Rockwood, Green, and Wilkins' Fractures, 10e Package

115

CHAPTER 5 • Classification of Fracture

a particular fracture to a group of similar injuries. Such assign ment expedites assessment of the fracture and naturally flows into treatment options. The ideal fracture classification system should help the surgeon visualize a mental picture of the frac ture prior to viewing any patient-specific imaging studies. The Schatzker classification of tibial plateau fractures is an excellent example of such a system where the fracture pattern type is easy to visualize based on its classification. 53 At its best, fracture classification easily communicates import ant information and clearly transmits detailed understanding of a clinical problem. Classification can consolidate a large amount of information and shared understanding into succinct bites. As an example, to experienced acetabular fracture surgeons, the terms “both column fracture” or “posterior wall fracture disloca tion” capture a large amount of information about fracture line location, involvement of the joint, possible surgical approaches, direction of displacement, and even potential outcomes. While more information is necessary, these few words consolidate a lot of complex concepts. TREATMENT AND PROGNOSIS Ideally, the classification system of a fracture would immediately confer whether the fracture is at risk for poor outcomes or a bad prognosis. Is this particular pattern notorious for associated complications and/or poor outcomes? Some clinicians already do this as a mental exercise by combining multiple categories of fractures into discrete groups. For example, while the Gustilo and Anderson classification of open fractures has several true cate gories, in practice, it is typically summarized as Gustilo IIIb/IIIc versus all other open fractures. 21,22 Infection rates for Gustilo types I, II, and IIIa are fairly similar and the soft tissue enve lope is adequate, while infection rates for IIIb and IIIc are much higher and often require soft tissue coverage. Similarly, intracap sular fractures of the hip are often collapsed into nondisplaced or valgus impacted compared to a displaced femoral neck fracture; pragmatically speaking, there is little practical value to the original four categories of Garden. Conversely, Pauwels’ femoral neck classification is directed toward prognosis, based upon predicting the biomechanical favorability at the fracture site for healing. Pauwel calculated that increasing fracture line inclination results in greater shear and varus stresses; thus, at an advanced type III fracture (>50 degrees of inclination of the fracture line from horizontal), there is a high risk of fracture displacement, and ultimately, nonunion. 40 Another example of how fracture classifications guide treat ment and prognosis is the Mason classification of radial head fractures. Published in 1954, it was created purely to determine whether an injury should be managed conservatively or opera tively. Later modifications by Hotchkiss (regarding mechanical block to motion) and Broberg/Morrey (adding the type IV radial head fracture with associated elbow dislocation) now allow us to determine whether nonoperative management, ORIF, radial head resection, or radial head arthroplasty is the best option for treatment. 26 Finally, in certain situations, combinations of fracture classi fications are used in conjunction to understand the mechanism

of injury (and therefore method of reduction/fixation), as well as considering which injuries require operative fixation. A clas sic example of this synergy is the Tile and Young–Burgess clas sification systems for evaluation of pelvic ring injuries. The Tile system emphasizes stability: type A injuries are stable, type B injuries are rotationally unstable but vertically stable, and type C injuries are both rotationally and vertically unstable. 62 Conversely, the Young and Burgess system is concerned with mechanism, with groupings based on a lateral compression force, anterior to posterior compression force, vertical shear, or combined mechanisms. 67 These systems were designed to be studied using radiograph imaging, but the advent of CT imag ing has helped to improve precise delineation of fracture mor phology even within these systems. Despite their extensive use, until recently, the inter-observer reliability of the Tile and Young–Burgess systems had not been studied. If agreement among surgeons is poor, it is difficult to determine the treatment and prognosis for patients with any degree of certainty. 62 In 2009, it was found that the overall Kappa values for the Young–Burgess system based on XR imag ing or XR + CT imaging, were among the highest reported for fracture classification systems at 0.62 and 0.72, respectively. There was a positive correlation between the level of agreement and expertise; the pelvis/acetabular specialists had the highest level of agreement, followed by the orthopaedic traumatologists and then senior orthopaedic trainees. A similar pattern was seen for the Tile classification system, with both the addition of CT imaging and the seniority of the evaluator improving accuracy. 30 RESEARCH Research is perhaps the number one reason why the classifi cation of fractures is important and will continue to develop. Clinical research on fractures and fracture care is virtually impossible without systematically grouping fracture charac teristics. Fractures in the same anatomic area differ from each other, and researchers must classify them to set study protocols, decide on inclusion and exclusion criteria, and analyze sub groups. A study that reports the results of volar plate fixation of distal radius fractures is of little meaning without specific information about the types of fractures included in the study. To compare two treatment techniques for femoral neck frac tures, the fractures must be classified since treatment and prog nosis vary based on well-known femoral neck subgroups. Issues with reliability of categorical classifications and problems with measurement present challenges for clinical research. Tradition ally, multiple reviewers examine each set of imaging studies. If there is difficulty reaching a consensus utilizing a classifica tion system, this particular scheme is ineffective in meeting the “reproducibility” aspect discussed previously and is unlikely to be helpful in furthering research in the orthopaedic canon. EDUCATION The education and training of orthopaedic surgeons relies foun dationally on the principles of fracture assessment and treat ment. Classification of fractures is part of the way experienced surgeons think, and education must impart this information