Archives of Orthopaedic and Trauma Surgery

, Volume 133, Issue 5, pp 595–602

Non-bridging external fixation employing multiplanar K-wires versus volar locked plating for dorsally displaced fractures of the distal radius


  • Georg Gradl
    • Department of Trauma and Reconstructive SurgerySurgical Clinic University of Rostock
    • Hand and Upper Extremity Service, Department of Orthopaedic SurgeryMassachussetts General Hospital
    • Department of Trauma and Reconstructive SurgeryUniversity of Aachen
  • Martina Wendt
    • Department of Trauma and Reconstructive SurgerySurgical Clinic University of Rostock
  • Thomas Mittlmeier
    • Department of Trauma and Reconstructive SurgerySurgical Clinic University of Rostock
  • Guenther Kundt
    • Department of Medical Informatics and BiometryUniversity of Rostock
  • Jesse B. Jupiter
    • Hand and Upper Extremity Service, Department of Orthopaedic SurgeryMassachussetts General Hospital
Orthopaedic Surgery

DOI: 10.1007/s00402-013-1698-5

Cite this article as:
Gradl, G., Gradl, G., Wendt, M. et al. Arch Orthop Trauma Surg (2013) 133: 595. doi:10.1007/s00402-013-1698-5



The aim of this study was to compare non-bridging external fixation to palmar angular stable plating with respect to radiological outcome, wrist function, and quality of life.


One hundred and two consecutive patients (mean age: 63 years) were enrolled in the study. Fifty-two patients were randomized for plate osteosynthesis (2.4 mm, Synthes), 50 patients received non-bridging external fixation (AO small fixator). Objective (range of motion, grip strength), patient rated outcomes (quality of life, pain), and radiological outcome were assessed 8 weeks, 6 months, and 1 year after surgery.


Loss of radial length of more than 3 mm was not detected in any group. Volar tilt was better restored by external fixation (7.2°) than by volar plating (0.1°). Wrist function was good in both groups. The external fixator was tolerated very well, and the quality of life assessment revealed comparable results in both groups. Osteoporosis was found in 54 % of patients and had no influence on radiological and functional outcome.


Non-bridging external fixation employing multiplanar K-wires is a suitable treatment option in intra- and extra-articular fractures of the distal radius even in osteoporotic bone.

Level of evidence

Prospective randomized trial, Level I.


Distal radius fractureExternal fixationNon-bridgingFunctional outcomeVolar plating


Treatment of unstable distal radius fractures has evolved over the last two decades from a primary non-operative treatment to operative treatment [1].

The goal of treatment for distal radius fractures is to obtain sufficient pain-free motion, allowing return to activities while minimizing the risk for future degenerative changes or disability [2].

At the same time, demographic changes with an aging population and an increase of fractures with concomitant osteoporosis [3] imply a need for more predictable operative techniques and stable implants [3]. Open reduction and internal fixation (ORIF) of dorsally displaced radial fractures employing locking plates through a volar approach has been advocated to be less invasive than dorsal plating [4, 5] and at the same time provide reasonable stability in osteoporotic bone [6], non-bridging external fixation on the other hand, offers a potential benefit in accelerating rehabilitation by allowing primary wrist mobilization and a minimally invasive operative procedure [7].

The main factor that has so far precluded its use is lack of space in the distal fragment, either because it is too small (<5 mm of volar cortex) or because in severe articular fractures the space is occupied by fixation for the joint surface (Hayes et al. [19]). We recently introduced a non-bridging external fixation technique, employing multiplanar K-wires, suitable for both extra- and intra-articular fractures [8]. In a biomechanical study, our group found higher fixation stability for this non-bridging technique in an intra-articular fracture model when compared to volar fixed-angle plating [9].

However, it remains unclear as to whether this treatment method will yield comparably good results in osteopenic bone.

We therefore evaluated differences in non-bridging external fixation and volar fixed-angle plate fixation with regard to functional and radiological otucome in patients with and without underlying osteopenia in a prospective randomized trial.

Materials and methods


Between January 2005 and May 2006 we prospectively evaluated 102 patients (89 women, 13 men; average age 63 years, range 18–88) with displaced fractures of the distal radius. The fractures were classified according to the AO classification [10]. Inclusion criteria were dorsally displaced (>20°) extra-articular A3 and intra-articular C1–C3 fractures. Patients that sustained a dorsal or volar sharing fracture, type B fracture or patients that presented with a history of previous wrist trauma were excluded from the study. Patients were randomly assigned to the two treatment groups by using established methods of complete randomization [11]. Group I (50 patients) was treated by non-bridging external fixation (AO small Fixator; Mathys Medical, Bettlach, Switzerland), Group II (52 patients) was treated by ORIF using a volar fixed angle plate (2.4 mm Synthes®, Mathys Medical, Bettlach, Switzerland). Institutional Review Board approval was granted before initiation of this study, and strict confidentiality guidelines were followed.

Surgical technique

Surgery was performed under general or brachial plexus anesthesia. The surgical technique of non-bridging external fixation was carried out as previously reported by our group [8]. Two Schanz screws (4.0 mm) were inserted in the distal radial shaft and one Schanz screw in the second metacarpal bone for a preliminary joint bridging construction. This was done to regain and maintain radial length. In a second step, three threaded K-wires (1.8 mm) were driven into the distal fragment in a semicircular manner and tensioned to a curved fixator bar (modified Ilizarov hybrid fixation technique). Prior to fixation, the K-wires served as joy-sticks to achieve precise reduction of the articular surface and restore volar tilt. The elements, bridging the radiocarpal joint, were removed once reduction had been achieved (Fig. 1).
Fig. 1

Clinical picture of a 69-year-old female 1 week after fixation of a distal radius fracture with the non-bridging external fixator (Small fixator, Synthes, USA)

Figure 2a, b give an example of a C2 distal radius fracture, treated by the novel non-bridging external fixation technique. Volar plate fixation (Fig. 3) was performed through a standard Henry approach [12]. Volar splints were applied in both groups for not longer than 3 days to allow for early functional after treatment. The external fixator was removed as an outpatient procedure 7 weeks after surgery. Implant removal after plate osteosynthesis was not encouraged, though carried out in 13 patients (25 %) due to symptomatic hardware.
Fig. 2

a 59-year-old female with an intraarticular distal radius fracture (AO type C2.2). bUpper panel 1 day after surgery with the AO small fixator (Synthes, USA) lower panel 1 year after surgery
Fig. 3

55-year-old female with an extraarticular distal radius fracture (AO type A3.3). Left panel at the time of injury; right panel 1 year after volar locking plate fixation (2.4 mm, Synthes, USA)

Follow up assessment

Follow up examinations at 8 weeks, 6 months, and 1 year after surgery included assessment of active range of wrist motion (extension, volar flexion, pronation, supination, radial deviation, ulnar deviation) using a goniometer. Grip power was measured with a dynamometer (Jamar; JA Preston Corp., Jackson, MI) at position 3, and an average of three trials. Values were expressed as a percentage of the uninjured side. Wrist pain was evaluated using the visual analog scale (VAS) (VAS 0 = no pain, VAS 10 = severe pain) at rest and during activity. Quality of life was measured with the Short Form 36 (SF-36) questionnaire 8 weeks after surgery [13]. All patients were evaluated by an independent investigator not involved in the patients’ treatment.

Standard posterior–anterior and lateral radiographs were taken 1 day, 8 weeks, and 12 months after surgery. All radiographs were digitally recorded and analyzed with respect to ulnar inclination, volar tilt (number of degrees from the neutral position), radial length, radioulnar congruity, and congruity of the articular surface. The measurements were performed relative to a central axis that was constructed on the lateral and posteroanterior films. Radial length was assessed by measuring the distance between the ulnar border of the distal radius and the distal articular surface of the ulna [14] all radiographic data were compared with the non-involved side. Subjective and objective data of wrist function and pain were summarized in the Gartland and Werley Score [15] and the Castaing Score [16], which comprises wrist function and radiographic data.

Bone density was determined using peripheral Quantitative Computed tomography (pQCT; Stratec XCT 900) measurements of the distal radius of the uninjured side at the time of hospital stay. A trabecular density between 120 mg HA/cm3 and 80 HA/cm3 was considered as ostopenic bone stock, <80 mg HA/cm3 was considered as osteoporotic bone loss [17].

Statistical analysis

All data were stored and analyzed using the SPSS statistical package 14.0 (SPSS Inc. Chicago, IL). Descriptive statistics were computed for variables of interest. The statistics computed included mean and standard errors of continuous variables, frequencies, and relative frequencies of categorical factors.

After testing for normality, either non-parametric paired Wilcoxon signed-rank-test or unpaired t test was chosen to assess differences between groups concerning the investigated continuous variables.

Testing for differences of categorical factors was realized by using Fisher’s exact test or χ2 test. Significance level of statistical tests was set at p ≤ 0.05.

A power analyses indicated that a sample of 90 patients total would provide 80 % statistical power, with α = 0.05, for a medium effect size d = 0.6.


One-hundred and two patients were enrolled in the study. There was no significant difference between groups regarding age or sex. Fractures were classified according to the AO comprehensive fracture classification and were equally distributed between groups (Table 1). Follow-up data could be obtained for 98 patients (96 %) after 8 weeks, for 97 patients (95 %) after 6 months, and for 92 patients (90 %) after 1 year. Patients were lost to follow-up because of death (n = 2), moving houses (n = 3), and non-respondence (n = 5).
Table 1

Fracture classification


Group I (n patients)

Group II (n patients)

AO Classification















The mean time elapsed between injury and surgery was 7.3 ± 0.3 days.

There was a significant difference (p < 0.05) between groups for surgery time and intraoperative fluoroscopic time. Mean surgery time in the external fixation group was 43.1 ± 1.5 min and mean intraoperative fluoroscopic time was 2.2 ± 0.2 versus 58.5 ± 2.9 and 1.4 ± 0.2 min, respectively, in the plate fixation group.

Clinical examination

Results of wrist function, pain and grip strength at 8 weeks, 6 months, and 1 year after surgery are presented in Table 2.
Table 2

Range of motion, grip strength, and pain at 8 weeks, 6 months, and 1 year


8 weeks

6 months

1 year

Group I

Group II


Group I

Group II


Group I

Group II



59.5 ± 3.6

66.5 ± 2.8


89.4 ± 1.7

86.3 ± 2.0


94.5 ± 1.3

92.2 ± 1.9


Volar flexiona

71.4 ± 2.9

65.9 ± 3.2


86.6 ± 2.3

79.5 ± 2.6


91.1 ± 1.9

86.6 ± 2.3



98.1 ± 0.7

98.2 ± 0.8


100 ± 0.0

99.8 ± 0.2


100 ± 0.0

100 ± 0.0



80.9 ± 3.9

94.7 ± 1.5


98.1 ± 0.8

98.6 ± 0.6


98.5 ± 0.7

99.0 ± 0.6


Radial deviationa

68.1 ± 4.1

77.9 ± 4.0


91.2 ± 1.9

92.9 ± 2.1


97.6 ± 1.3

93.9 ± 2.1


Ulnar deviationa

74.7 ± 3.2

76.3 ± 2.9


88.1 ± 2.4

85.1 ± 2.6


93.1 ± 2.0

92.3 ± 2.4



0.03 ± 0.03

0.3 ± 0.1


0.3 ± 0.2

0.2 ± 0.2


0.1 ± 0.1

0.0 ± 0.0


Grip strengtha

26.8 ± 3.4

50.6 ± 4.3


72.2 ± 3.8

80.2 ± 2.9


86.8 ± 2.8

84.1 ± 3.3


* Values are statiscally significant

aValues are reported as mean ± SEM in percentage of the uninjured side

At the 8-week follow-up examination, patients after non-bridging external fixation presented with significantly less pain and greater volar flexion (n.s.).

At the 6-month follow-up evaluation, patients in the ex fix group showed significantly greater volar flexion than in the plate group.

At the 1-year follow-up, no significant difference was found between treatment groups with respect to any of the assessed parameters.

One year after surgery, all patients treated with external fixation and 97.8 % of patients with volar plating had an excellent or good result in the Gartland and Werley Score [fix ex: excellent: n = 36 (81.8 %), good: n = 8 (18.2 %); volar plating: excellent: n = 37 (82.2), good: n = 7 (15.6 %; P = 0.568)]. The Castaing Score showed a perfect or good result in 93 % after external fixation and 95.6 % after volar plating [fix ex: excellent: n = 13 (30.2 %), good: n = 27 (62.8 %); volar plating: excellent: n = 11 (24.4 %), good: n = 32 (71.1 %; p = 0.689)] (Table 3).
Table 3

Castaing Score and Gartland and Werley Score at 1 year


Group I

Group II


Castaing Scorea


1.65 ± 0.28

1.71 ± 0.3


 Osteoporotic patients

1.37 ± 0.28

2.08 ± 0.9


Gartland and Werley Scoreb


1.18 ± 0.3

1.4 ± 0.35


 Osteoporotic patients

0.47 ± 0.25

2.42 ± 1.06


Values reported as mean ± SEM

a0 perfect; 1–5 good; 6–11 satisfactory; 12–15 fair; 16–25 poor; >25 very poor

b0–2 excellent; 3–8 good; 9–20 fair, >20 poor

The results of the SF-36 health survey 8 weeks after surgery are summarized in Fig. 4.
Fig. 4

Results of the SF-36 Health Survey 8 weeks after surgery

Apart from “social function”, which was significantly better in the plating group, patients tolerated the external fixator very well.

For all other parameters (bodily pain, physical function, mental health, physical role, general health, vitality, emotional role), no significant difference was found between groups.

Osteopenic or osteoporotic bone loss had no impact on functional parameters (multivariate analysis of variances, data not shown).

Radiological outcome

Radiographic outcome at 1 year is summarized in Table 4.
Table 4

Radiological outcome at 1 year


Group I

Group II

Radial shortening (mm)

0.5 ± 0.1

0.24 ± 0.1

≥2 mm: n = 3

≥2 mm: n = 2

≥3 mm: n = 2

≥3 mm: n = 2

Volar tilt (°)

7.2 ± 0.7

0.1 ± 0.5

Residual dorsal tilt (°)

≥5°: n = 0

≥5°: n = 6

≥10°: n = 0

≥10°: n = 2

Residual articular step off

n = 0

n = 1

Values are reported as mean ± SEM

All fractures united and there was no significant difference between groups with respect to time to union. Radial shortening of more than two mm was seen in four cases (ex fix: n = 2; volar plating n = 2), moderate shortening of two mm was registered in five cases (ex fix: n = 3; volar plating: n = 2) Restoration of volar tilt was achieved in all cases in the ex fix group (7.2 ± 0.7°) and in no case in the volar plating group (0.1 ± 0.5°). Six cases in the plating group featured a residual dorsal tilt of 5°, and two cases one of 10°. No residual articular step-off was observed in the ex fix group, compared to one case in the plating group (articular step-off 1 mm).

There was no significant difference between groups with respect to bone quality (p = 0.15).

Peripheral Quantitative Computed tomography measurements revealed a moderate reduction in bone quality in 21 patients (20.6 %) and a manifest osteoporosis in 34 patients (33.3 %).

There was no correlation between bone quality and radiological outcome (data not shown).


Five patients (10 %) treated with external fixation developed reddening around the proximal pin sites (Schanz screws) that resolved by local pin care and required no further treatment.

K-wire sites in the distal fracture fragments were not affected. There was no case of pin loosening, even in osteoporotic bone.

Four patients developed complex regional pain syndrome type 1 (ex fix: n = 2; volar plating: n = 2) that improved within 3 months of physical therapy and pain medication. Prolonged signs of swelling, inflammation and occasional pain were registered in an additional three cases with external fixation, but failed to qualify for the diagnosis of CRPS I.

Tendon complications were seen in six patients. Two patients treated by non-bridging external fixation had a rupture of the extensor pollicis longus (EPL) tendon; 1, 6 weeks after surgery and one, with an 11 months delay after surgery. The latter, was on high dose steroid medication for autoimmune vasculitis and had previously experienced spontaneous ruptures of the Achilles tendon. Both patients received reconstruction with an extensor-indicis proprius graft.

In the volar plating group, one rupture of the EPL tendon 4 months after surgery was found and reconstruction with an extensor-indicis proprius graft was performed. Symptomatic tendinitis of the EPL tendon due to dorsal protrusion of screw tips occurred in two cases and adhesion of the flexor pollicis tendon to the plate occured in one case. Both patients received implant removal after bony union was achieved.

Carpal tunnel syndrome (CTS) developed in four patients after volar plating (7.7 %). Implant removal, carpal tunnel release and neurolysis of the median nerve which was surrounded by scar tissue were performed in three cases. In one case, CTS resolved with temporary volar splinting. One patient in the external fixation group developed CTS (2 %) and received operative treatment.

The overall re-operation rate was 36.5 % (19 patients) following plate osteosynthesis compared to 6 % (three patients) following external fixation.


This prospective randomized trial evaluates the quality of fracture care using a novel non-bridging external fixator for the treatment of distal radial fractures in comparison to volar fixed-angle plating. There are several reports on non-bridging external fixation techniques for the treatment of distal radial fractures [1821]. The techniques described by McQueen et al. [22] Fischer et al. [23] and Gausepohl et al. [21] require a distal fragment that is broad enough to accommodate two Schanz screws with a diameter of 3.5 mm or more. Non-bridging external fixation was therefore considered difficult in displaced articular fractures and fractures with a small distal fragment, raising the question, whether these fracture characteristics represent a contraindication for its use [2123]. In this study, however, we could demonstrate that non-bridging external fixation led to fracture union even in displaced intra-articular fractures and was associated with only moderate fracture settling with subsequent loss of radial height after pin removal. Similar results have been reported by Atroshi et al. [24] and Hayes et al. [19] more than two-thirds of patients in our study population suffered from osteopenic bone loss or manifest osteoprosis. Although pathologic bone stock has been considered to be associated with a higher incidence of implant failure and secondary loss of reduction especially with external fixation [7, 22] there was no correlation between poor bone quality and maintenance of radial length and no case of implant failure in both treatment groups in our study. Multidirectional placement of multiple load carriers in the subchondral layer may have led to a broader load distribution along the fracture line, thus enhancing fixation stability in osteopenic bone stock.

Immediate functional treatment resulted in restoration of wrist function of up to 98 % of the uninjured side in both groups and did not result in loss of reduction.

Volar flexion was regained earlier in the external fixation group and remained substantially greater over time. This may be attributed to the meticulous restoration of volar tilt.

Radial deviation (62.7 % of the uninjured side) was significantly reduced in the external fixation group with the implant still in place, due to the prominent K-wires through the radial styloid but improved after implant removal.

Despite the fact that one might expect higher levels of pain in patients wearing an external fixator with the extensor tendons sliding along the distal pins, volar plating was associated with significantly more pain in the first 8 weeks after surgery. This may be related to the greater approach related morbidity associated with volar plating [25].

Differences in range of motion, grip strength and pain between groups became less pronounced over time and both groups showed almost full wrist function at the final follow up and reported no pain.

With good or excellent results in most cases according to the Gartland and Werley as well as to the Castaing Score our data compares favorably to other studies of non-bridging external fixation or volar plating [5, 2628]. The SF-36 health survey revealed good patient tolerance of the external fixator.

Patients in the volar plating group scored only higher at the “social function” parameter of the SF-36 1 week after surgery, suggesting that patients’ activities of daily living are not restricted to a higher extent by an external fixator.

The overall complication rate was similar in both groups (20 % in Group I versus 21 % in Group II) whereas the overall reoperation rate after plate fixation was much higher than for external fixation. Patients reported persistent discomfort at the wrist without underlying radiographic abnormalities and requested implant removal in 21 % of cases. Tendon irritation was encountered not only because of direct contact with the volar plate but also due to dorsal protrusion of screw tips. Extensor tendon complications such as tendonitis, attrition, and rupture have been described for both dorsal and volar plating as well as non-operative treatment [26, 29, 30] rupture of the EPL tendon following volar plating is reported to be as high as 9 % [26, 31]. Similar to our own results, Mckenney et al. [32] and Hayes et al. [19] reported a rate of 3 and 1 %, respectively, following external fixation.

Compared to the pin-site infection rate of up to 43 % of cases reported in the literature, our rate of pin site cellulitis was relatively low (10 %) and no further treatment was required [18, 33, 34].

Our study protocol involved regular visits at our outpatient facility that included thorough wound checks and pin care which might have reduced the rate of pin infections.

There are several limitations to our study. The volar plates used in this study feature a single distal screw row and fixed-angle locking. Currently, available volar locking plates that incorporate at least two distal screw rows and have a variable-angle locking mechanism might yield better results and might prevent secondary loss of reduction. A recent biomechanical study, however, reported no significant differences between single-row and double-row plates with regard to axial deformation, implant rigidity, or maximum displacement [35]. Although not significant, the rate of intraarticular fractures was slightly higher in the plating group (44 vs. 34 %). We tried to account for these differences by randomizing patients to receive either treatment.

In conclusion, non-bridging external fixation yields comparable good results with respect to functional outcome, fracture reduction and maintenance of reduction as volar angular stable plating in the treatment of distal radial fractures. Based on our results, we believe that non-bridging external fixation represents an alternative treatment option for displaced fractures of the distal radius. Osteopenic bone loss and articular involvement do not present contraindications for non-bridging external fixation.


The study was supported by a grant from the AO.

Conflict of interest

The authors declare that they have no conflict of interest.

Copyright information

© Springer-Verlag Berlin Heidelberg 2013