RESUMEN
INTRODUCTION: Displaced fractures of middle third of clavicle are traditionally managed non-operatively. Recently, the trend is towards surgical management considering functional deficits in some of the non-operatively managed patients. The purpose of the study was to examine the functional outcome of non-operative treatment in these injuries, while identifying the factors responsible for less than ideal outcome and determine the guidelines for patient counselling. PATIENTS AND METHODS: One hundred patients with displaced fractures of middle third of clavicle were prospectively evaluated clinico-radiologically for one year from injury. Risk factors for non-union were determined using Fisher's exact test. Logistic regression was used to identify factors contributing to functional outcome. Threshold values for the radiological displacements were estimated with the smooth threshold regression using the logistic transition function. RESULTS: Ninety-four out of 100 fractures united. The factors associated with non-union were: smoking, diabetes and poor soft tissue condition over the fracture. Ninety-two out of 94 patients who had their fractures united achieved "good" (Constant Score above 70) or "very good" (Constant Score above 85) functional outcome at one year, out of which only 49 belonged to "very good" category. The most important factors influencing functional outcome were fracture angulation and clavicular shortening. From the model, it was estimated that with one unit increase in degree of angulation or 1 mm increase in shortening, the odds of scoring above 85 reduces by around 14%. Based on this, an equation and a probability calculator were developed from which the probability of achieving a Constant Score above 85 can be calculated. Threshold analysis yielded 22.8° for angulation and 16.8 mm for shortening at which the probability of achieving "very good" function is only 0.23. CONCLUSION: Displaced midshaft clavicular fractures with the intent of achieving "good" outcome must be managed non-operatively. If the patient expectation is not to accept even minor functional deficits, the treatment objective must be raised to achieve a minimum Constant score of 86. Substituting the radiological displacements in the equation, the probability of the patient achieving this objective can be calculated. Non-operative management is offered if this probability figure is acceptable to the patient, otherwise counselled for surgery. Non-operatively managed patients with coexisting diabetes, smoking or poor soft tissue condition must be watched for the possibility of going for non-union.