“…Interpretation of these studies is limited due to methodological problems, including retrospective data collection, 4,10,13,14,27 poorly defined cohorts, 9,12 and high loss to followup. 4,13,14,27 Most studies fail to describe the predictive value of individual variables or combinations of variables. 4,6,10,13,14,27 Fracture classification is one variable that has been studied for its predictive value.…”