547
7. Stable temporary traction substitute wiih the Pinless external fixator
R. Babst, P. Regazzoni,
N.Renner,
R. Rosso, M.Heberer
Department of Surgery University of Basle CH-2000 Basle
7.1Mroduction and historyof skeletaltraction
caneal pin traction by a joint bridging pinless fixator (JBPF) inserted under local anaesthesia (Fig. 1). Producing the same effect as ligamentotaxis, this should offer more rotational stabiity and, in particular, should allow active mobilisation of the knee joint during the period before definitive internal fiation. Patient care is also expected to be easier and comfort improved.
The timing of open reduction and internal fixation (ORlF) of distal tibia1 and malleolar fractures essentially depends on the condition of the soft tissues (Muller, 1991). If in doubt, it is safer to postpone internal fixation for several days to avoid wound healing complications in these situations. Skeletal traction is indicated in unstable fractures to align the fracture and to avoid additional lesions by displaced bone fragments and to maintain length, thus making delayed fixation easier. The current concept of skeletal traction using nails and wires under tension dates back to the early years of this century: In 1903 Codivilla introduced transcalcaneal traction with a nail, whereas Steinmann contributed the idea of an additional half-ring. Klapp suggested the use of wires in 1912. Beck (1924) and Kirschner (1927) used halfrings to put the wrres under tension and increase their stability in order to decrease soft tissue irritation around the wires. Using nails, Lorenz Bijhler introduced revolving half-rings in 1916 to avoid relative movements of the nail in the bone (BBhler, 1943). Since these times the technique of skeletal traction has not changed fundamentally. Transcalcaneal Steinmann pins, introduced in local anaesthesia (L.A.), and half-rings are commonly used in those cases of malleolar and pilon fractures in which early ORIF is not possible for soft tissue reasons. The disadvantage, however, consists in rotational instability resulting in pain during changes of position of the limb, difficult patient care and the need for frequent readjustments. The patient is completely bedridden. This disadvantage can be overcome by the temporary placement of a joint bridging external fixator. This device, however, normally requires local anaesthesia for its placement. Pin track complications are a further drawback The pinless external fixator, introduced tice for open tibia1 fractures, suggested traction substitute because of its pinless this feasibility study was to replace the
into clinical pracitself for use as a frame. The aim of conventional cal-
Fig. 1: Model of a joint bridging pinless fiiator (JBPF) with a large asymmetrical clamp on the tibii and a large symmetrical clamp on the calcaneus.
7.2studydesip 10 patients with malleolar, pilon or distal tibia1 fractures treated by a JBPF at the Department of Surgery, University Hospital of Basle, from January to April 1992, were consecutively followed. Patients were included in the study if they had 1.
an acute fracture dislocation of the ankle, a pilon fracture or a fracture of the distal tibia that did not allow immediate ORlF for soft tissue reasons or because of concomitant injuries,
2.
the JBPF secondary
was used as a temporary internal fixation.
device
before
The post-operative protocol consisted of standard X-ray views (Fig. 6), of daily reviews of the stability of the frame, the clamps and the skin entrance points. Swab
samples were taken if local infection was suspected. Pain medication was given as a non-steroidal analgesic on a fixed schedule (three times 50 mg of Diclofenac). Opioid drugs were given only on demand. Data were recorded during placement of the frame, during daily interviews and by reviewing the charts. On a numeric analogue scale (Melzack, 1975) (0 points = no pain, 10 points = intolerable pain), pain perception was recorded for the placement of the device including the reduction manoeuvre, and also for the maximal pain sustained whilst the JBPF was in place. The time of additional opioid drug administration and all complications were recorded. After definitive treatment the patients were followed until discharge from the hospital. All patients were re-evaluated six weeks after the operation by physical and radiological examination.
73 Patients Six fracture dislocations of the ankle, three pilon and one second degree open fracture of the tibia were treated. There were five men and five women The age of the patients ranged from 53 to 88 (mean 61.3). Eight fractures involved the right, and two the left lower limb. Seven fractures occurred by stumbling on the floor or on a stair, two by falling 3 m or 1.5 m from a roof, and one in a pedestrian/automobile accident. Nine fractures were closed, one was a second degree open fracture (according to Gustilo’s classification). In this patient concomitant head and thomcic injuries did not permit immediate wound toilet.
Fig. 2: Elements of the JBPF: 4 rods (3 x 2&Y mm, 1 x 300 400 mm), 3 universal clamps, 3 tube-to-tube clamps, 1 symmetrical and 1 asymmetrical clamp. The calcaneal clamp was connected to a transverse carbon fibre rod of 200 mm. The traction and reduction manoeuvre was then Performed by an assistant using this rod as a handle. While the assistant maintained the limb in correct alignment this rod was connected in a triangular fashion to the proximal tibia1 clamp (Fig. 3, using three tube-to-tube clamps and three further carbon fibre rods (two times 200 mm, one time 300400 mm). In trimalleolar fractures with a large Volkmann’s fragment, the frame was suspended on a half-ring by a cord to prevent dorsal subluxation (Figs 4 and 5).
7.4 Technical aqxctm The patients were placed in normal beds in the supine position with the injured leg elevated on a frame and the knee flexed at 45O. For the JBPF, we used two types of large pinless clamps (Fig. 2). The asymmetrical clamp was placed 5 or 10 cm from the most proximal fracture line, not to interfere with the future skin incision on the tibia1 diaphysis. The symmetrical clamp was placed in the same position as a Steinmann pin (ca. 4 cm distal and 4 cm dorsal of the lateral malleolus) on the calcaneus. Nine JBPFs were mounted under local anaesthesia in the emergency room after a single dose of prophylactic antibiotic using third generation cephalosporin. One fixator was mounted in the intensive care unit (ICU) on an intubated polytraumatised patient.
Fig. 3: JBPF after application under local anaesthesia on a 45O padded splint (case 8).
s 49
Babst: Tractionsubstitute
Table 1: Pain rating. GSA?
Sex/Age (yea=)
Fracture
F, 83 M, 43
A0 type
Mechanism of injury
Complications
Trimall. dislocation
B 23
shrmbling
slight pes equinus
Pilon
c 3.3
fall 3 m
F, 58
2nd degree open tibia
A 3.2
pedestrian/
F, 61
Trimall. dislocation
B 3.3
stumbling
loosening tibia1 clamp
L.A.
M, 52
Trimall. dislocation
B 3.1
stumbling
erysipeloid, fever
LA.
F, 88 M, 63
Bimall. dislocation
B 2.3
stumbling
Pilon
c 1.1
stumbling (stairs)
pressure ulcer
L.A.
M, 45
Pilon
B 2.2
fall 1.5 m
loosening tibia1 clamp
L.A.
M, 52
Trimall. dislocation
B 3.3
stumbling
L.A.
F, 70
Bimall. dislocation
B 2.2
shrmbling
L.A.
automobile act.
Time of opioid application (days)
Pain during JBPF application (points)
DWatiO” Pain with JBPF of JBPF in place (days) (wink)
2
2
12
LA.
2
0
1
16
GA.*
G.A.*
Anaesthesis
L.A.
G.A.’
2
10
3
2
10
6
4
12
4
2
5
2
6
3
13
3
4
2
6
1
0
3
8
0
0
2
12
5
LA.
slight pes equinus
Fig. 4: Trimalleolar dislocation fracture of the ankle (case 9). Fig. 6: Case 9 standard X-my views after reduction and fixation with the JBPF. In all patients, definitive fracture treatment consisted of plate fixation after swelling had subsided. Before clamp removal and skin incision a single dose of third generation cephalosporin was given i.v. For the operation, the skin entrance points for the clamps were covered with sterile adhesive. 7.!S
Fig. 5: Case 9 on the 45O padded splint. The JBPF suspended by a cord to prevent dorsal subluxation.
Resuh
The average pain suffered during clamp placement, reduction manoeuvre and assembly of the fixator was rated at 2.8 points (range: min 0, max 6 points) by the patient (Table 1). The maximum pain during the time the JBPF was in place was rated at 2.2 points (range: rain 1, max 4 points). The need for opioid drugs averaged one a day (range: min 0, max 5 days). On average, the JBPF was left in place for 10 days (range: min 5, max 16 days). In two cases, the tibia1 clamp became loose on the 5th and the 6th day respectively. Two patients developed a slight equinus of the foot (case 1, 10). In five patients, skin entrance points had serous discharge on the first day, without clinical evidence of infection. In one case, a skin ulcer developed on the lateral calcaneal entrance point due to inadequate clamp placement (case 7). An erysipeloid skin redness with fever was observed 24 hours
S50
after placement of the JBPF in another patient (case 5) who had sustained a trimalleolar fracture with a bruise and skin necrosis on the medial malleolus. A swab taken from the tibia1 entrance point showed Clostridium perfringens after culture on broth. All fractures were adequately distracted and reduced. The definitive treatment was in no case limited by the previous placement of a JBPF. Eight patients had uneventful healing of their incisional wound with no signs of infection. After definitive internal fixation of two pilon fractures, skin necrosis developed at the incision site. In one case (case 3), skin necrosis was superficial with no need of further treatment. In the other case (case 2), the anterior cloverleaf plate had to be covered by a free rectus abdominis flap. Both complications were unrelated to the application of the fixator.
7.6 Ditimmion The aim of this pilot study was to analyse the feasibility of a stable pin traction substitute for fractures around the ankle that would not interfere with subsequent internal fixation. Conventional skeletal traction with a Steinmann pin is a well established and simple method which is usually performed under local anaesthesia in the emergency room. One of the prerequisites for a traction substitute is that it could be applied under the same conditions equally well. Our data show that this holds true for the JBPF. During the placement of the JBPF, maximal pain was produced by the reduction manoeuvre as reported by the nine patients undergoing this procedure under local anaesthesia. Only three patients felt slight pain (2 points) at the moment of clamp application All non-intubated patients were able to lift up their injured leg at the end of the procedure without further pain While the fiiator was in place, pain was reported at the beginning or at moments of inadvertent sudden rotational movement of the foot. The short period of additional opioid administration correlates well with measurement of subjective pain perception. For mechanical reasons, a two point clamp fixator can provide only limited rotational stability. Therefore, it is clear that minor readjustments of the construction need to be made in some of the cases. Due to sag&al instability, trimalleolar fracture dislocation with a large Volkmann’s fragment have a tendency to displace dorsally. A suspension cord attached to the calcaneus (Fig. 3) usually solves this problem. This displacement could also be prevented by the fitting of a second tibia1 clamp more proximally. It is obvious that conventional external fiiator frames with Schanz screws do provide better stability, but we think this is usually not required for traction purposes and therefore does not justify the additional risk of local anaesthesia and possible pin track complications. Compared to conventional skeletal traction the JBPF reduces pain, provides better stability and therefore improves mobility, increases patient comfort and simplifies both patient care and physiotherapy.
The absence of pain allowed us to wait for optimal soft tissue conditions. This explains the rather long interval between placement of the device (average 10 days) and final ONF. The seemingly high incidence of clamp loosening might have been prevented by a more medial placement of the dorsal arm of the tibia1 clamp. Ideally the dorsal arm should be placed as far medial on the dorsal aspect of the tibia as possible. This is in some cases limited by the design of the clamp, leading to undue pressure on the soft tissue. This might be improved by changes of clamp design. One clamp 1oosened in an elderly confused lady in the early stages of Altzheimer disease (case 4). This was easily corrected under local anaesthesia. The other was due to an uncontrolled movement during sleep (case 8). In this case, the frame was converted to a simple traction device through the calcaneal clamp. The equinus malreduction was due to positioning the clamps too anteriorly and could easily have been prevented by proper placement The Clostridium perfringens infection, discovered on the swab culture, resolved after appropriate antibiotic treatment without sequelae. In summary, we feel that the JBPF can replace transcalcaneal traction offering good patient comfort and easy care. Further studies, however, are necessary to evaluate potential advantages as compared to skeletal traction and conventional external fixators for this indication. 7.7 summaly The stable traction substitute with a JBPF seems a very promising indication for pinless clamps, offering good patient comfort and easy care. The implantation of a JBPF is possible under the same conditions as for conventional calcaneal traction. Although the stability provided is less than for conventional external fixators, it is sufficient for a temporary traction device. The JBPF does not affect secondary ORIF. 7.8
References
Bohler L (1943) Die Technik der Knochenbruchbehandlung. 9-11 edition, Verlag Wilhelm Maudrich, Wien Melzack R (1975) The McGill Pain Questionnairez Major properties and scoring methods. Pain 127 Miiller ME, Al&wet M, Schneider R, Willenegger H (1991) Manual of Internal Fixation, 3rd edition, Springer Verlag
7. Stable temporary traction substitute wiih the Pinless external fixator
R. Babst, P. Regazzoni,
N.Renner,
R. Rosso, M.Heberer
Department of Surgery University of Basle CH-2000 Basle
7.1Mroduction and historyof skeletaltraction
caneal pin traction by a joint bridging pinless fixator (JBPF) inserted under local anaesthesia (Fig. 1). Producing the same effect as ligamentotaxis, this should offer more rotational stabiity and, in particular, should allow active mobilisation of the knee joint during the period before definitive internal fiation. Patient care is also expected to be easier and comfort improved.
The timing of open reduction and internal fixation (ORlF) of distal tibia1 and malleolar fractures essentially depends on the condition of the soft tissues (Muller, 1991). If in doubt, it is safer to postpone internal fixation for several days to avoid wound healing complications in these situations. Skeletal traction is indicated in unstable fractures to align the fracture and to avoid additional lesions by displaced bone fragments and to maintain length, thus making delayed fixation easier. The current concept of skeletal traction using nails and wires under tension dates back to the early years of this century: In 1903 Codivilla introduced transcalcaneal traction with a nail, whereas Steinmann contributed the idea of an additional half-ring. Klapp suggested the use of wires in 1912. Beck (1924) and Kirschner (1927) used halfrings to put the wrres under tension and increase their stability in order to decrease soft tissue irritation around the wires. Using nails, Lorenz Bijhler introduced revolving half-rings in 1916 to avoid relative movements of the nail in the bone (BBhler, 1943). Since these times the technique of skeletal traction has not changed fundamentally. Transcalcaneal Steinmann pins, introduced in local anaesthesia (L.A.), and half-rings are commonly used in those cases of malleolar and pilon fractures in which early ORIF is not possible for soft tissue reasons. The disadvantage, however, consists in rotational instability resulting in pain during changes of position of the limb, difficult patient care and the need for frequent readjustments. The patient is completely bedridden. This disadvantage can be overcome by the temporary placement of a joint bridging external fixator. This device, however, normally requires local anaesthesia for its placement. Pin track complications are a further drawback The pinless external fixator, introduced tice for open tibia1 fractures, suggested traction substitute because of its pinless this feasibility study was to replace the
into clinical pracitself for use as a frame. The aim of conventional cal-
Fig. 1: Model of a joint bridging pinless fiiator (JBPF) with a large asymmetrical clamp on the tibii and a large symmetrical clamp on the calcaneus.
7.2studydesip 10 patients with malleolar, pilon or distal tibia1 fractures treated by a JBPF at the Department of Surgery, University Hospital of Basle, from January to April 1992, were consecutively followed. Patients were included in the study if they had 1.
an acute fracture dislocation of the ankle, a pilon fracture or a fracture of the distal tibia that did not allow immediate ORlF for soft tissue reasons or because of concomitant injuries,
2.
the JBPF secondary
was used as a temporary internal fixation.
device
before
The post-operative protocol consisted of standard X-ray views (Fig. 6), of daily reviews of the stability of the frame, the clamps and the skin entrance points. Swab
samples were taken if local infection was suspected. Pain medication was given as a non-steroidal analgesic on a fixed schedule (three times 50 mg of Diclofenac). Opioid drugs were given only on demand. Data were recorded during placement of the frame, during daily interviews and by reviewing the charts. On a numeric analogue scale (Melzack, 1975) (0 points = no pain, 10 points = intolerable pain), pain perception was recorded for the placement of the device including the reduction manoeuvre, and also for the maximal pain sustained whilst the JBPF was in place. The time of additional opioid drug administration and all complications were recorded. After definitive treatment the patients were followed until discharge from the hospital. All patients were re-evaluated six weeks after the operation by physical and radiological examination.
73 Patients Six fracture dislocations of the ankle, three pilon and one second degree open fracture of the tibia were treated. There were five men and five women The age of the patients ranged from 53 to 88 (mean 61.3). Eight fractures involved the right, and two the left lower limb. Seven fractures occurred by stumbling on the floor or on a stair, two by falling 3 m or 1.5 m from a roof, and one in a pedestrian/automobile accident. Nine fractures were closed, one was a second degree open fracture (according to Gustilo’s classification). In this patient concomitant head and thomcic injuries did not permit immediate wound toilet.
Fig. 2: Elements of the JBPF: 4 rods (3 x 2&Y mm, 1 x 300 400 mm), 3 universal clamps, 3 tube-to-tube clamps, 1 symmetrical and 1 asymmetrical clamp. The calcaneal clamp was connected to a transverse carbon fibre rod of 200 mm. The traction and reduction manoeuvre was then Performed by an assistant using this rod as a handle. While the assistant maintained the limb in correct alignment this rod was connected in a triangular fashion to the proximal tibia1 clamp (Fig. 3, using three tube-to-tube clamps and three further carbon fibre rods (two times 200 mm, one time 300400 mm). In trimalleolar fractures with a large Volkmann’s fragment, the frame was suspended on a half-ring by a cord to prevent dorsal subluxation (Figs 4 and 5).
7.4 Technical aqxctm The patients were placed in normal beds in the supine position with the injured leg elevated on a frame and the knee flexed at 45O. For the JBPF, we used two types of large pinless clamps (Fig. 2). The asymmetrical clamp was placed 5 or 10 cm from the most proximal fracture line, not to interfere with the future skin incision on the tibia1 diaphysis. The symmetrical clamp was placed in the same position as a Steinmann pin (ca. 4 cm distal and 4 cm dorsal of the lateral malleolus) on the calcaneus. Nine JBPFs were mounted under local anaesthesia in the emergency room after a single dose of prophylactic antibiotic using third generation cephalosporin. One fixator was mounted in the intensive care unit (ICU) on an intubated polytraumatised patient.
Fig. 3: JBPF after application under local anaesthesia on a 45O padded splint (case 8).
s 49
Babst: Tractionsubstitute
Table 1: Pain rating. GSA?
Sex/Age (yea=)
Fracture
F, 83 M, 43
A0 type
Mechanism of injury
Complications
Trimall. dislocation
B 23
shrmbling
slight pes equinus
Pilon
c 3.3
fall 3 m
F, 58
2nd degree open tibia
A 3.2
pedestrian/
F, 61
Trimall. dislocation
B 3.3
stumbling
loosening tibia1 clamp
L.A.
M, 52
Trimall. dislocation
B 3.1
stumbling
erysipeloid, fever
LA.
F, 88 M, 63
Bimall. dislocation
B 2.3
stumbling
Pilon
c 1.1
stumbling (stairs)
pressure ulcer
L.A.
M, 45
Pilon
B 2.2
fall 1.5 m
loosening tibia1 clamp
L.A.
M, 52
Trimall. dislocation
B 3.3
stumbling
L.A.
F, 70
Bimall. dislocation
B 2.2
shrmbling
L.A.
automobile act.
Time of opioid application (days)
Pain during JBPF application (points)
DWatiO” Pain with JBPF of JBPF in place (days) (wink)
2
2
12
LA.
2
0
1
16
GA.*
G.A.*
Anaesthesis
L.A.
G.A.’
2
10
3
2
10
6
4
12
4
2
5
2
6
3
13
3
4
2
6
1
0
3
8
0
0
2
12
5
LA.
slight pes equinus
Fig. 4: Trimalleolar dislocation fracture of the ankle (case 9). Fig. 6: Case 9 standard X-my views after reduction and fixation with the JBPF. In all patients, definitive fracture treatment consisted of plate fixation after swelling had subsided. Before clamp removal and skin incision a single dose of third generation cephalosporin was given i.v. For the operation, the skin entrance points for the clamps were covered with sterile adhesive. 7.!S
Fig. 5: Case 9 on the 45O padded splint. The JBPF suspended by a cord to prevent dorsal subluxation.
Resuh
The average pain suffered during clamp placement, reduction manoeuvre and assembly of the fixator was rated at 2.8 points (range: min 0, max 6 points) by the patient (Table 1). The maximum pain during the time the JBPF was in place was rated at 2.2 points (range: rain 1, max 4 points). The need for opioid drugs averaged one a day (range: min 0, max 5 days). On average, the JBPF was left in place for 10 days (range: min 5, max 16 days). In two cases, the tibia1 clamp became loose on the 5th and the 6th day respectively. Two patients developed a slight equinus of the foot (case 1, 10). In five patients, skin entrance points had serous discharge on the first day, without clinical evidence of infection. In one case, a skin ulcer developed on the lateral calcaneal entrance point due to inadequate clamp placement (case 7). An erysipeloid skin redness with fever was observed 24 hours
S50
after placement of the JBPF in another patient (case 5) who had sustained a trimalleolar fracture with a bruise and skin necrosis on the medial malleolus. A swab taken from the tibia1 entrance point showed Clostridium perfringens after culture on broth. All fractures were adequately distracted and reduced. The definitive treatment was in no case limited by the previous placement of a JBPF. Eight patients had uneventful healing of their incisional wound with no signs of infection. After definitive internal fixation of two pilon fractures, skin necrosis developed at the incision site. In one case (case 3), skin necrosis was superficial with no need of further treatment. In the other case (case 2), the anterior cloverleaf plate had to be covered by a free rectus abdominis flap. Both complications were unrelated to the application of the fixator.
7.6 Ditimmion The aim of this pilot study was to analyse the feasibility of a stable pin traction substitute for fractures around the ankle that would not interfere with subsequent internal fixation. Conventional skeletal traction with a Steinmann pin is a well established and simple method which is usually performed under local anaesthesia in the emergency room. One of the prerequisites for a traction substitute is that it could be applied under the same conditions equally well. Our data show that this holds true for the JBPF. During the placement of the JBPF, maximal pain was produced by the reduction manoeuvre as reported by the nine patients undergoing this procedure under local anaesthesia. Only three patients felt slight pain (2 points) at the moment of clamp application All non-intubated patients were able to lift up their injured leg at the end of the procedure without further pain While the fiiator was in place, pain was reported at the beginning or at moments of inadvertent sudden rotational movement of the foot. The short period of additional opioid administration correlates well with measurement of subjective pain perception. For mechanical reasons, a two point clamp fixator can provide only limited rotational stability. Therefore, it is clear that minor readjustments of the construction need to be made in some of the cases. Due to sag&al instability, trimalleolar fracture dislocation with a large Volkmann’s fragment have a tendency to displace dorsally. A suspension cord attached to the calcaneus (Fig. 3) usually solves this problem. This displacement could also be prevented by the fitting of a second tibia1 clamp more proximally. It is obvious that conventional external fiiator frames with Schanz screws do provide better stability, but we think this is usually not required for traction purposes and therefore does not justify the additional risk of local anaesthesia and possible pin track complications. Compared to conventional skeletal traction the JBPF reduces pain, provides better stability and therefore improves mobility, increases patient comfort and simplifies both patient care and physiotherapy.
The absence of pain allowed us to wait for optimal soft tissue conditions. This explains the rather long interval between placement of the device (average 10 days) and final ONF. The seemingly high incidence of clamp loosening might have been prevented by a more medial placement of the dorsal arm of the tibia1 clamp. Ideally the dorsal arm should be placed as far medial on the dorsal aspect of the tibia as possible. This is in some cases limited by the design of the clamp, leading to undue pressure on the soft tissue. This might be improved by changes of clamp design. One clamp 1oosened in an elderly confused lady in the early stages of Altzheimer disease (case 4). This was easily corrected under local anaesthesia. The other was due to an uncontrolled movement during sleep (case 8). In this case, the frame was converted to a simple traction device through the calcaneal clamp. The equinus malreduction was due to positioning the clamps too anteriorly and could easily have been prevented by proper placement The Clostridium perfringens infection, discovered on the swab culture, resolved after appropriate antibiotic treatment without sequelae. In summary, we feel that the JBPF can replace transcalcaneal traction offering good patient comfort and easy care. Further studies, however, are necessary to evaluate potential advantages as compared to skeletal traction and conventional external fixators for this indication. 7.7 summaly The stable traction substitute with a JBPF seems a very promising indication for pinless clamps, offering good patient comfort and easy care. The implantation of a JBPF is possible under the same conditions as for conventional calcaneal traction. Although the stability provided is less than for conventional external fixators, it is sufficient for a temporary traction device. The JBPF does not affect secondary ORIF. 7.8
References
Bohler L (1943) Die Technik der Knochenbruchbehandlung. 9-11 edition, Verlag Wilhelm Maudrich, Wien Melzack R (1975) The McGill Pain Questionnairez Major properties and scoring methods. Pain 127 Miiller ME, Al&wet M, Schneider R, Willenegger H (1991) Manual of Internal Fixation, 3rd edition, Springer Verlag
