TRAUMA

Inadequate three-point proximal fixation

predicts failure of the Gamma nail

S. G. F. Abram,
T. C. B. Pollard,
A. J. M. D. Andrade
From Royal
Berkshire NHS
Foundation Trust,
Reading, United
Kingdom

The Gamma nail is frequently used in unstable peri-trochanteric hip fractures. We

hypothesised that mechanical failure of the Gamma nail was associated with inadequate
proximal three-point fixation. We identified a consecutive series of 299 Gamma nails
implanted in 299 patients over a five-year period, 223 of whom fulfilled our inclusion criteria
for investigation. The series included 61 men and 162 women with a mean age of 81 years
(20 to 101). Their fractures were classified according to the Modified Evans classification
and the quality of fracture reduction was graded. The technical adequacy of three points of
proximal fixation was recorded from intra-operative fluoroscopic images, and technical
inadequacy for each point was defined. All patients were followed to final follow-up and
mechanical failures were identified. A multivariate statistical analysis was performed,
adjusting for confounders. A total of 16 failures (7.2%) were identified. The position of the
lag screw relative to the lateral cortex was the most important point of proximal fixation,
and when inadequate the failure rate was 25.8% (eight of 31: odds ratio 7.5 (95% confidence
interval 2.5 to 22.7), p < 0.001).
Mechanical failure of the Gamma nail in peri-trochanteric femoral fractures is rare (< 1%)
when three-point proximal fixation is achieved. However, when proximal fixation is
inadequate, failure rates increase. The strongest predictor of failure is positioning the lateral
end of the lag screw short of the lateral cortex. Adherence to simple technical points
minimises the risk of fixation failure in this vulnerable patient group.
Cite this article: Bone Joint J 2013;95-B:82530.

S. G. F. Abram, BA, BM, BCh,

FY2
T. C. B. Pollard, MD, FRCS(Tr
& Orth), Lower Limb
Arthroplasty Fellow
A. J. M. D. Andrade, M.Sc.,
FRCS(Tr & Orth), Consultant
Orthopaedic Surgeon
Royal Berkshire NHS
Foundation Trust, Department
of Trauma and Orthopaedics,
London Road, Reading RG1
5AN, UK.
Correspondence should be sent
to Mr A. J. Andrade; e-mail:
tony.andrade@royalberkshire.
nhs.uk
2013 The British Editorial
Society of Bone & Joint
Surgery
doi:10.1302/0301-620X.95B6.
31018 $2.00
Bone Joint J
2013;95-B:82530.
Received 30 September 2012;
Accepted after revision 25
February 2013

VOL. 95-B, No. 6, JUNE 2013

The incidence of hip fractures in the United

Kingdom is high, affecting approximately
70 000 patients and costing the NHS nearly
2 billion per year.1 If operative treatment fails,
in addition to the increased mortality and morbidity there is the financial burden of further
surgical treatment, which was estimated to be >
10 000 per patient in 2000 and 2010.2,3
Approximately 39% of hip fractures are
peri-trochanteric.1 Sliding hip screws and
cephalo-medullary implants are frequently
used in their operative treatment.4 Cephalomedullary nails provide a biomechanical
advantage over sliding hip screws for unstable
fracture patterns.5-9 However, a recent
Cochrane review noted a higher complication
rate associated with cephalo-medullary devices
and recommended further studies using more
contemporary implants.4 It has been suggested
that failure rates for cephalo-medullary nails
may be lower when a meticulous operative
technique is applied.10 Stability in unstable
peri-trochanteric fractures may be achieved
simply through adequate two-point fixation
between the head and the shaft, connected by a

fixed-angle device such as a sliding hip screw

and plate, or a cephalo-medullary nail. The
importance of the tipapex distance (TAD) for
fixation of the proximal fragment has been
documented for both devices.11,12 For cephalomedullary nails, three-point proximal fixation may confer additional stability and is one
of the design features of the Gamma nail
(Stryker, Newbury, United Kingdom).13 These
three points are the cortical contact point of
the lag screw at the lateral femoral cortex, the
cortical contact point of the nail (or end-cap)
at the greater trochanteric cortex, and the TAD
of the lag screw. Whereas high TAD has been
shown to predict failure rate in cephalo-medullary nails,12 the importance of the other two
points of proximal fixation has not yet been
demonstrated.
Given the accepted importance of the TAD,
our null hypothesis was that the other two
points of proximal femoral fixation were of no
importance in predicting mechanical failure of
the Gamma nail. We sought to establish the
mechanical failure rates of the Gamma nail by
means of a radiological analysis of the
825

826

S. G. F. ABRAM, T. C. B. POLLARD, A. J. M. D. ANDRADE

299 Gamma-3
Gamma 3 nails
implanted over a
5-year
year period

76 exclusio
exclusions:
ns:
20 - Inadequate radiographs for
analysis
19 - Prophylactic nail
14 - Death within 30 days
9 - Fracture not peri-trochanteric
8 - Pathological fracture
6 - Peri-prosthetic fracture

223 cases included in

multivariate analysis

16 mechanical failures:
12 - Fixation failure
3 - Nail fracture
1 - Nail subsidence

207 cases without

failure

Fig. 1
CONSORT flow diagram of the cohort.

adequacy of three-point proximal fixation in patients with

peri-trochanteric femoral fractures.

Patients and Methods

We identified a consecutive series of 299 Gamma 3 nails
(Stryker) implanted in 299 patients over a five-year period
between 1 January 2006 and 31 December 2010. A total of
76 patients were excluded (Fig. 1), leaving a total of
223 patients (61 male and 162 female) with a mean age of 81
years (20 to 101) for inclusion in the study. A total of 180 long
Gamma nails were used and 43 short versions. Post-operatively, all patients were allowed to partially bear weight, building to full weight-bearing over a period of six weeks.
Pre-and intra-operative radiographs were reviewed using
Centricity PACS software (General Electric Healthcare,
Chalfont St Giles, United Kingdom). Fracture patterns were
graded by the modified system14 of Evans15 (type 1, undisplaced trochanteric fracture; type 2, additional fracture of
lesser trochanter; type 3, additional fracture of greater trochanter with posteromedial comminution; type 4, intersubtrochanteric fracture, fracture of greater trochanter,
posteromedial comminution with subtrochanteric component; type 5 (or R), reverse obliquity fracture). The quality
of reduction was graded as good, acceptable or poor
according to Baumgaertner et al.11 These assessments were
undertaken by an orthopaedic fellow (TCBP).
Fixation at the femoral head was graded by the TAD
according to Baumgaertner et al,11 and classified as satisfactory if < 25 mm.12 The diameter of the lag screw
(10.5 mm) was used to correct for radiological magnification. The position of the lag screw at the lateral femoral
cortex was classified as satisfactory when the screw was
in contact with or protruding from the lateral femoral

cortex. Finally, the position of the tip of the proximal nail

(or end-cap if used) was classified as satisfactory if in contact with or protruding beyond the cortex at the greater trochanter. Examples of inadequate positioning for each of
these points are shown in Figure 2. These binary outcomes
of each of the three proximal femoral fixation points were
assessed by a second, single observer (SGFA).
Mechanical failure was defined as loss of fracture reduction with migration of the lag screw within the proximal
fragment (ultimately leading to screw cut-out); nail subsidence (defined by fracture of the distal locking bolts); or nail
fracture (fatigue failure due to non-union). Our institutions
policy is that all patients are routinely followed up at six
weeks, and those who are mobile and pain-free are discharged at that visit, although radiological union will not be
evident at this stage. Patients who subsequently develop pain
or complications are referred back to the department for further imaging and assessment. For the purposes of this study,
all cases were followed to the most recently recorded clinical
episode. The follow-up for each patient was thus determined
as being from the date of surgery to this most recent episode.
This study was considered as a service evaluation, and following the guidance of the National Research Ethics Service
did not require Research and Ethics Committee approval.
Statistical analysis. An independent statistician analysed
the data. A univariate logistic regression analysis of potential confounding factors (age, gender, fracture classification, reduction quality and implant type (short or long
nail)) was performed. Those factors that appeared to affect
failure rates, with a p-value < 0.2, were then included, and
adjusted for, in a multivariate regression analysis. Each of
the three proximal femoral fixation points was added into
this model in turn, to establish their effect on risk of failure.
THE BONE & JOINT JOURNAL

INADEQUATE THREE-POINT PROXIMAL FIXATION PREDICTS FAILURE OF THE GAMMA NAIL

Fig. 2b

Fig. 2a

827

Fig. 2c

Intra-operative fluoroscopic radiographs of inadequate proximal femoral fixation, a) with the lateral end of the lag screw positioned short of the
lateral femoral cortex, b) with a tip-apex distance > 25 mm, and c) with the end-cap of the nail positioned distal to the cortex of the greater trochanter.

Table I. Details of the 16 failures, listed in order of modified Evans grade for ease of interpretation (TAD, tipapex distance; GT, greater
trochanter)
Inadequate fixation points
Patient

Gender

Age (yrs)

Evans grade

Reduction

Nail type

Time to
failure (wks)

4
9
10
11
14
16
7
1
2
3
5
6
8
12
13
15

Female
Female
Male
Male
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Female
Male

82
85
75
88
87
82
70
84
58
85
88
84
92
88
60
41

2
3
3
3
3
3
4
4
5
5
5
5
5
5
5
5

Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Poor
Poor
Good
Good
Good
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable

Long
Long
Long
Long
Long
Long
Long
Short
Long
Short
Long
Short
Long
Long
Long
Long

70
6
30
3
27
14
8
10
3
2
2
12
4
7
46
5

In addition, the proximal femoral fixation points were

included in the model together, so that independence
between points could be established. Statistical significance
was ascribed when the p-value was < 0.05.
Inter-observer reproducibility and intra-observer repeatability for the two observers (TCBP and SGFA) classifying
three-point fixation (SGFA), fracture classification (TCBP)
and quality of reduction (TCBP) was evaluated by the
repeated assessment of a random sample of 20 cases, by
both observers, studied after an interval of four weeks.
Unweighted kappa coefficients were calculated to quantify
the reliability of these measurements.
VOL. 95-B, No. 6, JUNE 2013

n TAD

Lateral cortex

GT

Failure mode

1
1
1
1
1
1
1
1
1
2
1
2
1
1
0
3

Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes

Yes
Yes

Fixation failure
Fixation failure
Nail fracture
Fixation failure
Fixation failure
Nail subsidence
Nail fracture
Fixation failure
Fixation failure
Fixation failure
Fixation failure
Fixation failure
Fixation failure
Fixation failure
Nail fracture
Fixation failure

Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes

Results
Group demographics. The mean follow-up was 290 days
(44 to 1834). A total of 16 failures were identified at a
mean time post-operatively of 15.6 weeks (2 to 70): 12 due
to fixation failure, three to nail fractures and one due to
nail subsidence (Table I, Fig. 1). There were two deep infections, which were not counted as failures for the purposes
of the analysis.
There were seven late distal femoral fractures in the
43 patients treated with a short Gamma nail. In all these
patients the proximal fracture had united. Use of the short
nail in our unit was abandoned in 2008 because of the

828

S. G. F. ABRAM, T. C. B. POLLARD, A. J. M. D. ANDRADE

Table II. Failures rates by fracture pattern, quality of reduction, and number of adequate points of
fixation
Variable

Fractures (n)

Failures (n)

Failure rate (%)

Fracture pattern (Modified Evans grade14)

1
2
3
4
5

6
45
79
17
76

0
1
5
2
8

0.0
2.2
6.3
11.8
10.5

Quality of reduction11
Good
Acceptable
Poor

79
123
21

3
11
2

3.8
8.9
9.5

Adequate points of fixation*

3 of 3
2 of 3
1 of 3
0 of 3

132
75
15
1

1
12
2
1

0.8
16.0
13.3
100.0

* comprising the tipapex distance, the greater trochanter and the lateral cortex

higher complication rate.16 The seven patients with late

distal fractures were not excluded nor counted as failures
for the purpose of the analysis, as the original proximal
fracture had already united; the specific complication of
distal fracture was considered unique to the short nail and
was not relevant to the study question.
Measurement reliability. The kappa values for intraobserver repeatability and inter-observer reproducibility for
the fracture classification were 0.86 (95% confidence interval (CI) 0.67 to 1.0) and 0.71 (95% CI 0.46 to 0.96), respectively: 0.89 (95% CI 0.67 to 1.0) and 0.89 (95% CI 0.67 to
1.0) for the fracture reduction quality, 1.0 (95% CI 1.0 to
1.0) and 1.0 (95% CI 1.0 to 1.0) for both the TAD and lateral cortex points, and 1.0 (95% CI 1.0 to 1.0) and 0.64
(95% CI 0.44 to 0.86) for the greater trochanteric point:
kappa values between 0.81 and 1 are considered very good,
and between 0.61 and 0.8 are seen as good.17
Failures related to the proximal femoral fixation points. A
total of 31 Gamma nails (13.9%) were inserted with the lag
screw being short of the lateral cortex: of these, eight failed
(25.8%); 50 nails (22.4%) were inserted with a TAD
> 25 mm, and of these, nine failed (18.0%); 29 nails
(13.0%) were inserted with the top of the nail short of the
proximal femoral cortex: of these, two failed (6.9%)
(Table I). Given the small numbers within each failure subtype (Table I), it was not possible to perform a meaningful
statistical analysis on whether the type of failure was associated with any particular inadequacy of proximal fixation.
However, there did appear to be an association between the
Evans grade and which proximal fixation points were
inadequate. For Evans type 3 and 4 fractures that failed, the
TAD was usually inadequate in isolation (six out of seven
patients), whereas for type 5 fractures, seven of eight fractures had inadequate fixation of the lateral cortex point,
four of which were in isolation (Table I).

The failure rates by fracture pattern, quality of reduction

and number of adequate points of fixation are given in
Table II. Univariate regression analysis demonstrated a
trend towards increasing failure rate with increasing grade
of fracture, with an odds ratio (OR) of 1.53 (95% CI 0.98
to 2.39, p = 0.06). Therefore fracture pattern was included
in the subsequent multivariate analysis of the three points
of proximal fixation. Univariate statistical analysis of failure rates by the quality of reduction demonstrated a tendency towards increasing failure rate with worse quality
of reduction, with an OR of 1.73 (95% CI 0.77 to 3.91,
p = 0.19) (Table II). Accordingly the quality of reduction
was also included in the multivariate analysis. Age and
implant type (short or long nail) had no significant effect on
the outcome on univariate analysis (ORs of 0.98 (95%
CI 0.95 to 1.02, p = 0.34) and 0.96 (95% CI 0.26 to 3.54,
p = 0.96)) and were excluded from the multivariate analysis. Although female gender was associated with a higher
odds of failure (OR 1.69 (95% CI 0.46 to 6.14, p = 0.43),
this did not approach statistical significance and therefore
was also not included in the multivariate analysis.
The multivariate regression analysis, adjusting for modified Evans grade and quality of reduction, revealed ORs for
failure of 7.53 and 6.86 for the lateral cortex and TAD,
respectively, which were highly statistically significant
(p < 0.001 and p = 0.001) (Table III). The greater trochanteric point was not significantly associated with failure
(p = 0.92). In order to determine whether the TAD and lateral cortex points were independently associated with failure, the analysis was repeated including adjustment for
TAD when the lateral cortex point was the outcome, and
vice versa. From this analysis, the OR for failure for TAD
being > 25 mm was 9.44 (95% CI 2.63 to 33.94,
p = 0.001), and that for the lag screw being short of the lateral cortex was 10.84 (95% CI 2.98 to 39.40, p < 0.001).
THE BONE & JOINT JOURNAL

INADEQUATE THREE-POINT PROXIMAL FIXATION PREDICTS FAILURE OF THE GAMMA NAIL

Table III. Multivariate regression analysis for failure for

each point of proximal fixation, adjusted for fracture
grade and quality of reduction
Fixation point

Odds ratio (95% CI)

p-value

Lateral cortex
Tipapex distance
Greater trochanter

7.53 (2.50 to 22.71)

6.86 (2.27 to 20.73)
1.09 (0.23 to 5.20)

< 0.001
0.001
0.92

Fig. 3
Diagram showing ideal three-point proximal femoral
fixation. The two independent predictors of failure are
circled. Reproduced with modifications from
Gamma3 Trochanteric Nail 180 Operative Technique,13 with permission from Stryker UK Ltd.

When all three points of proximal fixation were positioned adequately, the failure rate was < 1% (Table II). The
addition of the number of failed points into the multivariate
analysis, with Evans grade and reduction quality, indicated
that the odds for failure with one or two missed points was
> 20 (p = 0.01). However, when the TAD and lateral cortex
points were also added into this model, the OR for one
inadequate point fell to 1.05 (95% CI 0.11 to 102.44,
p = 0.98) and for two inadequate points to 4.87 (95% CI
0.25 to 95.72, p = 0.30). This suggests that the TAD and
lateral cortex points are specific drivers of risk, rather than
simply the total number of failed points.

Discussion
Although cephalo-medullary nails provide a biomechanical
advantage over sliding hip screws for unstable fracture patterns,5-9 they may have a higher complication rate.4 Previously, the only operative factor that has been shown to
predict failure of the Gamma nail is the TAD,12 and our
findings confirm this. However, we also noted a high failure
rate when the lateral end of the lag screw was positioned
short of the lateral femoral cortex.
The multivariate analysis indicated that the TAD and the
lateral cortex points were independent predictors of failure,
with ORs of approximately 7. We did not find a significant
association between failure and the most proximal portion
VOL. 95-B, No. 6, JUNE 2013

829

of the nail lying distal to the cortex of the greater trochanter,

suggesting that this is the least important of the three points
of proximal fixation. This finding may, however, have been
subject to bias due to measurement error. The radiological
projection of the relative positions of the proximal end-cap
and the greater trochanter may have resulted in false positives for inadequacy, lowering the failure rate for this point.
Furthermore, the observer reliability of this point was not as
good as for the TAD and the lateral cortex points. Overall,
the lowest rate of mechanical failure in this series was when
all three points of proximal fixation were positioned correctly (< 1%). Therefore, despite the apparent lack of importance regarding the greater trochanteric cortical point, we
would emphasise the importance of aiming to achieve threepoint proximal fixation as a rule (Fig. 3).
Although three-point fixation is a mechanical principle,
we could find no biomechanical or clinical studies that have
investigated its role in this context. Most biomechanical
evaluations of the Gamma nail have focused on the inherent differences between it and a sliding hip screw, particularly with respect to the lever arm on the fracture, cut-out
from the head, stress-shielding at the calcar, and dynamic
sliding of the lag screw.5,6,10,12,16,18 Study of the individual
failures in Table I offers potential explanations for our findings. For modified Evans type 3 fractures the lateral femoral cortex is intact, and so one would expect TAD to be the
predictor of failure rather than the lateral cortical point.
For the uncommon modified Evans type 4 fractures the lateral cortex is not in continuity with the head fragment, so
again TAD would be the most important predictor of failure, and this again appeared to be the case. For modified
Evans type 5 fractures (reverse oblique or transverse) the
lateral cortex is likely to be in continuity with the head and
greater trochanter, therefore its fixation would enhance the
stability of the construct and prevent collapse into varus.
Indeed, the lateral cortex point was inadequate in seven of
the eight failures in type 5 fractures.
The strengths of this study include the large sample size
of 223 patients over a five-year period, the reliability of the
measurements made and the multivariate analysis, which
adjusted for important confounding variables. Age,10,19
gender,10,19 fracture classification,10,20 quality of fracture
reduction21 and implant type (long nail versus short)16 have
all been associated with inferior outcomes after fracture of
the proximal femur.
The study also has a number of limitations. We were
unable to accurately account for the grade of operating
surgeon. In addition the mechanism of injury was not
accounted for, but the vast majority (> 90%) of patients
were aged > 60 years. Therefore it is likely most injuries
resulted from low-energy injuries occurring in osteoporotic bone.1 Although it would be desirable to follow all
patients to radiological union, this was not possible
because of the volume of work and limited clinical
resources. However, the population in this study
was non-migratory, and given that all but one of the

830

S. G. F. ABRAM, T. C. B. POLLARD, A. J. M. D. ANDRADE

mechanical failures occurred within one year of surgery

(Table I), it is likely that all patients who had been discharged at six weeks and who then developed complications would have been referred back to our unit.
This study has demonstrated that mechanical failure of the
Gamma nail in peri-trochanteric femoral fractures is rare
(< 1%) when three-point proximal fixation is achieved
(Fig. 3). When proximal fixation is inadequate, failure rates
increase, independently of important confounding variables.
The strongest predictors of failure are positioning the lateral
end of the lag screw short of the lateral cortex, and having a
TAD > 25 mm. These findings are in keeping with the biomechanics of the fracture pattern and fixation construct.
This paper shows the importance of the position of the lateral end of the lag screw relative to the lateral cortex. Adherence to these simple technical points will minimise the risk of
fixation failure in this vulnerable group of patients.

Supplementary material
A further opinion by Dr Cyril Mauffrey is available
with the electronic version of this article on our website at www.boneandjoint.org.uk/site/education/further_op
No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.
This article was primary edited by S. Hughes and first-proof edited by G. Scott.

References
1. Currie C, Partridge M, Plant F, et al. The National Hip Fracture Database. National
Report 2011. http://www.nhfd.co.uk/003/hipfracturer.nsf/NHFD NationalReport
2011_Final.pdf (date last accessed 25 February 2013).

4. Parker MJ, Handoll HH. Gamma and other cephalocondylic intramedullary nails
versus extramedullary implants for extracapsular hip fractures in adults. Cochrane
Database Syst Rev 2010;9:CD000093.
5. Curtis MJ, Jinnah RH, Wilson V, Cunningham BW. Proximal femoral fractures: a
biomechanical study to compare intramedullary and extramedullary fixation. Injury
1994;25:99104.
6. Rosenblum SF, Zuckerman JD, Kummer FJ, Tam BS. A biomechanical evaluation
of the Gamma nail. J Bone Joint Surg [Br] 1992;74-B:352357.
7. Halder SC. The Gamma nail for peritrochanteric fractures. J Bone Joint Surg [Br]
1992;74-B:340344.
8. Butt MS, Krikler SJ, Nafie S, Ali MS. Comparison of dynamic hip screw and
gamma nail: a prospective, randomized, controlled trial. Injury 1995;26:615618.
9. Barton TM, Gleeson R, Topliss C, et al. A comparison of the long gamma nail with
the sliding hip screw for the treatment of AO/OTA 31-A2 fractures of the proximal part
of the femur: a prospective randomized trial. J Bone Joint Surg [Am] 2010;92-A:792
798.
10. Bojan AJ, Beimel C, Speitling A, et al. 3066 consecutive Gamma Nails: 12 years
experience at a single centre. BMC Musculoskeletal Disorders 2010;11:133.
11. Baumgaertner MR, Curtin SL, Lindskog DM, Keggi JM. The value of the tip-apex
distance in predicting failure of fixation of peritrochanteric fractures of the hip. J Bone
Joint Surg [Am] 1995;77-A:10581064.
12. Geller JA, Saifi C, Morrison TA, Macaulay W. Tip-apex distance of intramedullary devices as a predictor of cut-out failure in the treatment of peritrochanteric
elderly hip fractures. Int Orthop 2009;34:719722.
13. No authors listed. Stryker. Gamma3 Trochanteric Nail 180 Operative Technique.
http://www.osteosynthesis.stryker.com/medias/pdf/B030008_Rev_8_Gamma3_180.pdf
(date last accessed 16 April 2013).
14. Kyle RF, Gustilo RB, Premer RF. Analysis of six hundred and twenty-two intertrochanteric hip fractures. J Bone Joint Surg [Am] 1979;61-A:216221.
15. Evans EM. The treatment of trochanteric fractures of the femur. J Bone Joint Surg
[Br] 1949;31-B:190203.
16. Hesse B, Gchter A. Complications following the treatment of trochanteric fractures with the gamma nail. Arch Orthop Trauma Surg 2004;124:692698.
17. Petrie A. Statistics in orthopaedic papers. J Bone Joint Surg [Br] 2006;88-B:11211136.
18. Haynes RC, Pll RG, Miles AW, Weston RB. Failure of femoral head fixation: a
cadaveric analysis of lag screw cut-out with the gamma locking nail and AO dynamic
hip screw. Injury 1997;28:337341.
19. Zuckerman JD. Hip fracture. N Engl J Med 1996;334:15191525.

2. Thakar C, Alsousou J, Hamilton TW, Willett K. The cost and consequences of

proximal femoral fractures which require further surgery following initial fixation. J
Bone Joint Surg [Br] 2010;92-B:16691677.

20. Kim WY, Han CH, Park JI, Kim JY. Failure of intertrochanteric fracture fixation
with a dynamic hip screw in relation to pre-operative fracture stability and osteoporosis. Int Orthop 2001;25:360362.

3. Palmer SJ, Parker MJ, Hollingworth W. The cost and implications of reoperation
after surgery for fracture of the hip. J Bone Joint Surg [Br] 2000;82-B:864866.

21. Davis TR, Sher JL, Horsman A, et al. Intertrochanteric femoral fractures: mechanical failure after internal fixation. J Bone Joint Surg [Br] 1990;72-B:2631.

THE BONE & JOINT JOURNAL