Factores clínicos que predicen fracturas asociadas con una dislocación de hombro

"Factores clínicos que predicen fracturas asociadas con una dislocación anterior de hombro."

Marcel Émond, MD, MSc, Natalie Le Sage, MD, MSc, André Lavoie, PhD and Louis Rochette, MSc

From the Trauma Research Unit, Centre Hospitalier Affilié Universitaire de Québec (ME, NL, AL), Emergency Medicine Division, Department of Family Medicine (ME, NL), and Department of Social and Preventive Medicine (AL, LR), Faculty of Medicine, Laval University, Quebec City, Quebec, Canada

Objectives:
To identify risk factors for fractures associated with an anterior shoulder dislocation treated in an emergency department (ED).

Methods:
A retrospective case–control study over five years of patients with an anterior shoulder dislocation was accomplished in a university-affiliated ED. Chart review identified possible predictors of fractures. Comparing the profile of patients having a clinically important fracture associated with their shoulder dislocation (cases) with those sustaining a noncomplicated dislocation (controls) provided the outcome measure.

Results:
A total of 334 patients were included in the study. Eighty-five (25.5%) had a clinically important fracture-dislocation, and the remaining 249 (74.5%) sustained a noncomplicated shoulder dislocation. Chi-square, logistic regression, and recursive partitioning analysis showed three significant factors for the presence of fracture-dislocation: 1) age 40 years or older, 2) a first episode of dislocation, and 3) mechanism of injury (i.e., a fall greater than one flight of stairs, a fight/assault episode, or a motor vehicle crash). A multiple logistic regression model estimated the significant adjusted odds ratios (and their 95% confidence intervals [95% CIs]) for each of the three factors: 5.18 (95% CI = 2.74 to 9.78), 4.23 (95% CI = 1.82 to 9.87), and 4.06 (95% CI = 1.95 to 8.48), respectively. A predictive model using any one of the three factors reached a sensitivity of 97.7% (95% CI = 91.8% to 99.4%), a specificity of 22.9% (95% CI = 18.1% to 28.5%), and a negative predictive value of 96.6% (95% CI = 88.3% to 99.6%).

Conclusions:
Three risk factors predict clinically important fractures that are associated with shoulder dislocation: age, first episode, and mechanism of dislocation. A prospective validation may lead to standardized use of prereduction radiographs of the shoulder in the ED.

Academic Emergency Medicine Volume 11, Number 8 853-858, © 2004.