The Journal of TRAUMA威 Injury, Infection, and Critical Care

Femoral Shortening After Surgical Treatment of Trochanteric

Fractures in Nongeriatric Patients
Patrick Platzer, MD, Gerhild Thalhammer, MD, Gerald E. Wozasek, MD, and Vilmos Vécsei, MD

Background: Femoral shortening is a 3.2 years (2–5 years) after trauma. Frac- In patients with fracture types 31 A2 and
well-known clinical finding after surgical tures were classified by the AO/OTA sys- A3 femoral shortening was found to be
treatment of per- and intertrochanteric tem and divided into sub-types 31 A1, 31 more severe than in patients with fracture
fractures. Particularly, in geriatric pa- A2, and 31 A3. Two different implants type 31 A1. Additionally, in unstable frac-
tients with poor bone quality and unstable (dynamic hip screw and cephalomedullary ture types (31 A2 and A3) femoral short-
fracture types, secondary compression of nail) were used for operative treatment, ening was significantly larger, if patients
these fractures often leads to length in- mainly depending on type and stability of were treated by dynamic hip screw than
equality of the lower limbs. In younger the fracture. Femoral shortening was an- by a cephalomedullary nail.
patients with good bone quality and mo- alyzed by standardized lower extremity Conclusion: Femoral shortening af-
bilization with delayed weight bearing, radiographs measuring the distance from ter operative treatment of per- and inter-
limb length shortening is expected to be a the top of the femoral head to the center trochanteric fractures was found to be a
rare complication. The purpose of this of a line drawn between the most distal common clinical finding in nongeriatric
study was to analyze incidence and degree part of the medial and lateral femoral patients. Nearly half of them showed a
of femoral shortening in patients younger condyles. lower limb length inequality after fracture
than 60 years of age after fixation of dif- Results: Fifty-seven patients were fixation. The degree of the shortening was
ferent types of per- and intertrochanteric treated by a cephalomedullary nail, 38 by rather low and depended mainly on the
fractures. In addition, we compared the re- dynamic hip screw. Femoral shortening fracture type. Comparing the two differ-
sults of two different implants, which were was seen in 46 patients (48%) with a mean ent implants used for operative treatment,
used for operative treatment. value of 11 mm. Twenty-two patients had a cephalomedullary nail was more suc-
Methods: Ninety-five patients, younger a femoral inequality of less than 10 mm, cessful in preventing limb length discrep-
than 60 years of age, were evaluated for 17 patients an inequality between 10 mm ancy in unstable fracture types than
femoral shortening after surgical treat- and 20 mm, and 7 patients a shortening of dynamic hip screw.
ment of per- and intertrochanteric frac- more than 20 mm. Statistical analysis re- Key Words: Trochanteric fractures,
tures between 1997 and 2002. Follow-up vealed that fracture type and implant had Operative treatment, Femoral shortening,
examinations took place at an average of a significant influence on the shortening: Comparison of implants.
J Trauma. 2008;64:982–989.

F
emoral shortening is a well-known clinical finding after ture compression.7–16 Standard implants for operative treatment
surgical treatment of per- and intertrochanteric fractures.1– 4 of these fractures usually consists of a sliding hip screw in
Particularly in geriatric patients with poor bone quality and combination with a side plate or an intramedullary femoral nail.
unstable fracture types, secondary compression of these frac- The sliding hip screw combined with a side plate which is fixed
tures, resulting from early mobilization with immediate weight on the lateral cortex is commonly used for stabilization of
bearing, might lead to length inequality of the lower limbs.1,2,5,6 minimally displaced and stable fractures, whereas intramedullary
In younger patients with better bone quality and mobilization devices are preferred for patients with unstable fractures.4,7,8,10,12,17
with delayed weight bearing, limb length shortening is expected The purpose of this study was to analyze femoral short-
to be a rare complication. ening in nongeriatric patients after surgical treatment of per-
Several devices for fixation of per- and intertrochan- and intertrochanteric fractures. In detail, we retrospectively
teric fractures have been developed to ensure mechanical analyzed the incidence and degree of limb length shortening
stability to prevent secondary loss of reduction and frac- after fixation of different types of per- and intertrochanteric
fractures in patients younger than 60 years of age and com-
Submitted for publication July 30, 2006. pared the results of two different implants (dynamic hip
Accepted for publication February 6, 2007. screw and cephalomedullary nail), which were used for op-
Copyright © 2008 by Lippincott Williams & Wilkins erative treatment.
From the Department of Traumatology, University of Vienna Medical
School, Vienna, Austria.
Address for reprints: Gerald E. Wozasek, MD, or Patrick Platzer, MD, PATIENTS AND METHODS
Department of Traumatology, University of Vienna Medical School, Wae- This study reviewed admission data and trauma registry
hringer Guertel 18-20, A-1090 Vienna, Austria; email: gerald.wozasek@ of the Vienna General Hospital, University of Vienna Med-
meduniwien.ac.at or [email protected]. ical School and identified all trauma patients (n ⫽ 1,292)
DOI: 10.1097/TA.0b013e3180467745 with per- and intertrochanteric fractures of the hip that were

982 April 2008

 Femoral Shortening in Patients

admitted to this level I trauma center between 1997 and 2002. (intertrochanteric).18 Fifty-six patients sustained a fracture of
The trauma registry of the Vienna General Hospital is a sub-type 31 A1, 28 patients showed a fracture of sub-type 31
prospectively gathered database that was established for reg- A2, and 11 patients were identified with fracture type 31 A3.
istration of injury characteristics (type, mechanism, etc.) and Fractures of sub-type A1 were simple pertrochanteric frac-
demographic data of trauma patients. The criteria for inclu- tures and considered stable, whereas fractures of subtype A2
sion in our study were patients aged 60 years or younger after and A3 represented multifragmentary pertrochanteric and in-
surgical treatment of per- and intertrochanteric fractures of tertrochanteric fractures and were categorized as unstable.
the femur. Their dataset was examined for completeness and After the AO/OTA classification, we had 56 patients (59%)
accuracy. Collected data included variables, such as age, with stable fractures and 39 patients (41%) with unstable
gender, mechanism of injury, fracture type, type of implant fractures.
and operative techniques, intra- and postoperative complica- For operative treatment, two different implants (dynamic
tions, postoperative rehabilitation, as well as clinical and hip screw and cephalomedullary nail) were used, depending
radiographic follow-up examination. mainly on type and stability of the fracture (see Figs. 1 and 2).
Patients with incomplete dataset, patients with patho- Fifty-seven patients (60%) were treated by a cephalomedul-
logic fractures, as well as patients who had undergone sur- lary nail [Gamma nail (GN), 1st generation, Stryker], 38
gery of the ipsilateral or contralateral femur earlier on were patients (40%) by dynamic hip screw (DHS, Synthes). In
excluded from this study. Exclusion criteria also contained stable fractures (A1) both implants were used equally (29
patients who sustained combined cervical and pertrochanteric GN/26 DHS), whereas in unstable fracture types (A2 and A3)
fractures and patients with combined per- and subtrochanteric cephalomedullary nailing was used more commonly (28
fractures. GN/12 DHS) for fixation (see Table 1).
From a database of 114 patients aged 60 years or younger Operative techniques for insertion and fixation have been
after surgical treatment of per- and intertrochanteric fractures, well described for both implants. In patients treated by DHS,
five patients were excluded from this study because of surgery the length of the hip screw ranged from 90 mm to 100 mm,
of the ipsilateral or contralateral femur (n ⫽ 3) earlier on, due to and the side plate was 130° to 135° with four or six holes. In
a combined per- and subtrochanteric fracture (n ⫽ 1) or due to cases of cephalomedullary nailing, the hip screw measured 95
incomplete dataset (n ⫽ 1). mm to 105 mm and the nail had a diameter of 11 mm to 12
One hundred four patients met the criteria for inclusion mm. All cephalomedullary nails had a length of 180 mm
in this study and were invited to assess the femoral shortening
after fixation of different types of per- and intertrochanteric
fractures and to compare the results of two different implants
that were used for operative treatment. Ninety-five patients
answered our invitation and had a complete clinical and
radiologic follow-up examination, whereas nine patients did
not respond.
Ninety-five patients, who were finally enrolled and ex-
amined, showed an average age of 51.6 years (22– 60) with a
majority of men (65%). There were 33 women with an
average age of 50.9 years (22– 60) and 62 men with an
average age of 52.2 years (25– 60). Fifty-one patients sus-
tained a fracture of right femoral bone, 44 patients had a
fracture of the left femur. Follow-up examinations took place
at a mean time of 3.2 years (⫾0.8) after trauma.
Clinical records showed several mechanisms of injury:
The injuries resulted from sports-related injuries in 41% (n ⫽
39), motor-vehicle injuries in 32% (n ⫽ 30), falls in 18%
(n ⫽ 17), and other mechanisms in 9% (n ⫽ 9). In 73 patients
(77%), the per- or intertrochanteric fractures were diagnosed
as isolated injuries, whereas in 22 patients (23%) we found
significant concomitant injuries.

Fracture Classification, Implants, and Operative

Techniques
Fractures were classified according to the AO/OTA sys-
tem and divided into sub-types 31 A1 (pertrochanteric sim-
ple), 31 A2 (pertrochanteric multifragmentary) and 31 A3 Fig. 1. Gamma nail (Stryker).

Volume 64 • Number 4 983

 The Journal of TRAUMA威 Injury, Infection, and Critical Care

third day after operation. Walking with gradual weight bear-

ing was allowed according to the radiographic status. Patients
with stable fractures were generally mobilized with full
weight-bearing as tolerated, patients with unstable fracture
types were advised not to bear weight fully until a minimum
of 6 weeks.
Clinical and radiographic follow-up examinations were
performed 3, 6, and 12 months after surgery. For clinical
evaluation mobility, range of motion and local pain in the hip
were routinely assessed. For standard radiographic assess-
ment, accurate anteroposterior (AP) and lateral radiographs
of the injured hip and AP overview of the pelvis were ob-
tained at each follow-up control. Changes in the position of
the implants (eg, migration or sliding of the hip screw, plate
loosening, femoral malrotation, etc) and the extent of fracture
union were noted.

Radiographic Assessment of Femoral Shortening,

and Clinical Examination
Radiologic follow-up examination within our study-
protocol consisted of standard radiographs of the hip (AP,
lateral and AP overview of the pelvis) and additional radio-
graphs depicting both lower extremities from the pelvis to the
ankle. These lower extremity radiographs were performed
trying to keep both legs parallel in extension, parallel to the
Fig. 2. Dynamic hip screw (Synthes). film cassette, and in neutral rotation.
Femoral shortening was analyzed by measuring the dis-
tance from the top of the femoral head to the center of a line
Table 1 Method of Fixation by Fracture Type
drawn between the most distal part of the medial and lateral
(AO Form)
femoral condyles (see Fig. 3). The length of the uninjured
Fracture Type (AO form) contralateral femur was used as a reference. Assuming that
Method of Fixation
31 A1 31 A2 31 A3 both femurs had the same length before injury, femoral short-
Dynamic hip screw 26 9 3 ening was diagnosed, if the measured distance was shorter on
Gamma nail 29 19 9 the surgically stabilized side. To achieve exact measurement
Total (N ⫽ 95) 55 28 12 of femoral length, a measuring staff was integrated into
full-leg radiographs, so that the length of both femoral bones
could be defined exactly. Measurement of femoral length was
(short nails) and were distally locked. In 92 patients, fixation standardized by using a computerized digital photo system,
was performed after closed reduction of the fracture, whereas which allows exact adjustment of magnification of the radio-
in three patients open reduction was required. All operative graphs. Measurements of femoral length of all radiographs
procedures were performed either under general or regional were repeated by another surgeon, to assess interobserver
anesthesia with the patients in supine position on the operat- variability.
ing table. Clinical assessment within our study protocol included
evaluation of mobility, range of motion and presence of pain
Hospital Course, Rehabilitation, and Follow-up in the hip. Patients were additionally interviewed for limita-
Examinations tions or impairments in activities of daily living and if they
Patients were routinely given prophylactic antibiotic ther- had returned to their preinjury activity level. To quantify the
apy intraoperatively and in some cases (n ⫽ 19) for 3 days after clinical results, we used the Harris-Hip-Score (HHS) which
surgery. All patients were treated with low-molecular-weight allows to grade the functional outcome of the affected hip by
heparin as prophylactic anticoagulation from the day of admis- a maximal score of 100 points.19,20 Clinical outcome was
sion until advanced mobilization status. The incidence of any rated as “excellent” in patients with 90 points or more, as
postoperative complications was meticulously recorded and rea- “good” in patients with 80 points or more and as “satisfac-
sons for any reoperations were noted. tory” if the patients were satisfied and achieved at least two
Postoperative rehabilitation was identical for both im- thirds of points of the score. “Unsatisfactory” results were
plants and mobilization was routinely begun on the second or those that failed to meet these criteria.

984 April 2008

 Femoral Shortening in Patients

Fig. 3. Method of Measurement.

Statistical Analysis simple t tests. Statistical significance was defined as p ⬍

We analyzed the influence of different factors on femoral 0.05.
shortening after surgical treatment and considered the short-
ening with values of 0 cm, 0 to 1 cm, 1 to 2 cm, and more than RESULTS
2 cm as an ordered variable. The influencing factors within Ninety-five patients with an average age of 51.6 years
this study were age, gender, mechanism of injury, fracture (22– 60) were examined for femoral shortening at a mean
type, type of implant, and operative technique, intra- and time of 2.7 years (⫾0.8) after trauma. Fifty-seven patients
postoperative complications and postoperative rehabilitation. (60%) were treated by a cephalomedullary nail, 38 patients
For univariate analysis, we performed a simple regression (40%) by dynamic hip screw. Statistical analysis showed, that
model for age and a simple analysis of variance for all other both implants were used equally in stable fractures (A1),
variables using proc general linear model of SAS (SAS In- whereas in unstable fracture types (A2 and A3) cephalomed-
stitute, Inc., Cary, NC). To analyze differences between pa- ullary nailing was used significantly more often for fixation
tients treated by GN and those treated by DHS, we applied ( p ⬍ 0.05; see Table 1). Stable fractures were more common

Volume 64 • Number 4 985

 The Journal of TRAUMA威 Injury, Infection, and Critical Care

than unstable fractures, as we had 55 patients (58%) with by a cephalomedullary nail, 18 patients (39%) by dynamic
fracture type 31 A1 and 40 patients (42%) with fracture type hip screw. Statistical results revealed that the implant used for
31 A2 or 31 A3, but this was not significant statistically ( p ⬎ fixation also had a significant effect on the shortening: In
0.05). unstable fracture types (31 A2 and A3) femoral inequality
was significantly higher, if patients were treated by dynamic
Femoral Shortening hip screw than by a cephalomedullary nail ( p ⬍ 0.05). We
Femoral shortening after surgical treatment of per- and had a mean femoral shortening of 10.6 mm after fixation by
intertrochanteric fractures was found in 46 patients (48%) a cephalomedullary nail and a mean value of 22.2 mm after
with a mean value of 11 mm (⫾4.4 mm; range, 4 –32 mm). stabilization by dynamic hip screw. In stable fracture types
Twenty-two patients had a femoral inequality of less than 10 (31 A1), we found no significant differences between cepha-
mm, 17 patients an inequality between 10 mm and 20 mm lomedullary nailing and fixation by dynamic hip screw, as we
and 7 patients a shortening of more than 20 mm. Two of them had a mean femoral inequality of 5.5 mm in the GN group
showed a femoral length discrepancy of more than 30 mm and a mean shortening of 5.7 mm in the DHS group. The
after stabilization of a multifragmentary pertrochanteric frac- overall mean femoral shortening was 0.6 mm lower after
ture in both cases. In the remaining 49 patients, (52%) fem- fixation by a cephalomedullary nail (mean, 10.7 mm) than by
oral length discrepancy could not be analyzed (see Fig. 4). dynamic hip screw (mean, 11.3 mm). Statistical results also
Thirteen (28%) of the patients with femoral length in- revealed that the incidence of femoral shortening was signif-
equality showed stable fractures (31 A1), whereas 33 patients icantly higher in the DHS group than in the GN group ( p ⬍
(72%) had unstable fractures (31 A2 and A3). Statistical 0.05): In unstable fracture types, 92% (n ⫽ 11) of the patients
analysis revealed that the fracture type had a significant treated by dynamic hip screw and 78% (n ⫽ 22) of the
influence on the shortening: univariate analysis of variance patients stabilized by a cephalomedullary nail showed an
between AO sub-types showed that femoral shortening was inequality of femoral lengths. In stable fractures, femoral
more severe in patients with unstable fracture types (31 A2 length inequality was found in 27% (n ⫽ 7) of the patients
and A3) than in patients with stable fractures (31 A1) ( p ⬍ treated by dynamic hip screw and in 21% (n ⫽ 6) of the
0.05). The mean femoral shortening was 5.6 mm for sub-type patients fixed by a cephalomedullary nail.
31 A1, 15.9 mm for sub-type 31 A2, and 16.9 mm for fracture None of the other variables had any significant influence
type 31 A3. In the comparison of fracture types 31 A2 and 31 on femoral shortening. Repeated measurements of femoral
A3, femoral inequality was marginally higher in sub-type 31 length by a second investigator showed a mean variability of
A3, but remained on the same level of significance. Statistical 0.4 mm (range, 0 –1 mm).
results also showed that femoral shortening occurred more
frequently in patients with unstable fractures ( p ⬍ 0.05), as Functional Outcome, Radiographic Results,
83% (n ⫽ 33) of the patients with femoral length inequality and Complications
showed fractures types 31 A2 or 31 A3. In patients with Complete clinical and radiographic follow-up examina-
stable fractures, femoral length discrepancy was found in tions were achieved in 92 patients (97%) at 3, 6, and 12
only 24% (n ⫽ 13) of the patients. months after trauma. In the remaining three patients (3%), the
Regarding the type of implant that was used for fixation, assessment at 6 months was missing, as the patients did not
28 (61%) of the patients with femoral shortening were treated come for a follow-up examination.

60

49
50

40

30
22
20 17

10
5
2
0
< 10 mm 10 - 20 mm 20 - 30 mm > 30 mm
Femoral length equality Femoral length shortening
Fig. 4. Femoral shortening.

986 April 2008

 Femoral Shortening in Patients

Functional results revealed that 88 patients (93%) had was finally found in nine of them, but hip screw cutout was
returned to their preinjury activity level 1 year after trauma. not seen in any of these cases and no reoperations were
Seven patients (7%) complained about limitations in daily necessary, either.
living, particularly in performing certain sports activities Comparing radiographs between the day of discharge
(e.g., climbing, mountain-biking, or skiing). None of the and 1-year follow-up, a mild degree of migration or sliding
patients had to use any walking aids and we did not see any of the lag screw was noted in 56 patients (59%). The
relevant functional impairments in hip function. Average hip number of patients with changes in position of the lag
flexion was 120° (range, 90 –130°) without any significant screw was distributed similarly between GN- and DHS-
differences to the contralateral side. External or internal ro- group. (GN: n ⫽ 31, DHS: n ⫽ 25). Malalignment or
tation measured in supine position and average abduction or malrotation was not seen in any of the patients.
adduction measured in lateral position did not differ notably Complications of wound infection were found in four
between both hips, either. Sixteen patients (17%) complained patients (4%). Three of them had superficial infections and
of occasional pain during long distance walking (n ⫽ 6) or in were treated successfully by antibiotic therapy for an average
the change of weather (n ⫽ 10), but were generally satisfied of 7 days. One patient developed a deep wound infection with
with their treatment. persisting discharge, which finally healed after revision and
Using the HHS to quantify the clinical outcome of the temporary insertion of gentamicin impregnated beads. Fur-
patients, we had a mean outcome score of 93.6 points. ther postoperative complications were not seen in any of the
Seventy-nine patients (83%) achieved excellent results with patients.
90 points or more on HHS and 19 patients had good results
with 80 points or more. None of the patients had less than 80 DISCUSSION
points. Comparing the hip score results between patients with Femoral shortening after surgical treatment of per- and
femoral shortening and those without femoral length discrep- intertrochanteric fractures mainly results from secondary
ancy, we did not find any significant differences ( p ⬎ 0.05). compression and impaction of the stabilized fractures.1,2,4
Patients who were identified with a femoral length inequality Factors like bone quality, fracture type, method of fixation,
had a mean HHS of 91.9 points, whereas patients without and early postoperative mobilization have been described to
femoral shortening showed a mean HHS of 95.3 points. influence the degree of fracture compression and the inci-
Particularly in patients with a shortening of less than 1 cm or dence of lower limb length inequality.1,2,4,5 Patients with
2 cm, we had similar results compared with those without good bone quality and stable fracture types are expected to
femoral shortening. However, referring to the patients with a have a lower incidence of femoral shortening after fracture
shortening of more than 2 cm (n ⫽ 7) and more than 3 cm fixation, than patients with poor bone quality and unstable
(n ⫽ 2), we found a relevant decrease of functional results. fracture types. The method of fixation and postoperative
Compared with patients without femoral length inequality, mobilization may also play an important role in predicting
they had a significantly worse outcome with a mean HSS of increased risk for limb length shortening. Implants and oper-
86.3 points ( p ⬍ 0.05). Particularly, the two patients with a ative techniques are required to ensure mechanical stability,
length discrepancy of more than 3 cm had worse functional to allow early postoperative rehabilitation without secondary
results presenting a score result of 81 points in one case and loss of reduction and fracture compression.2,8,10,12,17
84 points in the other case. The main purpose of this study was to analyze incidence
Radiographic results revealed that all stabilized fractures and degree of femoral shortening in nongeriatric patients after
had healed uneventfully 3 months after surgery. We had no surgical stabilization of per- and intertrochanteric fractures.
cases of delayed union or nonunion in any of the patients. In planning this study, we assumed a low incidence rate and
Technical failures of fixation were found in five patients degree of this complication, as nongeriatric patients were
(5%). Two of them were treated by a cephalomedullary nail, expected to present good bone quality and predominantly
three of them by dynamic hip screw. In four patients the lag stable fractures. In cases of unstable fracture types, the pa-
screw had been placed too high in the femoral head and in the tients were mobilized with gradual weight bearing for a
fifth patient, we found an inaccurate placement of the side minimum of 6 weeks, to prevent secondary fracture compres-
plate. All five patients were mobilized without any weight sion and discrepancy of femoral lengths.
bearing for an average of 7 weeks. Relevant loss of reduction, However, the results of our study demonstrate that we
hip screw cutout or plate loosening was not seen in any of had a surprisingly high rate of limb length inequality after
these patients and no reoperations were necessary. surgical treatment of per- and intertrochanteric fractures, as
Incorrect anatomic reduction was noted in 17 patients femoral shortening was found in 48% (n ⫽ 46) of the exam-
(18%). Ten of them were fixed in varus position, four of them ined patients. A comparison of data in literature showed that
in valgus position, and three patients had a slight malrotation limb length discrepancy might be a common clinical finding
after surgical stabilization. Ten of them were treated by a after fixation of these fractures, but rather in geriatric patients
cephalomedullary nail, seven of them by dynamic hip screw. with poor bone quality.1,2,4,6,12 The high incidence rate of
Further loss of reduction and slight migration of the hip screw femoral shortening in our series might be attributed to the fact

Volume 64 • Number 4 987

 The Journal of TRAUMA威 Injury, Infection, and Critical Care

that 41% of the examined patients had sustained unstable In patients with stable fractures, the method of fixation did
fractures. An analysis of our results revealed, that the vast not reveal any relevant influence on femoral shortening. Both
majority of the patients (83%) with unstable fracture types groups had equal results relating to the degree of limb length
also showed an inequality of femoral lengths, whereas in inequality. We had a mean femoral shortening of 5.5 mm after
patients with stable fractures, femoral shortening was found fixation by a cephalomedullary nail and a mean femoral inequal-
in only 24%. ity of 5.7 mm after surgical treatment by dynamic hip screw.
The degree of femoral shortening was rather low accord- Regarding the incidence of femoral shortening in patients with
ing to our initial hypothesis. We found a length inequality stable fractures, we had a slightly lower rate of length discrep-
with a mean value of 11 mm. Half of the patients with ancy if patients were treated by a cephalomedullary nail, but this
femoral shortening showed an inequality of less than 10 mm, remained on the same level of significance.
one-third of the patients had an inequality between 10 mm In summary, we had good functional and radiographic
and 20 mm and only 15% showed a discrepancy of more than results with both implants. The vast majority of patients were
20 mm. The degree of the shortening was mainly influenced satisfied with their treatment and we did not see any relevant
by the stability of the fracture, as our analysis revealed. functional impairments of hip and knee function or compen-
Femoral length inequality was found to be more severe in satory strategies in the walking pattern.29 We had no cases of
patients with unstable fractures than in patients with stable delayed union or nonunion and no reoperations were neces-
fractures. In summary, referring to the low degree of limb sary due to technical or implant failures. The need for internal
length discrepancy, we had a notable difference comparing fixation and early mobilization of patients with per- and
patients with stable fractures to those with unstable fractures, intertrochanteric femoral fractures is generally accepted, but
as femoral shortening was on average 10.8 mm higher in any advocacy for one of the two implants is not sustainable
patients with fracture types 31 A2 or 31 A3. within this study, as this was not a prospective trial and we
The second purpose of this study was to compare the two did not have any control groups. We may also be criticized
different implants (cephalomedullary nail and dynamic hip for other potential weaknesses linked to a retrospective de-
sign, but as this study contains patients from a prospectively
screw) that were used for operative treatment, particularly in
gathered database with previously analyzed exclusion crite-
preventing limb length shortening after fixation of per- and
ria, and as prospectively randomized trials are extremely rare
intertrochanteric fractures. Several previous studies have re-
within these patients, thoroughly important and interesting
ported on advantages and disadvantages of both devices and the
information for clinical science can be given.
indications for operative treatment.4,7–11,17,21–26 In stable frac-
tures both implants are considered to produce equally reliable
results, whereas in unstable fracture types the intramedullary CONCLUSION
device is given a theoretical advantage.4,7,10,12,17,24 Particularly Femoral shortening after operative treatment of per- and
with the prospect of preventing severe fracture compression and intertrochanteric fractures was found to be a common clinical
shortening of the femoral length, cephalomedullary nails have finding in patients aged 60 years or younger, as nearly half of
been reported to be a stronger device, which allows more accu- them showed a lower limb length inequality after surgery.
rate and rigid fixation.4,7,10,24,27 However, other randomized The degree of the shortening was rather low and depended
controlled trials have not found any significant differences in mainly on the fracture type. Comparing the two different
functional or radiographic outcome when comparing the two implants used for operative treatment, a cephalomedullary
methods of fixation.7,10,12,13,25,28 nail was more successful in preventing limb length discrep-
In our series, both implants were used equally in stable ancy in unstable fracture types than dynamic hip screw. In
fractures, whereas in unstable fracture types, the majority of summary, we had a good clinical and radiographic outcome,
the patients were treated by a cephalomedullary nail, after the and based on the results of the HHS, we did not see any
department policy that requires intramedullary fixation of relevant influence on posture and gait.
unstable per- and intertrochanteric fractures. Regarding inci-
dence and degree of femoral length inequality, an analysis of
our results revealed that the method of fixation had a signif- REFERENCES
icant influence on the shortening, if patients had sustained 1. Bendo JA, Weiner LS, Strauss E, Strauss E, Yang E. Collapse of
intertrochanteric hip fractures fixed with sliding screws. Orthop Rev.
unstable fractures. In these patients, fixation by cephalomed-
1994;(Suppl):30 –37.
ullary nailing was more successful in preventing limb length 2. Olson O, Ceder L. Femoral shortening in intertrochanteric fractures.
discrepancy, than by dynamic hip screw. We found a mean A comparison between the Medoff sliding plate and the compression
femoral shortening of 10.6 mm after stabilization by a cepha- hip screw. J Bone Joint Surg Br. 2001;83:572–578.
lomedullary nail and a mean value of 22.2 mm after stabili- 3. Song KM, Halliday SE, Little DG. The effect of limb-length
inequality on gait. J Bone Joint Surg Am. 1997;79:1690 –1698.
zation by dynamic hip screw. In addition, we also had a lower 4. Tarantino U, Oliva F, Impagliazzo A, et al. A comparative prospective
incidence of femoral shortening, if patients were treated by a study of dynamic variable angle hip screw and Gamma nail in
cephalomedullary nail. intertrochanteric hip fractures. Disabil Rehabil. 2005;27:1157–1165.

988 April 2008

 Femoral Shortening in Patients

5. Davis TRC, Sher JL, Horsman A, Simpson M, Porter BB, Checketts 17. Schipper IB, Marti RK, van der Werken C. Unstable trochanteric
RG. Intertrochanteric femoral fractures: mechanical failure after femoral fractures: extramedullary or intramedullary fixation. Review
internal fixation. J Bone Joint Surg Br. 1990;72:26 –31. of literature. Injury. 2004;35:142–151.
6. Obrant JS, Bengner U, Johnell O, Nilsson BE, Sernbo I. Increasing age- 18. Mueller ME, Nazarian S, Koch P. The Comprehensive Classification
adjusted risk of fragility fractures: a sign of increasing osteoporosis in of Fractures of the Long Bones. Berlin: Springer-Verlag; 1987:120 –
successive generations? Calcif Tissue Int. 1989;44:157–167. 121.
7. Adams CI, Robinson CM, Court-Brown CM, McQueen MM. 19. WH Harris. Traumatic arthritis of the hip after dislocation and
Prospective randomized controlled trial of an intramedullary nail acetabular fractures: treatment by mold arthroplasty. An end-result
versus dynamic screw and plate for intertrochanteric fractures of the study using a new method of result evaluation. J Bone Joint Surg
femur. J Orthop Trauma. 2001;15:394 – 400. Am. 1969;51:737–755.
8. Baumgaertner MR, Curtin SL, Lindskog DM. Intramedullary versus 20. Shields RK, Enloe LJ, Evans RE, Smith KB, Steckel SD. Reliability,
extramedullary fixation for the treatment of intertrochanteric hip validity and responsiveness of functional tests in patients with total
fractures. Clin Orthop Relat Res. 1998;348:87–94. joint replacement. Phys Ther. 1995;75:169 –176; discussion 176 –
9. Bridle SH, Patel AD, Bircher M, Calvert PT. Fixation of 179.
intertrochanteric fractures of the femur. A randomized prospective 21. Chevalley F, Gamba D. Gamma nailing of pertrochanteric and
comparison of the Gamma nail and the dynamic hip screw. J Bone subtrochanteric fractures: clinical results of a series of 63
Joint Surg Br. 1991;73:330 –334.
consecutive cases. J Orthop Trauma. 1997;11:412– 415.
10. Butt MS, Krikler SJ, Nafie S, Ali MS. Comparison of dynamic hip
22. Habernek H, Schmidt L, Frauenschuh E. Sport related proximal
screw and gamma nail: a prospective randomized controlled trial.
femoral fractures: a retrospective review of 31 cases treated in an
Injury. 1995;26:615– 618.
eight year period. Br J Sports Med. 2000;34:54 –58.
11. Habernek H, Wallner T, Aschauer E, Schmid L. Comparison of
23. Hardy DC, Descamps PY, Krallis P, Fabeck L, Smets P, Bertens
Ender nails, dynamic hip screws, and Gamma nails in the treatment
CL. Use of an intramedullary hip-screw compared with a
of pertrochanteric femoral fractures. Orthopedics. 2000;23:121–127.
compression hip-screw with a plate for intertrochanteric femoral
12. Saudan M, Lübbeke A, Sadowski C, Riand N, Stern R, Hoffmeyer P.
Pertrochanteric fractures: is there an advantage to an intramedullary fractures: a prospective, randomized study of hundred patients.
nail? A randomized prospective study of 206 patients comparing the J Bone Joint Surg Am. 1998;80:618 – 630.
dynamic hip screw and proximal femoral nail. J Orthop Trauma. 2002; 24. Sommers MB, Roth C, Hall H, et al. A laboratory model to evaluate
16:386 –393. cutout resistance of implants for pertrochanteric fracture fixation.
13. Parker MJ, Handoll HH. Gamma and other cephalocondylic J Orthop Trauma. 2004;18:361–368.
intramedullary nails versus extramedullary implants for 25. Crawford CH, Malkani AL, Cordray S, Roberts CS, Sligar W. The
extracapsular hip fractures in adults. Cochrane Database Syst trochanteric nail versus the sliding hip screw for intertrochanteric hip
Rev. 2005;19:CD000093. fractures: a review of 93 cases. J Trauma. 2006;60:325–329.
14. Banan H, Al-Sabti A, Jimulia T, Hart AJ. The treatment of unstable, 26. Parker MJ, Gurusamy K. Modern methods of treating hip fractures.
extracapsular hip fractures with the AO/ASIF proximal femoral nail Disabil Rehabil. 2005;27:1045–1051.
(PFN)— our first 60 cases. Injury. 2002;33:401– 405. 27. Rosenblum SF, Zuckerman JD, Kummer FJ, Tam BS. A
15. Lee PC, Yu SW, Hsieh PH, Su JY, Chen YJ. Bridge-plating biomechanical evaluation of the gamma nail. J Bone Joint Surg Br.
osteosynthesis of 20 comminuted subtrochanteric fractures with 1992;74:352–354.
dynamic hip screw. Chang Gung Med J. 2002;25:803– 810. 28. Williams WW, Parker BC. Complications associated with the use of
16. Yoo MC, Cho YJ, Kim KI, Khairuddin M, Chun YS. Treatment of the Gamma nail. Injury. 1992;23:291–292.
unstable peritrochanteric femoral fractures using a 95 degrees angled 29. McCaw ST, Bates BT. Biomechanical implications of mild leg
blade plate. J Orthop Trauma. 2005;19:687– 692. length inequality. Br J Sports Med. 1991;25:10 –13.

Volume 64 • Number 4 989