Knowledge of changes in the management of this fracture may enable us to improve the management of our patients. We conducted a descriptive cross-sectional study in 6 hospitals representing the 4 districts of the city of Kinshasa in the DRC. The study used the records of adult patients who presented with ULF between January 2009 and December 2018. The delay in publishing the results of this work was due to the global health crisis linked to COVID-19 but we remain convinced of the importance of the epidemiological character of this study in that it constitutes a data base for future work.

Data were collected using a pre-established form: definitive diagnosis of ULF confirmed by radiography, CT scan and/or MRI. Sex, age, limb involved, fracture site, type of treatment, immobilization material used. Therapeutic course and prognosis were not analyzed in this study.

Given the convenience sampling, we believe that the results of this work cannot be generalized, as the sample is not representative.

The data were entered into sphinx V5 and exported to Excel 2010 for verification. After a quality and consistency check, the data were exported to IBM SPSS (statistical Package for social sciences), version 21.0. The results were presented in tabular form, expressed as a percentage, frequency, median and extremes.

This study was validated by the head of the integrated emergency service, followed by authorization from the Ethics Committee of the School of Public Health of the University of Kinshasa (Approval number: ESP/CE/083B/2016). The data were processed anonymously in accordance with the Helsinki declarations.

Between 2009 and 2018, we counted 852 ULF treated in 5 hospital structures in the city of Kinshasa. The age distribution of fractures revealed a very high rate in young patients (between 16 and 45 years) n = 571 (67%) with a median of 38 years and extremes of 16 and 85 years. The distribution of fractures in the age groups reveals a very high rate in young patients (age between 16 and 45 years) n = 571 (67%) with a median of 38 years and extremes of 16 and 85 years. This distribution tended to decrease after the age of 45, reaching its lowest rate at the age of 60 (n = 56, 6.5%).

The majority of patients were men (n = 528, 62%), compared with women (n = 324, 38%). (Table 1)

Of the 852 fractures recorded, the humeral diaphysis is the preferred fracture site n = 292 (34.3%); the distal radius comes second with 13.5%; the radial and ulnar diaphyses represent n = 79 (9.3%) and n = 74 (8.7%) respectively and the middle 3/5 of the clavicle n = 55 (6.5%). in total The humeral diaphysis is the preferred fracture site n = 292 (34.3%). The distal radius came second n = 115 (13.5%); the radial and ulnar diaphyses represented 9.3% (n = 79) and 8.7% (n = 74) respectively, and fractures of the middle third of the clavicle 6.5% (n = 55). The data in this table show that fractures of the upper limb were generally single n = 844 (99%) and their location in relation to the elbow was in almost equal proportions n = 425 (49.9%) for fractures located above the elbow and 424 (49.7%) for location below the elbow with a predilection for the left upper limb (Table 2).

These accounted for n = 63 (7.4%). According to Robinson’s classification, these fractures were more frequently observed in the middle third n = 55 (6.5%). According to the literature, this location is less resistant to flexion and torsion forces [10] . Fractures of the 3/5ths of the clavicle have the advantage of being able to be consolidated without surgery, with 3% to 7% of non-unions in simple fractures and 20% to 33% in complex fractures [11] .

There has been increasing surgical activity in CF. A Finnish study reported a 9-fold increase between 1987 and 2010 for no obvious reason [3] .

Since time immemorial, all CF have been considered good candidates for surgery [12] .

Current evidence recommends NOT for fractures with little or no displacement [13] . Fractures with a displacement of more than 15 mm are considered surgical in order to restore the length of the scapuloclavicular triangle. [11] . It is also practical to treat young, active patients surgically [14] .

Our study noted a relatively high rate of recourse to ST. Could this be explained by the high rate of young people in this study? A more in-depth study could enable us to explain this observation.

The results of ST and NOT in CF were compared: ST resulted in a significantly lower rate of non-unions, symptomatic malunions and earlier functional return. However, the functional difference appears to be minimal in the long term [3] [15] .

Other studies have compared the results of different osteosynthesis materials with variable results [11] [13] [16] .

Proximal humerus fractures accounted for n = 34 (4.0%) of all fractures. These fractures were treated conservatively in 38.2% of cases n = 13. The use of ST was observed in 62% n = 21 and the material used was dominated by plates n = 9 (26.5%) followed by EFs n = 7 (20.6%).

It is clear that PHFs have a high level of surgical activity. This observation should be put into perspective given the relatively small size of the sample.

The majority of PHF can be treated orthopaedically with acceptable results. Only 15% - 20% of fractures are displaced or unstable and require ST [1] .

Several authors have noted a very high level of surgical activity on PHFs over the last twenty years [1] [3] .

Other authors have evaluated the rates of use of different osteosynthesis implants. Open reduction and plate osteosynthesis were the most frequent procedures, with an increasing rate throughout the study period. The trend was particularly marked in women in the older age groups but the results of these techniques do not allow us to recommend one technique over the other. [17] [18] [19] [20] .

Beks also noted an increase in complications and a higher revision rate in the surgical [21] .

Other studies, on the other hand, have been able to demonstrate the ascendancy of ST over NOT [22] . Olerud found that patients treated surgically with hemiarthroplasty had a slight improvement (disabilities of the arm, shoulder and hand score (DASH) < 10) compared to those treated non-surgically [23] .

It is clear that several studies have not been able to globally demonstrate the superiority of one treatment over the others. High quality studies are needed to point the way.

Humeral shaft fractures had a very high frequency during our study period n = 292 (34.3%) of all fractures with high surgical activity n = 228 (78.1%). This activity was dominated by the external fixator n = 140 (47.9%). The use of the NOT was observed in a proportion of 23.6% (n = 69).

As with the other fractures previously studied, HDF was the subject of a high level of surgical activity during the period covered by our study. This finding appears to be consistent with the literature.

In the past, HDF was treated non-surgically with a high rate of consolidation (Sarmiento et al. 2000).

Huttunen 2014 [3] noted a steady increase in the ST level of HDF. Patients treated surgically had open surgery (n = 1268 or 98%) and EF was quite rare (n = 71.2%).

The use of EF in the treatment of DFH seems to be decreasing significantly and its indications remain limited [24] .

In our study, EF is still widely used. Is its use dictated by absolute indications? This study does not provide a precise answer to this question. It is possible that it is used more widely in our study, particularly in the case of closed fractures requiring nailing or plication. It is also likely that its use was dictated by its availability, surgeons’ habits and the ease of the technique.

To date, no high-quality studies have compared the results of NOT and ST traitements for HDF. Recent studies have shown very high healing rates and good functional results after ST [25] .

It is clear that HDF in general can be treated without surgery with good results, as demonstrated by case series [12] .

It is true that HDF are increasingly being treated with plaster casts and nailing, but the results after these procedures are not unequivocal.

Heineman et al. 2010 published a meta-analysis including 203 patients [25] . The authors found no difference in functional outcome between plating and nailing of HDFs.

Kurup, Hossain [26] , dans une autre étude portant sur 160 patients, ont comparé l’enclouage intramédullaire (IMN) et la mise en place d’une plaque; aucune différence nette n’a été observée entre les deux interventions en ce qui concerne les résultats fonctionnels. D’après les résultats de cette même étude, il existe une restriction de mouvement de l’épaule en cas d’IMN qui favoriserait la plaque.

Sur base de ces éléments, aucune conclusion ne peut être tirée quant au choix des options thérapeutiques (ST ou NOT).

These represented 13.5% (n = 115) of all fractures recorded. The treatment of these fractures was dominated by the NOT n = 76 (66.1%); the ST noted the use of plates, EF and pins in the order of n = 21 (18.3%); n = 10 (8.6%) and n = 8 (6.9%) respectively.

These data show a much greater use of the traditional NOT method. This result confirms the trend observed in our preliminary study carried out in the university clinics of Kinshasa; however, it should be noted that there is a high rate of recourse to ST with a predominance of plaques.

The management of DRF is becoming increasingly standardised, despite the lack of consensus between the various authors [27] . ST has increased too much with the extensive use of plaques [3] .

Studies have compared functional results between ST and NOT in patients aged 70. In these retrospective studies, the authors found that there was no difference in functional or pain results [28] .

Hector quoted by Gutierrez [29] noted good functional results at 1 year in favour of plates, but this difference was not clinically significant, suggesting that the two types of management are equally effective in patients over 60 with DRF.

Other randomised studies have demonstrated that the differences between the techniques are mainly in terms of complications [26] .

Yan et al. 2019 using Rikli and Regazzoni’s classification noted that DRF involving the orthopaedically treated intermediate spine had a detrimental effect on forearm rotation [30] .

It is clear that several studies have not been able to demonstrate the overall superiority of one treatment over the others. High-quality studies are needed to show the way.

Strength of the study

This study enabled us to determine the temporal evolution of the management of upper limb fractures in Kinshasa. It therefore provides a basis for detecting future changes.