FEMORAL NECK FRACTURE FIXATION

COMPARISON OF A SLIDING SCREW WITH LAG SCREWS

D. I. CLARK, C. E. CROFTS, M. SALEH

From the University ofSheffield and Royal Hallamshfre Hospital, Sheffield

The rigidity of a sliding compression screw and three cannulated lag screws in the treatment of subcapital
fractures was compared in five pairs of female cadaver femora. There were no significant differences between
the compressive sfrength, bone density, cortical thickness or Singh index of the bones in each pair. A
subcapital fracture was standardised using a perpendicular saw cut across the femoral neck. A urnaxial ‘load
test system’ with force and length measurement facifities was used to mimic cyclical stressing applied in vivo
at a frequency of 0.5 Hz from 0 to 3 times body-weight. There was no significant difference between the
fixation afforded by the sliding compression screw and three lag screws. Bone quality was the single most
important factor in the stabifity of the bone implant unit.

The importance of bone quality in the fixation of femora! heads from the same patient and (2) a dynamic loading
neck fractures has been emphasised by in vitro studies regime to simulate walking.
(Franke! 1960; Van Audekercke et a! 1979; Husby,
H#{248}iseth and F#{248}nstelien 1987). However, most in vitro
MATERIALS AND METHODS
studies have compared fixation devices using femora
from different subjects (Van Audekercke et a! 1979; Fourteen pairs of femora were obtained from females
MacKechnie-Jarvis 1983 ; Elmerson eta! 1987). The over the age of 60. Those with known hip pathology such
limitation of this technique is that the Yariation in the as a previous arthroplasty, a femoral fracture or a tumour
mechanical properties of bone from different individuals were excluded from the study. The specimens were
is not taken into account. The two reports which initially stored at - 20#{176}C.
Prior to testing, the mechanical
overcame this problem by using paired femora from the properties of the right and left femora were investigated
same patient employed a static loading system (Mizrahi by measuring bone density (Dal#{233}n, Hellstr#{246}m and
et a! 1980; Engesaeter et a! 1984). This method is Jacobson 1976), cortical thickness (Griffiths, Swanson
unsatisfactory since in life the fixation device must be and Freeman 1971), the Singh index (Khairi et a! 1976)
able to withstand the cyclical loading pattern associated and using a static compression test.
with walking. The frozen bones were radiographed in pairs. The
We report an in vitro study which investigated the Singh index was assessed for each proximal femur by
fixation offemoral neck fractures using (1) paired femora! three independent observers and the femora! shaft
cortical thickness measured 120 mm below the greater
trochanter using vernier calipers. After the bones had
D. I. Clark, MB, ChB, BMed Sci, Anatomy Demonstrator been radiographed they were returned to their storage
Department of Anatomy, University of Newcastle, Framlington Place,
Newcastle upon Tyne NE2 4HH, England. environment. Using a ND 2100 bone density scanner,
C. E. Crofts, BSc, PhD, University Lecturer readings were taken from the proximal two-thirds of
Department of Medical Physics and Clinical Engineering, University each femur, and from the femoral shaft sections. As soon
of Sheffield, Royal Hallamshire Hospital, Glossop Road, Sheffield 510
2JF, England. as the measurements had been made a!! the bones were
M. Saleh, MSc(Bio Eng), FRCS, FRCS Ed, SeniorLecturer, University again returned to storage.
ofSheffield, Honorary Consultant, Sheffield Health Authority Before mechanical testing, the bones were defrosted
Department of Orthopaedics, Northern General Hospital, Barnsley
Road, Sheffield 55 7AU, England.
for six hours at room temperature, carefully cleaned of
Correspondence should be sent to Dr D. I. Clark.
muscle and soft tissue, and rehydrated with warm saline.
The mechanical tests were performed on a Mayes
© 1990 British Editorial Society ofBone and Joint Surgery
030l-620X/90/S1SS $2.00 material testing machine (Fig. 1). Static compression
J Bone Joint Surg [Br] 1990; 72-B :797-800. tests were carried out on six pairs of femoral shaft

VOL. 72-B, No. 5, SEPTEMBER 1990 797

798 D. I. CLARK, C. E. CROFTS, M. SALEH

sections. After the bones had defrosted, the length and Failure of the system was defined as the point at which a
diameter of each specimen were measured. The length of deflection of 125% of the value recorded at 30 cycles was
each sample was adjusted using a band-saw to give a observed (a figure set for the proposed British Standard
length to diameter ratio of approximately 4 to 1 The . for femora! head prostheses BSI DD91 :1986). The
specimen was loaded to failure at a rate of 3 mm/mm. specimens were refrozen after testing and later radio-
The maximum load was recorded. graphed to determine the pattern of failure.
Cyclical compression tests were performed on five
pairs of randomly selected femora artificially fractured
RESULTS
and fitted with two different fixation devices. After the
bones had defrosted, one specimen from each pair was Bone quality. In order to be certain that there was no
selected at random for fixation with a sliding compression difference in the bone quality of the paired femora, four
screw, whilst the other was fixed with three self-tapping pretest investigations were performed : static compres-

c LLR
) r
a
a
-

. a a -

-jet fl, -

JJJ.JJ#{149}#{149}.

. rOr .. .

1E 0 :iu
i.:. iJ:i;i

i. d#{149}#{149}O

i:i’’
Fig. 1
I Fig. 2

Figure 1 - The Mayes material testing machine as used in our experiments. Figure 2 - The proximal femur held between the purpose-built
grips.

cannulated lag screws. In order to achieve an anatomical sion tests, bone density measurements, the Singh index
reduction, both devices were initially inserted into the and cortical thickness as measured radiographically.
unfractured femur. They were removed and a fracture Statistical analysis revealed no significant difference
artificially produced using a saw cut perpendicular to the between the right and left femora (Table I).
neck. Purpose-built grips to hold the bone at an Rigidity. The deflections recorded during the fatigue
anatomical angle of 20#{176}
from the vertical were fitted to testing of the five pairs of proximal femora are given in
the test machine (Fig. 2). The specimen was placed in Table II. There was no significant difference between the
the lower grip. Each bone was covered with damp paper deflections recorded after the first or 30th cycles for the
during testing. Each specimen was loaded sinusoidally bones fixed with compression or lag screws. In addition,
from 0 to 3 times body-weight (2. 1 kN) at a frequency of there was no significant difference between the number
0.5 Hz until fracture occurred. The load and displacement of cycles required to produce failure in the femora fixed
were recorded graphically and from a digital display. with a sliding compression screw and those fixed with
The deflections after 1 and 30 cycles were recorded. three lag screws.

THE JOURNAL OF BONE AND JOINT SURGERY

 FEMORAL NECK FRACTURE FIXATION 799

The results in Table III compare the age and bone first the fracture impacted with the screw sliding down
quality (Singh index and bone density) against deflection the barrel. The head then began to shear across the
of each pair of fixed bones. The pair with the largest fracture site due to the screw barrel cutting through the
deflection (number 4) were from the oldest subject (84 femora! neck trabecu!ar bone (Fig. 3a). During loading
years) and had the weakest bone (Singh index 3) and the the three cannulated lag screws were forced back down
lowest bone density of the five pairs. By contrast, the the drilling holes. A crush fracture of the inferior surface
smallest deflection occurred in bones from the youngest of the femora! neck then occurred. Finally the head
subject (65 years) with the best bone quality (Singh index sheared across the fracture line and the screw threads
6) and a high bone density. converged and rotated downwards (Fig. 3b). No plastic
Failure mode. The sliding compression screws did not deformation of any implant was observed.
bend in any of the tests. Failure was due to collapse of
the trabecular bone in the femora! neck and head. At
DISCUSSION

The rigidity ofthe Richards compression screw and three

Table I. Comparison of the bone quality in the right and left femora
from the same patients lag screws used currently in the treatment of femoral
neck fractures have been compared. In order to reflect
Pairs
Left
the population at risk, femora were obtained from
tested Right
females over the age of 60 (Lewinnek et a! 1980). To
Compressive strength(kN) 6 14.24±2.79 12.53 ±2.4
exclude bone quality as a source of error, paired femora
Bone density (g/cm2) 12 0.97 ± 0.06 0.96± 0.05 from the same subject were tested. The bones from each
Singhindex 14 5.2±0.26 5.2±0.30 pair were shown to be matched mechanically by four
pretesting investigations. In addition an attempt was
Cortical thickness (mm) 14 7.21 ±0.27 6.97 ± 0.36
made to reduce experimental variables to a minimum by
using a standardised fracture (Klenerman and Marcuson
1970) and ensuring an anatomical reduction. A saw cut
perpendicular to the neck was chosen in accordance with
Klenerman’s observations that subcapita! fractures have
a similar angle. During walking, the angle ofthe resultant
load on the femora! head has been shown to be 10#{176} to 15#{176}
from the midline (Inman 1947; Elson and Charn!ey
1968). The offset angle of the femora! shaft is 7#{176}
to 10#{176}
(Williams et a! 1989). The optimum loading angle for the
femora! head in vitro is therefore 17#{176}
to 25#{176}
from the
vertical; 20#{176}
was chosen in this study. A cyclical rather
than a static loading regime was applied in order to
simulate walking. A loading regime of 0 to 3 times body-
weight (2. 1 kN acting sinusoidally) was chosen for two
reasons. First, three times body-weight is within the peak
Fig. 3a Fig. 3b
theoretical load acting on the hipjoint (Rydell 1966 ; Paul
Mode of failure. Figure 3a - of the sliding screw. Figure 3b - of the lag 1967; Crowninshield, Brand and Johnston 1978). Se-
screws.
cond!y, the British Standards Institute recommends a

Table II. Deflections recorded during fatigue testing ofthe proximal Table III. Age, bony quality and deflection under testing of
femora each pair of fixed femora

Number Bone
Age Screw Deflection after Deflection after of cycles Age Deflection after Singh density
Pair (yr) used 1 cycle (mm) 30 cydes (mm) to failure Pair (yr) 3Ocycles(mm) index (g/cm2)

1 71 Compression 7.20 12.93 66 1 71 12.93 5 0.95

Lag 2.88 11.18 67 11.18 6 0.98

2 65 Compression 1.65 4.71 55 2 65 4.71 6 0.96

Lag 1.74 3.45 65 3.45 6 1.00

3 65 Compression 3.44 6.45 65 3 65 6.45 6 0.75

Lag 1.94 5.71 38 5.71 6 0.69

4 84 Compression 18.50 24.97 S 4 84 24.97 3 0.53

Lag 11.40 22.21 3 22.21 3 0.56

5 81 Compression 2.75 4.30 75 5 81 4.30 4 0.89

Lag 2.94 5.46 150 5.46 4 0.79

VOL. 72-B, No. 5, SEPTEMBER 1990

800 D. I. CLARK, C. E. CROFTS, M. SALEH

2. 1 kN sinusoidal loading pattern when fatigue testing shown that bone quality was the most important factor
femoral heads. in influencing the outcome of fixation.
The choice of the three lag screws was based on
No benefits in any form have been received or will be received from a
biomechanical investigations (Mizrahi et a! 1980 ; Rubin commercial party related directly or indirectly to the subject of this
article.
et a! 1981). Mizrahi found a triangular three-screw
orientation of two superior and one inferior to be the REFERENCES
optimum configuration, whilst Rubin showed the impor- Brodetti A. An experimental study on the use of nails and bolt screws
in the fixation of fractures of the femoral neck. Ada Orthop Scand
tance ofpara!le! screw placement. Both authors, however, 1961 ; 31 :247-71.
found no improvement in fixation if more than three
BSJ. Draft for development. Method for determination of endurance
screws were used. The compression hip screw has a properties of stemmed femoral components of hip joint prostheses
1986. BSI(DD91 :1986).
sliding barrel for controlled impaction at the fracture site
Oawn DK. Trochanteric fractures treated by the sliding screw plate
(Clawson 1964). Its blunt nose preventsjoint penetration.
fixation method. J Trauma 1964; 4:753-6.
Biomechanical testing has shown that a large threaded
Crowninshield RD, Brand RA, Johnston RC. The effects of walking
screw provides better internal fixation than a triflanged velocity and age on hip kinematics and kinetics. C/in Orthop 1978;
132:140-4.
nail (Brodetti 1961). The advantage of the side plate to
Dal#{233}nN, Hellstr#{246}m L-C, Jacobson B. Bone mineral content and
provide improved fixation to the femora! shaft was
mechanical strength of the femoral neck. Acta Orthop Scand 1976;
demonstrated by Haboush (1952). In both devices the 47:503-8.
initial event in failure was impaction ofthe femoral neck. ElmersonS, Andersson CB, Pope MH, Zetterberg C. Stability of fixation
in femoral neck fractures : comparison of four fixation devices in
This underlines the importance of parallel placement of
vivo and in cadavers. Acta Orthop Scand 1987 :8:109-12.
the lag screws allowing backing out, and the controlled
Elson RA, Charnley J. The direction of the resultant force in total
sliding action permitted by the compression hip screw. prosthetic replacement of the hip joint. Med and Biol Eng 1968;
6:19-27.
Franke! (1960) showed that in saw cut femoral neck
Engesaeter LB, Asserson 0, Molster A, et al. Stability of femoral neck
fractures fixed in vitro, the bone absorbed 75% of the
osteotomies fixed by Von Bahr screws or by compression hip screw.
load applied to the femoral head and the device the EurSurgRes 1984; l6suppl2:37-40.
remaining 25%. The results in Table III confirm Frankel’s Frankel VH. Mechanical factors for internal fixation of the femoral
neck. Acta Orthop Scand 1960; 29 :21-42.
findings and suggest bone quality to be the most important
Criffiths WEC, Swanson SAy, Freeman MAR. Experimental fatigue
factor in the rigidity ofthe bone implant unit. fracture of the human cadaveric femoral neck. J Bone Joint Surg
Using an unphysiological static load and femora! [Br] 1971 ; 53-B:136-43.
heads from different subjects, other workers have shown Haboush EJ. Photoelastic stress and strain analysis in cervical fractures
ofthe femur. Bu//Hosp Joint Dis 1952; 13:252-8.
that both the sliding compression screw and three lag
Husby T, Hoiseth A, F#{248}nstelien E. Strength of femoral neck fracture
screws have a superior holding power to the Smith fixation : comparison of six techniques in cadavers. Acta Orthop
Peterson nail (Brodetti 1961 ; Van Audekercke et a! Scand 1987; 58 :634-7.
1979). However, when comparing parallel lag screws to Inman VT. Functional aspects of the abductor muscles of the hip. J

the sliding compression screw there is disagreement. Van BoneJointSurg 1947; 29:607-19.
Khaki MRA, Cronin JH, Robb JA, et al. Femoral trabecular-pattern
Audekercke et a! (1979) and MacKechnie-Jarvis (1983)
index and bone mineral content measurement by photon absorption
found both devices equally effective. Husby et a! (1987) in senile osteoporosis. JBone Joint Surg [Am] 1976; 58-A :221-6.
showed that the sliding compression screw was superior Klenerman L, Marcuson RW. Intrascapular fractures of the neck of the
femur. J Bone Joint Surg [Br] 1970; 52-B :514-7.
to three lag screws, whilst Elmerson et a! (1987) found
Lewinnek CE, Kelsey J, White AA, Kreiger NJ. The significance and a
the reverse, three lag screws providing better fixation
comparative analysis of the epidemiology of hip fractures. C/in
than the Richards sliding compression screw. A!! these Orthop 1980; 152:35-43.
studies used unmatched femoral heads from a range of MacKechnie-Jarvis AC. Femoral neck fracture fixation : rigidity of five
techniques compared. J R Soc Med 1983 ; 76:643-8.
male and female subjects. This factor could have
Mizrahi J, Hurlin RS, Taylor JK, Solomon L. Investigation of load
significantly influenced the results since it does not allow
transfer and optimum pin configuration in the internal fixation, by
for possible variations in bone quality of the different Muller screws, of fractured femoral necks. Med Bio/ Eng Comput
1980; 18:319-25.
femora under test.
Paul JP. Forces transmitted by joints in the human body. Proc Inst
The only report to use paired femora! heads from
Mech Engrs 1967; 18l(3J):8-13.
the same subject compared two Von Bahr lag screws with
Rubin R, Trent P, Arnold W, Burstein A. Knowles pinning of
a sliding compression screw (Engesaeter et a! 1984). A experimental femoral neck fractures : a biomechanical study. J
Trauma 1981 ; 21 :1036-9.
static loading regime was used and no significant
Rydell NW. Forces acting on the femoral head-prosthesis : a study on
difference was found between the two devices. The
strain gauge supplied prostheses in living persons. Ada Orthop
results of our study, using a dynamic loading regime Scand 1966; 37 (Suppl 88):l-132.
confirm Engesaeter’s static test results and show that Van Audekercke R, Martens M, Muller JC, Stuyck J. Experimental
study on internal fixation of femoral neck fractures. C/in Orthop
three parallel screws provide as rigid a fixation as the
1979; 141 :203-12.
larger sliding compression screw. In addition, by quanti-
Williams PL, Warwick R, Dyson M, Bannister LH. Gray’s Anatomy.
fying the mechanical strength of the femora, we have 37th ed. Edinburgh, Churchill Livingstone, 1989 :435.

THE JOURNAL OF BONE AND JOINT SURGERY