79:413–16.

    Article  PubMed  Google Scholar 

  7. Choi J, Alawa J, Tennakoon L, Forrester JD, DeepBackRib. Deep learning to understand factors associated with readmissions after rib fractures. J Trauma Acute Care Surg. 2022;93:757–61.

    Article  PubMed  Google Scholar 

  8. Hamilton C, Barnett L, Trop A, Leininger B, Olson A, Brooks A, et al. Emergency department management of patients with rib fracture based on a clinical practice guideline. Trauma Surg Acute Care Open. 2017;2:e000133.

    Article  PubMed  PubMed Central  Google Scholar 

  9. Debray TP, Vergouwe Y, Koffijberg H, Nieboer D, Steyerberg EW, Moons KG. A new framework to enhance the interpretation of external validation studies of clinical prediction models. J Clin Epidemiol. 2015;68:279–89.

    Article  PubMed  Google Scholar 

  10. Debray TP, Damen JA, Snell KI, Ensor J, Hooft L, Reitsma JB, et al. A guide to systematic review and meta-analysis of prediction model performance. BMJ. 2017;356:i6460.

    Article  PubMed  Google Scholar 

  11. Battle C, O’Neill C, Newey L, Barnett J, O’Neill M, Hutchings H. A survey of current practice in UK emergency department management of patients with blunt chest wall trauma not requiring admission to hospital. Injury. 2021;52:2565–70.

    Article  PubMed  Google Scholar 

  12. Geersing GJ, Bouwmeester W, Zuithoff P, Spijker R, Leeflang M, Moons KG. Search filters for finding prognostic and diagnostic prediction studies in Medline to enhance systematic reviews. PLoS ONE. 2012;7:e32844.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  13. Moons KGM, de Groot JAH, Bouwmeester W, Vergouwe Y, Mallett S, Altman DG, et al. Critical Appraisal and Data extraction for systematic reviews of Prediction Modelling studies: the CHARMS Checklist. PLoS Med. 2014;11:e1001744.

    Article  PubMed  PubMed Central  Google Scholar 

  14. Moons KGM, Wolff RF, Riley RD, Whiting PF, Westwood M, Collins GS, et al. PROBAST: A Tool to assess risk of Bias and Applicability of Prediction Model studies: explanation and elaboration. Ann Intern Med. 2019;170:W1–33.

    Article  PubMed  Google Scholar 

  15. Pape HC, Remmers D, Rice J, Ebisch M, Krettek C, Tscherne H. Appraisal of early evaluation of blunt chest trauma: development of a standardized scoring system for initial clinical decision making. J Trauma Acute Care Surg. 2000;49:496–504.

    Article  CAS  Google Scholar 

  16. Aukema TS, Beenen LF, Hietbrink F, Leenen LP. Validation of the Thorax Trauma Severity score for mortality and its value for the development of acute respiratory distress syndrome. Open Access Emerg Med. 2011;3:49–53.

    PubMed  PubMed Central  Google Scholar 

  17. Daurat A, Millet I, Roustan J-P, Maury C, Taourel P, Jaber S, et al. Thoracic trauma severity score on admission allows to determine the risk of delayed ARDS in trauma patients with pulmonary contusion. Injury. 2016;47:147–53.

    Article  PubMed  Google Scholar 

  18. El-Aziz T, El Din Abdallah A, Abdo A, El-Hag-Aly M. Thorax trauma severity score and trauma injury severity score evaluation as outcome predictors in chest trauma. Res Opin Anesth Intensive Care. 2022;9:112–18.

    Article  Google Scholar 

  19. Esme H, Solak O, Yurumez Y, Yavuz Y, Terzi Y, Sezer M, et al. The prognostic importance of trauma scoring systems for blunt thoracic trauma. Thorac Cardiovasc Surg. 2007;55:190–95.

    Article  PubMed  CAS  Google Scholar 

  20. Kanake V, Kale K, Mangam S, Bhalavi V. Thorax trauma severity score in patient with chest trauma: study at tertiary-level hospital. Indian J Thorac Cardiovasc Surg. 2022;38:149–56.

    Article  PubMed  Google Scholar 

  21. Kim H, Lee MS, Yoon SY, Han J, Lee JY, Seok J. Validation of chest trauma scoring systems in polytrauma: a retrospective study with 1,038 patients in Korea Korean soc traumatol. 2024: https://doi.org/10.20408/jti.2023.0087

  22. Martínez Casas I, Amador Marchante MA, Paduraru M, Fabregues Olea AI, Nolasco A, Medina JC. Thorax Trauma Severity score: is it reliable for patient’s evaluation in a secondary Level Hospital? Bull Emerg Trauma. 2016;4:150–5.

    PubMed  Google Scholar 

  23. Mommsen P, Zeckey C, Andruszkow H, Weidemann J, Frömke C, Puljic P, et al. Comparison of different thoracic trauma Scoring systems in regards to Prediction of Post-traumatic complications and Outcome in Blunt chest trauma. J Surg Res. 2012;176:239–47.

    Article  PubMed  Google Scholar 

  24. Moon SH, Kim JW, Byun JH, Kim SH, Choi JY, Jang IS, et al. The thorax trauma severity score and the trauma and injury severity score: do they predict in-hospital mortality in patients with severe thoracic trauma? A retrospective cohort study. Medicine. 2017;96:e8317.

    Article  PubMed  Google Scholar 

  25. Seok J, Cho HM, Kim HH, Kim JH, Huh U, Kim HB, et al. Chest trauma Scoring systems for Predicting Respiratory complications in isolated Rib fracture. J Surg Res. 2019;244:84–90.

    Article  PubMed  Google Scholar 

  26. Battle C, Hutchings H, Lovett S, Bouamra O, Jones S, Sen A, et al. Predicting outcomes after blunt chest wall trauma: development and external validation of a new prognostic model. Crit Care. 2014;18:R98.

    Article  PubMed  Google Scholar 

  27. Buchholz CJ, D’Aquila ML, Lollar DI. External validation of novel revised intensity battle score and comparison of static rib fracture scoring systems. J Trauma Acute Care Surg. 2024;96:466–70.

    PubMed  Google Scholar 

  28. Callisto E, Costantino G, Tabner A, Kerslake D, Reed MJ. The clinical effectiveness of the STUMBL score for the management of ED patients with blunt chest trauma compared to clinical evaluation alone. Intern Emerg Med. 2022. https://doi.org/10.1007/s11739-022-03001-0.

    Article  PubMed  Google Scholar 

  29. Giamello JD, Martini G, Prato D, Santoro M, Arese Y, Melchio R, et al. A retrospective validation study of the STUMBL score for emergency department patients with blunt thoracic trauma. Injury. 2022. https://doi.org/10.1016/j.injury.2022.08.028.

    Article  PubMed  Google Scholar 

  30. Mukerji S, Tan E, May C, Micanovic C, Blakemore P, Phelps K, et al. Retrospective validation of a risk stratification tool developed for the management of patients with blunt chest trauma (the STUMBL score). Emerg Med Australas. 2021;33:841–7.

    Article  PubMed  Google Scholar 

  31. Easter A. Management of patients with multiple rib fractures. Am J Crit Care. 2001;10:320–27.

    Article  PubMed  CAS  Google Scholar 

  32. Fokin A, Wycech J, Crawford M, Puente I. Quantification of rib fractures by different scoring systems. J Surg Res. 2018;229:1–8.

    Article  PubMed  Google Scholar 

  33. Maxwell CA, Mion LC, Dietrich MS. Hospitalized injured older adults: clinical utility of a rib fracture scoring system. J Trauma Nurs. 2012;19:168–74.

    Article  PubMed  Google Scholar 

  34. Pressley CM, Fry WR, Philp AS, Berry SD, Smith RS. Predicting outcome of patients with chest wall injury. Am J Surg. 2012;204:910–4.

    Article  PubMed  Google Scholar 

  35. Chen J, Jeremitsky E, Philp F, Fry W, Smith RS. A chest trauma scoring system to predict outcomes. Surgery. 2014;156:988–93.

    Article  PubMed  Google Scholar 

  36. Harde M, Aditya G, Dave S. Prediction of outcomes in chest trauma patients using chest trauma scoring system: a prospective observational study. Indian J Anaesth. 2019;63:194–99.

    Article  PubMed  PubMed Central  Google Scholar 

  37. Ujjaneswari B, Kumari VM, Madhusudhan KV, Lokesh K, Lakshmipathi Naik KS. A study of role of chest trauma scoring system in blunt chest trauma. Eur J Mol Clin Med. 2023;10:42–66.

    Google Scholar 

  38. Chapman BC, Herbert B, Rodil M, Salotto J, Stovall RT, Biffl W, et al. RibScore: a novel radiographic score based on fracture pattern that predicts pneumonia, respiratory failure, and tracheostomy. J Trauma Acute Care Surg. 2016;80:95–101.

    Article  PubMed  Google Scholar 

  39. Schmoekel N, Berguson J, Stassinopoulos J, Karamanos E, Patton J, Johnson JL. Rib fractures in the elderly: physiology trumps anatomy. Trauma Surg Acute Care Open. 2019;4:e000257.

    Article  PubMed  PubMed Central  Google Scholar 

  40. Buchholz CJ, Jia L, Manea C, Petersen T, Wang H, Stright A et al. Revised intensity battle score (RIBS): development of a clinical score for Predicting Poor outcomes after Rib fractures. Am Surg. 2022:31348221123087.

  41. Bläsius FM, Wutzler S, Störmann P, Lustenberger T, Frink M, Maegele M, et al. Predicting tracheostomy in multiple injured patients with severe thoracic injury (AIS ≥ 3) with the new T3P-Score: a multivariable regression prediction analysis. Sci Rep. 2023;24:13:3260.

    Article  Google Scholar 

  42. Hardin KS, Leasia KN, Haenel J, Moore EE, Burlew CC, Pieracci FM. The Sequential Clinical Assessment of respiratory function (SCARF) score: a dynamic pulmonary physiologic score that predicts adverse outcomes in critically ill rib fracture patients. J Trauma Acute Care Surg. 2019;87:1260–68.

    Article  PubMed  Google Scholar 

  43. Nelson A, Reina R, Northcutt A, Obaid O, Castanon L, Ditillo M, et al. Prospective validation of the Rib Injury guidelines for traumatic rib fractures. J Trauma Acute Care Surg. 2022;92:967–73.

    Article  PubMed  CAS  Google Scholar 

  44. Wutzler S, Wafaisade A, Maegele M, Laurer H, Geiger EV, Walcher F, et al. Lung organ failure score (LOFS): probability of severe pulmonary organ failure after multiple injuries including chest trauma. Injury. 2012;43:1507–12.

    Article  PubMed  Google Scholar 

  45. Gonzalez KW, Ghneim MH, Kang F, Jupiter DC, Davis ML, Regner JL. A pilot single-institution predictive model to guide rib fracture management in elderly patients. J Trauma Acute Care Surg. 2015;78:970–75.

    Article  PubMed  Google Scholar 

  46. Choi J, Marafino BJ, Vendrow EB, Tennakoon L, Baiocchi M, Spain DA, et al. Rib fracture Frailty Index: a risk stratification tool for geriatric patients with multiple rib fractures. J Trauma Acute Care Surg. 2021;91:932–39.

    Article  PubMed  Google Scholar 

  47. Emond M, Guimont C, Chauny JM, Daoust R, Bergeron E, Vanier L, et al. Clinical prediction rule for delayed hemothorax after minor thoracic injury: a multicentre derivation and validation study. CMAJ Open. 2017;5:E444–53.

    Article  PubMed  PubMed Central  Google Scholar 

  48. Kishawi SK, Ho VP, Bachman KC, Alvarado CE, Worrell SG, Argote-Greene LM, et al. Are we underestimating the morbidity of single Rib fractures? J Surg Res. 2021;268:174–80.

    Article  PubMed  PubMed Central  Google Scholar 

  49. Li R, Grigorian A, Nahmias JT, Inaba K, Kuza CM. Development of a novel tool to predict pulmonary complications in trauma patients with and without chest injury. Am J Surg. 2022;224:64–8.

    Article  PubMed  Google Scholar 

  50. Baker JE, Millar DA, Heh V, Goodman MD, Pritts TA, Janowak CF. Does chest wall Organ Injury Scale (OIS) or abbreviated Injury Scale (AIS) predict outcomes? An analysis of 16,000 consecutive rib fractures. Surgery. 2020;168:198–204.

    Article  PubMed  Google Scholar 

  51. Bass GA, Duffy CC, Kaplan LJ, Sarani B, Martin ND, Ismail AM, et al. The revised cardiac risk index is associated with morbidity and mortality independent of injury severity in elderly patients with rib fractures. Injury. 2023;54:56–62.

    Article  PubMed  Google Scholar 

  52. Bass GA, Stephen C, Forssten MP, Bailey JA, Mohseni S, Cao Y, et al. Admission Triage with Pain, Inspiratory Effort, Cough score can predict critical care utilization and length of stay in isolated chest Wall Injury. J Surg Res. 2022;277:310–18.

    Article  PubMed  Google Scholar 

  53. Cinar E, Usul E, Demirtas E, Gokce A. The role of trauma scoring systems and serum lactate level in predicting prognosis in thoracic trauma. Ulusal Travma ve Acil Cerrahi Dergisi. 2021;27:619–23.

    PubMed  Google Scholar 

  54. Sayed MS, Elmeslmany KA, Elsawy AS, Mohamed NA. The validity of quantifying Pulmonary Contusion Extent by Lung Ultrasound score for Predicting ARDS in Blunt thoracic trauma. Crit Care Res Pract. 2022;3124966.

  55. Cornillon A, Balbo J, Coffinet J, Floch T, Bard M, Giordano-Orsini G, et al. The ROX index as a predictor of standard oxygen therapy outcomes in thoracic trauma. Scand J Trauma Resusc Emerg Med. 2021;29:81.

    Article  PubMed  PubMed Central  Google Scholar 

  56. Birse F, Williams H, Shipway D, Carlton E. Blunt chest trauma in the elderly: an expert practice review. Emerg Med J. 2020;37:73–8.

    Article  PubMed  Google Scholar 

  57. Bouwmeester W, Zuithoff NP, Mallett S, Geerlings MI, Vergouwe Y, Steyerberg EW, et al. Reporting and methods in clinical prediction research: a systematic review. PLoS Med. 2012;9:1–12.

    Article  PubMed  Google Scholar 

  58. Bhopal R, Steiner MF, Cezard G, Bansal N, Fischbacher C, Simpson CR, et al. Risk of respiratory hospitalization and death, readmission and subsequent mortality: Scottish health and ethnicity linkage study. Eur J Public Health. 2015;25:769–74.

    Article  PubMed  Google Scholar 

  59. Jennison T, Kulenkampff C, Lee J, Mahmood A. Is ethnicity a risk factor for mortality in major trauma? A single-centred cohort study. Ann R Coll Surg Engl. 2023:rcsann20220097.

  60. Hamidi M, Haddadin Z, Zeeshan M, Saljuqi AT, Hanna K, Tang A, et al. Prospective evaluation and comparison of the predictive ability of different frailty scores to predict outcomes in geriatric trauma patients. J Trauma Acute Care Surg. 2019;87:1172–80.

    Article  PubMed  Google Scholar 

  61. Braude P, Short R, Bouamra O, Shipway D, Lecky F, Carlton E, et al. A national study of 23 major trauma centres to investigate the effect of a geriatrician assessment on clinical outcomes in older people admitted with serious injury in England (FiTR 2): a multicentre observational cohort study. Lancet Healthy Longev. 2022;3:e549–57.

    Article  PubMed  Google Scholar 

Download references

Acknowledgements

Not applicable.

Funding

No funding was received for the completion of this work.

Author information

Authors and Affiliations

  1. Physiotherapy Dept, Morriston Hospital, Swansea Bay University Health Board, Swansea, Wales, SA6 6NL, UK

    Ceri Battle

  2. Swansea Trials Unit, Swansea University Medical School, Swansea University, Swansea, UK

    Ceri Battle & Kym Carter

  3. Centre of Trauma Sciences, Blizard Institute, Queen Mary University of London, London, UK

    Elaine Cole

  4. Emergency Dept, Kings College Hospital, London, UK

    Edward Baker

Authors
  1. Ceri Battle
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Elaine Cole
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Kym Carter
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Edward Baker
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

All authors designed this work. K.C. and C.B. undertook the searches. C.B., E.B. and E.C. completed the analysis and interpretation of data. C.B. drafted the article and K.T., E.B. and E.C. all revised it critically for important intellectual content. All authors approved the version to be published.

Corresponding author

Correspondence to Ceri Battle.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Supplementary Material 2

Supplementary Material 3

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Battle, C., Cole, E., Carter, K. et al. Clinical prediction models for the management of blunt chest trauma in the emergency department: a systematic review. BMC Emerg Med 24, 189 (2024). https://doi.org/10.1186/s12873-024-01107-6

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12873-024-01107-6

Keywords

BMC Emergency Medicine

ISSN: 1471-227X

Contact us