Singer et al studied 15 000 fractures in adults attending The Edinburgh Royal infirmary 1992-1993. (Singer)
The incidence of femoral diaphyseal fractures found by Singer et al is shown in table 1.
Table 1. The incidence of femoral diaphyseal fracture per 10000 population.
|
Age
(Yr) |
Male |
Female |
|
15-19 |
3.55 |
0.57 |
|
20-24 |
3.56 |
0.88 |
|
25-29 |
1.64 |
0.64 |
|
30-34 |
3.73 |
0.18 |
|
35-39 |
0.39 |
0.20 |
|
40-44 |
0.57 |
0.37 |
|
45-49 |
0.72 |
0.47 |
|
50-54 |
0.76 |
1.2 |
|
55-59 |
0.27 |
0.25 |
|
60-64 |
0.00 |
1.01 |
|
65-69 |
0.00 |
1.81 |
|
70-74 |
2.3 |
3.15 |
|
75-79 |
0.58 |
3.74 |
|
80-84 |
4.11 |
10.7 |
|
85-89 |
2.76 |
16.39 |
|
90-94 |
0.00 |
37.14 |
Similar figures in Stockholm were found in 1986 by Hedlund et al. (Hedlund) Femoral shaft fractures occur in a bimodal distribution. They are either high-energy injuries in young men or low-energy injuries in elderly women. (Singer, Taylor, Kolmert, Hedlund) The two distinct patient population groups need to be considered individually when comparing mechanism of injury, treatment, complications and outcome.
Several options exist for the treatment of femoral shaft fractures. Non-operative, including skeletal traction and cast bracing. Operative, including plate osteosynthesis, external fixation and intramedullary nailing. In the last ten years the indications for femoral nailing have broadened. Intramedullary nailing has thus become the most common form of treatment. Intramedullary nailing has helped improve the outcome following open and closed fractures. Complication rates following intramedullary nailing are quoted at 15-19 % (Alho, Hammacher, Braten). These figures however refer to studies predominantly of femoral shaft fractures in the younger patient. Complication rates as high as 45-60% (Bouchard, Boyd, Salimen) have been reported for Femoral shaft fractures in patients older than 65years, with a peri-operative mortality approaching 20% in the older patient. (Bouchard, Boyd)
Taylor et al in 1994 clearly showed the
bimodal distribution of age and mechanism of injury causing femoral shaft
fractures. Outcome has been shown to be dependent on the mechanism of injury. (Taylor)
Young adults normally fracture their femurs
in high-energy accidents involving motorcycles and pedestrian vehicle
accidents. Motorcycle and pedestrian
injuries are associated with a higher incidence of associated injuries than
falls. (Taylor) Medical and engineering
studies of 1074 car accidents involving 2520 vehicle occupants have shown: The
incidence of femoral fractures in car occupants is 1.7% and 0.8% for front and
rear seat passengers respectively.
¨ The incidence of femoral shaft fracture was not significantly different in unrestrained drivers, but the associated injuries were more severe
¨
The mean velocity change during the
collision required to fracture a femur is 42km/h (26 mph). There is a higher
incidence of femoral shaft fractures when velocity change was greater than
48km/h (30 mph).
¨
Estimates of forces needed to fracture
a femur were higher than those found in cadaver studies.
¨
Frontal impacts caused the majority of
femoral fractures. Rear impacts were never associated with femoral shaft
fracture.
¨
Side impact collisions causing femoral
shaft fractures, involved severe intrusion into the car body, over 0.4m (16
inches) in all instances.
(Rastogi)
Like most bones the femur fails under tensile strain. The commonest mechanism of injury is bending load causing a transverse fracture. It has been estimated that 250 Nm of bending moment is required to fracture a normal adult femur. (Kyle)
The incidence of femoral fractures
following low velocity gun shot wounds is increasing in urban centers. (Levy, Wright)
Low energy injuries tend to occur in the elderly female population. They account for between 25% and 56% of femoral fractures, depending on the resident population age profile. (Moran) The majority of these fractures are closed and not associated with significant concomitant injuries. Despite the fact that these are low-energy fractures with no associated injuries. These fractures have a high incidence of treatment complications. This is quoted at 45-60% with a peri-operative mortality approaching 20% at six months. (Bouchard, Boyd, Moran, Salimen)
Salimen et al studied 201 acute fresh fractures of the femoral shaft admitted over a 10-year period, 50 (25%) patients had fractures caused by low energy trauma. The mean age of the low energy group was 65 years. Two thirds of the patients had at least one local or general factor weakening the mechanical strength of the bone. These include; diabetes mellitus, chronic alcohol use, previous major fracture and neuromuscular disorders. The most frequent site and pattern was spiral mid shaft femoral fractures accounting for 2/3 rds of the fractures. (Salimen)
These occur in athletes and military recruits following an increase in training intensity. In a study of Seventy-one athletes with 74 stress injuries to the femur forty-three were females (27 yrs) and 28 were males (31 yrs). Running was the most common activity at the time of injury (89%) followed by triathlon (5%) and aerobic dance (5%). Thirty per cent of the runners had increased their training duration immediately prior to their first symptom. Anterior thigh pain was the most frequent site of exercise-induced pain (46%) followed by hip pain (27%) and groin pain (8%). (Clement) Stress fractures of the femoral condyles are uncommon, but should be included in the differential of knee pain. During the clinical examination, when asked to hop on the affected limb, 70% of the patients had pain reproduced in the hip, groin or anterior thigh.
Plain radiographs are reported as being
abnormal in only 24% of cases. Radioisotope scans have been shown to be the
most sensitive tests for early detection of these injuries. The mean time to
diagnosis and recovery has been quoted as 6.6 and 10.4 weeks respectively. (Clement) The
literature on femoral stress fractures is limited and suggests that this is a
rare injury. Femoral stress fractures account for 3.8%- 7.2% of all stress
fractures. (Matheson, Hulkko) An isolated prospective study of military recruits by Milgrom et al
in Israel, reported tibial (51.2%) and femoral (29.8%) stress fractures to be
more common than stress fractures of the feet (8.7%). (Milgrom)
Milgrom et al actively looked for signs and symptoms suggestive of a stress
fracture. All suspected stress fractures were confirmed on bone scan. It should
be noted that 69% of the femoral stress fractures in this study were
asymptomatic. They were picked up incidentally on bone scan. The clinical
importance of this finding is borne out in case reports of displaced stress
fractures of the femur in athletes. (Luchini,
Clement) Displaced stress fractures of the
femur do occur, albeit rarely. In case studies of displaced stress fractures of
the femoral shaft, the duration of symptoms prior to diagnosis has been 2-6
weeks. Most displaced stress fractures of the femoral shaft required surgical
repair. (Visuri)
Femoral shaft stress fractures have excellent healing potential when diagnosed early and treated conservatively.
There is no universally accepted classification system for femoral shaft fractures. The factors affecting outcome include age of patient, energy of injury, open or closed fracture, comminution, anatomical position and the associated injuries. As such there are three classification systems of importance.
Anatomical classification systems refer to
proximal, mid-shaft and distal femoral fractures. Distal femoral fractures are
further sub classified into: supracondylar fractures, which are within 9cm of
the distal femoral articular surface. The most widely used anatomical
classification system of distal femoral fractures is the AO/ ASIF system.
AO DIAGRAM
It is important to distinguish Supracondylar fractures from femoral shaft fractures as treatment, complications and outcome varies.
Winquist et al 1980 devised a fracture classification based primarily on the fracture pattern and its degree of stability, however it is now used more widely to define the degree of comminution. Segmental fractures are a double level fracture of the femoral shaft. The fracture pattern at each level must be studied in terms of its own individual stability to determine appropriate treatment.
I Very small fragment, not affecting fracture stability
II At least 50% cortical contact, preventing translation and shortening.
III Less than 50% cortical contact, or poor purchase of nail on cortex i.e. Proximal or distal femoral fractures
IV Circumferential loss of stable cortical contact. (Winquist)
Open fractures are generally classified using the Gustillo-Anderson classification. The fracture should be graded at the time of debridement.
I. The wound is less than 1cm long. It is usually a moderately clean puncture, through which a spike of bone has pierced the skin. There is little soft-tissue damage and no sign of crushing injury. The fracture is usually simple, transverse, or short oblique, with little comminution.
II. The laceration is more than 1 cm long, and there is no extensive soft-tissue damage, flap, or avulsion. There is slight or moderate crushing injury, moderate comminution of the fracture, and moderate contamination.
III. These are characterized by extensive damage to soft-tissues, including muscles, skin, and neurovascular structures, and a high degree of contamination. The fracture is often caused by high velocity trauma, resulting in a great deal of comminution and instability.
III A – Soft tissue coverage of the fractured bone is adequate
III B – Extensive injury to, or loss of soft tissue, with periosteal stripping and exposure of bone, massive contamination, and severe comminution of the fracture. After debridement and irrigation a local or free flap is needed for coverage.
III C – Any open fracture that is associated with an arterial injury that must be repaired, regardless of the degree of soft tissue injury. (Gustillo)
There is some degree of inter-observer discrepancy when applying the Gustillo-Anderson classification. As demonstrated in tibial fractures where the average agreement was only 60%, however this has not been formally studied in femoral shaft fractures to date. (Brumback)
The diagnosis of femoral shaft fractures is generally straightforward. Nevertheless there are recorded cases of delayed diagnosis. It should also be borne in mind that stress fractures might be asymptomatic.
Assesment of the associated injuries at the time of presentation is of particular importance.
Vascular injury following femoral shaft fracture is rare. The incidence is quoted as 0.7-2% of all femoral shaft fractures. (Howard, Kluger, Barr) Penetrating injury is the leading cause of vascular injury in the USA. Experience in Europe shows blunt trauma to be leading cause of vessel injuries.
Lower extremity vascular injuries are rare and easy to miss, especially when encountered in hypotensive patients with multiple injuries. (Kluger, Barr) In a study by Kluger et al 1994 of 765 consecutive patients with closed midshaft fractures 10 (1.3%) had an acute vascular injury. In 3 out of 10 patients with an acute vascular injury dorsalis pedis and posterior tibial pulses were present initially. The pulses disappeared early in the course of initial resuscitation and treatment. Thus the presence of peripheral pulses does not exclude vascular injury. Fracture configuration was the same in the group with and without vascular injury.
Active bleeding, expanding haematoma and bruit, or a thrill over an injured vessel are absolute indications for angiography or surgical exploration. Of the 765 patients reviewed 31 patients had an angiogram. Eleven were on the grounds of loss of pulse, of these 11 patients with no palpable pulse 9 had a positive angiogram for vascular injury. Twenty patients had an angiogram for large haematoma, only 1 of these had a positive angiogram for vascular injury. Of the 765 patients, 734 had no clear indication for an angiogram, of these only 1(0.1%) presented late, with an AV fistula. (Kluger)
The important message is repeated examination and a high index of suspicion.
Primary nerve injury following femoral shaft fracture is rare. Isolated case reports exist in the literature. (Spiegel) Most nerve injuries are traction injuries associated with operative treatment of femoral shaft fractures. (See complications) (Brumback) Clawson et al 1960 reviewed the late consequences of sciatic nerve injury from various aetiologies. They found it was generally true that good neurological recovery and good function went together. Although on occasion there were remarkable exceptions. One third of patients showed vasomotor and trophic disorders. Pressure sores were the most serious consequence of sciatic nerve injury and at some time or other were present in 14% of patients. The cause of the pressure sores was deformity rather than insensibility. Of note when amputation was performed for fixed deformity with secondary ulceration the result was satisfactory. When it was done for pain there was no relief. Amputation is therefore avoidable provided that vigorous steps are taken to prevent or correct deformity; it should not be done for pain relief. (Clawson) Kline et al 1989 reported on sciatic nerve injuries over a 24-year period. Regardless of the mechanism of injury, patients treated conservatively because of partial deficit and/or improvement in function had a slightly greater than 80% chance of useful function in their tibial division distribution and approximately 60% chance of ultimately regaining useful peroneal distribution function. In surgically treated patients, recovery of useful function was greatest after neurolysis if associated with a positive nerve action potential, intermediate after suture, and less likely after nerve grafting. Recovery in the tibial nerve far exceeded that in the peroneal innervated structures. Femoral shaft fractures most commonly caused traction injuries at the thigh level. Injuries at thigh level had a slightly better prognosis compared to those in the buttock region. (Kline)
The incidence of ipsilateral intracapsular femoral neck fracture with concomitant femoral shaft fractures ranges from 2.5%-5%. A delay in diagnosis of the femoral neck fracture is quoted as 19% - 30%. (Swiontkowski, Alho) This delay is usually due to inadequate initial Radiographs. Avascular necrosis of the femoral head complicating this injury occurs in 3%-5% of reported cases. (Swiontkowski, Alho) The incidence of avascular necrosis is less than that reported for age related isolated femoral neck fractures, where it is reported to be 10%. (Gerber) Chapman has attributed this to the majority of energy being dissipated in the femoral shaft fracture. (Chapman) Alternatively this incidence may be erroneously low as greater than 60% of cases have less than 1 year follow up, and avascular necrosis can occur up to 3 years post injury. The true incidence of avascular necrosis of the femoral head in these injuries therefore probably approaches 10-15%. (Swiontkowski)
Most authors agree that rigid internal
fixation of both fractures is indicated. The argument exists as to which
fractures treatment should take priority. Alho et al in a meta-analysis of 659
cases of concurrent ipsilateral fractures of the hip and femoral shaft made a
general observation that complications and reduced function resulted from the
shaft fracture component. In addition the low incidence of avascular necrosis
of the femoral head suggests outcome depends on the treatment of the femoral
shaft fracture. (Alho) However Swiontkowski et al in1987 has made the point that, the
complication of symptomatic avascular necrosis of the femoral head is much
harder to deal with than mal/non union of the femoral shaft. As such the neck
fracture deserves priority in treatment. (Swiontkowski)
Intertrochanteric (extracapsular) fractures account for 28% of ipsilateral hip fractures, they do not pose as much of a diagnostic dilemma and are not associated with the above complications of avascular necrosis and non-union. (Alho)
Ipsilateral knee injury is easily missed during initial treatment of the femoral shaft fracture. The reported incidence of ipsilateral knee ligament injury with femoral shaft fracture is between 17% and 52%. (Moore, De Campos, Vangsness) This wide variation in frequency is caused by several factors;
· Variations in recording the incidence due to an increased awareness of ligament injuries currently as opposed to historically.
· Data gathered acutely as opposed to chronically is more likely to miss the diagnosis, as examination of the knee is difficult in the acute setting.
· Variations in the mechanism of injury. High-energy injuries are more likely to be associated with ligamentous injury.
Arthroscopic evidence of ipsilateral knee injury following midshaft femoral fractures is common reported as 55%(22/40) by De Campos et al 1993. (De Campos) Blacksin et al 1998 reported MRI derangements of the knee associated with femoral shaft fractures. Blacksin reported a meniscal injury rate of 27%, with the posterior horn of the medial meniscus most frequently torn. The most frequent site of ligamentous injury was the medial collateral ligament (38%). (Blacksin) The presence of an intra-articular lesion does not imply progression to a symptomatic or degenerative problem or exclude its pre morbid presence. It should be borne in mind that in cadaver studies meniscal abnormalities are seen in 8%-57% of knees with little or no correlation with symptoms on history. (Fahmy)
In Arthroscopic studies of the knee very
little correlation exists between ligamentous laxity and meniscal pathology.
Knee effusion was associated with a high incidence knee pathology 13 out of 40
(54%), however the absence of a knee effusion did not rule out significant
intra articular pathology, 11 out of 40 (28%). (De Campos)
Table 2. Ipsilateral knee injury
|
|
De
Campos 1994 Arthroscopy |
Blacksin
1998 MRI |
|
Number
patients |
40 |
34 |
|
Total
abnormal |
22 (55%) |
34 (100%)+ |
|
PCL |
3 (7.5%) |
7 (21%) |
|
ACL |
13 (32.5%) |
2 (6%) |
|
LCL |
5 (12.5%) |
2 (6%) |
|
MCL |
11 (27.5%) |
13 (38%) |
|
A-
scope ACL |
2 complete 19 partial |
|
|
A-
scope PCL |
1 complete |
|
|
Total
meniscus |
22 (55%) knees |
9 (27%) |
|
Lateral
meniscus |
13 (33%) |
6 (18%) |
|
Medial
meniscus |
13 (33%) |
4 (12%) |
|
Bone
bruise |
|
32% 1 (3%) occult tibial plateau fracture |
+Magnetic resonance imaging of the ipsilateral knee was performed on patients with closed femoral shaft fractures, fixed with intramedullary nailing. Only knees with an effusion or clinical abnormality on examination under anaesthesia were included in this study.
Ipsilateral tibial and femoral shaft fractures the “floating knee” are uncommon. These complex injuries are associated with a high morbidity and prolonged rehabilitation. Treatment is also associated with significant complications. (See treatment and “the floating knee” )
The current favoured treatment of femoral shaft fractures is with statically locked reamed intramedullary nailing. Alternatives include: traction with or without cast bracing, plate osteosynthesis and external fixation. These will be dealt with in turn.
In a study of 2805 femoral shaft fractures between 1989 and1992, 31% were treated non-operatively; these were associated with a higher mortality and morbidity. (Fakhry) Traction, with or without cast bracing yields acceptable functional results. Union of the fracture usually occurs within thirteen to fifteen weeks with most non-operative forms of treatment. Common sequelae included shortening, malrotation and contractures of the knee.
Table3. Non-operative management of femoral shaft fractures
|
|
Hardy
1983 Cast bracing |
Suman
1981 Cast bracing |
Buxton
1981 Perkins traction |
|
Fractures |
108 |
117 |
50 |
|
Patients |
106 |
117 |
50 |
|
Infection |
29/106 (27%) skin ulcers from cast |
2 (1.7%) pressure sores, 1 (0.8%) infected Tibial pin track |
18 (36%) pin track problems. No pressure sores |
|
Union |
10-12 wks-22 (20%), 12-16, wks-74 (69%), 16-24 wks-24 (22%) |
15 weeks (11-24) |
47 (94%) clinical and radiological union at 12 wks. |
|
Delayed
union |
9 (8.3%) |
|
|
|
Non-
union |
0 |
|
|
|
Shortening |
> 2cm-11/92 (12%) |
>2cm-15% |
>2cm-3 (6%) |
|
Rotation |
>10 degrees-18/95 (20%) >20 degrees- 4/95 (4%) |
|
>10 degrees- 0 |
|
Angulation
>15 degrees |
Varus/valgus-0 anterior bowing-2 (2%) posterior bowing-1 (1%) |
Varus/valgus-6 (5%) |
Varus/valgus-2/46 (4%) |
|
Re-fracture |
2 (2%) |
|
4 (8%) |
|
Knee
movement |
<90 degrees flexion-3 (3%) 90-110 degrees-2 (2%), >110 degrees but lacking last 10 degrees flexion-16 (15%) |
(15 excluded), < 90 degrees-13/102 (13%), 90-120 degrees-28/102 (27%) |
<120 degrees flexion-0 |
Several modifications of skeletal traction have been advocated to try and improve outcome. For example Perkins traction with a split bed allowing knee motion helped to reduce the incidence of knee stiffness. Non-operative treatment requires close supervision of the healing fracture configuration to ensure acceptable results. Several reports of the results of traction achieved in the early 1970’s exist; (Mooney et al 1970, Connolly et al 1973; Brown and Preston 1975; Wardlaw 1977, Lesin et al 1977). There is very little published data available from the last decade on treatment with traction. Mooney et al 1970 in a prospective study of one hundred and fifty patients with fractures of the distal part of the femur, treated with traction and cast bracing, reported no non-unions or re-fractures and a mean healing time of 14.5 weeks. (Mooney) Thomas et al 1981, in a comparative study of 81 patients with fractures of the distal half of the femur, treated with: 1.Traction 2. Traction followed by cast brace application at 5 to seven weeks 3. Open reduction and intramedullary nailing. They found that prolonged traction led to slower bony union and a prolonged hospital stay. Traction and cast bracing averaged 7 weeks in hospital and 15 weeks to union, no delayed union and no cases of deep infection. Intramedullary nailing averaged 5 weeks in hospital and bony union at 13 weeks. However this was associated with one case of deep infection which took 20 weeks to unite, 3 sequestrectomies and 3 years of treatment. Traction with timely hinged knee bracing provides a safe reliable method for the treatment of distal femoral fractures, combining the advantages of non-operative management with early mobilisation and knee movement. (Thomas)
The principles of anatomical reduction and rigid fixation of femoral shaft fractures followed by early limb rehabilitation have gained widespread acceptance.
Dissatisfaction with the results of non-operative treatment of femoral shaft fractures, because of prolonged hospitalization, high costs, fracture shortening, malunion, delayed union and joint stiffness has led to the development of various operative techniques for the management of femoral shaft fractures. In the last ten years intramedullary nail design and insertion techniques have been pre-eminent especially in femoral shaft fractures, with associated multiple injuries, including head injuries and chest injuries.
Plate osteosynthesis is an important technique in the treatment of femoral shaft fractures. It is advantageous in situations where intramedullary nailing is not ideal. These include:
· Adult and paediatric polytrauma with concommitant head trauma or pulmonary compromise
· Ipsilateral femoral neck and shaft fractures
· Open fracture with a vascular injury
· Fracture location in the proximal or distal femoral shaft and excessively narrow intramedullary canals
· Where the equipment necessary for intramedullary nailing is not available. (Rozbruch)
The principles of accurate reduction and rigid internal fixation of femoral fractures with plates should be followed. Despite the improvements in anatomical reduction, earlier mobilization, decreased pulmonary morbidity and earlier discharge from hospital, plating has brought with it an array of new complications. The problem of knee movement has not been completely addressed. This is a bit skewed in that plating is more likely to be used in the distal third (supracondylar) fractures where knee movement is more likely to be affected. Malunion, infection, mechanical failure of fixation, delayed union, non-union, re-fracture and re-operations are all possible complications following plate fixation of femoral fractures. Accurate open reduction and plate fixation requires considerable experience and operative expertise to ensure satisfactory results. Recent evolution in plating techniques has lead to the concepts of careful tissue dissection, epiperiosteal exposure of bone, and indirect reduction of fractures to minimize stripping and devascularization of bone fragments. Reduced soft tissue dissection also leads to less quadriceps tethering and improved knee range of motion. Reduction of the fracture to achieve anatomic alignment of intra-articular fractures and optimal rather than maximal stability is the goal. Routine bone grafting of the medial cortical defect as advocated in the 1980’s (Loomer, Magerl, Ruedi) is no longer advocated. (Rozbruch) The greater initial soft tissue dissection in those earlier series likely compromised the biologic healing potential, then mandating a biologic stimulus for healing, namely bone grafting. Plate construction and design has also evolved, limited contact dynamic compression plates composed of pure titanium provide improved tissue compatibility and improved blood supply to the plated bone segment. Longer plates relative to the fracture length are thought to be particularly important, increasing the lever arm of the plate. Optimal use of screws through the plate as opposed to maximal use of screws the length of the plate has also been suggested. The use of unicortical screws at the plate’s periphery and multiple inter-fragmentary lag screws outside of the plate have been abandoned. The latest study by Rozbruch et al 1998 clearly shows the improvement in outcome with new biological internal fixation. (Rozbruch) The improved figures for 1993-1994 surely also reflect the fact that plating requires experienced hands to achieve acceptable outcomes.
Table 4. Plate osteosynthesis of femoral shaft fractures.
|
|
Rozbruch
1972-1973 |
Rozbruch
1982 |
Rozbruch
1993-1994 |
Geissler
1980-1987 |
|
Number
fractures |
25 |
30 |
25 |
71 |
|
Number
patients |
23 |
30 |
23 |
69 |
|
Age
(years) |
32.8 (12-64) |
26.7 (14-58) |
49.5 (20-86) |
26 (8-88) |
|
Open |
8 (32%) |
10 (33%) |
3 (12%) |
13 (18%) |
|
Bone
graft |
4 (16%) |
11 (30%) |
1 (4%) |
41 (69%) |
|
Infection-deep |
1 (4,8%) |
1 (3.4%) |
1 (4.3%) |
0 |
|
Union |
4.9 months |
4.98 months |
3.38 months |
16 weeks (23-72) |
|
Delayed
union |
3 (14.3%) |
2 (6.9%) |
0 |
0 |
|
Non-
union |
2 (9.5%) |
1 (3.4%) |
1 (4.3%) |
2 (3%) |
|
Implant
failure |
4 (19%) |
3 (10.3) |
1 (4.3%) |
1 (1.4%) |
|
Implant
loosening |
2 (9.5%) |
0 |
2 (8.7%) |
0 |
|
Successes
|
13 (62%) |
24 (84%) |
20 (87%) |
66 (93%) |
|
Re
operations |
9 (43%) |
9 (31%) |
3 (13%) |
3 (4%) |
As the indications for closed locked intramedullary nails expand, indications for external fixation diminish. In certain circumstances for instance the unstable polytraumatized patient and unstable patients with head injuries, external fixation allows for rapid rigid fixation. (Alonso) Operating time and the ease of external fixator application in the polytrauma patient have been cited as indications for external fixation. Volker et al 1995 in a study of 18 open femoral shaft fractures showed that, soft tissue debridement, reduction, and application of external fixator took a mean time of 73 minutes. (Volker) Van den Bosche et al 1995 showed that it was safe to stabilize fractures temporarily with external fixation and convert to intramedullary nailing when the patient’s condition was stable. (Van Den Bosch) The external fixator is a useful adjunct to fractures about the knee that include both the supracondylar area and the proximal tibia. Generally external fixation is reserved for the more complex femoral fracture, including severe comminution and open fractures with extensive soft tissue injury and/or vascular injury. This selection bias is reflected in the relatively high complication rates.
Table 5. External fixation of femoral shaft fractures
|
|
Volker
1995 (1985-1989) Open |
Gottschalk
1985 (1981-1983) severely comminuted |
Alonso
1989 (1983-1986) 14 patients temporary external fixation |
Murphy
1988 (1979-1986) Complex fractures |
|
Fractures |
18 |
7 |
24 |
34 |
|
Patients |
18 |
7 |
24 |
34 |
|
Open |
18 (100%) |
3 (43%) |
13 (54%) |
26 (76%)^ |
|
Pin
tract infection |
0 |
1 (14%) |
3 (12.5%) |
6 (19%) |
|
Infection
deep |
2 (11%) |
0 |
1 (4.2%) |
2 (6%)^^^ |
|
Union |
5.5 months (1.8-9.6) |
6 months (4.5-8) |
|
4.8 months (2-9) |
|
Delayed
union |
|
|
2 (8%) |
4 (13%) |
|
Non-
union |
0 |
|
1 (4%) |
0 caution, re-operation |
|
Shortening |
>2cm-1/15 (7%) |
0 |
>2cm-2 (8%) |
> 2 cm- 5 (16%). |
|
Rotation
>15 degrees |
0 |
|
|
1 (3%) |
|
Angulation
>15 degrees |
Valgus-1 (7%) |
|
|
|
|
Re-fracture |
0 |
|
|
|
|
Knee
movement |
3 (20%)- flexion deficit>10 degrees |
All 7 (100%) had almost full ROM |
Average loss of motion 56 degrees |
Average ROM 91 degrees, <90 degrees 14 (44%) |
|
Re-operations
|
|
|
8- converted to dynamic compression plating or casting, 6 delayed intramedullary nailing. |
4 (13%) |
Closed intramedullary nailing has many theoretical and practical advantages compared with other treatments for fractures of the femoral shaft. It is the most commonly used treatment for femoral fractures. Femoral nailing gives predictable realignment of bone, rapid healing and early functional use of the limb. The advent of static locking to prevent shortening and malrotation has extended the indications of intramedullary nailing to severely comminuted fractures, subtrochanteric fractures and supracondylar fractures. There is mounting evidence to show that early intramedullary nailing of open fractures including Gustillo grade III, yields acceptable results. (Rutter, Williams, Nicholas)
Closed intramedullary nailing should be attempted, but inadequate reduction and eccentric reaming of comminuted fractures can lead to inferior results. Open reduction during intramedullary nailing is controversial. Some authors contend that open reduction is not associated with a significant increase in complications, and therefore if a satisfactory closed reduction cannot be obtained the fracture should be opened and held reduced. (Winquist, Rokkanen, Leighton) Green et al 1987 tempered this, in an article dealing with the consequences of chronic sepsis in the femur following intramedullary nailing. The perceived benefits derived from open reduction are rapidly lost in the presence of chronic deep infection. In his study several of the patients referred with deep-seated infection were initially closed injuries, which were opened to allow reduction. (Green) In a comparative study by Rokkanen et al 1969, of open or closed intramedullary nailing versus conservative management. They found both early and late results of treatment were better after intramedullary nailing than after conservative management. The difficult fractures in the older patient benefited most from intramedullary nailing, whereas low energy injuries in young people healed equally well regardless of the method of treatment. (Rokkanen) Controversy exists over reamed versus unreamed nailing. With improved implant technology, smaller diameter, closed section, thicker walled nails are available. They theoretically could provide adequate support to allow uncomplicated union. Smaller unreamed intramedullary nailing has the theoretical advantages of less soft tissue and bony trauma, lower intramedullary pressure and less fat embolism risk, lower incidence of infection and shorter operating times. The theoretical disadvantages are loss of mechanical strength and stability leading to implant failure, higher rates of non-union or delayed union. Theoretically they have advantages in open fractures, head injured patients and patients with pulmonary compromise. Conflicting articles exist in the literature. Boyer et al 1996 concluded smaller diameter nails had not been shown to increase the rates of mal or non-union. (Boyer) Kropfl et al 1995 showed superior results with unreamed smaller diameter nails compared with the published results for reamed intramedullary nailing. Kropfl et al quoted non-union figures of 0% and infection rates of 0%. (Kropfl) Hammacher et al 1998 was unable to confirm these findings in a multicenter experience of 129 unreamed intramedullary nails he quoted non-union figures of 5.1% and infection rates of 2.9%. Of note in the Hammacher study 26% of the cases were open fractures as opposed to 10% in the Kropfl study and open reduction was performed more frequently in the Hammacher series. (Hammacher) Tornetta et al 1997 in a prospective randomized study of 81 fractures concluded that reamed canal preparation led to faster healing of distal fractures treated with statically locked intramedullary nails and that their was no advantage to nail insertion without reaming. (Tornetta) Clatworthy et al 1998 in a prospective randomised trial of 48 patients with 50 fractures, concluded that; reaming aids fracture healing and suggested the use of reamed nails. (Clatworthy) The latest largest study available to date by Wolinsky et al 1999 reporting on 551 femoral shaft fractures treated with reamed intramedullary nails, reported 94% of fractures healing after primary nailing, 4.7% needed a second procedure to heal the fracture, 0.5% needed two procedures and 1% did not heal. Six fractures (1%) became infected. All six fractures that became infected, healed after adequate treatment and at follow up were infection free. Only 1 nail (0.2%) broke after 17 months, in a case with segmental bone loss and non-union. Thirteen locking bolts broke with no effect on treatment or outcome. Malunion was rare. All fractures healed with less than 10 degrees of angulation in any plane. Careful attention must be paid to reduction and nail placement when treating distal third fractures. (Wolinsky)
Table 6. Intramedullary nailing of femoral shaft fractures
|
|
Wolinsky
1999 (1986-1996) Reamed |
Braten
1995 (1982-1991) Reamed |
Winquist
1980 (1968-1979) Comminuted |
Kropfl
1995 (1992-1993) Unreamed |
Hammacher
1998 (1994-1996) Unreamed |
|
Fractures |
551 |
120 |
245 |
81 |
129 |
|
Patients |
515 |
116 |
|
75 |
122 |
|
Infection
(superficial) |
3 (0.5%) |
4 (3%) |
|
0 |
2 (1.5%) |
|
Infection
(deep) |
3 (0.5%) |
1 (0.8%) |
1 (0.4%) |
0 |
2 (1.5%) |
|
Union |
545 (99%) |
100% |
243 (99%) |
81 (100%) |
118 (91%) |
|
Delayed
union |
29 (5%) |
6 (5%) |
|
0 |
4 (3%) |
|
Non-
union |
6 (1%) |
0 |
2 (0.8%) |
0 |
7 (5%) |
|
Shortening |
5 noted on gait analysis |
>2cm-1 (0.8%) |
>2cm-7 (3%) |
>2 cm-0 |
N/A |
|
Rotation
|
None requiring correction |
> 15 degrees 23/116 (20%) |
>10 degrees 16 (7%) |
|
>20 degrees 1/ 125 (1%) |
|
Angulation
> 5-10 degrees |
44/418-(11%) |
0 |
2 (1%) |
7 (9%) |
N/A |
|
Angulation
> 10 degrees |
0 |
1 (0.8%), |
3 (1%) |
0 |
N/A |
|
Implant
failure |
1 (0.2%) |
2 (1.6%) |
|
0 |
1 (0.8%) nail |
|
Knee
movement |
N/A |
Full ROM |
<125 degrees 6/201 (3%) |
N/A |
N/A |
|
Re
operation |
29 (5%) |
1 (0.8%) |
|
0 |
9 (6.6%) |
Several studies have shown that intramedullary nailing of open fractures is safe if thorough debridement and irrigation are performed prior to nailing. (Rutter, Williams, Grosse, Brumback, Van den Bossche) Controversy still exists regarding the intramedullary nailing of grade III open fractures. Below are various advocates of intramedullary nailing various grades of open femoral fractures.
· All grades: Rutter et al 1994, Grosse et al 1993
· Grade I,II and IIIA: Van den Bossche 1995
· Grade I, II and selected grade III: Brumback et al 1989,
Williams et al
1995
· Gunshot wounds: Nicholas et al 1995, Levy et al 1993
Proponents of the more conservative approach of only nailing selected grade III fractures cite the high incidence of local complications following intramedullary nailing of open grade III B and III C fractures. They advocate external fixation for grade IIIB and IIIC fractures.
Green et al 1991 reported a high incidence 55% (6/11) of infection for grade III fractures versus a low incidence 4.5% (1/22) for grade I and II femoral shaft fractures treated with intramedullary nailing. (Green) In a prospective study by Grosse et al 1994, of 115 consecutive open fractures, including 12 grade IIIB and 5 grade IIIC fractures treated with intramedullary nailing, only 3 patients (2.6%) developed serious infections. (Grosse) In a prospective randomized trial by Williams et al 1995, forty-two patients with open femoral shaft fractures of all Gustillo-Anderson grades, were randomized to primary nailing (at the time of initial debridement) and delayed nailing (two to five days post initial debridement). This included 4- grade IIIA, 3- grade IIIB and 2- grade IIIC fractures in the primary nailing group. He reported an infection in 1 patient (2.4%). This infection was in the group treated by delayed intramedullary nailing, Gustillo-Anderson grade II. This patient was later found to be HIV positive. (Williams) Further evidence to suggest the low complication rate of intramedullary nailing in Gustillo-Anderson IIIB and IIIC is accruing, Rutter et al 1994 reported 0% infection in 5- grade IIIB and 1- grade IIIC femoral fracture treated with intramedullary nailing. (Rutter) Recent reports on the use of intramedullary nailing for femoral shaft fractures caused by gunshots also support the use intramedullary nailing in grade I to grade IIIC open fractures. (Wright, Nicholas)
Although studies are starting to show that Intramedullary nailing produces acceptable results in Gustillo-Anderson grade IIIC fractures, controversy still exist about the optimal form of treatment for femoral fractures with associated vascular injuries. Some authors believe that short term temporary external fixation of severe open fractures followed by conversion into an intramedullary nail can be done without increased risk of deep infection. (Van den Bossche)
Table 7. Infection rates in open femoral shaft fractures.
|
|
Brumback 1989 (nail) |
Grosse 1991 (nail) |
Green 1991 (mixed)* |
Rutter 1994 (nail) |
Williams 1995 (nail) |
Van Den Bossche 1995
(mixed)*** |
Total (Mixed) |
|
Open fractures |
89 |
115 |
53 |
28 |
42 |
57 |
384 |
|
Infection rate |
3 (3.4%) (all IIIB) |
3 (2.6%) |
9 (16%) (2 type II, 7 type
III) |
2 (7%) (IIIB,II) |
1 (2.4%) (II) ** |
0 (0%) |
18 (4.7%) |
Table 8. Infection rates according to Gustillo grade of open femoral shaft fractures
|
|
Brumback
1989 (nail) |
Rutter
1994 (nail) |
Williams
1995 (nail) |
Van
Den Bossche 1995 (mixed)*** |
Total (Mixed) |
|
Open
fractures |
89 |
28 |
42 |
57 |
216 |
|
Grade
I |
27 (30%) |
10 (36%) |
12 (29%) |
18 (32%) |
67 |
|
Grade
II |
16 (18%) |
9 (32%) |
16 (38%) |
11 (19%) |
52 |
|
Grade
III |
46 (52%) |
9 (32%) |
14 (33%) |
28 (49%) |
97 |
|
Grade
III A |
19 (21%) |
3 (11%) |
6 (14%) |
21 (37%) |
49 |
|
Grade
III B |
27 (30%) |
5 (18%) |
5 (12%) |
5 (9%) |
42 |
|
Grade
III C |
0 (0%) |
1 (3%) |
3 (7%) |
2 (3%) |
6 |
|
Infection
rate |
3 (3.4%) |
2 (7%) |
1 (2.4%) ** |
0 (0%) |
6 (2.7%) |
|
Gustillo
grades |
I=0 II=0 III=3 (3.5%) IIIA=0 IIIB=3 (3.4%) IIIC=0 |
I=0 II=1(3.5%) III=0 III=1 (3.5%) IIIA=0 IIIB=1(3.5%) IIIC=0 |
I=0 II=1(2.4%) III=0 IIIA=0 IIIB=0 IIIC=0 |
I=0 II=0 III=0 IIIA=0 IIIB=0 IIIC=0 |
I=0 II=2 (4%) III=4 (4%) IIIA=0 IIIB=4 (10%) IIIC=0 |
* Forty-two had internal fixation, 1 external fixation, 10 Traction and cast bracing.
** This patient was seropositive for the human immunodeficiency virus.
*** Twenty-six treated with plating, 11 intramedullary nailing, 20 external fixation (External fixator used for all 7 IIIB and IIIC fractures, and 9 IIIA fractures). The external fixator was routinely converted to internal fixation on average 21 days after initial injury.
Fraser et al 1978 reported on 222 cases of ipsilateral tibial and femoral shaft fractures. They divided these into three groups; group 1 had operative fixation of both fractures, group 2 had operative fixation of one fracture and in, group 3 both fractures were treated non-operatively. They found the best overall results in group 1 and the worst in group 3. The most alarming finding was a 30% incidence of osteomyelitis in patients who had operative fixation of both fractures. Other authors have reported figures of deep infection of 7% and 10% among their patients treated operatively. (Karlstrom, Fraser) Group 3, the non-operative group, was associated with a 30% rate of delayed union or non-union. Despite this high incidence of osteomyelitis in the operative group, patients treated non-operatively had the worst functional results. Fraser et al recommended intramedullary nailing of the femoral fracture and the use of rigid external fixation or cast bracing in the management of the tibial fracture, In an attempt to reduce the morbidity associated with osteomyelitis. (Fraser) Karlstrom et al 1977 reporting on a series of 32 cases in 31 patients noted, patients that had their femoral fracture internally fixed had better outcomes than those that did not. (Karlstrom) McAndrew et al 1988 reported long term follow up results on 29 out of 45 patients (64% follow up) with ipsilateral tibial and femoral diaphyseal fractures. They also found improved outcome in, patients who had intramedullary nailing of their femoral fracture. At follow up, a mean time of 6.5 years the 29 patients were functioning well, all had returned to work. Forty eight per cent were heavy labourers and 31% had sedentary occupations. Seventeen patients (59%) limited their sporting activities. (McAndrew)
De Campos et al 1993 found no complication directly related to pneumatic tourniquet use in knee arthroscopy following reamed intramedullary nailing of femoral fractures. (De Campos) Pollack et al 1997 in a retrospective review found that the use of a pneumatic tourniquet in patients with polytrauma who have also undergone reamed intramedullary nailing is associated with increased length of ventilator dependence and increased length of stay in the intensive care. (Pollack) Moore et al 1987 described compartment syndrome complicating surgical treatment of ipsilateral femur and ankle fractures associated with pneumatic tourniquet usage. (Moore)
The long term sequelae and poor functional outcomes associated with chronic sepsis are devastating. Green et al 1987 reporting on chronic sepsis following intramedullary nailing observed: shortening (average 4.3cm), thigh atrophy and substantial limitation of knee movement (mean knee flexion 68 degrees). Treatment varied, Green et al removed the nail when radiographs demonstrated unattached, apparently nonviable, bone sequestra at the fracture site requiring debridement. The nail was left in place if it was stabilizing the fracture and not showing evidence of sequestration. They found healing to be prolonged. Healing defined as the ability to ambulate pain free without an external support. For patients in whom the intramedullary nail was left in place, fracture healing occurred between 19 months and 10 years. Excluding the 10-year case the average time to fracture healing in the retained rod group was 33 months. In six patients the intramedullary nail was removed as part of the debridement and an external fixator used to provide stability and maintain length. Fracture healing time in this group ranged from 14 months to 53 months (average, 31 months). The twelve patients in this series of Green et al required four surgical procedures on average, to obtain fracture union. Green et al reported two cases of non-union, one in a case lost to follow up after 1 year and the second in a wheel chair bound patient with a traumatic through knee amputation. Eleven patients (91%) in this series were initially managed by exposing the fracture site combined with retrograde nailing of the proximal femur, either at the time of injury or shortly thereafter. The septic outcome following placement of a femoral rod was especially tragic for the six patients (50%) who started out with closed fractures. (Green)
Non-operative treatment has the least potential for deep infection. Hardy et al reporting on 108 open and closed fractures and Moulton et al reporting on 45 closed fractures, reported no deep bone infections. (Hardy, Moulton) Mooney et al 1970 in a personal communication with P W Brown stated that in over 1000 documented cases treated non-operatively, there were no non-unions or persistent infections, despite an approximately 30 per cent incidence of open fractures. (Mooney) Pressure areas and skin ulcers vary from 0%- 27%, these however do not significantly contribute to final outcome or deep infection. Traction pin sepsis can cause problems with an incidence of 0-16%. (Hardy, Suman, Buxton)
Infection is frequently quoted as the major complication of plating femoral shaft fractures. Ruedi et al 1989 has published the largest series with 131 femoral shaft fractures managed by compression plating. (Ruedi) His infection rate was 6%, Loomer et al 1980 reported an infection rate of 7%. (Loomer) Geissler 1995 et al pointed out that most of the reported series quoting infection rates around 5-7% following femoral shaft plating, were prior to the routine use of preoperative antibiotics for all cases, not just open fractures. Studies of cases routinely using prophylatic preoperative antibiotics for all fractures quote 0%-3% as the incidence of infection. (Geisler, Thompson, Magerl)
External fixation, pin tract sepsis occurs in 13%-19% of patients, deep infection ranges from 4%-11%. (Volker, Alonso, Murphy) It should be borne in mind however these studies involved small numbers and were of complex comminuted open fractures. The incidence of infection following closed intramedullary nailing of femoral fractures has been generally low, ranging from 0-2% in published series. (Wolinsky, Braten, Murphy, Alho, Winquist, Kropfl, Hammacher) Sepsis following intramedullary nailing of open femoral nailing is more variable occurring in 2%-5% of patients. (Brumback, Williams, Grosse, Rutter) Treatment of open fractures has a higher rate of infection overall, irrespective of treatment method, deep infection occurred in 18 out of 384 open fractures (4.7%). (Brumback, Grosse, Green, Rutter, Williams, Van Den Bosche)
Compartment syndrome
Compartment syndrome is defined as an elevation of the interstitial pressure in a closed osseofascial compartment that results in microvascular compromise. The diagnosis of compartment syndrome differs depending on the awareness and ability of the patient to co-operate. In the awake patient, pain and the presence of tension and swelling of the thigh raise the suspicion. The clinical examination includes pain on passive stretching of the muscles of the thigh, decreased sensation, in the area of the distribution of the nerves in the compartment involved, weakness of the muscles and diminution or absence of distal pulses. Measurement of compartment pressures in the awake patient is only an adjunctive diagnostic role. In patients unable to co-operate with the physical examination compartment pressures should be measured with little hesitation in patients with tense swelling of the thigh. The critical threshold of compartment pressure above which tissue is irreversibly damaged is controversial. The overall clinical picture must help the physician decide the critical threshold for each patient. Thirty millimeters of mercury has been suggested as the low threshold at which the diagnosis of compartment syndrome needs to be considered. Whitesides et al 1975 believed that fasciotomy was indicated if tissue pressure rose within 10-30 mmHg of the diastolic pressure i.e. 40-45 mmHg in a patient with a diastolic blood pressure of 70 mmHg. (Whitesides) In borderline cases repeated measurement should be obtained and a low threshold for the performance of decompressive fasciotomy should be used. Compartment syndrome of the thigh following femoral shaft fracture is uncommon. Scwartz et al 1989 identified 5 cases of compartment syndrome of the thigh in a period where more than 370 patients had intramedullary nailing of the femur giving an approximate incidence of 1%. (Schwartz) Tarlow et al 1986 reviewed the orthopaedic literature and found no cases of compartment syndrome in 1,311 reported cases of femoral shaft fracture. (Tarlow) Schwarz et al 1989 reviewed 21 cases of compartment syndrome of the thigh. They identified several risk factors for the development of compartment syndrome i.e. systemic hypotension, vascular injury, high-energy blunt trauma, external compression including the application of military anti shock trousers, coagulopathy, multiple trauma and infusion of large volumes of fluid to maintain systemic blood pressure. (Schwartz) It is difficult to correlate outcome from compartment syndrome of the thigh as it rarely occurs as an isolated injury. The diagnosis of compartment syndrome in the multiply injured patient is at least a measure of the severity of the patient’s injuries and portends a poor prognosis. Schwartz et al 1989 found a high mortality rate of 47% associated with compartment syndrome of the thigh. Of the 9-surviving patients with 10 compartment syndromes, 6 patients developed local wound infections. In 6 of the 9 patients (7 compartment syndromes) who survived there was no demonstrable sequel of compartment syndrome. Two patients used crutches for persistent quadriceps and hamstring muscle weakness, accompanied by sensory loss distal to the knee and one patient who had an ipsilateral open fracture of the tibia had decreased flexion of the knee. (Schwarz)
Compartment syndrome of the uninjured leg complicating intramedullary nailing and the use of the hemi-lithotomy position has been reported. (Anglen, Morrow, Dugdale, Carlson). The risk of compartment syndrome complicating the use of the lithotomy and hemi-lithotomy position has been previously noted in urologic, gynaecological, anaesthetic, and general surgical literature following positioning in the lithotomy position for 6 to 10 hours. Dugdale et al 1989 reported two cases, which took of 5 hours 45 minutes and 6 hours respectively. Anglen et al 1994 reported two cases taking 6 hours and 6 hours 15 minutes respectively. A special case is the presence of bilateral femoral fractures. Carlson et al 1995 reported 2 cases and Morrow et al 1994 reported one case of compartment syndrome following intramedullary nailing of bilateral femoral fractures in all three cases the duration in the hemilithotomy position was less than 3 ˝ hours. (Carlson, Morrow)
Unfortunately different authors have used different definitions of delayed union when reporting their findings. One definition of delayed healing is healing taking longer than 6 months or requiring a second operation to speed up healing. Delayed healing is still a problem in fractures treated with plating occurring in 0%-7% of cases. (Rozbruch, Geissler, Ruedi) Delayed healing is especially a concern in plated fractures due to the relatively high incidence and the consequences of implant failure. In an attempt to decrease the incidence of delayed and non-healing grafting of the medial femoral defect was advocated. This decreased the rate of delayed/ non-union. Lately however it has been suggested that with the new biological methods of fixation involving less soft tissue dissection, grafting of the medial femoral defect is not essential. (Rozbruch) Intramedullary nailing has a lower incidence 3.4% (14/416) of delayed union and the consequences of early implant failure are not as common. Static locking was considered a risk for increased delayed and non-union, this has not been clearly shown and routine dynamisation to prevent delayed union is no longer considered necessary. Delayed healing is highest in external fixation 8%-13% this can be explained partly on the basis of the complex fracture patterns treated with external fixation. (Alonso, Murphy)
Alho et al defined non-union as any fracture not healed at 12 months, or requiring a second operation to achieve healing. (Alho) Non-union rates following plating vary from 1.3%- 6.9%; the highest value (6.9%) was in a series with 100% comminution and an infection rate of 6.1%. (Ruedi) Intramedullary nailing has a very low reported non-union rate around 1%. The majority of cases of non-union were in a study by Hammacher et al 1998 all involving unreamed intramedullary nails. In contrast Kropfl et al reported a 0% non-union rate in 81 fractures treated with unreamed intramedullary nailing. The largest latest study by Wolinsky et al 1999 on 551 cases treated with intramedullary nailing reported a non-union rate of 6 cases (1%). Once again the varying definitions of non-union complicate the matter Wolinsky et al did not include patients undergoing a second procedure (nail dynamization or nail exchange) to attain healing in their figures for non-union. These included 26 patients requiring one further operation and the 3 cases requiring two further operations to achieve union. Giving a comparable non-union rate for the Alho definition of 6.4% (35/551). (Wolinsky) The importance of delayed and nonunion following operative intervention is the risk of implant failure.
Implant failure is a significant problem following plating of the femoral shaft occurring in 1%-11% of cases (35/417=8.4%). (Rozbruch, Geissler, Magerl, Loomer, Ruedi, Thompson) Ruedi et al 1979 noticed a large decrease in the incidence of implant failure following routine bone grafting of the medial femoral defect. (Ruedi) When looking closely at the figures of Rozbruch et al the incidence of implant failure in 1972 prior to routine bone grafting was 19%, following increased bone grafting of the medial femoral defect in 1982 he quoted an implant failure rate of 10%, in 1993 with the advent of biological fixation, and less grafting of the medial defect he quoted an implant failure rate of 4%. Implant failure is rare following intramedullary nailing. It is common for locking screws to fail, but this is not associated with any significant clinical consequence. Franklin et al studying reports of intramedullary nailing after 1980 found 7 cases in 1268 nails (0.5%). (Franklin) Alho et al reported an implant failure incidence of 3% following intramedullary nailing, this was a study of only comminuted fractures i.e. increasing the time to healing and placing increased mechanical stress on the nail. (Alho)
Varus or valgus angulation less than 10 degrees does not lead to clinical deformity or limitation. The exact amount of malalignment acceptable to prevent future degenerative arthritis is not positively known. (See arthritis) In the earlier studies of non-operative treatment malunion was reported if it measured greater than 15 degrees. In the operative treatment group angulation greater than 5 degrees was considered significant.
Kootstra et al 1973 assessed patients with external rotation deformity of 20 degrees and found 8 out of 23 patients (35%) were symptomatic. Braten et al 1995 in a study of 116 patients found rotational differences less than 15 degrees were not associated with any significant symptoms, whereas almost half the patients with greater than 15 degrees were troubled by this condition. (Braten)
Relatively little information is available about the effect of angular malalignment following a femoral fracture on the development of degenerative arthritis in the hip, knee or ankle. The concern is that abnormal forces may lead to degenerate arthritis. At present there is no data to support the position that malalignment always leads to degenerate arthritis. Kettelkamp et al 1988 discussed degenerative arthritis of the knee secondary to fracture malunion. They studied 15 knees in 14 patients who developed unicompartmental degenerative arthritis of the knee after angular malunion of femoral or tibial fractures. Average interval from fracture presentation to presentation for care of degenerative arthritis of the knee was 31.7 (10-60) years. They found that, the knee was more tolerant of increased valgus force than increased varus force. Nine femoral fractures had varus angulation of 3 to 25 degrees (average12.4 degrees), and all developed degenerative genu varum in an average of 28.7 years (range, 10-49 years). The medial plateau force was calculated to be 1.87 times more than the medial plateau force of the uninjured extremity. Two femoral fractures had valgus angulation of 8 and 14 degrees respectively and developed genu valgum in 60 and 37 years respectively. The lateral plateau force averaged 1.78 times more on the fractured side than on the normal extremity. (Kettelkamp) Egund et al 1982 studied 62 distal femoral fractures, treated non-surgically in 35 and surgically in 27 cases. They evaluated angulation and displacement of the fracture in relation to the development of gonarthrosis. Unfortunately their study has a short follow up time of 2- 10 years. As would be expected gonarthosis developed more commonly in displaced intercondylar fractures with a diastasis or planar difference in the articular surface exceeding 3 mm. There seemed to be less connection between angular deformity and gonarthrosis. Gonarthosis was mostly confined to the patello- femoral articulation. Arthrosis of the tibio-femoral articulation was rarely seen. (Egund)
Table 9. Incidence of arthritis following Supracondylar fractures. Egund et al
|
|
Undisplaced Supracondylar |
Displaced Unicondylar |
Undisplaced
Intercondylar |
Displaced intercondylar
(diastasis) |
|
Number |
9 |
10 |
17 |
18 |
|
Gonarthrosis |
1 (11%) |
2 (20%) |
1 (9%) |
10 (55%) |
Volpin et al studied degenerative arthritis of the knee following 31 intra-articular fractures of the knee, including 3 undisplaced and 7 comminuted displaced distal femoral fractures. Their follow up was from 6 to 22 years (average 14). Six of the 10 femoral fractures were treated conservatively and 4 surgically. They found no significant differences between results after surgical or conservative treatment. Patients who had early active and passive knee movement did better. There was no significant correlation between degeneration and the initial mechanism of injury, nor with the type and localisation of the fracture except that fractures caused by great violence were more susceptible. Degenerative changes developed mainly within the first six to eight years after initial injury. Later deterioration seemed to be rare; it seems that if signs and symptoms of degenerative arthritis have not developed by 10 years they are unlikely to appear later. The probability of degenerative change increased significantly with greater age at time of injury (older than 45 years). Radiological changes of severe degenerative arthritis are of little significance unless they are correlated with clinical signs. A number of patients in this study had very good knee function with marked radiographic change. (Volpin)
The main factors in preventing degenerative changes in surrounding joints appears to be the early restoration of joint congruity, realignment to the normal anatomical axis, joint stability and early movement.
Leg length inequality of 1cm or more is found in 7% (4%-8%) of the normal adult population. This incidence in the normal adult population is thought to be due to out of phase growth. (Giles) Patients with shortening of 1cm have a normal gait and most patients with less than 2 cm shortening have a normal gait and are free of symptoms. Shortening is most common in severely comminuted fractures. With the advent of statically locked intramedullary nailing the incidence and severity of shortening has significantly reduced. In a meta analysis of 655 fractures treated by intramedullary nailing, shortening greater than 2 cm was present in 19 fractures (3.4%). (Braten, Murphy, Alho, Winquist, Kropf) Of those 19, 16 were in two large studies of only comminuted fractures including 368 fractures. (Alho, Winquist) As opposed to studies where approximately only 55-60% of fractures are comminuted only 3 out of 287 (1%) fractures had shortening greater than 2cm. (Braten, Murphy, Kropfl) This problem of shortening in comminuted fractures has been further addressed by increased use of static locking. Non-operative treatment had the highest incidence of shortening 12%, greater than 2 cm, in a meta analysis of 259 fractures. (Hardy, Suman, Buxton) External fixation also has a high incidence of shortening, this in part is due to the complex nature of fractures treated with external fixation. Gibson et al 1983 reported the influence on the spine of leg length discrepancy after femoral fracture. Forty patients with acquired leg length discrepancy of at least 1.5 cm, as a result of femoral shaft fractures sustained after skeletal maturity but below 21 years of age were studied. Patients with painful joints of the lower limb were excluded leaving fifteen patients. During the ten year period between fracture and review only one patient had worn a shoe raise for a short period of time. Follow up at least ten years later of these fifteen patients showed that in this group acquired leg length discrepancy produced little permanent structural abnormality in the lumbar spine and no degenerative change in the 10 year period after the fracture that led to shortening. No patient complained of low back pain nor had any had significant back pain during the previous ten years. The patients were still young adults at the time of examination and it is possible that structural or degenerative changes would develop. (Gibson) Leg length inequality and the development of chronic low back pain is controversial. Giles et al 1981 studied low back pain and limb length inequality in 217 patients. They found the percentage of control cases with 10 mm or more leg length inequality was 8 %, whereas the percentage of chronic low-back-pain patients with 10 mm or more leg length inequality was 18.3%. They presented this as evidence of the importance of limb length inequality, as an aetiological factor in the production of low-back pain. Only 4 out of the 217 patients having a leg length inequality of greater than 9 mm had previously broken a femur or tibia. Clinical and radiological follow up showed that postural scoliosis did become structural with increasing age. Correction of limb inequality with a shoe raise resulted in a variable correction of the scoliosis. This was age related. A proportion of patient’s symptoms improved with a shoe raise alone. The authors suggested that shoe raise therapy plus manipulation might result in quicker remission of symptoms. (Giles)
Reoperation to achieve healing
Reoperation to achieve healing excluding implant removal and treatment for infection. Rozbruch et al showed a dramatic reduction in reoperation on fractures treated with plating reducing from 43% in the 1970’s to 13% in the 1990’s. This reflects the evolution in plating techniques and improved outcome as experience is gained in a particular technique. Similar trends are seen in the intramedullary nailed group where the overall rate of reoperation on 1004 fractures was 51 (5%). (Wolinsky, Braten, Alho, Kropfl, Hammacher) Later studies in the 90’s show much lower reoperation rates approaching 1%. (Braten) This low reoperation rate following reamed intramedullary nailing can be attributed to lower rates of mal/non-union, routine static locking and less routine dynamization of intramedullary nails.
In the past 90 degrees of knee flexion has been considered an “acceptable” result. The average range of motion required for sitting is 93 degrees, for climbing stairs is 100 degrees, for tying shoes is 106 degrees and for squatting to lift an object is 117 degrees. These figures are averages, so minimal standards must be higher to express a desired functional result for 70-90% of the population. Expressed in convenient ranges:
· 125 degrees or more of flexion enable a patient to squat and carry on most normal daily activities
· 110-124 degrees would permit shoe tying in most patients
· 100-109 degrees generally allows patients to handle stairs and sit comfortably
· Less than 100 degrees often means difficulty with sitting and significant functional loss. (Laros & Spiegel)
The highest incidence of restricted knee movement less than120 degrees is in the non-operative group, 18% (range 0- 40%). (Hardy, Suman, Buxton) Of the non-operative treatment modalities Perkins traction has the lowest knee morbidity with 0% having less than 120 degrees of knee flexion. (Buxton)
Intramedullary nailing has been shown to give a consistently good range of knee movement with only 2% (range 0-3%) having less than 120 degrees knee flexion. (Braten, Grosse, Winquist) In a retrospective comparative study by Rowntree et al of 85 patients treated by three methods: 1. Open intramedullary fixation, 2. Skeletal traction with early knee motion, 3. Skeletal traction with immobile knee. The mean follow up was 5 years (range 2-13.5 years). Thirty per cent of patients had minor knee symptoms. In all cases knee flexion was more than 90 degrees, no statistically significant difference could be demonstrated in the effect on final knee function between intramedullary fixation and skeletal traction. (Rowntree)
Refracture of the femoral shaft is rare. The incidence varies from 2% to10%. (Breederveldt) Theoretically, refracture can only be defined as such when there has been consolidation of the primary fracture. For practical reasons refracture is defined as any fracture in the same bone as the primary fracture without adequate trauma and independent of the primary fracture line or radiographic degree of consolidation. Breederveldt in a retrospective study of 148 femoral shaft fractures presented 19 refractures, mostly after plate fixation. (Breederveldt) Intramedullary nailing leads to less reduction in cortical density than following plating. A mean cortical reduction in bone density of 11% in the plated segment has been reported by Terjesen et al 1985. (Terjesen) Braten et al 1992 reported a mean cortical density reduction of 7 % in the proximal diaphysis, and 6% in the distal diaphysis after reamed intramedullary nailing, only the proximal reduction was statistically significant. (Braten) Intramedullary reaming reduces torsional strength of the femur; even reaming to only 12mm reduces torsional strength by 37.5%. There also appears to be a sharp transition between femurs reamed to 14 mm and 15 mm respectively. A more reliable formula is to try and keep reamed diameter divided by bone diameter (Rd/Bd) less than 0.45. (Pratt)
Table 10. Meta-analysis of complications where information available in respective articles.
|
|
Non-operative
275 Hardy,
Suman, Buxton |
Plate
481 Rozbruch,
Geissler, Magerl,Loomer,Ruedi,Thompson |
External
fixation 98 Volker, Gottschalk, Alonso, Rooser, Murphy |
Intramedullary
nailing 1349 Wolinsky,
Braten, Murphy, Alho, Winquist, Kropfl, Hammacher |
|
Infection
deep |
1/275- (0.4%) *, closer to 0 Hardy, Suman, Buxton |
16/481- (3.3%) Rozbruch, Geissler, Magerl, Loomer, Ruedi,Thompson |
8/98- (8%) Volker, Gottschalk, Alonso, Rooser, Murphy |
10/1349- (0.74%) Wolinsky, Braten, Murphy, Alho, Winquist, Kropfl, Hammacher |
|
Delayed
union |
9/108-(8.3%) Hardy |
14/206-(6.8%) Rozbruch, Ruedi |
6/58-(10%) Alonso, Murphy |
14/416- (3.4%) Braten,Murphy, Hammacher |
|
Non-
union |
0/108-(0%) ** Hardy |
8/274-(3%) Rozbruch, Geissler, Loomer, Thompson |
1/57-(1.7%)** Volker, Alonso, Rooser |
15/1349- (1.1%) ** Wolinsky, Braten, Murphy, Alho, Winquist, Kropfl, Hammacher |
|
Angulation |
>15 degrees- 8/275 (3%) Hardy, Suman, Buxton |
>5 degrees-3/144-(2%) Magerl, Thompson |
>15 degrees- 3/33 (10%) Volker, Rooser |
>5degrees- 60/906= (6.6%), >10 degrees 7/1029- (0.68%) Wolinsky, Braten, Murphy, Alho, Winquist, Kropfl |
|
Rotation |
>20 degrees 4/158 (2.5%) hardy, Buxton |
N/A |
>15 degrees- 1/34 (3%) Volker, Murphy |
>15 degrees 23/473- (4.8%), >20 degrees 6/535- (1%) Braten,Murphy, Alho,Winquist, Hammacher |
|
Shortening |
>2cm- 30/259 (12%) Hardy,Suman, Buxton |
> 2cm- 2/144 (1.4%) Magerl, Thompson |
>2cm- 8/93- (8.6%) Volker, Gottschalk, Alonso, Rooser, Murphy |
>2cm- 19/655 (3%), Braten, Murphy, Alho, Winquist, Kropfl |
|
Re-operation |
N/A |
40/368-(10.9%) Rozbruch, Geissler, Ruedi, Thompson |
4/34- (13%) Murphy |
51/1004-(5%) Wolinsky, Braten, Alho, Kropfl, Hammacher |
|
Knee
movement |
<120 degrees- 46/259 (18%) Hardy, Suman, Buxton |
<120 degrees- 5/215- (2.3%) Geissler, Magerl, Thompson |
<120 degrees- 12/39- (30%) Gottschalk, Volker, Rooser |
<120 degrees- 6/321(1.8%) Braten, Winquist |
|
Re-fracture |
6/275 (2.2%) Hardy, Suman, Buxton |
8/387- (2%) Geissler, Magerl, Loomer, Ruedi, Thompson |
0/18- (0%) Volker |
0/53- (0%) Braten, Kropfl |
|
Mortality |
|
6/251- (2.4%) Loomer, Ruedi, Thompson |
2/54- (3.7%) Volker, Murphy |
14/564- (2.5%) Kropfl, Hammacher, Alho, Braten |
*1 infected tibial pin tact with sequestrum
** Does not include fractures requiring additional procedures to heal.
The true incidence of deep venous thrombosis following fractures of the lower extremity is difficult to assess. Rokkanen et al 1969 reported deep venous thrombosis in 5.8 % of cases. (Rokkanen) Braten et al 1995 reported the incidence of clinical thromboembolic events to be 4.2% (5/120). (Braten) Abelseth et al 1996 in a prospective study demonstrated that the incidence of clinically occult deep venous thrombosis was 28% in lower extremity fractures. He quoted an incidence of 40% (8/20) following femoral shaft fracture. This was however a small study, actively looking for venous thrombosis. Fortunately there are very few clinical episodes of pulmonary emboli. (Abelseth) There is a concern that patients may develop post phlebitic syndrome many years later with its associated morbidity. Aitken et al 1987 was not able to demonstrate a significant incidence of associated post phlebitic syndrome after femoral shaft fractures, however their study included very small numbers. Most of the literature has concentrated on tibial fractures. The postphlebitic limb syndrome takes a long time to develop 13% at 3years, 35% at 9 years, 39% at 14 years. Therefore patients who develop a deep venous thrombosis should be of be cautious accepting compensation concluded in “ full and final assessment”, as the syndrome may take years to fully develop. (Aitken)
Heterotopic ossification about the hip after intramedullary nailing of femoral shaft fractures is rarely commented on. The incidence is seldom quoted in articles on femoral shaft fractures. Mild to moderate heterotopic ossification has been shown to be compatible with excellent function of the hip. Brumback et al 1990 studied 100 consecutive patients with unilateral femoral shaft fractures, in a prospective randomized trial to delineate the incidence and factors predisposing to heterotopic ossification about the hip. Twenty patients were excluded leaving 80 patients. In 32 (40%) no heterotopic ossification developed, whereas mild ossification developed in 27 patients (34%). Moderate ossification developed in 12 patients (15%) and severe ossification, in 9 patients (11%). They were unable to identify any specific factors related to the patient, fracture, timing or type of intramedullary nailing and development of heterotopic ossification. They randomized the patients into two groups, in group 1, the operative incision was irrigated with 250 milliliters of saline, and group 2 the operative incision was irrigated with 3000 milliliters of normal saline with use of pulsatile lavage. Irrigation with pulsatile lavage did reduce the amount of mild and moderate heterotopic ossification. It did not however reduce the only clinical significant ossification, namely severe heterotopic ossification. (Brumback)
Pudendal nerve palsy is noted as an infrequent complication of traction, compression following intramedullary nailing of the femoral shaft. Presenting with numbness of the penis and scrotum or Labia and rarely erectile dysfunction in men. When specifically looked for the incidence is as much as 10%-17%. In a prospective study of 106 patients 10 (9%) (6 men, 4 women) had a pudendal nerve palsy. Only one of the men complained of erectile dysfunction. The symptoms resolved completely in 9 out of 10 patients within 1-11 weeks (average 4 weeks). One man still complained of altered sensation in the penis and scrotum at 6 months.
The duration of procedure did not predict onset of symptoms. Statistical analysis revealed the magnitude of total traction forces for the duration of the procedure was significantly higher in the group who developed a pudendal nerve palsy. However some patients who received high pressures for the duration of the procedure did not get palsies, similarly some patients who received very low traction pressures for the duration developed palsies. It is thought that the anatomy and thin post play an important role. Hip adduction time should also be limited as much as possible, i.e. relaxing traction and abducting the hip while closing the wounds may reduce the incidence of these iatrogenic injuries. (Brumback)
Carlson et al 1995 has reported peroneal
nerve palsy and compartment syndrome in bilateral femoral fractures. This is a
rare complication related to patient positioning. Placing one of the fractured
limbs in the hemilithotomy position, usually the left leg. It is thought the
posterior angulation and or added mobility of the leg may stretch the peroneal
nerve. The calf supports all the weight of the leg, because the fractured femur
cannot provide support. (Carlson)
Rokkanen et al 1969 comparing open and
closed intramedullary nailing to conservatively treated case, showed earlier
return to normal function and work with operative treatment. They showed the
greatest benefit to be gained with operative intervention was in the older age
groups with complicated fractures. Only a marginal benefit was derived with
closed versus open intramedullary nailing. They also showed a curious
difference of two weeks, in fractures treated within 6 days; interval for
return to work was 32 weeks (39 cases) as opposed to 30 weeks in 49 cases when
the nailing was performed 7 to 21 days post fracture. (Rokkanen)
In a study by Bednar et al 1993 of 47 patients with isolated femoral shaft
fractures treated by intramedullary nailing. Forty one patients were available
for analysis of function: 33 (80%) had returned to full time occupation in
their original occupation, 4 (10%) were working in other full time employment;
3(7%) were working part time and only 1 (2%) was disabled and unable to return
to work. (Bednar)
Benirschke et al 1993 reported on a subset of 56 patients out of 144 available for analysis (treated with closed locked intramedullary nailing). They found at a minimum of 12 months follow up that 39% had some limitation in ability to ambulate or stand and 9 % had to obtain new employment or seek job modifications. (Benirschke)
Alho et al 1991 reported on 120 patients
with 123 fractures treated with intramedullary nailing, with a mean follow up
of 20 months (range, 12-60 months). He found the working ability of 18(15%)
patients was reduced at follow up examination. Four patients (3%) had stopped
their sports activity and 31(26%) had various degrees of reduction of activity.
(Alho)
Kropfl et aI 1995 documented a subset of patients with workers compensation treated with plate osteosynthesis, all patients returned to their preoperative level of activity within a mean period of 3.25 months. (Kropfl) Loomer et al 1980 reporting on 45 patients with 46 fractures treated with plating stated that the average patient healed their fracture in 7 months, returned to work at 12 months and returned to sports activities at 15 months. (Loomer)
Table 11. Employment following femoral shaft fracture
|
|
Hardy
1983 |
Mira A
J 1980 |
Bednar
D A 1993 |
|
Number
patients |
106 |
29 |
41 |
|
Age |
77%<30 years |
25.1 (16-64) |
22 |
|
Treatment |
Cast bracing |
Mixed |
Nailing |
|
Open |
16 (15%) |
6 (21%) |
9 (19%) |
|
Comminuted |
85 (79%) |
12 (41%) |
N/A |
|
Non
operative |
106 (100%) |
19 (66%) |
- |
|
Plating |
- |
2 (7%) |
- |
|
External
fixation |
- |
- |
- |
|
Intramedullary
nailing |
- |
8 ( 27%) |
22 (100%) |
|
Follow
up |
|
37.3 months (16-69) |
34.5 months |
|
%
return to work |
75 (71%) |
* |
33 (80%) |
|
Time
to work |
30 weeks (6.9 months) |
7.5 months (3-19) |
4 months |
|
|
~ |
* |
+ |
~ Seventy-five (71%) returned to work within 30 weeks, fourteen were retired or unemployed. The other seventeen were either living in institutions or attending a school or university where they were continuing their studies by correspondence, or had abandoned their studies until the start of the next academic year.
* This study was of patients who had returned to work. The only significant difference in time of return to work was in simple proximal (5.4 months) versus comminuted distal (7.8 months) and between treatment by open reduction and internal fixation (5.3 months) versus a spica cast (8.2 months) (Mira)
+ Of the patients not returned to preinjury employment, 4 patients (10%) were working full time in different employment, 3 (7%) were working part time. Only 1 (2%) remained disabled at the time of follow up. (Bednar)
Thigh muscle function after operative fixation of femoral shaft fractures varies depending on the operative form of fixation. Studies where the uninjured side is used as control need to be viewed with caution, as it has been shown that extension power and flexion power in the uninjured leg increase on average by between 11% and 17% after two years. Similarly in a group of normal volunteers the average quadriceps strength between one side and the other differed by 6%, with a standard deviation of 4%. This suggests a significant mean variation from normal needs to be more than 10%. (Mira) Zdravkovic 1978 showed that closed intramedullary nailing versus open intramedullary nailing resulted in better quadriceps function. (Zdravkovic) Finsen et al 1993 reported no loss in knee extension strength at least 23 months post intramedullary nailing, compared with a 12% reduction in knee extension strength of patients treated with plate osteosynthesis. (Finsen) Hamstring power is also significantly reduced after operative fixation of femoral fractures, this appears to be unaffected by the type of operation. (Finsen)
Published complications following
intramedullary nailing are low. In 1988 Christie et al reported their results
of intramedullary nailing of femoral fractures. They reported 4 perioperative
deaths, but stated that 3 of these patients were elderly. In 1984, Winquist et
al published a major study on 500 fractures treated with closed intramedullary
nailing; the average age was 30 years, with a wide range 10 to 92 years. The
nonunion rate was low 0.9%. However, of the 4 nonunions, 3 of them occurred in
patients older than 60 years of age. Bouchard et al 1996 retrospectively
reviewed the outcome of 138 femoral shaft fractures in the elderly, average age
79 (range 65-98). Low velocity injuries
caused 92% of the fractures, only 10% had associated injuries. Concomitant
musculoskeletal illnesses were found in 62% of cases, cardiopulmonary disease
in 44% of patients and neuromuscular conditions in 22% of patients. Only 12
patients (9%) were in perfect health. Treatment varied, 29 non-operative and
109 were treated surgically. Sixty-three patients had complications related to
the fracture or the treatment for a complication rate of 46%. Twenty-eight
patients died within 6 months giving a perifracture mortality rate of 20%. Most
deaths occurred early in the first month. The age and mental status on
admission to hospital were the main determinants for patient survival. The
number and type of medical illnesses did not influence the survival of the
patient nor did the type of treatment. Ninety patients were able to ambulate on
average 16 days after admission to the hospital. Forty-eight patients (35%)
were unable to ambulate after the fracture and remained bedridden. (Bouchard)
High-energy injuries carry increased morbidity due to the higher incidence of open fractures, higher incidence of associated injury, and higher incidence of ipsilateral injuries. Taylor et al 1994 showed union time to be affected by mechanism of injury irrespective of treatment. Motorcycle and car accidents were the slowest to unite; taking 14.3 weeks to unite compared with 11.1 weeks following falls. Motorcycle, car accidents and lower third fractures had the worst knee flexion. Shortening was not statistically affected by the initial mechanism, but was worst with traction. (Taylor)
Non-operative treatment by traction does not risk infection, but results in longer hospitalization, slower ambulation, more angulation and a higher frank non-union rate. Cast bracing shortens hospital stay and time to ambulation, but has the highest incidence of limb shortening
Operative treatment offers significant benefits in terms of mortality and morbidity in the critically ill multiply injured patient. (Fakhry)
Internal fixation with plate osteosynthesis is especially useful in proximal and distal fractures where intramedullary nailing is less effective. Plating provides rigid fixation and anatomic reduction, good motion, low risk of shortening and angulation, but risks the highest incidence of infection and implant failure.
External fixation in the adult is generally reserved for complex, high energy injuries, prone to complications.
Closed reamed intramedullary nailing provides short hospitalization, early ambulation and return to work, good maintenance of bone length, minimal angulation and a high percentage of patients with good knee flexion, but risks infection.
A number of studies have supported early aggressive surgical management of the patient with multiple injuries and long bone fractures. A strong relationship exists between early fixation of femoral fractures and reduction in the number of days in the intensive care unit; the number of days in hospital; and the rates of adult respiratory distress syndrome, pneumonia, infection and death. (Charash, Bone, Johnson, Behrman) Studies in the 1960’s and 1970’s by Charnley, Guindy and Wilber, suggested that fractures operated on early (0-6 days after fracture) compared with those operated on late (more than 6 days) had a higher incidence of non-union, delayed union and re-fracture. Modern day studies refute these claims and show no benefit from delaying surgery. (Pahud) In fact to the contrary they show significant benefit in early surgery (within 24-48 hours). The patient who has the most to gain is the critically ill multiply injured patient. (Bone, Behrman)
In a prospective randomized study by Bone et al 1989 of 178 patients who were randomized into an early fixation group (the first 24 hours) or a late stabilization group (more than 48 hours after injury). They found that delayed fixation greater than 48 hours, increased hospital stay and intensive care stay by 5 and 10 days respectively. (Bone) Johnson et al 1985 documented a five fold increased incidence of adult respiratory distress syndrome if internal fixation was delayed by 24 hours. (Johnson) Recent literature has attempted to further break down who does, or does not benefit from early surgical intervention. Pelias et al 1992 in a retrospective study of 130 consecutive patients with major blunt chest injury showed that patients with major chest injury and long bone fractures had a high propensity toward pulmonary complications. They also failed to show any significant improvement in the high pulmonary morbidity despite optimal early operative fixation. (Pelias) Pape et al 1993 challenged the philosophy of early fixation of femoral fractures using reamed intramedullary nails in the multiply injured patient. Concluding that in the presence of a thoracic injury, primary intramedullary nailing with reaming of the femur causes additional pulmonary damage and may trigger the development of adult respiratory distress syndrome. (Pape) Charash et al 1994 repeated the study design of Pape et al with similarly sized groups of patients who had a femoral fracture and a thoracic injury, and they reported contradictory findings. Charash et al reported a 48% rate of pneumonia in twenty-five patients who were managed with delayed fixation, compared with a 14% rate of pneumonia in fifty-six patients who were managed with early fixation. The overall rate of pulmonary complications (pneumonia, adult respiratory distress syndrome, fat embolism, and pulmonary embolism) was 56% in the patients who had delayed fixation, compared with 16% in those who had early fixation. (Charash) These findings of lower pulmonary morbidity following early fixation were supported by later retrospective studies done by Van der Made et al 1996. (Van der Made) Bosse et al 1997 strongly supported early operative fixation of femoral fractures even in the presence of major thoracic injury (Bosse) and Carlson et al 1998 concluded that reamed intramedullary femoral fixation did not increase pulmonary morbidity in chest injured patients. (Carlson)
Intramedullary nailing of fractured femurs has not been shown to be associated with any increase in neurological disability. (Mckee, Starr) Provided careful attention to prevention of secondary insults to the brain is applied during surgery, i.e. avoiding intraoperative hypoxia, hypotension and maintaining adequate cerebral perfusion pressure. Townsend et al 1998, showed patients having intramedullary nailing 0-2 hrs post admission, were 8 times more likely to become hypotensive during femur repair than patients in the >24 hr group. He therefore suggested a short delay to allow for adequate resuscitation and stabilization. (Townsend) Starr et al 1998 showed that delay in stabilization appeared to increase the risk of pulmonary complications and that early fracture stabilization did not increase the prevalence of CNS complications. (Starr)
Intramedullary nail removal is performed in 20% to 38% of fractures treated with intramedullary nailing. The most common indication for removal is pain (63%). The suggested time prior to removal is at least 18-24 months and it is important to protect the femur from heavy load bearing (sporting activities) for three months after removal. (Alho, Wolinsky, Breederveld) Braten et al found an increased incidence of hip and knee pain in patients, where the nail had not been removed; they also quote no re fracture in 47 femurs after nail removal. (Braten) Miller et al 1992 reported on the removal of 60 intramedullary rods in 58 patients. All the fractures had healed at the time of rod extraction. The implant was removed from 34 asymptomatic patients (36 femurs). Twenty-four patients had preoperative symptoms attributed to the to the femoral rod. Problems encountered after extraction were broken rods in 5 femurs and 6 patients required further hospital care for postoperative haematomas. There were no re fractures in this group. Miller et al advised against rod removal in asymptomatic patients. (Miller)
See Separate sheets.