ISSN 0972-978X 

  About COAA








Segmental humeral fractures: prevalence of non-union and functional outcome

Simon Fogerty1, Elefterios Tsiridis2, George Kontakis3, Robert West4,
Hans Christoph Pape5,Peter Giannoudis6

1 Specialist Registrar, York Hospital, Yorkshire
2 Consultant / Senior Lecturer, Academic Department of Trauma &
   Orthopaedic Surgery, School of Medicine, University of Leeds.
3 Associate Professor, University, of Crete, Greece.
4 Reader in Health Statistics, University of Leeds .
5 Professor, University of Pittsburgh Pittsburgh, USA
6 Professor, Academic Department of Trauma & 
   Orthopaedic Surgery, School of Medicine, University of Leeds.

Address for Correspondence:

P.V. Giannoudis Professor Trauma & Orthopaedic Surgery, 
Academic Department of Trauma & Orthopaedics,
The General Infirmary at Leeds, LS1 3EX, UK
Phone:  0044-113-3922750
Fax    :  0044-113-3923290



The aim of this study was to review a multicentre experience in treating segmental humeral fractures and report related complications and outcomes.
From 1996 to 2005 adult patients with segmental humeral fractures were identified. Demographic details, method of fixation, time to union and complications were recorded. At follow up all patients were assessed in terms of radiological result, range of movement and pain. Pain and range of movement were combined to give a final outcome rated as good, moderate or poor.
There were 63 patients. 22 non-operative, 11 intramedullary nailing, 17 plating, 2 Rush pins, 6 Enders nails and 5 external fixation. The non-union rate was 22.2% (14 patients). 64.2% (9 patients) of the 14 non-union patients went on to union in one, but not both, fracture sites. The Pearson chi-squared test is highly significant (p=0.002). It is clear that some treatments give better results than others with nails showing better outcomes.
Despite the severity of this injury a satisfactory outcome can be expected. The most successful method of surgical treatment was found to be the intramedullary fixation.

J.Orthopaedics 2009;6(2)e6


Segmental; humeral fracture; multicentre experience.


Segmental humeral fractures have been successfully treated using closed reduction and splinting followed by functional bracing21. Sarmiento et al found consistent restoration of shoulder and elbow movements with this method of treatment and similar results were confirmed by others3,5. Severely comminuted or segmental fractures however treated non-operatively may predispose to an unsatisfactory outcome13. Bone distraction and loss of fracture reduction could lead to delayed or non-union of the fracture15. Surgical stabilisation is advocated in certain circumstances including failure of acceptable closed reduction13,6,17, open fracture with nerve injury4, the multiply injured patient6, pathological fracture14 and the uncooperative patient6. Different stabilisation techniques have been described including intramedullary nailing13,15,6,14,19, open reduction and internal fixation (ORIF)6,4,16,25 or external fixation16. Intramedullary fixation offers the advantages of biomechanical load-sharing6, closed insertion technique6, decreased soft-tissue disruption13,14 and preservation of the extramedullary blood supply6. However, this method has been implicated in the development of shoulder and elbow dysfunction6,17 malrotation14 and radial nerve palsy15. Chen et al6 stated that plate osteosynthesis allows early active mobilisation of the shoulder and elbow but may require extensive soft tissue exposure and periosteal stripping13,6,14,15,25 thus disturbing the local biological substrate of the fracture. External fixation also plays its part particularly, although certainly not exclusively, in stabilising open fractures16 but pin site infection, non-union and patient compliance may limit their use16. Currently, the literature with regard to the management and outcomes of segmental humeral fractures remains poor. 

The aim of this study therefore was to review a multicentre experience in treating segmental humeral (AO type 12B or C) fractures. We wished to investigate the experience of our institutions in managing these injuries and report related complications and outcomes.

Materials and Methods:

A retrospective study of all adult patients treated with humeral fractures in three trauma centres between 1996 and 2005 was performed using clinical data and radiographs. Inclusion criteria for participation in the study were the presence of either a closed or an open fracture of the humerus with a segmental component. Patients with impending or pathological fractures were excluded. A segmental fracture was defined as a two-level humeral fracture with at least one intermediate free segment.

Fractures were classified according to the AO system and type 12B or C were included in the study20. Demographics, injury severity score (ISS)2, mechanism of injury, whether the fracture was closed or open (Gustilo Classification12), presence of associated injuries and method of stabilisation were recorded. Post-operative data such as the time to union, incidence of delayed union and non-union, other complications and range of movement were also recorded. Clinical healing was defined as pain-free functional range of movement and radiological union was defined as the evidence of callus formation in both the antero-posterior and lateral radiographs. Non-union was defined as a fracture not conforming to these characteristics by 6 months post-injury. At the final follow-up the patients were asked to complete a visual analogue scale for pain (table 1) and examined for range of movement of the upper limb (table 2).  These were combined to give an overall outcome score of good, moderate or poor (table 3).

Pain level on visual analogue scale (1-10)












Table 1: Grading system for pain

Range of movement


100% to 50% of the normal range


Between 50% and 25% of the normal range


Less than 25% of the normal range


Table 2: Grading system for range of movement.

Combination of pain grade and range of movement grade

Outcome, for the purposes of results and statistical analysis

Good and good


Good and moderate


Moderate and moderate


Moderate and poor


Poor and poor


Good and poor


Table 3: How the pain and range of movement grades were combined to give a final outcome.

The results were analysed using the Kruskal–Wallis test for one-way non-parametric ANOVA (analysis of variance) using SPSS, R18 and Stata23 packages.

Results :

Out of 810 humeral shaft fractures identified from our prospectively documented databases, 66 were found to fulfil the inclusion criteria (8.1%). Out of 66 patients identified, there were 3 deaths all unrelated to fracture treatment leaving 63 cases for the final analysis (Figure 1a,b,c,d). There were 33 females and 30 males with a mean age of 51.1 years (range 18–95) and a mean ISS 14.1 (range 9–41). There were 29 type 12B (12B1, 9B2, 8B3) and 34 type 12C (16C1, 5C2, 13C3) fractures. The mean length of the intermediate segment was 7.9 cm (range 3.5–17.5). The causes of injury included 23 vehicular accidents, 22 falls from standing (elderly), 13 falls from heights, 3 skiing injuries, 1 cyclist and 1 climber. Thirty patients had associated injuries including head, chest, pelvis and other long bone fractures. The mean follow up was 36 months (range 24–60).

Figure 1a and b
Radiographs showing a segmental humeral fracture in a 55 year old female following a motorcycle collision.

Figure1 c and d
Antero-posterior and lateral radiographs showing union of the fractures at the 5 month follow up having been originally stabilised with plating.

Demographics, fracture classifications, Injury Severity Scores, methods of treatment, complications and outcome measures are all presented in table 4. We correlated total range of movement, complications and time to union in all patients. For the purposes of analysis, due to small numbers we combined Enders and rush pins to make one group of ‘pins’. The overall non-union rate was 22.2% (14 patients). In 9 out of 14 cases (62.4%) of non-union, one site united but the other did not. The details are discussed in the complications section.   

Non-union column
B=Non-union both proximally and distally
P=proximal non-union only
D=distal non-union only

Table 4 : Results

Table 5 shows a simple comparison of outcome with treatment group. The Pearson chi-squared test for association between treatment type and outcome is highly significant (chi-square = 17.03 with 4 degrees of freedom and p=0.002).  Thus it is clear that some treatments give better results than others with nails showing more good outcomes and pins much less so.  



Good or moderate/poor







Treatment group








% within Treatment group





ex fix







% within Treatment group












% within Treatment group












% within Treatment group












% within Treatment group










% within Treatment group




Table 5: Treatment group and good or moderate/poor cross tabulation

The above analysis is unadjusted for other factors such as sex and age.  Logistic regression was used to adjust for further factors. A moderate/poor outcome can be predicted for pins and so these patients were dropped from the analysis. Sex, ISS, open/closed were all found to be not significant but age and therapy type remained of interest since they were significant at the 10% level in this exploratory model.  A moderate or poor outcome is considered bad and the adjusted results are: Age OR=1.032 per year increase in age (95% CI is (0.999, 1.066), p-0.059) so that the risk of bad outcome is increased with age; compared to non-operative treatment, the OR for External Fixation is 5.12 (95% CI (0.43, 6.47), p=0.196), the OR for Nail is 0.14 (95% CI (0.01, 1.33), p=0.087), the OR for the Plating group is 1.82 (95% CI (0.47, 6.98), p=0.385).

Figure 2 demonstrates that range of movement varies considerably with treatment.  This is confirmed by Kruskal–Wallis test for equality of median range of movement which is highly significant (Chi-square=43.9; df=4, p=0.0001). Nails and external fixation provided greater range of shoulder movement whilst Enders/Rush pins and plates were more restrictive than non-operative treatment.  

Figure 2: Total range of movement against method of treatment.  A Kruskal–Wallis test for equality of median range of movement yields c2= 43.9 with 4 d.f. (p= 0.0001).

Time to union varies with type of treatment. The variation was small however, with most fractures healing between 12 and 19 weeks. The non-operative, plate and Rush/Enders pins groups did have non-unions as detailed in the complications section of this paper. Since these patients required revision surgery to achieve union they did not have a true time union, therefore an analysis would be unhelpful.

Complications and outcomes:

a) Non-operative treatment
Complications are summarised in Table 4. Of the twenty-three patients who had non-operative management, six progressed to non-union. Four had united at the proximal but not the distal site. Two uni-nonunion and two bi-nonunion patients were operated on with plating, 2 with iliac crest bone grafting. These patients went on to bony union. The other two uni-nonunion patients developed painless pseudoarthroses but were left alone due to acceptable function and co-morbidities. Another patient’s fracture displaced after a week of conservative management and required a manipulation in theatre but went on to union. Another patient developed a malunion but this was left alone as she was functionally acceptable. There were 2 injury related radial nerve palsies which both recovered within 4 months fully.

b) Intramedullary nails
Eleven patients had intramedullary nail fixation. Only one reported on-going shoulder discomfort post-operatively (antegrade nail). There were no infections reported and no non-unions. One case of injury-related radial nerve palsy was noted which recovered within 8 weeks of the nailing.  

c) Plate fixation
Eighteen patients had plate fixation of their humeral fracture. Four patients went on to non-union, three of whom had united at one site but not the other (one ununited proximally and two distally). Three were revised with further plating and bone grafting from the iliac crest. The first of these patients had a broken plate 8 weeks post initial operation. The second patient was an open fracture which was revised with an LC-DCP at 20 weeks with bone grafting. The third non-union went on to have a long-stem hemiarthroplasty inserted with a poor eventual outcome. The fourth patient had a painless non-union and was left alone. There were 3 radial nerve palsies. Two were injury related and one iatrogenic intraoperatively. One patient whose injury caused the palsy, fully recovered and in the other two only partial recovery was noted.

d) External fixation
One of the 5 patients in the external fixation group progressed to a malunion and was taken down and revised with a plate. All the other four externally fixated fractures (including one open fracture) went on to satisfactory union. There was one case of injury-related radial nerve palsy which resolved gradually over 1 month to full recovery.

e) Pins
Two patients were treated with Rush pins and both progressed to non-unions. One patient united distally but not proximally and developed a painless pseudoarthrosis and was left alone. The other  was revised with double recon plates and iliac crest autograft which then went on to union. 

Six patients had Enders nails inserted, 4 antegrade and 2 retrograde. Four went on to union. There were 2 cases of non-union (both antegrade) and one of these became infected. The infected case was later revised with a long stem hemiarthroplasty with eventual outcome as moderate. The other case of non-union (united proximally but not distally) had revision with a plate.  One injury related radial nerve palsy occurred in the antegrade approach group which recovered.

Discussion :

Our data suggests that patients who have fixation of their segmental humeral fracture using an intramedullary nail, have a greater chance of a good outcome than when treated by the other reported methods.

There is obscurity in the literature with regards to best treatment method for the segmental humeral fractures. Balfour et al16 and Sarmiento et al21 stated that when uncomplicated diaphyseal fractures of the humerus are treated conservatively by reduction and subsequent immobilisation of the arm, successful healing occurs in 95% to 98% of cases. However, these studies are collective retrospective studies for one level, low energy fractures. In these circumstances non-operative treatment may be a good option but in our study we are dealing with segmental fractures which are usually associated with high energy trauma where significant soft tissue disruption has occurred. The non-union rate for our non-operative group was 27.2% (6 cases). In many cases one site of the segmental fracture went on to union but the other site did not. This can be attributed to compromised biology or biomechanics of attempted simultaneous healing in two sites.

Closed locked nailing has been reported to yield satisfactory results for femoral and tibial segmental fractures13. Intramedullary nailing has the advantages of being relatively minimally invasive, offers axial and rotational stability (when locked) and good alignment. The literature suggests the union rate for intramedullary nailing of straight forward humeral shaft fractures is 95% with a low risk of problems such as mal-union, delayed or non-union, infection or radial nerve palsy9. In the herein study, locked nailing of segmental humeral fractures yielded a 100% union rate in a mean time of 18.5 weeks. Amjal1 reported a 30% non-union rate when 33 humeri with non-segmental but comminuted fractures were nailed using the antegrade Russell-Taylor nail. He reported no significant correlation between fracture comminution and healing time. All the nailings in this study were reamed. One study that compared nailing with and without reaming was unable to document differences in union rates or complications9. Recent interest has been directed towards the issue of axial loading on surgically stabilised humeral diaphyseal fractures, with controversy over the ideal method of fixation when it is necessary6. When compared to plate fixation, for example, it is generally well accepted that patients can load more confidently through a nail than a plate4. We had one case of shoulder pain post-operatively which was associated with one of the proximally inserted nails. There were no other differences between the antegrade and the retrograde nail group. We had no cases of rotator cuff malfunction nor adhesive capsulitis.

Plate fixation is an alternative method of treating segmental humeral fractures4,25. The literature reports plating of non-segmental humeral shaft fractures has a non-union rate of 3.6%14,10 but it traditionally requires an extensive open operation with a variable degree of stripping of the soft tissues from the bone. It also provides less secure fixation in osteoporotic bone or if crutch walking is required14. An argument for plating against nailing is the accurate fixation under direct vision and control of malalingment and malrotation. Plating is a very reasonable method of fixation for fractures that cannot be nailed, for example, ones that are comminuted close to the metaphysis (meaning that either the entry point or the distal locking is compromised) or fractures associated with radial nerve palsy which require exploration. The results of plating in this study showed a 23.5% non-union rate. As with the non-operative group, the relatively high rate of non-union could be explained by considering that fact that one of the fractures may not unite as can be the case in segmental fractures in all long bones. There was also a case of compartment syndrome post-operatively and one iatrogenic radial nerve palsy.

External fixation is well recognised as a quick and easy way to stabilise an extremity fracture whether open or closed. In our study the results showed a 100% union rate albeit one patient had a malunion. Care must be taken with pin sites although we had no cases of deep infection in our study.

Rush and Enders pins are used less frequently now as compared to other intramedullary nailing fixation techniques. Here 4 of the 8 patients treated with these methods developed a non-union, 2 due to infection. The reason behind the high rate of non union is possibly the poor stability that pins and nails provide to the fracture site and in particular in segmental fractures. The number of cases also is very small and a type II error could be present.

We are aware that this study has a number of limitations including the retrospective nature, the non-randomisation process, a small number of cases and the personal experience of four surgeons. Furthermore the decision to operate or not and the type of fixation was purely based on the surgeon’s decision. Despite all of the above, the number of patients treated (63) provides valid information compared to the current available literature with scarce numbers of segmental humeral fractures 15,6,17,25. Strengths of this study include its focus on this specific subgroup of patients (segmental fractures and not humeral fracture treatment in general), analysis of the different methods of fixation (including non-operative treatment), the multiple outcome measures analysed and the fact that there is an element of randomisation as supported by the surgeon’s preference in terms of management decision. The findings therefore in the herein study can provide valuable information when such rare cases are considered for stabilisation. 

In conclusion, this study noted consistently good results with intramedullary nailing. The other management options may also be considered but a higher rate of complications may be encountered.

Reference :

  1. Ajmal M, O’Sullivan M, McCabe J, Curtin W. Antegrade locked intramedullary nailing in humeral shaft fractures. Injury 2001;32:692–4.

  2. Baker S et al. The Injury Severity Score: a method for describing patients with multiple injuries and evaluating emergency care. J Trauma 1974;14:187-196.

  3. Balfour G, Mooney V, Ashby M. Diaphyseal fractures of the humerus treated with a ready-made fracture brace. J Bone Joint Surg [Am] 1982;64-A:11.

  4. Bell M, Beauchamp C, Kellam J, McMurtry R. The results of plating humeral shaft fractures in patients with multiple injuries. J Bone Joint Surg [Br] 1985;67-B No.2:293-296.

  5. Camden P, Nade S. Fracture bracing the humerus. Injury 1992;23: 245.

  6. Chen AL, Joseph TN, Wolinsky PR et al. Fixation stability of comminuted humeral shaft fractures: locked intramedullary nailing versus plate fixation. J Trauma, Injury, Infection and Critical Care 2002; 53(4):733-737.

  7. Cole P, Wijdicks. The Operative Treatment of Diaphyseal Humeral Shaft Fractures. Hand Clinics 2007; 23(4): 437-448. 

  8. Court-Brown CM, Garg A, McQueen MM. The translated two-part fracture of the proximal humerus. J Bone Joint Surg.2001;83-B(6):799-804.

  9. Crates J, Whittle AP. Antegrade interlocking nailing of acute humeral shaft fractures. Clin Orthop. 1998;(350):40-50.

  10. Gregory PR, Sanders R. Compression Plating versus intramedullary fixation of humeral shaft fractures. J Am Acad Orthop Surg. 1997;5:215-223.

  11. Gregory P, Sanders R. Compression plating versus intramedullary fixation of humeral shaft fractures. J Am Acad Orthop Surg. 1997;5: 215-223.

  12. Gustilo R, Anderson J. Prevention of infection in the treatment of one thousand and twenty-five open fractures of long bones: retrospective and prospective analyses. J Bone Joint Surg [Am]. 1976;58:453-458.

  13. Ikpeme J. Intramedullary interlocking nailing for humeral fractures: experiences with the Russell-Taylor humeral nail. Injury 1994;Sep(25):447-455.

  14. Ingman A, Waters D. Locked intramedullary nailing of humeral shaft fractures: implant design, surgical technique and clinical results. J Bone Joint Surg [Br]1994; 76-B:23-9.

  15. Lin J, Hou S. Locked nailing of severely comminuted or segmental humeral fractures. Clin Orthop and Rel Res 2002;406,195-204.

  16. McCormack R, Brien D, Buckley R et al. Fixation of fractures of the shaft of the humerus by dynamic compression plate or intramedullary nail: a prospective, randomised trail. J Bone Joint Surg [Br] 2000; 82,Iss 3; 336-340.

  17. Modabber M, Jupiter J. Operative management of diaphyseal fractures of the humerus: plate versus nail. Clin. Orthop 1998;347:93-104.

  18. R Development Core Team (2006). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0.

  19. Remiger A, Miclau T, Lindsey R, Blatter G. Segmental avascularity of the humeral diaphysis after reamed intramedullary nailing. Journal of Orthop Trauma 1997;Vol 11(4):308-314.

  20. Ruedi T, Murphy W et al. (2000). Principles of Fracture Management. Davos: AO Publishing & Stuttgart New York: Georg Thieme Verlag.

  21. Sarmiento A, Kinman P, Galvin E et al. Functional bracing of the humerus. J Bone Joint Surg [Am] 1977;59-A:596.

  22. Stannard J, Harris, McGwin G, Volgas D, Alonso J. Intramedullary nailing of humeral shaft fractures with a locking flexible nail. J Bone Joint Surg. 2003; 85: 2103-2110.

  23. Stata 9.2. Copyright 1984-2006. Statistics/Data Analysis Stata Corp, 4905 Lakeway Drive, College Station, Texas 77845. USA (800-STATA-PC) 979-696-4600,

  24. Webb J. Distal humeral factures in adults. J Am Acad Orthop Surg.1996;4:336-344.

  25. Yang K. Helical plate fixation for treatment of comminuted fractures of the proximal and middle one-third of the humerus. Injury 2005;36:75-80.

This is a peer reviewed paper 

Please cite as: Simon Fogerty: Segmental humeral fractures: prevalence of non-union and functional outcome.

J.Orthopaedics 2009;6(2)e6





Arthrocon 2011

Refresher Course in Hip Arthroplasty

13th March,  2011

At Malabar Palace,
Calicut, Kerala, India

Download Registration Form

For Details
Dr Anwar Marthya,
Ph:+91 9961303044



Powered by



© Copyright of articles belongs to the respective authors unless otherwise specified.Verbatim copying, redistribution and storage of this article permitted provided no restrictions are imposed on the access and a hyperlink to the original article in Journal of Orthopaedics maintained. All opinion stated are exclusively that of the author(s).
Journal of Orthopaedics upholds the policy of Open Access to Scientific literature.