HOW EFFECTIVE ARE OSTEOTOMIES FOR UNSTABLE

INTERTROCHANTERIC FRACTURES?

M. F. GARGAN, R. GUNDLE, A. H. R. W. SIMPSON

From the John Radcliffe Hospital, Oxford, England

Osteotomy has been used in the treatment of unstable load and lowered the tensile strain on the plate, compared
intertrochanteric hip fractures in an attempt to with medial displacement osteotomy. Rao Ct a! (1983)
increase the stability of the fracture fragments. We also found no benefit from medial displacement osteotomy
have assessed this stability in a randomised prospective in a retrospective study of 39 patients, but Harrington and
trial on 100 consecutive patients, all having fixation by Johnston (1973), in another retrospective study, supported
an AO dynamic hip screw, comparing anatomical the use of medial displacement osteotomy with a sliding
reduction with two types of osteotomy. hip screw. Clark and Ribbans (1990), after a prospective
The groups were similar in terms of age, gender, study, suggested that anatomical reduction was the
mental test score, and fracture configuration. There treatment of choice, although they had more failures of
were more failures of fixation in the osteotomy groups, fixation in this group.
and the operations took longer. We found no clear Our study examines first whether osteotomies achi-
benefit from osteotomy and therefore recommend eve their objective of a more stable configuration and
anatomical reduction and fixation by a sliding hip secondly, by a prospective randomised trial, assesses the
screw in most cases. Rarely, a fracture configuration need for osteotomies in the treatment of unstable
which does not allow load-sharing between the fracture intertrochanteric fractures.
fragments and the device may benefit from an
osteotomy or the use of an alternative implant.
PATIENTS AND METHODS
J BoneJoint Surg [Br] 1994; 76-B:789--92.
A consecutive series of 100 patients with unstable
Received 4 August 1992; Accepted after revisions 31 March 1994
intertrochanteric fractures of the femur was treated with
AO dynamic hip
screws (DHS) at the John Radcliffe
Sliding hip screws have been shown to give better results Hospital, Oxford, over a two-year period. After excluding
than fixed devices for the treatment of trochanteric those with multiple injuries, pathological fractures or
fractures ofthe femur (Bannister and Gibson 1983; Heyse- severely arthritic hips, patients were randomly allocated
Moore, MacEachern and Jameson Evans 1983; Esser, to receive either anatomical reduction and fixation or an
Kassab and Jones 1986). Before these screws were osteotomy and then fixation. Both groups were followed
introduced some surgeons used either a media! displace- prospectively. The study was approved by the Central
ment osteotomy (Dimon and Hughston 1967; Wolfgang, Oxford Research Ethics Committee.
Bryant and O’Neill 1982) or a valgus osteotomy (Sar- Fractures were defined as unstable by the presence
miento and Williams 1970) for unstable fractures to of one or more of the following:
convert them to more stable configurations. There are 1) four parts (Dimon and Hughston 1967),
opposing views on the need for osteotomy, especially 2) media! cortical comminution (Evans 1949),
with the use of sliding devices. Den Hartog, Bartal and 3) reverse obliquity of the main fracture line (Evans
Cooke (1991), in a cadaver study, found that a valgus 1949),
osteotomy increased the mean load to failure, but Chang 4) a large and separate posterior trochanteric fragment
et a! (1987), also using cadavers, reported that anatomical (Dimon and Hughston 1967), and
reduction increased the transmission of medial cortical 5) subtrochanteric extension (Sarmiento and Williams
1970).
The home circumstances, level of mobility, medical
M. F. Gargan, MA, FRCS, Senior Orthopaedic Registrar
history and Abbreviated Mental Test Score (Hodkinson
R. Gundle, MA, FRCS, Clinical Lecturer in Orthopaedic Surgery 1972) of all the patients were recorded before and after
A. H. R. W. Simpson, DM, FRCS, Clinical Reader in Orthopaedic Surgery
The Nuffield Orthopaedic Centre NHS Trust, Windmill Road, Headington,
the operation, and all were followed up until union of the
Oxford 0)3 7LD, UK. fracture or failure of the fixation. Randomisation was
Correspondence should be sent to Mr R. Gundle. achieved by means of sealed envelopes held in the
©1994 British Editorial Society of Bone and Joint Surgery operating theatre, the type of operation to be performed
0301-620X194/5571 $2.00 being determined from a table of random numbers. The

VOL. 76-B, No. 5, SEPTEMBER 1994 789

790 M. F. GARGAN, R. GUNDLE, A. H. R. W. SIMPSON

surgeon performed a valgus or medial displacement osteotomy group (p < 0.001). Mean blood transfusions
osteotomy according to personal preference; there was no were i.59 units in the anatomical group and 2.03 units in
difference in the proportions of different fracture patterns the osteotomy group (NS). Senior registrars performed
in the two osteotomy groups. The grade of the surgeon, 36% of the osteotomies and 24% of the anatomical
the operating time and the amount of blood transfused fixations; the remaining operations were done by regis-
were recorded. Patients were mobi!ised weight-bearing as trars.
soon as possible, usually on the second or third post- Nine patients died in hospital, four in the anatomical
operative day. group and five in the valgus osteotomy group. Two other
Radiographs were taken preoperatively, on the first patients in the anatomical group died at home before
postoperative day, at six weeks and at subsequent reviews. union. This left 89 surviving patients for full analysis of
The pattern of the fracture and the Singh osteoporosis the results. Twelve of these patients (13.5%) had failure
grade (1 to 3 or 4 to 6) were recorded (Singh, Nagrath of the fixation before union, four out of 49 in the
and Maini 1970). Postoperative radiographs were meas- anatomical group and eight out of 40 in the osteotomy
ured to determine the sliding distance available in each group. The odds ratio for failure of fixation in the
fixation, by subtracting the known length of the threaded anatomical group compared with the osteotomy group is
part of the screw with its accompanying shoulder from 0.36 (95% confidence interval 0.08 to 1.46). In nine cases
the length of screw protruding from the barrel. On the screw had penetrated the joint, in two it had cut out of
postoperative and review films the length of screw the neck, and in one case the plate had pulled off the
protruding from the barrel and the distance of the tip of femora! shaft. Other postoperative complications were
the screw to the joint line were compared to determine similarly frequent in the two groups.
the actual distance of screw slide and the amount by At review we found no difference in pain, mobility,
which the screw had cut into the head. Magnification and limping or walking distance between the two groups.
rotational errors were corrected by a factor obtained by Radiographs, however, showed differences in the amount
measuring the apparent length of the plate and barrel of of sliding within the implant during union, and in the
the fixation deviceand comparing it with the known amount by which the screws had cut into the femoral
dimensions of the implant. heads. The sum of sliding and cutting-in was recorded as
During most of the study period, standard devices ‘settling’ (Fig. 1). This differed in the operation groups
were used, but late in the series a short-barrelled device (Fig. 2). Settling of over 10 mm was seen in 50% of the
( 25 mm barrel) became available and was used when the anatomical group, 35% of the valgus osteotomy group
screw length required was less than 85 mm. In the last 20 and 10% of the medial displacement osteotomy group.
cases two of these were used in anatomical fixations and This may suggest that greater stability was achieved after
nine after osteotomies. medial displacement osteotomy.
In both osteotomy groups, all the failures occurred
after a fixation in which the amount of slide available was
RESULTS
less than 10 mm, due to effective shortening of the
Of the 100 patients, 55 had an anatomical reduction and femoral neck. Figure 3 shows the distribution of slide
DHS fixation. The other 45 had an osteotomy followed available by operation type and shows the large percentage
by DHS fixation; 25 valgus and 20 medial displacement of osteotomy patients whose fracture fixation allowed less
osteotomies were performed. The mean age ofthe patients than 10 mm of slide. After a short-barrel device became
was 82 years (60 to 100); there were 15 men and 85 available a subgroup of 20 patients was treated. Nine of
women. There was no significant difference between the 1 1 patients treated by osteotomy needed this short barrel
groups treated by osteotomy or by anatomical fixation in to allow sufficient slide since the screw length selected
terms of age, gender, degree of independence, mental test was less than 85 mm, but the short barrel was needed in
score and medical or drug history. There were 91 four- only two of nine treated without osteotomy. There was
part fractures, and nine three-part; 86 showed media! only one failure of fixation in this subgroup; this was after
cortical comminution, 34 had a large separate posterior osteotomy.
trochanteric fragment and 15 had subtrochanteric exten- There were seven failures in the 34 patients who had
sion. Only one showed reverse obliquity of the fracture an osteotomy before the introduction of the short-barrel
line. Singh osteoporosis grades were 1 to 3 in 77. There device but only one failure in 1 1 treated after this. The
were no significant differences between the two groups in use of a shorter barrel, when screws less than 85 mm are
pattern of fracture or osteoporosis grading. selected, increases the slide available and may reduce the
In all the anatomical fixations 135#{176}
DHS plates were failure rate.
used; five of the osteotomies were secured with 150#{176}
plates and 40 with 135#{176}
plates. The mean screw length in
DISCUSSION
the anatomical group was 90 mm and in the osteotomy
group 68 mm (p < 0.001). The mean operating time was The groups were generally well matched, their fracture
0.78 hours in the anatomical group and 1.16 hours in the configurations were similar, and the functional and

THE JOURNAL OF BONE AND JOINT SURGERY

 HOW EFFECTIVE ARE OSTEOTOMIES FOR UNSTABLE INTERTROCHANTERIC FRACTURES? 791

3i
Fig. 2

Settling (mm)

Fig. 1 Distribution of settling in each type of operation.

The settling (sum of sliding and cutting-in) in mm (mean ± SEM) for each
type of operation.

medical outcome was also similar. Technical failure of

the fixation is a serious problem in these frail elderly
people, and almost always necessitates further surgery
and prolonged rehabilitation with less chance of regaining
an independent existence. Osteotomy did not appear to
reduce this.
The aim of an osteotomy is to obtain a more stable
configuration, which will allow less movement of the
fracture fragments under physiological load. Our results
show that many osteotomies did not achieve this objective
(Fig. 2). Failures in the osteotomy group were attributable
to the reduction of the amount of sliding available, Fig. 3
because of the shortening effect of the osteotomy on the
femoral neck. The use of a shorter barrel, later in the
series, increased the sliding potential and appeared to
reduce the failure rate. The addition of an osteotomy
increased the operating time, the technique is more
difficult and there is a longer learning curve.
Successful fixation requires sharing of the load
between the fracture fragments and the implant. Al! four
failures in the anatomical group had an identifiable cause:
two had subtrochanteric extensions, one had reverse
obliquity of the fracture line and in one the screw was too Dto39 >40
Slide available (mm)
high in the head. Thus there may be some fractures, such
as those with reverse obliquity, which cannot be made Distribution of slide available for each type of operation.

VOL 76-B. No. 5. SEPTEMBER 1994

792 M. F. GARGAN, R. GUNDLE, A. H. R. W. SIMPSON

stable by an anatomical DHS. For this small group an when the AO dynamic hip screw was used and therefore
osteotomy or the use of a different implant may be recommend anatomical reduction and fixation by a sliding
appropriate. hip screw in most cases, taking care to choose an implant
We conclude that many osteotomies do not achieve which will allow sufficient slide.
their objective of putting the fracture into a more stable No benefits in any form have been received or will be received from a
configuration. We found no clear benefit from osteotomy
commercial party related directly or indirectly to the subject of this article.

REFERENCES

Bannister GC, Gibson AGF. Jewett nail plate or AO dynamic hip screw Hamngton KD, Johnston JO. The management of comminuted unstable
for trochanteric fractures?: a randomised prospective controlled trial. intertrochanteric fractures. J Bone Joint Surg [Am] 1973; 55-A:
J Bonefoint Surg [Br] 1983; 65-B:218. 1367-76.
Chang WS, Zuckerman JD, Kummer FJ, Frankel VH. Biomechanical Heyse-Moore GH, MacEachern AG, Jameson Evans DC. Treatment of
evaluation of anatomic reduction versus medial displacement osteo- intertrochanteric fractures of the femur: a comparison of the Richards
tomy in unstable intertrochanteric fractures. C/in Orthop 1987; 225: screw-plate with the Jewett nail-plate. J Bone Joint Surg [Br] 1983;
141-6. 65-B:262-7.
Clark DW, Ribbans WJ. Treatment of unstable intertrochanteric fractures Hodkinson HM. Evaluation of a mental test score for assessment of mental
of the femur: a prospective trial comparing anatomical reduction and impairment in the elderly.AgeAgeing 1972; 1:233-8.
valgus osteotomy. Injury 1990; 21:84-8.
Rao JP, Banzon MT, Weiss AB, Rayhack J. Treatment of unstable
Den Hartog BD, Bartal E, Cooke F. Treatment of the unstable intertrochanteric fractures with anatomic reduction and compression
intertrochanteric fracture: effect of the placement of the screw, its hip screw fixation. C/in Orthop 1983; 175:65-71.
angle of insertion, and osteotomy. J Bone Joint Surg [Am] 1991; 73-
Sarmiento A, Williams EM. The unstable intertrochanteric fracture:
A:726-33.
treatment with a valgus osteotomy and I-beam nail-plate. J Bone Joint
Dimon JH, Hughston JC. Unstable intertrochanteric fractures of the hip. Surg[Am] 1970; 52-A:1309-18.
J Bone Joint Surg [Am] 1967; 49-A:440-50.
Singh M, Nagrath AR, Maini PS. Changes in trabecular pattern of the
Easer MP, Kassab JY, Jones DHA. Trochanteric fractures of the femur: a upper end of the femur as an index of osteoporosis. J Bone Joint Surg
randomised prospective trial comparing the Jewett nail-plate with the [Am] 1970; 52-A:457-67.
dynamic hip screw. J BoneJoint Surg [Br] 1986; 68-B:557-60.
Wolfgang GL, Bryant MH, O’Neill JP. Treatment of intertrochanteric
Evans EM. The treatment of trochanteric fractures of the femur. J Bone fracture of the femur using sliding screw plate fixation. C/in Orthop
JointSurg[BrJ 1949; 31-B:190-203. 1982; 163:148-58.

THE JOURNAL OF BONE AND JOINT SURGERY