Demographic Characteristics of Paediatric Pelvic Fractures : 10-Years ' Experience of Single Paediatric Orthopaedics Clinic

Objective: To evaluate the demographic characteristics of paediatric pelvic fractures. Materials and Methods: Retrospective analysis of 26 patients who were hospitalized with the diagnosis of paediatric pelvic fracture between 2000 and 2010 was performed. Age, gender, hospitalization time, mechanism of injury, fracture type, associated injuries, haemoglobin level drop in the first 24 hours, management and blood transfusion requirement, injuries time (month) information were gathered from hospital records. Results: There were 16 male and 10 female patients. Average age was 10.5 (2-16). Average hospitalization time was 3.5 days (1-17). Average haemoglobin level drop in the first 24 hours was 1.51 (0.3-3.6) gr/dL. Mechanisms of the injuries were as following; 14 patients were struck by a car, 10 patients fell from height and 2 patients involved in a vehicle traffic accident. According to the classification of Torode and Zeig; there was 1 type 2, 22 type 3 and 3 type 4 injuries. Injuries’ occurrence season were; 12 in spring, 7 in summer and 7 in autumn. All of the patients had been managed conservatively. Conclusion: These injuries are rarely seen in children. Their management can be mostly conservative and even with a simple and stable pelvic injury marked bleeding can occur.


Introduction
Trauma is the leading cause of death and disability in children [1].Paediatric pelvic fractures are rare; their incidence has been estimated between 2.4-7.5% among all children's fractures [2].Pelvic fractures are consequences of trauma and are typically the results of high-energy forces such as motor vehicle collisions, falls from height or crush injuries [3].Although adult pelvic fractures were well documented previously, there are only few data about the paediatric pelvic fractures [4].
There is not enough data about paediatric pelvic fractures for our country population.Our hospital is a reference centre for paediatric trauma patients in a city where 4 million people live in.These patients are referring to our hospital especially out of the work hours.Our study questions were: How often did our clinic encounter these injuries?Which injuries were accompanying paediatric pelvis fractures?Which treatment modalities were performed about these injuries?The objectives of this study include understanding the patterns of paediatric pelvic fractures, their demographic characteristics and observing the injuries associated with these paediatric pelvic fractures.

Materials and Methods
All children which were hospitalized with a diagnosis of pelvic fracture over a 10-year period (2002-2012) were evaluated retrospectively from the hospital medical records.This study was approved by the Local Ethical Committee.
Inclusion criteria were: patients who were under 16 years old, who were hospitalized and who had pelvic fracture.All of the hospitalized patients' files were evaluated for pelvic fracture diagnosis.Patients' medical history and their radiographs were reviewed.Data recorded included: age, gender, hospitalization time, mechanism of injury, fracture type (according to Torode and Zeig) [5], associated injuries, haemoglobin level decrease in the first 24 hours, management and blood transfusion requirement, and injuries time (month).Number of patients who were admitted with trauma etiology to the emergency department during this period was gathered from hospital records.Detailed diagnosis of these patients was not available.

Statistical Analysis
Statistical Package for the Social Sciences (SPSS) for Windows IBM® (New York, United States of America) was used for descriptive statistical analysis.We decided to compare haemoglobin level decrease and grade of the injury according to Torode and Zeig at first; however, since it would be confusing because 2 of 3 patients with type 3 injury had received red blood cell transfusion on the first day, we decided not to perform comparative statistical analysis.
Average hospitalization time was 3.53±3.52days (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17).The patient who was hospitalized for 17 days had type 4 pelvic injury and additional femur and tibia fractures (patient number 1, Table 1).Mechanisms of the injuries were as following; 14 patients were struck by a car (53.8%), 10 patients fell from height (38.4%) and 2 patients involved in a vehicle traffic accident (7.8%).The fracture types, associated injuries, haemoglobin level decrease in the first 24 hours, and red blood cell transfusion rates are listed in Table 1.Mean decrease of haemoglobin level in the first 24 hours was 1.51±0.94gr/dL (0.3-3.6).During this time period, 3264 paediatric patients were found to have been admitted to emergency department with an etiology of trauma (major or minor trauma).According to the classification of Torode and Zeig; there were one type 2 (3.8%), twenty-two type 3 (84.6%),and finally three type 4 injuries (11.6%).All of the patients were treated conservatively except for one on whom the surgery was performed only for the fractures of tibia and femur (titanium elastic nail).The patient had also symphysis pubis separation, and was treated conservatively by pelvic sheet.The occurrence times of the injuries are provided in Table 2 as months.Analgesic suspension or pills were given as pain killer according to the patients' age and tolerance.The patient with type 2 fracture was treated conservatively by bed rest for 3 weeks.Then, progressive load bearing with crutches was allowed as the patient tolerated.Patients with type 3 injury were treated as type 2 injury but those patients who had posterior injury or bilaterally injury were allowed to bear weight after 6 weeks.Simple pelvic sheet had been applied to patients with type 3 injury.There were no patients who were treated by skeletal traction.Radiographic examination of type 4 and type 3 injuries are shown in Figures 1 a,

Discussion
Paediatric pelvic fractures are rarely seen entities.There are not enough data about these injuries in our country   As our clinic usually deals with paediatric trauma, we decided to review our hospitalized paediatric patients with a diagnosis of pelvic fracture.As a result we recognized that we had very rarely hospitalized these patients and we had treated whole of them conservatively.
It is difficult to determine the incidence of paediatric pelvic fractures.Incidence varies from 0.2% to 7.5% in all paediatric injuries [6].As our unit is one of the biggest centres dealing with paediatric trauma, for a 10-year period, 26 paediatric patients were hospitalized with a diagnosis of pelvic fracture.During this period, total traumatized patient number admitted to emergency department was 3264.Therefore, the incidence was 0.79%.This incidence is significantly lower than the previous data.During 1970's, Watts stated that 10 pelvic fractures per year could be expected in large children's hospitals [7].The incidence was 3.5% (39/1129) in Leonard et al. 's study group [4].Bond et al. [8] had reported 2.4% incidence in their series.
The most common mechanism of injury was pedestrian hit by car (53.8%) followed by fall from a height (38.4%), in our study group this finding is similar with the literature [9].
There has to be large amount of energy to form fracture at paediatric pelvis because of greater cartilaginous volume and bony plasticity [10].Some additional injuries usually accompany these fractures.In 10 of our patients (38.4%), there were no associated injuries, however these fractures were mostly minor-stable injuries.Only three of our patients had unstable pelvic injuries (11.6%) (Patient 1, 8 and 23 -shown in Table 1) and most serious associated injuries were tibia and femur fractures (Patient 1).
In the literature, it is stated that unlike adult pelvic fractures, severe haemorrhage from the injury side is uncommon for paediatric patients [11,12].Our study group demonstrated different findings about haemorrhage.The mean decrease in haemoglobin levels was 1.5 gr/dL (0.3-3.6).This decrease is significant for the paediatric patients because of their relatively small amount of blood volume.If these patients have anaemia, blood transfusion would be required.In our patient group, 3 patients required red blood cell transfusion.First patient had tibia and femur fractures as associated injuries and pelvic injury was unstable, his initial haemoglobin level was 12.7 gr/dL.Second patient had urethral injury as associated injury and bilaterally pubic ramus fracture, and her initial haemoglobin level was 11.2 gr/dL.Third patient had an unstable pelvic injury and his initial haemoglobin level was 9.8 gr/dL.The patients who have unstable pelvic injuries, low initial haemoglobin levels and/or associated severe injuries more commonly seem to require blood transfusion.
It has been observed that paediatric pelvic fractures had a seasonal trend [12].Children may be much likely to sustain a pelvic injury in summer and spring months, because they are school free and are usually outside because of good weather and long day-time in our country.Our study group's injury occurrence time is similar with the previous data; none of the injuries had occurred in winter time.

a b
Management of pelvic fractures in children has traditionally been non-operative [5,7,13,14].Because of the remodelling potential of the immature skeleton, most pelvic deformities could be expected to heal conservatively.This remodelling potential has been questioned [7,15] and surgical treatment had been advised for some circumstances.This newer operative concept is controversial.Blasier et al. [16] reported equivalent results for operative and non-operative management.We managed all of the injuries conservatively, there were three unstable injuries and they did not have any complaints at their clinic visits.We are aware that this is very limited number of patient to state that Torode-Zeig type 4 patients have to be managed conservatively.
There were one type 2 (3.8%), twenty-two type 3 (84.6%)and three type 4 injuries (11.6%) according to the classification of Torode and Zeig among our study group.In 2012 Shore et al. [17] modified the classification system of Torode and Zeig by separating type 3 into A and B. If there was a stable anterior pelvic ring injury they classified it as type 3A, and if both anterior and posterior pelvic ring fractured but pelvic ring remained stable this injury was classified as type 3B.If we had used this new classification only 3 patient's classification would have been changed into type B among patients with type 3 injury (patient number 22, 24 and 26-Table 1).In Shore et al. 's study group, there were 4% type 1, 14% type 2, 71% type 3 (37% type A, 34% type B) and 10% type 4 injuries (totally 115 patients in 8 years) [17].In Niedzielski et al. 's study group, there were 40.7 type 1, 25.9% type 2 and 3 and 34.4% type 4 injuries (totally 27 patients in 8 years) [6].
This study has several limitations.First of all, this study includes only retrospective data, and this is the most important limitation of the study.The clinical outcome measures could not be obtained.The number of the patients which were included in this study can be determined as limited but the patients with type 1 injuries could not be included in this study because of being treated as out-patient clinic and not hospitalized.Similar articles in recent literature have limited number of patients, too [6].
In conclusion; our experiences about paediatric pelvic fractures suggest that, these injuries are rarely seen and conservative management of these fractures can be acceptable.We should keep in mind that; even with a simple and stable paediatric pelvic injury, marked bleeding could be seen.Overall demographic characteristics of our country population are similar with other reports of these injuries.
b and 2 a, b.

Figure 1 .
Figure 1.a, b.Case number 1, symphysis pubis and left sacroiliac joint separation (a) Case number 1, end of three years follow-up (b).