BACKGROUND The Chest Wall Injury Society (CWIS) proposals for standardized nomenclature for multiple rib fracture (MRF) classifications were derived by international expert Delphi consensus. This study aimed to validate the CWIS taxonomy using a single-instituion clinical database. METHODS Computed tomography (CT) scans, of 539 consecutive patients with MRFs admitted to a regional major trauma center over a 33-month period, were reviewed (blinded for clinical outcomes). Every rib fracture in every patient was assessed according to each of the CWIS criteria (the degree of displacement, characterization of the fracture line, location of each fracture, and the relationship to neighboring fractures). The clinical significance of the proposed CWIS definitions were determined from independently coded, routinely collected Hospital Episodes Statistics data. RESULTS The radiologic aspects of 3,944 individual rib fractures were assessed. Indicators of injury severity (severe displacement greater series length, and flail segment) were positively associated with other fractures (p < 0.001), hemopneumothorax (p < 0.001), pulmonary complications (p = 0.002), adverse outcomes (p = 0.006), mechanical ventilation (p < 0.001) and prolonged hospital and intensive therapy unit length of stay (p = 0.006, p = 0.007 respectively). Four of the CWIS-proposed definitions were correlated with pulmonary complications and adverse outcomes: the categories of displacement, the definition of individual fracture characterization, the presence of a flail segment. Two definitions for which there was CWIS consensus were not correlated with clinical outcomes: the definition of a series to describe associated fractures on neighboring ribs, the inclusion of a paravertebral sector for fracture localization. CONCLUSION The CWIS rib fracture taxonomy demonstrates clinical relevance. There were associations between the severity of category groups within three of the proposed definitions, based on the clinical outcomes observed. Clinical outcome assessment proved inconclusive for four agreed definitions. Comprehensive, multiinstitutional data collection would be required to provide validation for all the CWIS-proposed definitions. LEVELS OF EVIDENCE Level IV.

Background The accepted classification for multiple rib fractures is binary: flail chest or not. There is a wide spectrum of morphology with subsequent variation in the impact on chest wall mechanics and clinical outcomes. As the practice of surgical stabilisation of rib fractures evolves, there is a need for a better taxonomy. The aim of this study was to create a data-driven radiological classification system for multiple rib fractures, prognostic of both complications and surgical stabilisation of rib fracture. Methods The radiological pattern of injury was assessed for cases undergoing surgical stabilisation of rib fracture (n = 48) over a five-year period and a consecutive sample of non-operative controls (n = 48). Every rib fracture (n = 1032) was assessed on CT scans for location, displacement and comminution. An iterative classification system was developed and tested for inter-observer agreement and outcome prediction. Results The fractures occurred in a ‘series’ (≥3 consecutive ribs at a similar location) in 72% of cases: these were more likely to be displaced (p < 0.001). Variables included in the classification were the anatomical pattern (presence, length and overlap of series) and degree of displacement. The classification was prognostic for complications (p < 0.001), discriminated for fixation (C = 0.907) and had acceptable inter-observer agreement (k = 0.50). Conclusions The Sheffield Multiple Rib Fracture Classification derived categories of short/long series, and short/long flail chest, with sub-division according to the presence of displacement. It was prognostic for clinical outcomes and of surgical fixation. It may facilitate communication, comparison of outcomes and selection for management protocols.

Whilst surgical stabilization of rib fractures (SSRF) results in better outcomes, selection algorithms are lacking. We aimed to validate the Rib Fracture Management Guideline proposed by Bemelman. From a cohort of 792 patients with multiple rib fractures, 2 sequential cohorts were selected: 48 patients who underwent SSRF and 48 patients who managed conservatively. Admission computed tomography scans and records were reviewed by an investigator blinded to the SSRF outcome. Adherence to the Bemelman guideline, revised to take account of consensus rib fracture definitions, was tested. Fifty-seven patients had multiple rib fractures only, and 39 patients also had a flail segment. Thirty-nine patients with flail segment underwent SSRF, and 18 patients were managed conservatively. Of the patients that the guideline predicted should have received surgery, 87% did. Of those that it predicted should not receive SSRF, 98% did not. The guideline displayed a sensitivity (95% confidence interval) and specificity for predicting the fixation of 0.98 (0.89–0.9995) and 0.83 (0.70–0.93), respectively. The positive and negative predictive values for surgical fixation were 0.87 (0.76–0.92) and 0.98 (0.85–0.99), respectively. The Bemelman guideline was thus a good predictor of SSRF in retrospective cohort but should be used in conjunction with clinical judgement. Further validation is indicated in a prospective study.