Proximal humerus fractures are rare in children, constituting less than 5% of paediatric fractures
[1]. Most of these fractures occur through the growth plate due to its relative weakness [2], with
Salter-Harris type I common in younger children and type II in adolescents; types III and IV are
uncommon [3]. Shoulder dislocations are also infrequent in children, with only 8 out of 500 cases
reported in those under 10 years of age [4]. The combination of a proximal humerus fracture and
shoulder dislocation is extremely rare, affecting less than 2% of the paediatric population [5]. The first
case was described in 1982 [6]. Proximal humeral metaphysis fracture-dislocations not involving the
growth plate are exceptionally unusual [7], often resulting from high-velocity trauma and typically
seen in children aged 5-12 years [8]. In children under 3 years with an unclear injury history, nonaccidental
injury should be suspected. Anterior dislocations are more common than posterior or
inferior ones. These injuries are challenging to treat, with few published cases on their management
[9]. We conducted a systematic review of reported cases from 1980 to 2024 to summarize existing
evidence and recommend optimal treatments for achieving favourable clinical and radiological
outcomes.

This systematic review compiles and analyses all published studies on shoulder fracture-dislocations in children. Following PRISMA guidelines, we searched databases including PubMed, Scopus, and ScienceDirect as outlined in Figure 1. Articles on traumatic shoulder fracture-dislocation in children were evaluated, and key data were extracted based on predefined criteria, including patient age, sex, diagnosis, treatment method, and clinical and radiological outcomes.

Authors’s comments on the search method We include a limited number of databases due to the following reasons. While we are open to all databases, we primarily focused on Scopus, PubMed, and ScienceDirect because they are subscribed to by our institution and readily available to us. These databases are renowned for their comprehensive and high-quality data on the specific type of fracture we were studying. Given the rarity of this fracture, we prioritized obtaining reliable and relevant data, which sometimes meant emphasizing quality over quantity. Expanding the number of databases could have introduced variability and potential inconsistencies, potentially detracting from the clarity and focus of
our analysis. For example, a search on Google Scholar returned 9,890 articles, most of which were irrelevant. Using Scopus, PubMed, and ScienceDirect, we aimed to ensure that our review was based on the most pertinent and trustworthy sources.

The search and selection process are described in Table 1, resulting in 18 studies as outlined in Table 2 after applying the following inclusion criteria.
 

Inclusion criteria

The inclusion criteria involved all case reports from 1980-2024 of children who underwent closed or open reduction for traumatic fracture-dislocation of the shoulder, with no exclusions based on geography or language. We included only cases with imaging evidence of a fracture involving the epiphysis or metaphysis of the humerus, accompanied by a shoulder joint dislocation following a traumatic episode.

In recent years, there have been reported cases in the literature that were not included due to our strict inclusion criteria. For instance, in 2022, Fettah et al. [24]. reported a fracture-dislocation of the left shoulder in a 14-year-old girl from group A. The treatment was successfully performed using a combination of closed intramedullary K-wire fixation for the fracture and open reduction for the dislocation. However, the follow-up data was missing, so we did not include it in our review.

In 2024, Navaeifar et al. [25]. reported a case of anterior dislocation in a 6-year-old child with a fracture. Since the fracture involved the clavicle, it was not considered in our review.

The patient cohort consisted of 10 girls and 8 boys, with a median age of 6 years. The most common injury mechanisms were falls from height (9 cases) and road traffic accidents (3 cases). The cases were categorized into two groups based on epiphyseal involvement as shown in Tables 3 and 4.

• Group A (with epiphyseal involvement): 11 cases (61%), comprising 6 boys and 5 girls, with 7 left-sided and 4 right-sided injuries. Of these, 8 (72%) were anterior dislocations, 2 (18%) were posterior, and 1 was (9%) inferior.
• Group B (without epiphyseal involvement): 7 cases (39%), comprising 5 girls and 2 boys, with 5 right-sided and 2 left-sided injuries. Six (85%) were anterior dislocations and 1 (15%) was posterior. The median age in this group was 6 years.

Treatment

In Group A, three children with SH type I injuries underwent open reduction and two underwent closed reduction, all with K-wire fixation. Those treated with closed reduction regained full movement in 6 weeks, earlier than those with open reduction. For SH type II, two children had open reduction and one had closed reduction with K-wire fixation. All three children with SH type III injuries were treated with open reduction and K-wire fixation.

In Group B, four cases had open reduction with K-wire or elastic stable intramedullary nail (ESIN) fixation, while three cases had closed reduction with intramedullary and percutaneous fixation, performed through a posterolateral approach at the distal humerus. All intramedullary devices passed through the epiphysis, and no cases of early epiphyseal closure or avascular necrosis (AVN) were reported.

Trauma to the shoulder can cause various injuries, including rotator cuff strain, glenohumeral subluxation, proximal humerus fractures, and joint dislocations. Paediatric proximal humerus fractures are rare, accounting for 2% of paediatric fractures and 3-6.7% of physeal fractures [1,3]. Combined fractures and shoulder dislocations in children are extremely rare, typically resulting from high-energy trauma. Over the past 40 years, only 18 cases have been reported, with an increase over the last 10 years [26]. This trend is illustrated in Figure 2.

When the injury history is vague, non-accidental injury (NAI) should be suspected. A thorough examination and skeletal surveys are recommended to rule out NAI, as seen in two reported cases [6,14].

Proximal humerus fractures in children can be categorized based on growth plate involvement. Younger children typically sustain Salter-Harris (SH) type I fractures, while older children (5-12 years) often have metaphyseal fractures without physis involvement.

Five cases of SH type I injuries, all involving anterior dislocation, have been reported. Two cases were suspected of non-accidental injury (NAI). Closed reduction succeeded in two children due to a partially intact periosteum and capsule, while the other three required open reduction [13]. The closed reduction cases achieved full movement more quickly.

Three cases of SH type II injuries occurred in older children. One case, managed with closed reduction, took 8 months for fracture remodelling and 76 weeks for full recovery. The other two, treated with open reduction and K-wire fixation, recovered fully in 12 and 24 weeks [15].

SH type III injuries are rare, occurring in adolescents and requiring open reduction due to complex fracture patterns. The incidence of avascular necrosis (AVN) is high (66.6%) but was not seen when smooth K-wires were used [18].

In Group B, seven cases were involved shoulder dislocation with metaphyseal fractures but no epiphysis involvement. Non-operative treatment is generally recommended due to children's high remodelling capacity. However, for combined dislocations and fractures, most authors recommend reduction and internal fixation to prevent complications [5]. Successful closed reduction and intramedullary fixation were achieved in some cases [21,22], while others required open reduction [7,9,10,23]. There was no significant difference in radiological union and functional recovery between closed and open reductions for these injuries.

For injuries affecting the epiphysis, open reduction with pin fixation is recommended to achieve an anatomical position of the epiphyses and humeral head. In cases without epiphyseal injury, closed reduction with intramedullary fixation allows for a quicker return to full shoulder movement compared to open reduction. Proper pre-operative imaging is essential to accurately assess the humeral head position before surgery to ensure normal anatomical restoration.

Limitations

One of the primary limitations of our mini- review is using only three databases: PubMed, Scopus, and ScienceDirect. This selective approach might have restricted the comprehensiveness of our review by potentially excluding relevant studies that are indexed in other databases.

By not including additional databases such as Embase, Cochrane Library, or Web of Science, we may have missed pertinent research, leading to potential publication bias.

Our choice was influenced by the availability of subscriptions and the known reliability and relevance of these databases to our specific research topic. While these databases provided high-quality and comprehensive data on the rare fracture type we studied, future reviews could benefit from a broader database search to ensure an even more exhaustive collection of relevant literature. This broader approach could help mitigate the risk of missing valuable studies and provide a more balanced and comprehensive understanding of the topic.