Factors Affecting Outcome in Tibial Plafond Fractures Todd M. Williams, MD*; James V. Nepola, MD*; Thomas A. DeCoster, MD†; Shepard R. Hurwitz, MD‡; Douglas R. Dirschl, MD§; and J. Lawrence Marsh, MD* To determine what fracture- and patient-specific variables affect outcome, 29 patients with 32 tibial plafond fractures were evaluated at a minimum of 2 years from the time of injury (range, 24–129 months; average, 46.5 months). The rank order method was used to assess severity of injury and accuracy of articular reduction on radiographs and agree- ment among the five surgeons was excellent with intraclass correlation coefficients of 0.93 and 0.94. Outcome was as- sessed by four independent measures: a radiographic arthro- sis score, a subjective ankle score, the Short Form-36 (SF- 36), and the patient’s ability to return to work. The four outcome measures did not correlate with each other. Radio- graphic arthrosis was predicted best by severity of injury and accuracy of reduction. However, these variables did not show any significant relationship to the clinical ankle score, the SF-36, or return to work. These outcome measures were more influenced by patient-specific socioeconomic factors. Higher ankle scores were seen in patients with college de- grees and lower scores were seen in patients with a work- related injury. The ability to return to work was affected by the patient’s level of education. This study highlights the difficulties of predicting patient outcome, after these severe articular fractures. The factors that most predictably affect patient outcome after high-energy fractures of the tibial plafond have not been determined. Knowledge of these factors and how they interact would help to determine optimum treatment and to counsel patients about prognosis. Many clinicians think that the severity of articular injury is the most critical determinant of outcome 2,5,8 whereas others have consid- ered that if an anatomic reduction of the articular surface is achieved, a good outcome can be expected. 14,15 How- ever, another study done by some of the current authors has shown that neither injury severity nor quality of re- duction consistently predicted clinical ankle scores. 3 This result led the authors of that study to speculate that there might be other important factors that affect patient out- come. For instance, the effect of patient demographic vari- ables such as age, gender, or socioeconomic status at the time of injury is unknown. It is possible that these previ- ously unassessed factors overshadow the effect of injury and reduction. To additionally address controversies, to confirm our previous work, and to identify other important outcome determinants, the goals of this study were to determine the severity of the fracture pattern, the quality of the articular reduction, and other factors associated with the patients and their injuries, and determine their effect on outcome. In response to the poor interobserver reliability of the stan- dard classification systems for assessing injury sever- ity 4,11,18 and quality of reduction, 7,10 a rank order system was used to allow greater stratification of these two im- portant variables. This method has been shown to lead to excellent agreement between experienced observers. 3 MATERIALS AND METHODS The study group consisted of 32 fractures (29 patients) chosen from a larger group of tibial plafond fractures treated by five of us (JLM, JVN, TAD, DRD, SRH). A uniform technique of cross- ankle external fixation and limited internal fixation had been used to treat all of the ankles. The results of this treatment technique have been published previously and the technique has been successful at avoiding major complications. 8 All patients signed an IRB approved informed consent document. Thirty-two fractures were selected to be divided into two groups of 16, the number that we thought would be optimal for the rank order technique to be described. Fractures were selected by one of the authors (TMW) who did not participate in the From the *University of Iowa Hospitals and Clinics, Iowa City, IA; †Uni- versity of New Mexico, Albuquerque, NM; ‡University of Virginia School of Medicine, Charlottesville, VA; and §Oregon Health Sciences Institution, Portland, OR. Research was supported by grant AR489389 from the National Institute of Arthritis and Musculoskeletal and Skin Diseases and the Orthopaedic Re- search Educational Fund. Correspondence to: J. L. Marsh, MD, Department of Orthopaedics, UIHC 200 Hawkins Drive, Iowa City, IA 52242. Phone: 319-356-0430; Fax: 319- 353-6754; E-mail: j-marsh@uiowa.edu. DOI: 10.1097/01.blo.0000127922.90382.f4 CLINICAL ORTHOPAEDICS AND RELATED RESEARCH Number 423, pp. 93–98 © 2004 Lippincott Williams & Wilkins 93