- Email: [email protected]
Femoral neck fracture: sliding screw plate versus sliding nail plate—a randomized trial
Injury (1985) 16,449-454
Printedin
Great Britain
449
Femoral neck fracture: sliding screw plate versus sliding nail plate-a randomized trial Peter Nordkild, Department
Stig Sonne-Holm
of Orthopaedic
Surgery,
and Jgrgen Gentofte
S. Jensen
Hospital,
In a randomized trial 49 patients with fracture of the neck of the femur and an age of less than 70 years or a high level of physical activity were allocated to treatment with a sliding screw plate or a sliding nail plate fixation. The patients were followed for 2-5 years. At follow-up the union rate was found to be 86.2 per cent of the fractures in the screw plate group and 73.7 per cent in the nail plate group (KO.3). Necrosis of the femoral head was encountered in respectively 10 and 21 per cent. Hip replacement was necessary in respectively 23.3 and 31.6 per cent. The nail slid out of the femoral head, resulting in recurrence of the fracture’s displacement in three fractures with a sliding nail plate, and in none with a sliding screw plate (P=O.O53). Secondary loss of the femoral neck’s angle was more frequently seen in the sliding nail plate group
(RO.01). In conclusion, the sliding screw plate gives better fixation of fractures of the neck of the femur and was followed by a lower frequency of reoperation than after an unthreaded device. INTRODUCTION FRACTURE of the neck of the femur is one of the most common fractures in the aged. As a consequence of the increasing number of old people in our society and the rising incidence of these fractures, which cannot be explained by an increase in the number of old people alone (Frandsen and Kruse, 1983), we must expect a substantial rise in the number of these fractures. Fifty years after Smith-Petersen et al. (1931) introduced a nail for internal fixation of fractures of the neck of the femur, the main complication is still the high incidence of necrosis of the head of the femur. The optimal method for fixation of these fractures is still being sought, and the fact that no single universally accepted device exists is reflected in the number of different methods of fixation; a recent review listed 76 (Tronzo, 1974). The principle of a sliding type of implant was introduced in the 1940s and this principle became very popular with the Pugh nail (Pugh, 1955). The sliding nail plate was further developed by Clawson (1964), who introduced the sliding screw plate in 1964. In this paper sliding nail plate fixation has been compared with sliding screw plate fixation in a randomized and prospectively conducted study including a follow-up of up to 5 years.
PATIENTS
All patients with non-pathological fractures of the neck of the femur and an age of less than 70 years or a high
Copenhagen,
Denmark
level of physical activity treated in the orthopaedic departments at the County Hospital in Gentofte in the period from January 1978 to December 1980 were included in this study. Patients older than 70 years and in poor general condition were treated with a Moore’s prosthesis. There were 49 patients with fracture of the neck of the femur including basal cervical fracture. For the allocation, odd and even numbers were used. Thirty patients were allocated to the operation using sliding screw plate fixation and 19 patients were allocated to sliding nail plate fixation (unfortunately the allocation resulted in an uneven distribution). There were 16 men with a median age of 60 years at the time of fracture (range 32-86) and 33 women with a median age of 63 years at the time of fracture (range 49-73). The male: female ratio was 1: 2.1. In Table 1 the distribution of preoperative variables in the two groups is presented. There were no significant differences in the distribution of sex or fracture stage. There was a significant difference in the distribution of age groups (RO.025. Table /), with more patients under 60 years of age in the sliding nail plate group than expected. Taking into account the fact that younger people have a higher incidence of fracture of the neck of the femur of Garden’s stages 3 and 4 (Nordkild and Sonne-Helm, 1984), this uneven distribution could have produced a bias, but there was no Table 1. Distribution of preoperative variables in 30 fractures of the neck of the femur operated on with sliding screw plate and 19 fractures operated on with sliding nail plate Sliding
screw plate Male Female Age (60 years Age >60 years Garden’s stages 1 and 2 Garden’s stages 3 and 4 Basal and transcervical fractures Subcapital fractures
Sliding nail plate
7
9 21 8 22 4t 23t 19t
12*
8t
5$
12 12 7 2*
15*
P value*
<0.8 CO.99
l Chi-squared test, DF 1. t In three fractures (11.1 per cent) the preoperative radiographs
were not available at follow-up. $ In two fractures (10.5 per cent) the preoperative radiographs were not available at follow-up.
450
Injury: the British Journal of Accident
significant difference stages. METHODS Classification
in the distribution
of Garden’s
The fractures were classified according to Garden’s classification (1961): stage 1, incomplete fracture; stage 2, complete fracture without displacement; stage 3, complete fracture with partial displacement; stage 4, complete fracture with full displacement. The fractures were further divided into subcapital, transcervical and basal cervical types. Time of operation
Operation was not performed as an emergency, but as soon as possible and no later than 48 hours after admission to the hospital. Surgical
technique
and postoperative
regimen
The implants were manufactured by Howmedica from a cobalt-chromium-molybdenum alloy (Fig. 1). On admission to the hospital traction was applied through the upper tibia to all displaced fractures. The patient was placed on the fracture table in the supine position and final reduction was made under general/ spinal anaesthesia with the assistance of an image intensifier. The operations were carried out by a number of surgeons. The implant was inserted through a lateral incision at the proximal end of the femur. As steep an angle as possible was attempted with support by the calcar. An attempt was made to place the proximal tip of the implant in the central third of the head of the femur in the frontal as well as the lateral view, and between 0.5 and 1 cm from the articular surface of the femoral head. After the operation early walking was encouraged. Follow-up
At the time of follow-up the patients were examined in the outpatient clinic, and this included an X-ray examination of the hip in the frontal and lateral views. The patients were questioned about pain and walking ability. The range of movement in the affected hip was examined by the authors according to Merle d’AubignC’s hip evaluation scales (Table II). Patients
(1985) Vol. 16/No. 7
ending up with a hip prosthesis were not examined. The median time of follow-up was 40 months (range l-64). Examination
of fractures
Surgery
of radiographs
The radiographs were examined by the authors simultaneously without knowledge about the clinical result of the matching patient. The reduction of the fracture was regarded as good if the angle between the medial trabeculae and the medial femoral cortex (normally approximately 160”) in the frontal view was between 155 and 165” and the head of the femur was erect in the lateral view. Acceptable reduction was defined as the head of the femur being displaced in only one view, and unacceptable reduction as the head of the femur being displaced in both views. The position of the tip of the implant in the head of the femur was good when it was in the central third in both views. An unacceptable position was when the implant was in the upper and/or anterior third of the head of the femur. The remaining positions were acceptable. It was recorded whether or not the tip of the implant had perforated the articular surface. The group without perforation was divided into two groups, according to whether the tip of the implant was more or less than 1 cm from the articular surface. The fixation index (defined as the length of the part of the implant positioned in the capital fragment in relation to the length of the capital fragment) was recorded: good, >OS; acceptable, 06-0+3; unacceptable, <0.6. The telescoping effect, judged by the postoperative films and the most recent radiographs, was estimated with correction for enlargement of the radiographs and divided into three groups: (1) a telescoping of less than 1 cm; (2) a telescoping between 1 and 2 cm; (3) a telescoping of more than 2 cm. Migration of the implant, vans of more than lo” and failure of screws in the cortex were noticed. Necrosis of the head of the femur was noted, when there was at least one of the following changes: (1) subtle diffuse mottling of the bone; (2) isolated patches of sclerosis; (3) irregular discrete rarefaction with fuzzy marginal sclerosis; (4) subchondral cortical depression; (5) generalized subchondral fragmentation; or (6) periosteal bone apposition along the femoral neck. These definitions are modified after Martel and Sitterley (1969). The pre- and postoperative radiographs and the films obtained at the end of follow-up were examined. In the patients receiving a hip prosthesis, the last films obtained before reoperation were examined. In six patients who had died during the period of follow-up the most recent films obtained before death were examined.
Statistical
methods
The chi-squared test with one degree of freedom (DF 1) was used for all statistical testing except in testing the distribution of cases, where the implant telescoped enough to come out of the head of the femur. In this case Fisher’s exact probability test was used. RESULTS Fig. 1. The two Howmedica implants used in this series are shown: the sliding screw plate (top) and the sliding nail plate
(bottom).
There was no significant difference in distribution of postoperative variables such as reduction of fracture, position of fixation appliance in the head of the femur,
451
Nordkild et al.: Femoral neck fractures
Table II. Merle d’Aubign6 (1954) method Revue d’orthopkdie (1949) 35, 542)
Pain is intense
7
Pain is severe even at night
2
Pain
3
and permanent
is severe when walking; prevents any activity Pain is tolerable with limited activity
value of hip (Adapted
Ability
Mobility
Pain 0
of grading functional
Ankylosis with bad position of the hip No movement; pain or slight deformity Flexion under 40 Flexion between
40 and 60
4
Pain is mild when walking; it disappears with rest
Flexion between 60 and 80”; patient can reach his foot
5
Pain is mild and inconstant; normal activity No pain
Flexion between 80 and 90”; abduction of at least 15” Flexion of more than 90”; abduction to 30”
6
including the distance between the tip of the implant and the articular surface, and the fixation index (Table rrr>. We found union, implying bone union of the fracture with radiographically visible trabeculae across the fracture line, in 86.2 per cent (25/29) of the fractures with a sliding screw plate and in 73.7 per cent (14/19) of the fractures with a sliding nail plate (RO.3, Table ZV). One patient in the screw plate group died on the 16th day after fixation. In the group of ununited fractures, we encountered three hips, all with a sliding nail plate, in which the sliding nail had slid out of the head, resulting in redisplacement of the fractures. No such technical failure was encountered in the sliding screw plate group (P=O.O53, Table IV). Ten per cent (3/30) of the fractures with a sliding screw plate developed necrosis of the head of the femur with a median time after fixation of 33 months (range
l
Chi-squared test, DF 1.
to walk
None Only with crutches Only with canes With one cane, less than 1 hour; very difficult without a cane A long time with a cane; short time without cane and with limp Without cane but with slight limp Normal
33-34). All three of these fractures had previously shown bone union. In the sliding nail plate group 21 per cent (4/19) developed necrosis of the head of the femur with a median time after fixation of 42 months (range 2-50). Seventy-five per cent (3/4) of these fractures had previously shown bone union. This difference was not significant (Table IV). During the period of follow-up, seven patients (23-3 per cent) from the sliding screw plate group received prosthetic hips. In four of these patients the indication for reoperation was non-union and in the remaining three patients capital necrosis. Six patients (31.6 per cent) with a sliding nail plate had a further operation. This was due to non-union in four, necrosis of the head of the femur in one and the combination of capital necrosis and non-union in one. In the group with a sliding screw plate 75.9 per cent (22/29) of the fractures healed without complications; 52.6 per cent (1009) of the fractures with a sliding nail
Table 111.Distribution of postoperative variables in 30 fractures of the neck of the femur operated on with sliding screw plate and 19 fractures operated on with sliding nail plate
Good reduction Fair and not acceptable reduction Good position of implant Fair and not acceptable position of implant Distance from articular surface of 1 cm or less Distance from articular surface of more than 1 cm or with primary perforation Good fixation index Fair and not acceptable fixation index
from Table I,
Sliding screw plate
Sliding nail plate
11 19
7 12
CO.99
22 8
14 5
co.99
21
11
9
8
20 10
8 11
P value *
452
Injury: the British Journal of Accident Surgery (1985) Vol. 16/No. 7
Table IV. Results and with
sliding
of operation with sliding screw plate in 30 fractures nail plate in 19 fractures of the neck of the femur
Sliding Sliding screw p/are nail plate Non-union of fracture Union of fracture Implant telescoping out of femoral head Implant remaining in femoral head Necrosis of the femoral head No capital necrosis Alloplastic operation No alloplastic operation Fractures with complications Fractures without complications Varus >lO” No varus or varus
Chi-squared
test,
DF
4 25t 0
5 14 3
30
16
3 27 7 23 7t 22 2 28 3 27
4 15 6 13 9 IO 7 12 4 15
P value* CO-3 0.053
1.
t One patient was excluded because of death 16 days postoperatively.
healed without complications (P~0.1, Table IV). During the period of follow-up two fractures from the sliding screw plate group and seven fractures with a nail plate developed a varus deformity of more than 10”. This difference was highly significant (P
plate
postoperative pain than were the patients in the nail plate group (RO.005, Table V). The hips’ mobility was equally good in both groups. There was no significant (P
Table V. Distribution of clinical results according to the hip evaluation scales of Merle d’Aubign6 (Table II) in 30 fractures of the neck of the femur operated on with sliding screw plate and 19 fractures operated on with sliding nail plate Sliding screw plate* Pain score of less than 6 Pain score of 6 Hip mobility score of less than 6 Hip mobility score of 6 Walking ability score of less than 6 Walking ability score of 6
Sliding nail platet
4 16 0
0
20 7
12 2
13
10
P valueS CO.005
* Five patients who had died during the period of follow-up and five patients who had had an alloplastic operation were not examined. t One patient who had died during the period of follow-up, five patients who had had an alloplastic operation and one patient who refused to participate in this examination were not examined. $ Chi-squared test, DF 1.
Nordkild
et al.: Femoral
neck fractures
453
a
Fig. 2. a, Anteroposterior radiograph of the pelvis of a 64-year-old man showing a Garden’s stage 3, subcapital fracture of the right femur. 6. Postoperative anteroposterior radiograph of the right hip showing the fracture reduced with the sliding nail plate inserted in an accept ble position. c, Anteroposterior radiograph of the right hip obtained 40 days postoperatively. The nail has telescoped out o% the head of the femur resulting in recurrence of the fracture displacement.
454
years, we have been able to find only one previous study (Frandsen and Andersen, 1981) with a random allocation of patients to two different types of fixation. In this series we found union of 86.2 per cent of the fractures with a sliding screw plate, whereas only 73.7 per cent of the fractures from the nail plate group showed union. In an earlier report (Nordkild and Sonne-Holm, 1984) union was found in 87.0 per cent of fractures operated with a sliding screw plate. Barnes et al. (1976) found union in 75.6 per cent using the sliding nail plate. This compares well with a 75 per cent rate of union in fractures of Garden’s stages 3 and 4 with a sliding nail plate (Frandsen and Andersen, 1981). Although the difference in frequency of union found in this series is not statistically significant (RO.3, Table IV), it shows a tendency towards the screw plate having a higher rate of union. In three out of five ununited fractures operated on with sliding nail plates, the reason for non-union was a serious technical failure. In these fractures the nail telescoped and slid out of the head of the femur, resulting in redisplacement of the fracture (Fig. 2). The time for these events was approximately 1 month. This technical failure appeared in all three cases within 3 months of the operation. No such technical failure was found in the screw plate group in this series, but it has been reported earlier (Nordkild and Sonne-Holm, 1984) with a frequency of 2.7 per cent. Although this difference is only of doubtful significance (P=O.O53, Table IV), it suggests that the sliding nail plate should be abandoned. The frequency of secondary varus displacement of the fracture was significantly higher in the nail plate group (RO.01, Table IV), and in addition to the rate of technical failure in the nail plate group mentioned above, this shows that the sliding nail plate does not provide reliable fixation for fractures of the neck of the femur. Necrosis of the head of the femur is still a feared and unexplained complication of these fractures. In this series, necrosis of the head was encountered in 12 per cent of the united fractures with a sliding screw plate. Necrosis of the head was found in 21.4 per cent of the healed fractures from the nail plate group. In an earlier report on sliding screw plates (Nordkild and SonneHolm, 1984) there was necrosis of the head in 16.3 per cent of healed fractures. Barnes et al. (1976) found that 24 per cent of fractures with a sliding nail plate, which had previously shown bone union, developed necrosis of the head of the femur within 3 years. Frandsen and Andersen (1981) found necrosis in 21 per cent of fractures of Garden’s stages 3 and 4 with a sliding nail plate. An explanation for this difference could be that there is less damage to the blood vessels of the head of the femur following screwing than after nailing (Stramquist, 1983).
Injury: the British Journal of Accident Surgery (1985) Vol. 16/No. 7
The mobility of the hip with healed fractures and without necrosis of the head was the same in the two groups. The difference in walking ability was not significant. We found that patients with sliding nail plates were significantly (P
REFERENCES
Barnes R., Brown J. T., Garden R. S. et al. (1976) Subcapital fractures of the femur. A prospective review. J. Bone Joint Surg. 58B. 2. Clawson D. K. (1964) Trochanteric fractures treated by the sliding screw-plate fixation method. J. Trauma 4, 737. d’Aubign6 M. and Pastel M. (1954) Functional results of hip arthroplasty with acrylic prosthesis. 1. Bone Joini Surg. 36A, 451.
Frandsen P. A. and Andersen P. E. (1981) Treatment of displaced fractures of the femoral neck. Smith-Petersen osteosynthesis versus sliding nail-plate osteosynthesis. Acta Orthop. &and.
52, 547.
Frandsen P. A. and Kruse T. (1983) Hip fractures in the county of Funen, Denmark. Implications of demographic ageing and changes in incidence rates. Acta Orthop. &and. 54, 681.
Martel W. and Sitterley B. H. (lY6Y) Roentgenological manifestations of osteonecrosis. Am. J. Roentgenol. 106, 509.
Nordkild P. and Sonne-Holm S. (1984) Sliding screw-plate for internal fixation of femoral neck fractures. Actu Orthop.
Stand.
55, 616.
Pugh W. L. (1955) A self-adjusting nail-plate for fractures about the hip joint. J. Bone Joint Surg. 37A, 1085. Smith-Petersen M. N., Cave E. F. and Vangorder G. W. (1931) Intracapsular fractures of the neck of the femur. Arch.
Surg. 23, 715.
Striimquist B. (1983) Femoral head vitality after intracapsular hip fracture. Acta Orthop. Stand. Suppl. 200. Tronzo R. G. (1974) Hip nails for all occasions. Orthop. Clin. North Am. 5, 479.
Paper accepted 8 November
1984.
Reyuesrs for reprints should be addressed lo: Stig Sonne-Holm, L. E. Bruunsvej 27, DK-2900 Charlottenlund,
Denmark
Printedin
Great Britain
449
Femoral neck fracture: sliding screw plate versus sliding nail plate-a randomized trial Peter Nordkild, Department
Stig Sonne-Holm
of Orthopaedic
Surgery,
and Jgrgen Gentofte
S. Jensen
Hospital,
In a randomized trial 49 patients with fracture of the neck of the femur and an age of less than 70 years or a high level of physical activity were allocated to treatment with a sliding screw plate or a sliding nail plate fixation. The patients were followed for 2-5 years. At follow-up the union rate was found to be 86.2 per cent of the fractures in the screw plate group and 73.7 per cent in the nail plate group (KO.3). Necrosis of the femoral head was encountered in respectively 10 and 21 per cent. Hip replacement was necessary in respectively 23.3 and 31.6 per cent. The nail slid out of the femoral head, resulting in recurrence of the fracture’s displacement in three fractures with a sliding nail plate, and in none with a sliding screw plate (P=O.O53). Secondary loss of the femoral neck’s angle was more frequently seen in the sliding nail plate group
(RO.01). In conclusion, the sliding screw plate gives better fixation of fractures of the neck of the femur and was followed by a lower frequency of reoperation than after an unthreaded device. INTRODUCTION FRACTURE of the neck of the femur is one of the most common fractures in the aged. As a consequence of the increasing number of old people in our society and the rising incidence of these fractures, which cannot be explained by an increase in the number of old people alone (Frandsen and Kruse, 1983), we must expect a substantial rise in the number of these fractures. Fifty years after Smith-Petersen et al. (1931) introduced a nail for internal fixation of fractures of the neck of the femur, the main complication is still the high incidence of necrosis of the head of the femur. The optimal method for fixation of these fractures is still being sought, and the fact that no single universally accepted device exists is reflected in the number of different methods of fixation; a recent review listed 76 (Tronzo, 1974). The principle of a sliding type of implant was introduced in the 1940s and this principle became very popular with the Pugh nail (Pugh, 1955). The sliding nail plate was further developed by Clawson (1964), who introduced the sliding screw plate in 1964. In this paper sliding nail plate fixation has been compared with sliding screw plate fixation in a randomized and prospectively conducted study including a follow-up of up to 5 years.
PATIENTS
All patients with non-pathological fractures of the neck of the femur and an age of less than 70 years or a high
Copenhagen,
Denmark
level of physical activity treated in the orthopaedic departments at the County Hospital in Gentofte in the period from January 1978 to December 1980 were included in this study. Patients older than 70 years and in poor general condition were treated with a Moore’s prosthesis. There were 49 patients with fracture of the neck of the femur including basal cervical fracture. For the allocation, odd and even numbers were used. Thirty patients were allocated to the operation using sliding screw plate fixation and 19 patients were allocated to sliding nail plate fixation (unfortunately the allocation resulted in an uneven distribution). There were 16 men with a median age of 60 years at the time of fracture (range 32-86) and 33 women with a median age of 63 years at the time of fracture (range 49-73). The male: female ratio was 1: 2.1. In Table 1 the distribution of preoperative variables in the two groups is presented. There were no significant differences in the distribution of sex or fracture stage. There was a significant difference in the distribution of age groups (RO.025. Table /), with more patients under 60 years of age in the sliding nail plate group than expected. Taking into account the fact that younger people have a higher incidence of fracture of the neck of the femur of Garden’s stages 3 and 4 (Nordkild and Sonne-Helm, 1984), this uneven distribution could have produced a bias, but there was no Table 1. Distribution of preoperative variables in 30 fractures of the neck of the femur operated on with sliding screw plate and 19 fractures operated on with sliding nail plate Sliding
screw plate Male Female Age (60 years Age >60 years Garden’s stages 1 and 2 Garden’s stages 3 and 4 Basal and transcervical fractures Subcapital fractures
Sliding nail plate
7
9 21 8 22 4t 23t 19t
12*
8t
5$
12 12 7 2*
15*
P value*
<0.8 CO.99
l Chi-squared test, DF 1. t In three fractures (11.1 per cent) the preoperative radiographs
were not available at follow-up. $ In two fractures (10.5 per cent) the preoperative radiographs were not available at follow-up.
450
Injury: the British Journal of Accident
significant difference stages. METHODS Classification
in the distribution
of Garden’s
The fractures were classified according to Garden’s classification (1961): stage 1, incomplete fracture; stage 2, complete fracture without displacement; stage 3, complete fracture with partial displacement; stage 4, complete fracture with full displacement. The fractures were further divided into subcapital, transcervical and basal cervical types. Time of operation
Operation was not performed as an emergency, but as soon as possible and no later than 48 hours after admission to the hospital. Surgical
technique
and postoperative
regimen
The implants were manufactured by Howmedica from a cobalt-chromium-molybdenum alloy (Fig. 1). On admission to the hospital traction was applied through the upper tibia to all displaced fractures. The patient was placed on the fracture table in the supine position and final reduction was made under general/ spinal anaesthesia with the assistance of an image intensifier. The operations were carried out by a number of surgeons. The implant was inserted through a lateral incision at the proximal end of the femur. As steep an angle as possible was attempted with support by the calcar. An attempt was made to place the proximal tip of the implant in the central third of the head of the femur in the frontal as well as the lateral view, and between 0.5 and 1 cm from the articular surface of the femoral head. After the operation early walking was encouraged. Follow-up
At the time of follow-up the patients were examined in the outpatient clinic, and this included an X-ray examination of the hip in the frontal and lateral views. The patients were questioned about pain and walking ability. The range of movement in the affected hip was examined by the authors according to Merle d’AubignC’s hip evaluation scales (Table II). Patients
(1985) Vol. 16/No. 7
ending up with a hip prosthesis were not examined. The median time of follow-up was 40 months (range l-64). Examination
of fractures
Surgery
of radiographs
The radiographs were examined by the authors simultaneously without knowledge about the clinical result of the matching patient. The reduction of the fracture was regarded as good if the angle between the medial trabeculae and the medial femoral cortex (normally approximately 160”) in the frontal view was between 155 and 165” and the head of the femur was erect in the lateral view. Acceptable reduction was defined as the head of the femur being displaced in only one view, and unacceptable reduction as the head of the femur being displaced in both views. The position of the tip of the implant in the head of the femur was good when it was in the central third in both views. An unacceptable position was when the implant was in the upper and/or anterior third of the head of the femur. The remaining positions were acceptable. It was recorded whether or not the tip of the implant had perforated the articular surface. The group without perforation was divided into two groups, according to whether the tip of the implant was more or less than 1 cm from the articular surface. The fixation index (defined as the length of the part of the implant positioned in the capital fragment in relation to the length of the capital fragment) was recorded: good, >OS; acceptable, 06-0+3; unacceptable, <0.6. The telescoping effect, judged by the postoperative films and the most recent radiographs, was estimated with correction for enlargement of the radiographs and divided into three groups: (1) a telescoping of less than 1 cm; (2) a telescoping between 1 and 2 cm; (3) a telescoping of more than 2 cm. Migration of the implant, vans of more than lo” and failure of screws in the cortex were noticed. Necrosis of the head of the femur was noted, when there was at least one of the following changes: (1) subtle diffuse mottling of the bone; (2) isolated patches of sclerosis; (3) irregular discrete rarefaction with fuzzy marginal sclerosis; (4) subchondral cortical depression; (5) generalized subchondral fragmentation; or (6) periosteal bone apposition along the femoral neck. These definitions are modified after Martel and Sitterley (1969). The pre- and postoperative radiographs and the films obtained at the end of follow-up were examined. In the patients receiving a hip prosthesis, the last films obtained before reoperation were examined. In six patients who had died during the period of follow-up the most recent films obtained before death were examined.
Statistical
methods
The chi-squared test with one degree of freedom (DF 1) was used for all statistical testing except in testing the distribution of cases, where the implant telescoped enough to come out of the head of the femur. In this case Fisher’s exact probability test was used. RESULTS Fig. 1. The two Howmedica implants used in this series are shown: the sliding screw plate (top) and the sliding nail plate
(bottom).
There was no significant difference in distribution of postoperative variables such as reduction of fracture, position of fixation appliance in the head of the femur,
451
Nordkild et al.: Femoral neck fractures
Table II. Merle d’Aubign6 (1954) method Revue d’orthopkdie (1949) 35, 542)
Pain is intense
7
Pain is severe even at night
2
Pain
3
and permanent
is severe when walking; prevents any activity Pain is tolerable with limited activity
value of hip (Adapted
Ability
Mobility
Pain 0
of grading functional
Ankylosis with bad position of the hip No movement; pain or slight deformity Flexion under 40 Flexion between
40 and 60
4
Pain is mild when walking; it disappears with rest
Flexion between 60 and 80”; patient can reach his foot
5
Pain is mild and inconstant; normal activity No pain
Flexion between 80 and 90”; abduction of at least 15” Flexion of more than 90”; abduction to 30”
6
including the distance between the tip of the implant and the articular surface, and the fixation index (Table rrr>. We found union, implying bone union of the fracture with radiographically visible trabeculae across the fracture line, in 86.2 per cent (25/29) of the fractures with a sliding screw plate and in 73.7 per cent (14/19) of the fractures with a sliding nail plate (RO.3, Table ZV). One patient in the screw plate group died on the 16th day after fixation. In the group of ununited fractures, we encountered three hips, all with a sliding nail plate, in which the sliding nail had slid out of the head, resulting in redisplacement of the fractures. No such technical failure was encountered in the sliding screw plate group (P=O.O53, Table IV). Ten per cent (3/30) of the fractures with a sliding screw plate developed necrosis of the head of the femur with a median time after fixation of 33 months (range
l
Chi-squared test, DF 1.
to walk
None Only with crutches Only with canes With one cane, less than 1 hour; very difficult without a cane A long time with a cane; short time without cane and with limp Without cane but with slight limp Normal
33-34). All three of these fractures had previously shown bone union. In the sliding nail plate group 21 per cent (4/19) developed necrosis of the head of the femur with a median time after fixation of 42 months (range 2-50). Seventy-five per cent (3/4) of these fractures had previously shown bone union. This difference was not significant (Table IV). During the period of follow-up, seven patients (23-3 per cent) from the sliding screw plate group received prosthetic hips. In four of these patients the indication for reoperation was non-union and in the remaining three patients capital necrosis. Six patients (31.6 per cent) with a sliding nail plate had a further operation. This was due to non-union in four, necrosis of the head of the femur in one and the combination of capital necrosis and non-union in one. In the group with a sliding screw plate 75.9 per cent (22/29) of the fractures healed without complications; 52.6 per cent (1009) of the fractures with a sliding nail
Table 111.Distribution of postoperative variables in 30 fractures of the neck of the femur operated on with sliding screw plate and 19 fractures operated on with sliding nail plate
Good reduction Fair and not acceptable reduction Good position of implant Fair and not acceptable position of implant Distance from articular surface of 1 cm or less Distance from articular surface of more than 1 cm or with primary perforation Good fixation index Fair and not acceptable fixation index
from Table I,
Sliding screw plate
Sliding nail plate
11 19
7 12
CO.99
22 8
14 5
co.99
21
11
9
8
20 10
8 11
P value *
452
Injury: the British Journal of Accident Surgery (1985) Vol. 16/No. 7
Table IV. Results and with
sliding
of operation with sliding screw plate in 30 fractures nail plate in 19 fractures of the neck of the femur
Sliding Sliding screw p/are nail plate Non-union of fracture Union of fracture Implant telescoping out of femoral head Implant remaining in femoral head Necrosis of the femoral head No capital necrosis Alloplastic operation No alloplastic operation Fractures with complications Fractures without complications Varus >lO” No varus or varus
Chi-squared
test,
DF
4 25t 0
5 14 3
30
16
3 27 7 23 7t 22 2 28 3 27
4 15 6 13 9 IO 7 12 4 15
P value* CO-3 0.053
1.
t One patient was excluded because of death 16 days postoperatively.
healed without complications (P~0.1, Table IV). During the period of follow-up two fractures from the sliding screw plate group and seven fractures with a nail plate developed a varus deformity of more than 10”. This difference was highly significant (P
plate
postoperative pain than were the patients in the nail plate group (RO.005, Table V). The hips’ mobility was equally good in both groups. There was no significant (P
Table V. Distribution of clinical results according to the hip evaluation scales of Merle d’Aubign6 (Table II) in 30 fractures of the neck of the femur operated on with sliding screw plate and 19 fractures operated on with sliding nail plate Sliding screw plate* Pain score of less than 6 Pain score of 6 Hip mobility score of less than 6 Hip mobility score of 6 Walking ability score of less than 6 Walking ability score of 6
Sliding nail platet
4 16 0
0
20 7
12 2
13
10
P valueS CO.005
* Five patients who had died during the period of follow-up and five patients who had had an alloplastic operation were not examined. t One patient who had died during the period of follow-up, five patients who had had an alloplastic operation and one patient who refused to participate in this examination were not examined. $ Chi-squared test, DF 1.
Nordkild
et al.: Femoral
neck fractures
453
a
Fig. 2. a, Anteroposterior radiograph of the pelvis of a 64-year-old man showing a Garden’s stage 3, subcapital fracture of the right femur. 6. Postoperative anteroposterior radiograph of the right hip showing the fracture reduced with the sliding nail plate inserted in an accept ble position. c, Anteroposterior radiograph of the right hip obtained 40 days postoperatively. The nail has telescoped out o% the head of the femur resulting in recurrence of the fracture displacement.
454
years, we have been able to find only one previous study (Frandsen and Andersen, 1981) with a random allocation of patients to two different types of fixation. In this series we found union of 86.2 per cent of the fractures with a sliding screw plate, whereas only 73.7 per cent of the fractures from the nail plate group showed union. In an earlier report (Nordkild and Sonne-Holm, 1984) union was found in 87.0 per cent of fractures operated with a sliding screw plate. Barnes et al. (1976) found union in 75.6 per cent using the sliding nail plate. This compares well with a 75 per cent rate of union in fractures of Garden’s stages 3 and 4 with a sliding nail plate (Frandsen and Andersen, 1981). Although the difference in frequency of union found in this series is not statistically significant (RO.3, Table IV), it shows a tendency towards the screw plate having a higher rate of union. In three out of five ununited fractures operated on with sliding nail plates, the reason for non-union was a serious technical failure. In these fractures the nail telescoped and slid out of the head of the femur, resulting in redisplacement of the fracture (Fig. 2). The time for these events was approximately 1 month. This technical failure appeared in all three cases within 3 months of the operation. No such technical failure was found in the screw plate group in this series, but it has been reported earlier (Nordkild and Sonne-Holm, 1984) with a frequency of 2.7 per cent. Although this difference is only of doubtful significance (P=O.O53, Table IV), it suggests that the sliding nail plate should be abandoned. The frequency of secondary varus displacement of the fracture was significantly higher in the nail plate group (RO.01, Table IV), and in addition to the rate of technical failure in the nail plate group mentioned above, this shows that the sliding nail plate does not provide reliable fixation for fractures of the neck of the femur. Necrosis of the head of the femur is still a feared and unexplained complication of these fractures. In this series, necrosis of the head was encountered in 12 per cent of the united fractures with a sliding screw plate. Necrosis of the head was found in 21.4 per cent of the healed fractures from the nail plate group. In an earlier report on sliding screw plates (Nordkild and SonneHolm, 1984) there was necrosis of the head in 16.3 per cent of healed fractures. Barnes et al. (1976) found that 24 per cent of fractures with a sliding nail plate, which had previously shown bone union, developed necrosis of the head of the femur within 3 years. Frandsen and Andersen (1981) found necrosis in 21 per cent of fractures of Garden’s stages 3 and 4 with a sliding nail plate. An explanation for this difference could be that there is less damage to the blood vessels of the head of the femur following screwing than after nailing (Stramquist, 1983).
Injury: the British Journal of Accident Surgery (1985) Vol. 16/No. 7
The mobility of the hip with healed fractures and without necrosis of the head was the same in the two groups. The difference in walking ability was not significant. We found that patients with sliding nail plates were significantly (P
REFERENCES
Barnes R., Brown J. T., Garden R. S. et al. (1976) Subcapital fractures of the femur. A prospective review. J. Bone Joint Surg. 58B. 2. Clawson D. K. (1964) Trochanteric fractures treated by the sliding screw-plate fixation method. J. Trauma 4, 737. d’Aubign6 M. and Pastel M. (1954) Functional results of hip arthroplasty with acrylic prosthesis. 1. Bone Joini Surg. 36A, 451.
Frandsen P. A. and Andersen P. E. (1981) Treatment of displaced fractures of the femoral neck. Smith-Petersen osteosynthesis versus sliding nail-plate osteosynthesis. Acta Orthop. &and.
52, 547.
Frandsen P. A. and Kruse T. (1983) Hip fractures in the county of Funen, Denmark. Implications of demographic ageing and changes in incidence rates. Acta Orthop. &and. 54, 681.
Martel W. and Sitterley B. H. (lY6Y) Roentgenological manifestations of osteonecrosis. Am. J. Roentgenol. 106, 509.
Nordkild P. and Sonne-Holm S. (1984) Sliding screw-plate for internal fixation of femoral neck fractures. Actu Orthop.
Stand.
55, 616.
Pugh W. L. (1955) A self-adjusting nail-plate for fractures about the hip joint. J. Bone Joint Surg. 37A, 1085. Smith-Petersen M. N., Cave E. F. and Vangorder G. W. (1931) Intracapsular fractures of the neck of the femur. Arch.
Surg. 23, 715.
Striimquist B. (1983) Femoral head vitality after intracapsular hip fracture. Acta Orthop. Stand. Suppl. 200. Tronzo R. G. (1974) Hip nails for all occasions. Orthop. Clin. North Am. 5, 479.
Paper accepted 8 November
1984.
Reyuesrs for reprints should be addressed lo: Stig Sonne-Holm, L. E. Bruunsvej 27, DK-2900 Charlottenlund,
Denmark