Conventional fracture fixation techniques and hemiarthroplasty traditionally have been recommended for

displaced, acute fractures of the femoral neck. Total hip arthroplasty has been perceived as excessively
costly with a higher risk of dislocation. Recent evidence suggests the management of acute femoral neck
fractures should be reevaluated. A number of randomized controlled trials are available comparing internal
fixation, hemiarthroplasty, and total hip arthroplasty. In a multicenter study of 450 patients, Rogmark et al.
reported 43% failures with internal fixation compared with 6% with arthroplasty. Arthroplasty patients also
had better walking ability and less pain, but dislocation occurred in 8%. Blomfeldt et al. randomized 120
cognitively normal patients to either hemiarthroplasty or total hip arthroplasty. Blood loss and operative
time were higher in the total hip group, but there were no differences in complication rates or mortality.

Harris hip scores were higher in the total hip group at both 4 and 12 months. All patients had the same
anterolateral approach, and there were no dislocations in either group. In a 4-year follow-up of this same
series of patients, Hedbeck et al. found that these differences persisted and increased, favoring total hip
replacement. Macaulay et al. compared results of 40 patients randomized to hemiarthroplasty and total hip
replacement. Operative time was only 7 minutes longer in the total hip group, although all participating
surgeons were specialized in hip arthroplasty. At 24 months, total hip patients had significantly less pain
than those with hemiarthroplasties and had significantly better SF-36 mental health and WOMAC function
scores. Only one patient with a total hip arthroplasty (5.8%) experienced dislocation requiring revision. Iorio
et al. carried out a costeffectiveness analysis of treatment methods for displaced femoral neck fractures
accounting for initial hospital costs, rehabilitation, and costs of reoperations and complications. They
concluded that cemented total hip replacement was the

most cost effective and that internal fixation was the most expensive. Although this topic remains
controversial, total hip arthroplasty is an acceptable option for treatment of acute, displaced fractures of the
femoral neck in patients who are living independently, fully ambulatory, mentally lucid, and living an active
lifestyle. Those who are less healthy, institutionalized, cognitively impaired, or require assistive devices for
ambulation are better suited for hemiarthroplasty. The selection process also depends on availability of
experienced personnel at a given institution. The risk of complications in patients with acute femoral neck
fractures is higher than in patients with osteoarthritis. Using information from the National Hospital
Discharge Survey (NHDS) over the years 1990 to 2007, Sassoon et al. found that patients having total hip
replacement for acute fracture had higher rates of mortality and pulmonary embolism compared with
osteoarthritis patients. Rates were also higher for hematoma formation, infection, and dislocation, although
these differences decreased over time. There was no difference in dislocation rate between groups during the
most recent period. Fracture patients had a longer length of stay and discharge to a rehabilitation facility at
all time periods. Specific measures can be taken to reduce the incidence of dislocation in this population. In
a series of 372 acute femoral neck fractures, Sköldenberg et al. reduced the risk of dislocation from 8% to
2% by switching from the posterolateral to the anterolateral approach. The use of anterior approaches
appears justified. If a posterior approach is used, consideration should be given to use of a larger-diameter
head and careful repair of the posterior capsule and short external rotators. Using such an approach, Konan
et al. had no dislocations in a series of 20 patients. Cementless dual-taper femoral components were used in
this series with no stem loosening and no fractures reported. Using a single taper cementless stem in 85
patients, Klein et al. reported two intraoperative fractures requiring cerclage cables and one postoperative
fracture requiring revision of the femoral component. The utility of modern cementless femoral components
in this population is not well established.

Other indications are displaced fractures of the femoral neck in patients with Paget disease (see later under
“Paget Disease) and some displaced neck or trochanteric fractures in patients with a preexisting painful
arthritic hip. Femoral neck fractures are uncommon in most patients with arthritis of the hip, however,
except for patients with rheumatoid arthritis. These patients usually have osteoporosis, and internal fixation
of displaced fractures often is unsatisfactory. For this reason, a total hip arthroplasty may be considered

FAILED HIP FRACTURE SURGERY

Internal fixation of proximal femoral fractures may fail because of nonunion, malunion, osteonecrosis, or
posttraumatic osteoarthritis. Patients with pain caused by destruction of the femoral head and acetabulum as
a result of intrusion of an internal fixation device are best treated with total hip arthroplasty. This is common
in nonunion of trochanteric and femoral neck fractures (Figs. 3-91 and 3-92).

Total hip arthroplasty for painful posttraumatic osteonecrosis of the femoral head or nonunion usually is
possible, but certain technical points are noteworthy.

Bleeding may be more extensive than usual because of the increased vascularity of the subsynovial tissue,
which is part of a reactive process secondary to the avascular bone in the head. Considerable soft-tissue
release may be necessary for exposure and restoration of limb length. If evidence of active infection is noted
at the time of removal of internal fixation devices, the total hip procedure should be abandoned.

With nonunion of a femoral neck fracture, a portion of the femoral neck may be eroded but reconstruction
usually can be accomplished easily using a standard femoral component with a long neck.

In contrast, with trochanteric nonunions, the length of the femur generally cannot be restored with a standard
implant, and a calcar replacement stem often is required (see Fig. 3-25C). Because a calcar replacement
stem simplifies the operation, allows early weight bearing, and obviates the need for union at a graft-host
junction, it is preferable to bone grafting.

Plates and screws in the proximal femur may be covered with bone and difficult to remove. Removal of
broken screws may leave a large defect in the femoral cortex that can give rise to fracture. In these cases, a
longer stem is required to bypass screw holes by approximately two bone diameters (see Fig. 3-91). Cortical
bone beneath a femoral side plate may become markedly porotic and easily perforated by reamers and
broaches. We have used cortical strut grafts over the site of the plate in these cases for protection from
fracture (see Fig. 3-92). When a cemented stem is used, an attempt should be made to occlude femoral screw
holes during cementation. Still, the cement mantle quality is typically inferior compared with intact femurs.
Additionally, with previously unstable trochanteric fractures, malunion with medial displacement at the
fracture site may produce distortion of the proximal femur and make preparation of the femur more
hazardous. Hospitalization and subsequent rehabilitation of patients with total hip arthroplasty for sequelae
of failed fixation of a hip fracture are more prolonged than for similar patients with arthroplasty for arthritic
conditions. Many patients with hip fractures have been nonambulatory for a period after the initial fracture
and become very debilitated. Complications

are more common, and the overall mortality is higher than for patients who undergo total hip arthroplasty
for arthritic conditions. Archibeck et al. reported an early complication rate of 11.8% including dislocations
and periprosthetic fractures.

ACETABULAR FRACTURES

Fractures of the acetabulum with or without dislocation of the hip, although they can become painful later,
usually are treated initially by open reduction and internal fixation. A united fracture provides better bone
for support of the acetabular component if arthroplasty should become necessary. Occasionally, primary
arthroplasty is indicated in an older osteoporotic patient who has an acetabular fracture combined with an
unreconstructable fracture of the femoral head or neck, marked articular surface impaction or comminution,
or a previously arthritic joint (Fig. 3-93).

Mears and Shirahama reported a technique of acetabular fracture fixation with braided cables combined
with acute total hip arthroplasty. Fracture union occurred in 19 patients, and there were no loose implants. In
a larger series of elderly patients with acute acetabular fractures treated with arthroplasty, Mears and
Velyvis found good or excellent outcomes in 79% of cases. Small degrees of acetabular component
migration occurred within the first 6 weeks, but no acetabular component had evidence of late radiographic
loosening. Both the fracture reduction and arthroplasty can be carried out though a single posterior
approach.

High rates of blood transfusion and prolonged operative times have been reported in other series.
Chakravarty et al. described a technique of percutaneous column fixation combined with total hip
arthroplasty to reduce operative time and blood loss associated with traditional acetabular fracture fixation
techniques. This procedure is best accomplished with collaboration of surgeons familiar with techniques of
open reduction of acetabular fractures and of complex hip arthroplasty.

In old fractures of the acetabulum treated nonoperatively, residual pelvic deformity and areas of nonunion
are common

A significant bony defect may be present posteriorly, especially if there has been a previous fracture of the
posterior wall. Judet views (see Chapter 56) of the acetabulum and a CT scan show the extent of the defect
and detect areas of nonunion not identified on plain radiographs. Failure to recognize these posterior
deficiencies often leads to placement of the acetabular component in retroversion, with subsequent
dislocation.

At the time of arthroplasty, either the acetabulum must be deepened so that the posterior edge of the cup is
supported by bone or the posterior wall must be extended by a graft consisting of part of the excised femoral
head and neck or an allograft anchored with several screws or a buttress plate.

For smaller contained defects, morselized autograft from the femoral head is adequate.

In patients with nonunion of a displaced transverse acetabular fracture or patients with extremely irregularly
shaped defects, an antiprotrusio cage with bone grafting can be considered.

If open reduction of the acetabulum has been done previously, extensive soft-tissue scarring can be
expected and exposure may be difficult.

Heterotopic ossification complicates exposure further and can cause impingement after the components are
placed. Excision of heterotopic bone is laborious and increases operative time and blood loss. Efforts to
prevent recurrence of heterotopic bone are warranted (see section on heterotopic ossification).
Previously placed internal fixation devices can be exposed during the process of reaming the acetabulum,
and it may be necessary to remove them to implant the acetabular component properly. Considerable
additional exposure may be required to remove screws and plates, risking injury to the sciatic nerve within
scarred soft tissues.

The ready availability of metal-cutting tools and screw removal instruments facilitates extraction of
previously placed implants from the interior of the acetabulum without added extraarticular exposure. The
acetabular component often can be implanted with removal of only a portion of the hardware, leaving the
remainder undisturbed. Makridis et al. conducted a meta-analysis of published series of total hip
replacement in patients with acetabular fractures. Heterotopic ossification, infection, dislocation, and nerve
injuries were the most commonly reported complications. Acetabular component survival was only 76% at
10 years. Yuan, Lewallen, and Hanssen reported no aseptic loosening with a highly porous metal acetabular
component (see Fig. 3-31), but failures related to infection were reported

PROXIMAL FEMORAL OSTEOTOMY AND DEFORMITY

Several problems may be encountered in inserting the femoral stem for arthroplasty after proximal femoral
osteotomy or when the proximal femur is otherwise deformed. Anatomic distortion and scarring from
previous surgery make the surgical exposure more hazardous. The displacement of the fragments and the
dense, cancellous bone in the femoral canal at the level of the healed osteotomy require careful reaming to
broach the obstruction and to avoid cortical perforation or fracture. A high-speed burr may be required to
remove dense intramedullary bone. Implant malposition and bony impingement resulting from the distorted
femoral architecture can lead to hip instability.

Previously placed internal fixation

devices often are covered with bone, and their removal alone constitutes a significant operation. Broken
screws are common, and the femur is prone to fracture after their removal. The femoral cement-bone
interface often is imperfect when there are multiple cortical perforations, and the durability of fixation is
compromised. If removal of hardware is complex, a staged procedure is appropriate, with the arthroplasty
being done after the soft tissues and any femoral cortical defects have healed.

Deformity may be present either in the proximal metaphyseal area or distally in the diaphysis. The location,
type, and degree of deformity are important factors in preoperative planning.

A metaphyseal valgus deformity produces a femur with a straight medial border, and conventional
metaphysealfilling cementless implants are unsuitable (Fig. 3-94). In this instance, a small cemented stem
provides a simple solution. We also have used modular stems effectively in this situation (see Fig. 3-94).
Alternatively, the deformed neck segment may be resected and replaced with a calcar replacement type of
stem (see Fig. 3-25C).

When there has been a previous metaphyseal varus osteotomy, conventional implants generally can be used,
although overhang of the greater trochanter may require trochanteric osteotomy to avoid fracture or varus
stem alignment.

Metaphyseal rotational deformities usually can be managed by cementing a slightly smaller stem in the
proper rotational alignment or by use of a cementless stem with diaphyseal fixation alone.

Repeat osteotomy at the metaphyseal level should be avoided because the proximal fragment would be
small, and it would be difficult to achieve stability at the osteotomy site. The application of supplemental
plates and strut grafts to the metaphysis often is technically unsatisfactory, and the added bulk increases the
risk of bony impingement and dislocation. If repeat osteotomy is required to manage a metaphyseal
deformity, it generally should be done at the subtrochanteric level where fixation is more reliable.

Diaphyseal deformities generally have a more substantial effect on implant placement.

For deformities in the distal part of the diaphysis, a short stem can be used and the deformity need not be
directly treated

. If the deformity is in the subtrochanteric area, however, careful preoperative planning is mandated.

Minor angular and translational deformities usually can be negotiated with a cemented stem of a size
smaller than usual to preserve an adequate circumferential cement mantle. If angular deformity is
significant, however, or translation of greater than 50% is present, repeat osteotomy is needed (Fig. 3-95).
Surgery can be done in two stages, although the introduction of cementless stems has simplified the
operation and made union predictable with a single-stage procedure. Osteotomy also provides direct access
to dense intramedullary bone at the previous surgical site, simplifying its removal. Stable fixation must be
obtained at the osteotomy site for union to occur. A fluted or extensively porous-coated stem is needed to
achieve distal fixation, and a precise fit in both fragments must be obtained to provide rotational stability. If
this cannot be achieved with the stem alone, a cortical strut or a plate must be added. An oblique or step-cut
osteotomy is intrinsically more stable than a transverse one, although the procedure is more complex. This is
particularly true if there is rotational malalignment that must be corrected. Cement can be used for stem
fixation, but the cement inevitably extrudes into the osteotomy site and jeopardizes

union. For this reason, we prefer cementless femoral components when a femoral osteotomy is necessary.
Mortazavi et al. reported a series of 58 patients with proximal femoral deformity who had hip arthroplasty.
Nonprimary femoral components were used in 25%, and 23% required femoral osteotomy. Cementless
fixation provided reliable fixation in this technically challenging situation

ACETABULAR OSTEOTOMY

With a resurgence of interest in pelvic and periacetabular osteotomy, the need for later hip arthroplasty in
these patients is likely to become more common. Parvizi, Burmeister, and Ganz reported results in 41
patients undergoing total hip procedures after prior periacetabular osteotomy. Because the initial
osteotomies were performed through a Smith-Petersen

approach, the procedures were done through virgin lateral soft tissues. There were no acetabular column
defects, but retroversion of the acetabulum was a common finding. Careful attention is needed when
positioning the acetabular component. The prior osteotomy was not thought to compromise the results of the
arthroplasty. In a series of patients with prior Chiari osteotomy, Hashemi-Nejad et al. found less acetabular
augmentation was needed compared with dysplastic hips without prior osteotomy.

ARTHRODESIS AND ANKYLOSIS

With the widespread media attention of the success of hip replacement, patients often are unwilling to accept
arthrodesis as a primary treatment option and likewise request conversion of an existing arthrodesis to
restore motion. The effects of hip fusion on other joints are significant. Often the ipsilateral knee is limited
in motion, with a variable degree of ligamentous laxity, and has a tendency for valgus malalignment.

Pain caused by arthritis or other conditions of the lumbar spine can increase significantly when sitting with
the spine partially flexed because of a fused hip. Care must be taken, however, to determine whether back
and leg pain may be caused by a herniated lumbar disc or some other condition that may not be improved by
total hip arthroplasty. If the hip is fused in poor position (i.e., flexed > 30 degrees, adducted > 10 degrees, or
abducted to any extent), osteotomy to correct the position should be considered, especially in younger
patients. Arthrodesis of one hip also applies greater mechanical stress to the opposite hip. Total hip
arthroplasty may be indicated if a fused hip causes severe, persistent low back pain or pain in the ipsilateral
knee or contralateral hip or if a pseudarthrosis after an unsuccessful fusion is sufficiently painful (Fig. 3-96).
The history of the initial reason for the arthrodesis is important. Patients with prior infection require a
thorough evaluation to rule out persistence. A careful assessment of the

function of other joints, especially the lumbar spine,

should
be done, and leg-length discrepancy should be measured. Preoperative metal-subtraction CT can be helpful
in determining the adequacy of bone stock and the presence of a pseudarthrosis.

Function of the abductors is difficult to evaluate before surgery, but in some patients active contraction of
these muscles can be palpated. Examination of the hip with the knee flexed helps differentiate the tensor
fasciae latae from the abductor muscles.

If the hip has been fused since childhood, and the trochanter appears relatively normal, the abductor
muscles are probably adequate. If the bone around the hip has been grossly distorted by disease or by one or
more fusion operations, the abductor muscles may be inadequate.

The utility of electromyographic testing of abductor function or imaging modalities such as MRI has not
been established.

Weak abductor musculature is associated with poorer functional outcome

. At surgery, a variety of screwdrivers, metal cutters, and other extraction instruments should be available to
remove antiquated fixation devices. The conversion of a fused hip to a total hip arthroplasty is safer and
easier if the trochanter is osteotomized. Complete mobilization of the femur without trochanteric osteotomy
is difficult, and the resulting inadequate exposure predisposes to component malposition, errors in femoral
reaming, and fractures. In addition, the limb often is fixed in external rotation, and consequently the
trochanter is posterior, overhanging the hip joint. Osteotomy of the neck can be difficult through a posterior
approach unless the trochanter is osteotomized.

The sciatic nerve often is displaced closer to the hip because the head-neck length is shorter than normal and
the nerve may be fixed in scar tissue; for this reason, special care is taken to avoid damage to the nerve.
Careful monitoring of tension on the nerve is necessary, and neurolysis may be indicated if the extremity is
significantly lengthened

After the femoral neck has been exposed, it is divided with a saw. The location of the osteotomy is
determined from bony landmarks or the position of previous fixation devices. The neck should not be
divided flush with the side of the ilium because sufficient bone must be left to cover the superior edge of the
cup (Fig. 3-97).

After the neck has been divided, release of the psoas tendon, gluteus maximus insertion, and capsulotomy
are necessary to mobilize the proximal femur. Usually the pelvic bone is sufficiently thick to cover the cup
adequately if the site for acetabular preparation is chosen carefully. Distortion of the normal bony
architecture may cause difficulty in locating the appropriate site for acetabular placement. Usually the
anterior inferior iliac spine remains intact and serves as a landmark. Additionally, a retractor can be placed
in the obturator foramen. Acetabular preparation is performed with conventional reamers, centering within
the available bone to preserve the anterior and posterior columns. An intraoperative radiograph is helpful
early in the acetabular preparation to ensure that the position of the reamer is as expected. The femoral canal
is prepared in the usual manner, taking into account any deformity from prior disease or femoral osteotomy.
Trochanteric fixation is accomplished by standard techniques (see Figs. 3-69 to 3-71). If the abductors are
markedly atrophic or deficient, a constrained acetabular component should be considered. Additionally, the
tensor fasciae latae may be attached to the trochanter to supplement the abductors. After the procedure has
been completed, the patient is placed supine. If the hip cannot be abducted 15 degrees because the adductors
are tight, a percutaneous adductor tenotomy is done through a separate small medial thigh incision. The
extremity usually is lengthened by the procedure and correcting prior flexion deformity. Lengthening
usually is desirable because in most instances the limb has been

shortened by the original disease, by the procedure for fusing the hip, or by the flexion deformity. The
postoperative treatment is routine, but the hip should be protected for at least 3 months by use of crutches
and then by use of a cane while the hip abductors and flexors are being rehabilitated. Patients rarely regain
flexion to 90 degrees, but they achieve sufficient motion to relieve back symptoms and permit sitting and
walking and tying shoes. Walking ability usually is improved, but in patients with inadequate abductor
function the gait pattern may worsen, and the support of a cane or walker may be required even if the patient
did not use one before conversion to arthroplasty. Most patients have some degree of residual abductor
weakness and limp, although this tends to improve over several years. The complication rate for conversion
of an arthrodesis to an arthroplasty is high. In the Mayo Clinic series of Strathy and Fitzgerald, 33% of
patients experienced failure within 10 years because of loosening, infection, or recurrent dislocation.
Patients with a spontaneous ankylosis fared much better than patients who had a prior surgical arthrodesis.
Stürup, Pedersen, and Jensen found a similar failure rate and recommended the procedure only for disabling
pain. Kim et al. reported revision in 16% at 10 years and a higher than average polyethylene wear rate with
osteolysis in more than half of patients. A 10-year survivorship of 74% was reported by Richards and
Duncan. Complications were noted in 54%, and functional scores were inferior to those of revision surgery