Latin American View of the Radiographic Study of Ankle Fractures Type AO 44-B1 e B2

Objective: to assess whether there is a type of image that guides the surgeon to indicate surgical treatment for ankle fractures and to determine whether the surgeon’s decision is associated with the age and experience of the evaluator, considering surgeons of different nationalities. Material and Methods: Three different presentations of radiographic images of ankle fractures were analyzed in 36 patients, in anteroposterior and lateral view, true anteroposterior and lateral view, and the three images combined. The images were evaluated randomly and independently by 89 professionals of different nationalities, ages, and lengths of experience. Results: For professionals from Ecuador, Argentina, Mexico, and Chile, the indications for surgery are more frequent when the images are presented in true anteroposterior and lateral view. For professionals from Peru, Venezuela, Bolivia, Paraguay, Brazil, Colombia, Uruguay, and Guatemala, as in the overall assessments, the indications for surgery are more frequent when images of the three positions combined are presented. the In the Conclusion: indications when images of the three the highest frequencies of indicated surgery were observed when the image was presented only in the true anteroposterior and lateral view. The surgical indications are not associated with the length of medical experience of the evaluating physician.


Introduction
Ankle fractures are among the most frequent injuries of the musculoskeletal system and correspond to approximately 9% of all fractures 1 and account for 10% of all x-rays performed on patients in emergency departments, resulting in a significant expense [1][2][3][4][5][6][7][8]. In ankle fractures, radiographic assessment is essential. The Ottawa ankle rules represented an attempt to reduce the indication of study by radiographs. However, the low radiation dose and the low cost of x-ray are much less important than the possibility of failure in diagnosis [6]. The exam with conventional radiographs for the diagnosis of ankle fractures consists of anteroposterior, anteroposterior with 20° internal rotation (true), and lateral images. Various studies have contested the need for three radiographic images for the diagnosis of ankle fractures [2,4,9,10].
Countless authors have suggested that two images can replace the three images without affecting patient results [2,4,9,11,12]. Some studies suggest that anteroposterior and lateral images could be used for the diagnosis of fractures [2,10]. However, when using two images, various authors recommend doing the anteroposterior with 20° internal rotation (true) and lateral image as they would provide, "with 95% accuracy", diagnosis of ankle fractures, when compared to the series of three views [4,9]. In most of these studies, the combination of two images was tested, comparing the standard series of three images, selected based on the researcher's preference (anteroposterior or true anteroposterior) [11] compared the images for the diagnosis of ankle fractures and combined the individual results to compare the performance between two and three images. This study aims to assess whether there is a type of image that guides the surgeon to indicate surgical treatment for ankle fractures and to determine whether the surgeon's decision is associated with the age and experience of the evaluator, considering surgeons of different nationalities.

Material and Methods
This cross-sectional observational study was carried out      of these professionals. The graph in Figure 1 shows the mean age of the professionals in descending order by country. In terms of means, the countries Peru, Ecuador, Chile, and Brazil had the oldest evaluators, with mean ages over 40 years, while Guatemala had a 28-year-old professional.  Table 2 shows the frequency distribution of the country in which the evaluators practice and the main statistics on the time    Table 3 shows the frequency distribution of the professionals'

Decisions of the Professionals
indications for surgery in the cases evaluated and the relative rate of surgical indication, by country, type of image presented, and overall. Figure 3 shows

Correlation coefficient between table row and column variables
(significance p-value of the correlation coefficient)  Table 3, it could be seen that the countries with higher relative rates of surgical indication are Ecuador, Peru, Venezuela, Bolivia, Argentina, and Paraguay, which have rates higher than the overall rates. The countries with lower relative rates of surgical indication are Guatemala, Uruguay, Mexico, Chile, Colombia and Brazil, which have rates lower than the overall rates of surgical indication. To determine whether the professional's decision to operate is associated with the professional's age and time since graduation, for each assessment the distributions of age and time since graduation of the professionals who did not indicate surgery and those who did indicate surgery were compared. Since the variables age and time since graduation did not follow a normal distribution (p-value of the Shapiro-Wilk test less than 0.001 for both variables), the distributions were compared using the nonparametric Mann-Whitney test.

Evaluator age Evaluator Time since Gradua-tion
In only two of the 108 assessments (1.9% of the assessments) was there a significant association of the decision to operate in relation to the age of the evaluating physician, that is, a p-value of less than 5% was found when comparing the distribution of the age of those who decided to operate with the distribution of the age of those who did not decide to operate. In both cases, the doctors who decided on surgical treatment were significantly older than the doctors who decided on conservative treatment. In the other assessments and overall, there was no significant association of the decision to operate with the age of the evaluating physician (p-values greater than 5% from the Mann-Whitney test comparing the distribution of the age of those who decided to operate with the distribution of the age who did not decide to operate). As the association of the decision to operate with the age of the evaluating physician occurred in a non-relevant frequency of cases, in only 1.9% of the assessments, it is concluded that the decision to operate is not significantly associated with the age of the evaluating physician.
In only five of the 108 assessments (4.6% of the assessments) was there a significant association between the decision to operate and the time since the medical examiner graduated, that is, a p-value of less than 5% was found in the comparison of the distribution of the time since graduating for those who decided to operate with the distribution of the time since graduating of those who did not decide to operate. In these five cases, the doctors who decided on surgical treatment were significantly more experienced than the doctors who decided on conservative treatment. In the other assessments and overall, there was no significant association between the decision to operate and the time since graduation of the evaluating physician (p-values greater than 5% from the Mann-Whitney test comparing the time since graduation of the evaluators who would not operate with the time since graduation of the evaluators who would operate).
As the association of the decision to operate with the time since graduation of the evaluating physician occurred in a nonrelevant frequency of cases (4.6% of assessments), it is concluded that the decision to operate is not significantly associated with the time since graduation of the evaluating physician. Additionally, we investigated whether the total number of the professional's surgical indications, in the three presentations and overall, is correlated with the professional's age and time since graduation.
According to the literature, various studies have shown how many images would be necessary to diagnose ankle fractures [1,4,9,10,16]. Cockshott et al. [2] observed that the fracture was present in only 29% of the x-rayed patients, Auletta et al. [13] in 30% of the patient x-rays, and Brandser et al. [11] found a 28%