JBJS Classics: Understanding Proximal Humeral Fractures
The contributions to the field of shoulder surgery from Dr. Charles Neer are too numerous to document in any one commentary. A partial list would include shoulder arthroplasty (both hemi and total), the concept of impingement and acromial pathology, multidirectional instability, and the role of the AC joint in rotator cuff pathology.
Dr. Neer also made numerous contributions to the understanding of fracture care, including the distal femur and clavicle. But no area of fracture management was of greater interest to him and his colleagues at Columbia than the proximal humerus. This classic manuscript has been cited thousands of time and remains the seminal piece in the foundation of understanding fracture patterns in the proximal humerus—and the attendant treatment implications.
Dr. Neer introduced the concept of the four parts of the proximal humerus in this manuscript, and with it the implication of isolating the humeral-head blood supply in a four-part fracture. The impetus to understand the complication of avascular necrosis of the humeral head began with this manuscript, as did the critical debates regarding surgical versus nonsurgical intervention and replace-or-fix. An important area of ongoing debate is Neer’s definition of a “displaced” fracture in the proximal humerus as having > 1 cm of displacement. The orthopaedic community to this day is wrestling with this definition and its relevance to treatment and outcomes.
This classic manuscript also helped launch a decades-old conversation about the role of fracture or musculoskeletal-disease classification systems. Subsequent publications by Zuckerman and Gerber identified issues with inter- and intra-rater reliability when applying the Neer classification system to a set of radiographs. The reliability debate surrounding this classification system led us to understand the issue of forcing continuous variables (fracture lines are infinite in their trajectory and displacement) into dichotomous variables (a classification system). Because of Dr. Neer’s work and subsequent research, our community understands that when we make these classification designations, we will agree about 60% of the time (kappa statistic of 0.6). That level of agreement is not reflective of a “good” or “bad” classification system; rather, it’s a consequence of moving a continuous variable to a dichotomous variable.
So we remain indebted to Dr. Neer not only for laying the foundation for the treatment of patients with proximal humeral fractures, but also for vastly expanding our knowledge regarding the role, strengths, and weaknesses of disease and fracture-classification systems.
Marc Swiontkowski, MD