Dislocation After Total Hip Arthroplasty:

Implant Design and Orientation

Robert L. Barrack, MD

Abstract
Implant design and positioning are important factors in maintaining stability
and minimizing dislocation after total hip arthroplasty. Although the advent of
modular femoral stems and acetabular implants increased the number of head,
neck, and liner designs, the features of recent designs can cause intra-articular
prosthetic impingement within the arc of motion required for normal daily
activities and thus lead to limited motion, increased wear, osteolysis, and subluxation or dislocation. Minimizing impingement involves avoiding skirted
heads, matching a 22-mm head with an appropriate acetabular implant, maximizing the head-to-neck ratio, and, when possible, using a chamfered acetabular
liner and a trapezoidal, rather than circular, neck cross-section. Computer
modeling studies indicate the optimal cup position is 45 to 55 abduction.
Angles <55 require anteversion of 10 to 20 of both the stem and cup to minimize the risk of impingement and dislocation.
J Am Acad Orthop Surg 2003;11:89-99

Dislocation is a frequent early complication of total hip arthroplasty1

(THA) and is associated with a
higher mortality rate compared
with THA patients who do not sustain a dislocation.2 Dislocation is
the second most common cause for
revision surgery, after loosening.3
The incidence of dislocation after
primary THA varies from 0.6% to
7%; one review of 16 large series
documented 804 dislocations in
35,894 THAs (2.24%).4 Most published studies are from high-volume
medical centers, yet most hip replacements are done by surgeons
who perform a lesser volume of hip
arthroplasties. Because of evidence
that dislocation rate may be associated with surgeon experience,5 the
incidence of dislocation overall
may be higher than is reported
from large centers. For example,
Fender et al6 reviewed reports of

Vol 11, No 2, March/April 2003

hip replacement across a region in

England and found a dislocation
rate of 5%.
If increased experience is associated with a lower dislocation rate,
then as volume and experience with
THA increase, the incidence of dislocation should decrease. However,
this does not seem to be true, and
some believe that the incidence of
dislocation may be increasing. 7
There are several potential explanations for this phenomenon. Indications for THA have been expanded
to include patients who may be at
higher risk for dislocation. Also,
most centers have decreased their
average length of stay after THA
from more than 10 days to 5 days or
fewer.8 In addition, implant design,
especially modularity, has been
associated with dislocation.9 The
popularity of cementless fixation in
the late 1980s led to most femoral

stems becoming modular (as with

most acetabular implants), largely
to limit inventories because cementless implants were offered in a larger
number of sizes. This led to changes
in neck and head design as necks
became larger and more circular
(presumably to allay concerns about
strength) and as acetabular liner
options increased. Early studies
suggested that such design features
could limit motion and thus lead to
dislocation as a result of impingement;10,11 subsequent studies indicated that an association exists
between implant design and the
optimal position for implantation.12
Choice of design and positioning of
implants are two factors under the
direct control of the surgeon; it is
therefore important to understand
the role of implant design and orientation in optimizing range of motion,
function, and stability after THA.

Dr. Barrack is Professor of Orthopaedic

Surgery and Director, Adult Reconstructive
Surgery, Tulane University School of
Medicine, New Orleans, LA.
The author or the department with which he is
affiliated has received something of value from a
commercial or other party related directly or
indirectly to the subject of this article.
Reprint requests: Dr. Barrack, 1430 Tulane
Avenue, New Orleans, LA 70112.
Copyright 2003 by the American Academy of
Orthopaedic Surgeons.

89

Dislocation After Total Hip Arthroplasty

Femoral Implant Design

The variable factors in head and neck
design related to impingement and
dislocation are head size, modular
head design, femoral offset, and neck
geometry. Clinical and experimental
data are available on the effect of
each variable on hip stability.

Head Size
Clinical Studies
The effect of head size on the incidence of dislocation is undetermined. Woo and Morrey13 reviewed
more than 10,000 procedures and
reported dislocation rates of 2.9%
and 3.3% with 22- and 32-mm implants, respectively. More recently,
no clinical correlation was shown
between hip dislocation and the use
of the 22-mm head size compared
with 32-mm.4
Hedlundh et al 14 reported on
3,197 cases using Charnley stems
(DePuy, Warsaw, IN) with 22-mm
heads compared with 2,875 cases
using Lubinus stems (Waldemar
Link, Hamburg, Germany) with 32mm heads; dislocation rates at 1 year
were 2.4% and 2.5%, respectively
(not statistically significant). Late
dislocation occurred more frequently
with the Charnley stems (3.7% ver-

sus 2.9%). (Late dislocation is more

common with 22-mm heads because
of their higher rate of head penetration and liner wear.) When surgeon
experience and hip replacement for
fracture nonunion were factored in,
the difference in late dislocation was
not statistically significant. However, the risk of recurrent dislocation
was 2.3 times higher with the 22-mm
heads.
There seems to be an association
between surgical approach, head
size, and hip stability. In a review
of a large series of procedures,
Morrey15 reported that the posterior
approach had the highest dislocation rate of three approaches evaluated, regardless of head size (22 mm,
28 mm, or 32 mm), and that the 32mm head was associated with the
lowest dislocation rate (Table 1).
This implies that with a posterior
approach, a 32-mm head may be
preferable. Modifications of the posterior approach, including in formal
capsular repair, have been associated with a lower incidence of dislocation than the 6% reported by Morrey15 when used with a 28-mm head.
Pellicci et al16 reported a 0% incidence of early dislocation with 26and 28-mm heads in 395 hips, indicating that a large head (32 mm) is
not necessary to achieve a low early

dislocation rate with the posterior

approach.
Head size is not the only important variable in determining stability.
Hedlundh and Fredin2 reported a
dislocation rate of 0.4% in 4,706 total
hip arthroplasties performed with
22-mm heads between 1972 and
1975. The procedures were done
with a transtrochanteric approach,
allowing for trochanteric advancement during closure to improve stability. The low positioned (inferior and medial) socket was used routinely with a long posterior wall cup
design. The relationship between
head size, surgical approach, socket
position, and liner design is important. Using a 22-mm head can
result in a very low dislocation rate,
but its use with the wrong combination of socket type, orientation, and
approach can lead to a high dislocation rate. A 22-mm head also allows
less room for error during surgery.2
Concerns about the potential for
dislocation with a 22-mm head led
to the development of the 32-mm
head. This effectively increased the
ratio of the diameter of the head to
that of the neck (head-to-neck ratio),
increasing the arc of motion achieved
before impingement occurs. Fraser
and Wroblewski17 redesigned the
Charnley prosthesis to improve the

Table 1
Relationship of Surgical Approach, Implant Head Size, and Hip Dislocation
Head Size
22 mm
Surgical
Approach
Anterior
Lateral
Posterior

28 mm

32 mm

Total

No. of
Hips

Dislocation
(%)

No. of
Hips

Dislocation
(%)

No. of
Hips

Dislocation
(%)

No. of
Hips

Dislocation
(%)

571

2.6

151

1.3

48

2.1

770

2.3

1,251

2.7

295

4.1

352

3.4

1,898

3.1

88

6.8

511

6.0

86

3.5

685

5.8

2.9

957

4.7

486

3.3

3,353

3.5

All approaches 1,910

(Adapted with permission from Morrey BF: Instability after total hip arthroplasty. Orthop Clin North Am 1992;23:237-248.)

90

Journal of the American Academy of Orthopaedic Surgeons

Robert L. Barrack, MD

head-to-neck ratio for increased

stability. The smaller neck crosssection requires a metal of greater
strength to minimize the risk of
fracture; with the introduction of a
stainless steel alloy, the diameter of
the neck was reduced 20%, from
12.5 to 10 mm.2,17 Most investigators agree that the head-to-neck
ratio is more important than head
size alone in determining hip stability. This may explain why the use
of a 22-mm modular head may predispose to dislocation. Kelley et al7
reported a 35% dislocation rate
(5/14) with the 22-mm modular
head compared with 0% (0/17) for
the 28-mm modular head in a small
prospective randomized study.
Similarly, Heithoff et al18 reported
that, in a series of 4,164 primary
THAs, the mean dislocation rate of
7.2% approximately doubled during
the 2 years modular 22-mm heads
were used (P < 0.001).
The head-to-neck ratio can be
maximized when a larger head
diameter is used. Until recently, the
use of 32-mm heads was considered
generally inadvisable because of the
reported higher rate of volumetric
wear. With the recent advances in
alternative bearing surfaces, such as
metal-to-metal, ceramic-to-ceramic,
and cross-linked polyethylene, interest has been renewed in the use of
large head sizes because the problem of increased volumetric wear
may not apply to these surfaces.

Laboratory Studies
Because of the controversy regarding the effect of head diameter
on stability, recent laboratory studies have used cadaveric models,
implant retrieval studies, finite element analysis, and virtual reality
computer animation to study the
effect of head size on range of motion and stability. Bartz et al19 used
a cable system in six fresh cadaveric
specimens to simulate muscle action
of the major muscle groups. The
three-dimensional position of the

Vol 11, No 2, March/April 2003

femur relative to the acetabulum

was recorded electronically at the
point of impingement and dislocation. Different head sizes were compared, including 22-, 26-, 28-, and
32-mm heads. The same cementless
femoral prosthesis was used; it had
a cylindrical neck diameter of 11.8
mm and a neck shaft angle of 132.
There was an association between
head size and the degree of flexion
at dislocation in 10, 20, and 30 of
adduction (P = 0.01 for all three
positions). The most dramatic increase in arc of flexion related to
impingement occurred when increasing from a 22- to a 28-mm
head; the range of flexion increased
a mean of 5.6 before impingement
and 7.6 before dislocation. The site
of impingement varied with the
head diameter, with the 22-mm
head impinging between the neck of
the femoral prosthesis and the
acetabular liner. The 32-mm head
usually impinged between the
femur and the osseous bony femur
and the pelvis.19
Yamaguchi et al20 examined 111
retrieved acetabular implants of a
single manufacturer. By combining
the spatial orientation of acetabular
implants measured radiographically
with the location of impingement of
the retrieved implants, the location
of impingement against the pelvis
was determined. There was gross
surface damage consistent with impingement in 39% of cases. The
most important factors for predicting impingement were the femoral
head size and the head-to-neck ratio.
Implants with evidence of impingement had a smaller mean head-toneck ratio compared with implants
without evidence of impingement
(1.95 versus 2.21 [P < 0.0001]).
Scifert et al21 used finite element
analysis to predict factors that predispose to total hip dislocation. Threedimensional finite element analysis
was used to simulate certain activities associated with posterior dislocation, such as leg crossing in an erectly

seated position (hip flexed 90, adducted 0, and externally rotated 0).
The values measured included the
peak intrinsic moment resisting dislocation, the range of motion before
impingement of the neck on the
acetabular liner, and the range of
motion before dislocation. A number
of specific liner design features as
well as femoral head diameter were
studied. Increased head size did
indeed lead to improvement in the
peak intrinsic moment resisting dislocation, but when the head-to-neck
ratio remained constant, there was
no notable improvement in implant
range of motion.
Barrack et al12 used virtual reality
computer modeling to simulate the
range of motion of ideally positioned total hip implants. Implants
were digitized and animated through
a range of motion; the range of motion until impingement occurred
between the neck and liner was
quantified in every direction, and a
composite arc of motion was calculated. The authors determined that
changing from a 28- to 32-mm head
increased the arc of flexion by 6
(Fig. 1). This technique did not sim-

Abduction

Flexion

Extension

Adduction
Figure 1 Composite cone of motion produced by computer animation. Increasing
from a 28-mm (dark shading) to 32-mm
(light shading) head size increased the arc
of flexion to impingement by approximately 6. (Adapted with permission from
Barrack RL, Thornberry RL, Ries MD,
Lavernia C, Tozakoglou E: The effect of
component design on range of motion to
impingement in total hip arthroplasty.
Instr Course Lect 2001;50:275-280.)

91

Dislocation After Total Hip Arthroplasty

ulate impingement of the bony
femur on the pelvis but only the
neck of the femoral implant on the
liner. Bartz et al19 indicated that
changing from a 28- to a 32-mm
head might not achieve the entire
increase in motion because in some
cases, impingement of the femur on
the pelvis might occur before impingement of the implant neck on
the liner.
Because impingement is likely to
occur in some cases, especially if the
implant position is suboptimal, it
seems prudent to polish the neck
and to avoid a roughened surface
on the neck where impingement is
likely to occur. With a metal neck
on polyethylene, increased rim wear
may occur until a groove is worn in
the polyethylene to accommodate
the neck (Fig. 2), as seen in numerous cases by Yamaguchi et al.20 This
could lead to more serious consequences with a metal or ceramic
liner.

Modular Head Design

Some modular heads incorporate
an extended flangereinforced neck
(the skirted neck). In most modular systems, the longer neck lengths
(> +6 or +8 mm) require the addition of a skirt to the modular head.
Krushell et al10,11 measured the point
at which impingement and subluxa-

Figure 2 An implant retrieved at surgery

indicated that the skirted head was
impinging on the elevated rim liner posteriorly, leading to dislocation as well as to
excessive polyethylene wear (arrow).

92

tion first occur in maximum flexion

and extension with hip components
implanted in a standard position.
The passive range of motion decreased dramatically with the longer
modular heads that had a flange.
This finding became more prominent with the combination of a
smaller head diameter with a skirt,
which produces a low head-to-neck
diameter ratio. Changing from a
nonmodular stem to one with a
skirted modular head decreased the
flexion arc from 152 to 117. The
addition of a 22-mm long modular
head further decreased the flexion
arc to 106.
Urquhart et al 22 examined the
clinical effect of a modular femoral
head with that of an extended
flangereinforced neck in THA. The
mean polyethylene wear rate was
significantly (P = 0.009) higher with
the skirted modular heads (0.17 versus 0.11 mm/yr). The dislocation
rate also was higher in the group
with the flange extension (9% [1/11]
versus 4% [2/55]); however, because
of the small numbers, the difference
was not statistically significant. The
design of the taper and modular
head can lead to a very high dislocation rate, as reported by Hedlundh
and Carlsson,23 who found a 10%
dislocation rate in two centers using
a 28-mm modular Lubinus SP-2
implant. This rate was attributed to
the combination of an extended
flangereinforced modular head
with a large (14/16) taper (ie, the
cross section of the cone at one end
of the taper is 14 mm and the cross
section at the other end, 16 mm).
They discovered that there were two
designs of modular head and neck
being used in the same system: the
combination of the skirted 28-mm
head with a large taper resulted in a
flexion extension arc of motion of
only about 60, compared with a
flexion extension arc >90 for the 32mm modular head with a large taper
or a 28-mm head with a smaller
taper.23

Although skirted heads are associated with notable disadvantages

(Fig. 3), they are occasionally necessary to restore length and offset and
to provide appropriate soft-tissue
tension. In some systems, offset can
be increased by the use of high-offset stems without resorting to a
skirted modular head. These stems
require greater strength.

Femoral Offset
Femoral offset can affect stability. Fackler and Poss24 as well as
Morrey15 both described a positive
correlation between decreased offset
and dislocation, which may be the
result of several factors. Decreasing
offset reduces soft-tissue tension,
leading to a propensity for dislocation. Decreasing offset also reduces
the clearance between the femur
and the pelvis, which may lead to
dislocation through bony impingement.2 In addition, decreasing offset has biomechanical consequences
such as increasing the joint reaction
force; this can increase the forces at
the bone-cement or head-liner interfaces, leading to higher wear rates.
Care should be taken if offset is
reproduced by use of long modular
heads; as noted, the longer modular heads with flanges can decrease
passive range of motion. In addition, longer modular heads increase length as well as offset and
can result in unintentional lengthening.
Dual-offset implants that use the
same neck/shaft angle but a more
medial takeoff point for the neck of
the implant allow the addition of
several millimeters of offset without changing length. Increased
femoral offset increases the rotational forces on femoral implants;
whether this affects implant loosening rates is not known. A lateralized liner also can be used to increase abductor muscle tension
without resorting to a skirted modular head, although this method
increases length and offset to ap-

Journal of the American Academy of Orthopaedic Surgeons

Robert L. Barrack, MD

Figure 3 A, Anteroposterior radiograph showing anterior dislocation that occurred with extension and external rotation consistent with
impingement of the skirted head on the elevated rim liner. Intraoperative photographs confirm impingement (arrow) of the skirted head
on the elevated rim liner (B), which was not present with a neutral liner (C). (Reproduced with permission from Barrack RL, Thornberry
RL, Ries MD, Lavernia C, Tozakoglou E: The effect of component design on range of motion to impingement in total hip arthroplasty.
Instr Course Lect 2001;50:275-280.)

proximately the same degree. Lateralized liners, however, increase

the body weight moment arm and
potentially could increase the joint
reaction force, which has some of
the same potential negative effects
as decreasing the femoral offset.

Neck Geometry
Amstutz and Kody25 emphasized
the importance of neck design on
the stability of the THA, using a
pelvis-mounted apparatus to evaluate the effect of head size and neck
geometry on arc of motion in various planes. The 22-mm Charnley
prosthesis has a head-to-neck ratio
of 1.74; it impinged at 80 of flexion.
A trapezoidal-28 (T-28) prosthesis
has a trapezoidal neck design,
which resulted in a variable headto-neck ratio of 1.97 to 2.97; it
achieved motion and flexion to
114 before impingement occurred.
Compared with the Charnley, the
T-28 allowed 36 more internal rotation at 90 of flexion and 32 more
external rotation in extension. Most
modular heads were combined with
a circular neck, which sacrificed the
advantages of a trapezoidal neck
demonstrated by the T-28.25
In a recent study of dislocations
after revision THA, the effect of

Vol 11, No 2, March/April 2003

neck geometry on stability was examined using a computer model,

then correlated with clinical findings.26 A virtual reality computer
animation was used to compare two
commonly used revision implant
neck designs: one large (14/16)
taper with a circular cross-section,
one smaller (12/14) with a trapezoidal neck cross-section. Computer modeling verified that the
cross-sectional area of the larger circular taper was 30% larger than that
of the smaller trapezoidal taper.
The animation study predicted a
total arc of motion that was 46% less
for the larger compared with the
smaller taper. A clinical study was
undertaken to evaluate the dislocation rate after revision hip replacement using two stems whose major
design difference of size and geometry of the femoral neck could potentially affect stability. When
patients with major risk factors for
dislocation were excluded (eg,
absent abductor muscles, revisions
for recurrent dislocation), the dislocation rate with the stem with a
large, circular cross-section neck
was more than three times higher
compared with the smaller, trapezoidal neck: 15.4% (8/52) versus
4.3% (2/46) (P = 0.07).26

Acetabular Implant Design

Clinical Studies
A number of design features of
acetabular implants can affect stability and range of motion. In all-polyethylene implants, the depth and
location of the bore into the polyethylene for the femoral head can affect
stability. Eftekhar3 advocated lower
placement of the bore to increase tissue tension and improve stability.
Letournel and Lagrange27 described
an acetabular implant that was 3 mm
larger than half a sphere to capture
the femoral head to prevent dislocations. Although this design may
reduce dislocation under certain conditions, it effectively reduces range
of motion and may increase the rate
of acetabular loosening by increasing
the stress on the implant because of
impingement. Brien et al 28 compared the dislocation rates of the
Charnley femoral implant combined
with an acetabular cup either larger
than or equal to a hemisphere. A
Charnley stem used with a Charnley
high-posterior-wall acetabular
implant in 60 hips had a dislocation
rate of 3% (2/60). The Charnley
stem combined with a hemispheric
implant (the Tibac cup; Zimmer,
Warsaw, IN) had a dislocation rate

93

Dislocation After Total Hip Arthroplasty

of 11.4% (8/70). This dislocation rate
caused the researchers to return to
using the Charnley high-posteriorwall cup, after which only 1 of 67
hips dislocated (1.5%). The authors
concluded that, to provide adequate
stability, a 22-mm head should be
combined with an acetabular cup
that is larger than a hemisphere.
Amstutz and Kody25 emphasized
the importance of the shape of the rim
of the socket. For larger head sizes,
they advocated the use of an angle
beveled from the hemispheric depth
line to the periphery. The T-28 socket
has a 15 bevel, which is thought to
promote relocation of the femoral
head should subluxation occur.
With the advent of modular cementless implants, a variety of liner
options has become available. The
use of elevated rim liners is now
common and has led to controversy
regarding their efficacy and indications. According to Amstutz and
Kody,25 the use of a 20 elevated rim
liner can potentially increase the
leverage applied to the cup should
dislocation occur. They therefore
recommended limiting the degree
of augmentation to 10 or even 0.25
Cobb et al29 studied whether elevated
rim liners do, in fact, improve postoperative stability after THA. They
compared 2,469 acetabular implants
with a 10 elevated rim liner with
2,698 implants with a standard
liner. The 2-year probability of dislocation was 2.19% with the elevated
rim liner compared with a probability of 3.85% with a standard liner (P
= 0.001). The difference was statistically significant at 2 years but not at
5 years because of the smaller sample size. In a subsequent study,30 no
difference was evident in the 5-year
loosening rate of elevated rim liners
compared with standard liners.

Laboratory Studies
Elevated rim liners have been the
subject of a number of laboratory
testing protocols. Krushell et al10
used a three-dimensional protractor

94

to quantify the effect of elevated rim

liners on range of motion and described two types (Fig. 4). With type
A, the liner reorients the axis of the
acetabular implant and, when placed
posteriorly, effectively increases
flexion and internal rotation while
decreasing extension and external
rotation. This type of liner does not
provide additional support once
impingement and subluxation begin
to occur. With type B, an extended
wall does provide support once subluxation starts. The point at which
subluxation begins does not change;
however, the point at which dislocation occurs is different. The authors
determined that elevated rim liners
do not improve range of motion but
rather reorient motion, increasing it
in one direction while decreasing it
by the same amount in the opposite
direction. They discouraged the use
of elevated rim liners on a routine
basis and recommended them for
dislocation caused by prosthetic malposition when it would be difficult to
change the acetabular implant. They
recommended that, in such cases, a
liner that reoriented the implant axis
(type A) be used.
Maxian et al31 used finite element
modeling to predict dislocation

B
Figure 4 Elevated rim liners. A, Liner
type A reorients the axis of rotation,
increasing flexion but not providing additional support once subluxation is initiated.
B, Type B does not increase motion to
impingement but does provide support
from dislocation once subluxation begins.
(Adapted with permission from Krushell
RJ, Burke DW, Harris WH: Elevated-rim
acetabular components: Effect on range of
motion and stability in total hip arthroplasty. J Arthroplasty 1991;6[suppl]:S53-S58.)

propensity with different types of

acetabular liners and femoral neck
diameters. They compared a chamfered nonlipped liner, a nonchamfered liner, and an extended lip
liner. The extended lip liner had
only slightly higher angles to dislocation and moments than did the
chamfered standard implant. The
nonchamfered standard lip implant
had the worst performance. The alternative neck diameters (16.3 versus
15.5 mm) had little effect on dislocation propensity in this experimental
model, although much smaller neck
diameters are in clinical use.
The interaction of different liner
types with skirted heads was studied by Barrack et al12 using virtual
reality computer animation. Two
types of liners were tested, one with
a high-angle, narrow chamfer zone,
another with a low-angle, wider
chamfer zone (Fig. 5). These were
combined with either a circular or a
trapezoidal cross-section neck. The
optimal combination for maximizing
range of motion to impingement was
a wide chamfer zone and trapezoidal neck. The worst combination
was a narrow chamfer zone and circular neck, which had a total arc of
motion of only 57% of the optimal
combination (Fig. 5, C). However,
the authors evaluated only range of
motion free of intra-articular impingement and not the clinically
important variable of hip range of
motion free of hip dislocation. The
two may not be the same. The impingement findings were consistent
with those of Yamaguchi et al,20 who
documented gross rim wear and
impingement in a high percentage of
retrieved implants that combined a
large circular neck with a narrow,
chamfered, extended wall liner.

Implant Position
Implant orientation is among the
most critical factors in assuring stability. 4 Acetabular orientation is

Journal of the American Academy of Orthopaedic Surgeons

Robert L. Barrack, MD

Abduction
Narrow chamfer

Wide chamfer
Flexion

Extension

Adduction

Figure 5 Comparison of a liner with a high-angle, narrow chamfer zone (A) with a liner with a low-angle, wide chamfer zone (B).
C, Comparison of the cone of motion for the combination of a trapezoidal neck with a wide chamfer zone (light shading) and a circular neck
with a narrow chamfer zone (dark shading). (Adapted with permission from Barrack RL, Thornberry RL, Ries MD, Lavernia C, Tozakoglou E:
The effect of component design on range of motion to impingement in total hip arthroplasty. Instr Course Lect 2001;50:275-280.)

particularly difficult to achieve consistently and is the most sensitive

variable predisposing to dislocation 9,15 (Fig. 6). Barrack et al 32
defined an acceptable range of 45
10 abduction and 20 10 anteversion (Fig. 7). Lewinnek et al33 defined a safe zone of 40 10 abduction and 15 10 anteversion. The
dislocation rate for implants outside
this range was four times higher
than for those within the range (6%
versus 1.5%). Coventry et al34 reported that 50% of posterior dislocations were associated with cup retroversion of 7 to 10. Fackler and
Poss24 identified excessive femoral
anteversion as the most common
implant malposition. In their study,
implant malposition was present in
44% of patients with dislocations
(15/34) but in only 6% of those without dislocation (3/50) (P < 0.05).24
Stem anteversion or retroversion
>20 may predispose to dislocation.2
The optimal implant position for
stability remains undetermined.
McCollum and Gray35 recommended
cup anteversion of 20 to 40 rather
than the 5 to 25 proposed by
Lewinnek et al.33 Optimal implant
orientation probably is associated
with surgical approach. Greater acetabular anteversion is advisable with
a posterior approach to allow more

Vol 11, No 2, March/April 2003

flexion before impingement occurs;

the recommendation of McCollum
and Gray35 of 20 to 40 is more appropriate for a posterior approach,

whereas the 5 to 25 recommended

by Lewinnek et al33 is more compatible with an anterolateral or direct
lateral approach.

Figure 6 A, Anteroposterior radiograph showing horizontal cup placement with 30

abduction. In spite of adequate anteversion on shoot-through lateral radiograph (B), posterior dislocation occurred with deep flexion (C). (Reproduced with permission from
Barrack RL, Lavernia C, Ries M, Thornberry R, Tozakoglou E: Virtual reality computer
animation of the effect of component position and design on stability after total hip arthroplasty. Orthop Clin North Am 2001;32:569-577.)

95

Dislocation After Total Hip Arthroplasty

High Dislocation

Anteversion

30
Acceptable Range

20
Target

10
35

45

55

Abduction
Figure 7 Implants positioned in the safe
zone are less likely to dislocate than those
outside the safe zone. (Adapted with permission from Barrack RL, Lavernia C, Ries
M, Thornberry R, Tozakoglou E: Virtual
reality computer animation of the effect of
component position and design on stability
after total hip arthroplasty. Orthop Clin
North Am 2001;32:569-577.)

In spite of the recommendations

for optimal implant positioning,
plain radiographs have been poor
predictors of propensity to dislocate.
Paterno et al 1 did radiographic
analysis of 32 dislocated hips and
compared them with 32 controls
matched for prosthesis type and surgical approach; they found no association between abduction (range,
38 to 57) or degree of anteversion
and incidence of dislocation. Pollard
et al36 compared 7 dislocating hips
with 90 controls and also reported
no difference in inclination abduction or anteversion between the two
groups. Pierchon et al37 did computed tomography (CT) scans of 38
dislocated THAs and compared
them with 14 controls; they found
no difference in alignment of the
prosthetic implants. Of the seven
cases requiring revision surgery, CT
identified pathology in only two
(one cup retroversion and one protruding osteophyte). The authors
concluded that muscle imbalance
rather than implant malposition was
the major contributing factor to dislocation.37

96

Besides implant malposition, factors such as patient compliance and

soft-tissue status are important to
stability. Although malposition
makes dislocation more likely, most
malpositioned implants will not dislocate. Conversely, many patients
with apparently well-positioned
implants will experience a dislocation. However, certain designs and
positions clearly lead to a higher
incidence of impingement and dislocation. When a metal neck contacts a plastic liner, a number of
potentially adverse consequences
can occur, including limited motion
and function; increased stress on the
liner, resulting in a modular liner
dislodgement or accelerated acetabular implant loosening; liberation of
metal debris from the femoral neck;
generation of rim wear, potentially
increasing the risk of osteolysis; and
subluxation or dislocation (Fig. 3).
Recent studies have attempted to
define the effect of various implant
positions on the range of motion
possible before impingement or dislocation. DLima et al38 used computer modeling to generate a threedimensional model of a generic hip
prosthesis. At 35 of acetabular
abduction, there were no zones of
excellent stability. Femoral anteversion 10 was required to permit

shoelace tying or stooping (Table 2).

Acetabular abduction of 45 resulted in better, but still suboptimal,
results. When one implant was
anteverted <15, the other had to be
anteverted >15 to remain outside a
zone of poor stability. The highest
zone of excellent stability occurred
when the cup was abducted 55.
Poor stability in this setting resulted
only when both implants were
anteverted >20 or <7. At an
acetabular abduction angle of 35,
activities of daily living were possible only with certain combinations
of femoral and acetabular anteversion (Table 2). Acetabular angles of
45 and 55 did not have such limitations, which is at odds with the
commonly held belief that a more
horizontal cup position is inherently
more stable.25
A similar study by Barrack et al32
used software that simulated sitting
and stooping rather than stooping
and shoelace tying. Six different
combinations of cup abduction, cup
anteversion, and stem anteversion
were tested. The optimal combination was 45 cup abduction, 20 cup
anteversion, and 15 stem anteversion (Table 3). Similar to the findings of DLima et al,38 a horizontal
cup position of 25 almost invariably resulted in unsatisfactory

Table 2
Combinations of Prosthetic Orientations That Permit Tying a Shoelace
and Stooping
Acetabular Abduction

Femoral Anteversion

Acetabular Anteversion

35
35
35
35
45
55

0
10
20
30
All positions
All positions

No position
10
All positions
10
All positions
All positions

(Reprinted with permission from DLima DD, Urquhart AG, Buehler KO, Walker RH,
Colwell CW Jr: The effect of the orientation of the acetabular and femoral components
on the range of motion of the hip at different head-neck ratios. J Bone Joint Surg Am
2000;82:315-321.)

Journal of the American Academy of Orthopaedic Surgeons

Robert L. Barrack, MD

Table 3
Implant Orientations and Summary of Results
Cup
Cup
Stem
Case Abduction Anteversion Anteversion
1
2
3
4
5
6
7

45
45
45
45
25
25
25

20
0
10
10
20
0
0

15
15
15
45
15
15
0

Sitting

Stooping

Satisfactory
Marginal
Unsatisfactory
Marginal
Satisfactory
Unsatisfactory
Unsatisfactory

Satisfactory
Unsatisfactory
Unsatisfactory
Marginal
Unsatisfactory
Unsatisfactory
Unsatisfactory

Marginal = impingement occurred in spite of any modifications; satisfactory = no

impingement encountered; unsatisfactory = position could be accomplished with some
modification of order, eg, abducting before flexing.
(Adapted with permission from Barrack RL, Lavernia C, Ries M, Thornberry R, Tozakoglou
E: Virtual reality computer animation of the effect of component position and design on
stability after total hip arthroplasty. Orthop Clin North Am 2001;32:569-577.)

motion to achieve sitting and stooping. Robinson et al39 and Seki et al40
also used computer modeling and
demonstrated less flexion before
impingement with low cup angles.
Clinical examples of dislocation with

a horizontal cup placement have

been described that anecdotally confirm the computer modeling predictions (Fig. 5).
Impingement can lead to other
adverse outcomes, including sub-

luxation, accelerated polyethylene

wear with or without osteolysis
(Fig. 8), liner dislodgement, and
implant loosening. Schmalzried et
al41 reported a statistically significant (P < 0.0001) correlation between cup angles >50 and a higher
incidence of osteolysis. Again, the
use of alternative bearing surfaces
might allow lower contact areas
without a substantial increase in
wear rate, but this awaits confirmation. These other potential sequelae make the correlation between
frank dislocation and implant malposition less strong. Nevertheless,
avoiding impingement and the
associated negative effects is desirable for optimal function and
longevity of a THA.

Summary
Implant design and orientation affect the arc of motion achieved
before implant impingement. Max-

Figure 8 Massive pelvic osteolysis noted radiographically (A) was associated with gross rim wear evident in retrieved liners (B and C).

Vol 11, No 2, March/April 2003

97

Dislocation After Total Hip Arthroplasty

imizing motion to impingement
while minimizing risk of dislocation is necessary for optimal function. A number of implant design
features are consistent with this
goal. Modular skirted heads should
be avoided when possible; highoffset stems and lateralized liners
minimize the need for skirted
heads. Maximizing the head-toneck ratio with a trapezoidal rather
than circular neck notably improves motion. The head-to-neck
ratio also can be increased by using
a larger head diameter, although
this may shift the area of impingement to the bony femur/
pelvis with implant diameters >28
mm. The use of larger heads can

improve motion and stability in

some cases, but there are concerns
about increased volumetric wear
seen with 32-mm heads and standard polyethylene. If an elevated
rim liner is used, a wide chamfer
zone allows greater clearance of the
neck and thus greater range of motion before impingement.
Alternative bearing surfaces
allow the use of larger femoral
head sizes without concerns about
increased wear. However, these
surfaces should be tested for the
effects of impingement. Even with
optimal design features, a substantial percentage of hips will experience implant impingement because
of a greater-than-average range of

motion or suboptimal implant position.

The optimal implant position for
stability and function remains controversial. Computer modeling
indicates that cup abduction of 45
to 55 is the most desirable. Robinson et al39 confirmed this but found
lower contact areas at abduction
angles >50, which could result in
higher wear rates. A cup angle of
45 is a compromise position. In
this position, 15 of stem and cup
anteversion appears to be optimal
for maximizing motion. A lesser
degree of anteversion of one implant can be compensated to some
degree by additional anteversion of
the other implant.

8. Barrack RL, Hoffman GJ, Tejeiro WV,

Carpenter LJ Jr: Surgeon work input
and risk in primary versus revision
total joint arthroplasty. J Arthroplasty
1995;10:281-286.
9. Barrack RL: Modularity of prosthetic
implants. J Am Acad Orthop Surg 1994;
2:16-25.
10. Krushell RJ, Burke DW, Harris WH:
Elevated-rim acetabular components:
Effect on range of motion and stability
in total hip arthroplasty. J Arthroplasty
1991;6(suppl):S53-S58.
11. Krushell RJ, Burke DW, Harris WH:
Range of motion in contemporary total
hip arthroplasty: The impact of modular
head-neck components. J Arthroplasty
1991;6:97-101.
12. Barrack RL, Thornberry RL, Ries MD,
Lavernia C, Tozakoglou E: The effect
of component design on range of
motion to impingement in total hip
arthroplasty. Instr Course Lect 2001;50:
275-280.
13. Woo RY, Morrey BF: Dislocations
after total hip arthroplasty. J Bone Joint
Surg Am 1982;64:1295-1306.
14. Hedlundh U, Ahnfelt L, Hybbinette
CH, Wallinder L, Weckstrom J, Fredin
H: Dislocations and the femoral head
size in primary total hip arthroplasty.
Clin Orthop 1996;333:226-233.
15. Morrey BF: Instability after total hip
arthroplasty. Orthop Clin North Am
1992;23:237-248.
16. Pellicci PM, Bostrom M, Poss R: Pos-

terior approach to total hip replacement

using enhanced posterior soft tissue
repair. Clin Orthop 1998;355:224-228.
Fraser GA, Wroblewski BM: Revision
of the Charnley low-friction arthroplasty for recurrent or irreducible dislocation. J Bone Joint Surg Br 1981;63:
552-555.
Heithoff BE, Callaghan JJ, Goetz DD,
Sullivan PM, Pedersen DR, Johnston
RC: Dislocation after total hip arthroplasty: A single surgeons experience.
Orthop Clin North Am 2001;32:587-591.
Bartz RL, Noble PC, Kadakia NR,
Tullos HS: The effect of femoral component head size on posterior dislocation of the artificial hip joint. J Bone
Joint Surg Am 2000;82:1300-1307.
Yamaguchi M, Akisue T, Bauer TW,
Hashimoto Y: The spatial location of
impingement in total hip arthroplasty.
J Arthroplasty 2000;15:305-313.
Scifert CF, Brown TD, Pedersen DR,
Callaghan JJ: A finite element analysis
of factors influencing total hip dislocation. Clin Orthop 1998;355:152-162.
Urquhart AG, DLima DD, VennWatson E, Colwell CW Jr, Walker RH:
Polyethylene wear after total hip
arthroplasty: The effect of a modular
femoral head with an extended flangereinforced neck. J Bone Joint Surg Am
1998;80:1641-1647.
Hedlundh U, Carlsson AS: Letter:
Increased risk of dislocation with collar reinforced modular heads of the

References
1. Paterno SA, Lachiewicz PF, Kelley SS:
The influence of patient-related factors
and the position of the acetabular component on the rate of dislocation after
total hip replacement. J Bone Joint Surg
Am 1997;79:1202-1210.
2. Hedlundh U, Fredin H: Patient characteristics in dislocations after primary
total hip arthroplasty: 60 patients compared with a control group. Acta
Orthop Scand 1995;66:225-228.
3. Eftekhar NS: Dislocation and instability, in Eftekhar NS (ed): Total Hip
Arthroplasty. St. Louis, MO: MosbyYear Book, 1993, pp 1505-1553.
4. Morrey BF: Dislocation, in Morrey BF,
An KN (eds): Reconstructive Surgery of
the Joints, ed 2. New York, NY: Churchill-Livingstone, 1996, vol 2, pp 12471260.
5. Hedlundh U, Ahnfelt L, Hybbinette
CH, Weckstrom J, Fredin H: Surgical
experience related to dislocations after
total hip arthroplasty. J Bone Joint Surg
Br 1996;78:206-209.
6. Fender D, Harper WM, Gregg PJ:
Outcome of Charnley total hip replacement across a single health region in
England: The results at five years from
a regional hip register. J Bone Joint Surg
Br 1999;81:577-581.
7. Kelley SS, Lachiewicz PF, Hickman
JM, Paterno SM: Relationship of
femoral head and acetabular size to
the prevalence of dislocation. Clin
Orthop 1998;355:163-170.

98

17.

18.

19.

20.

21.

22.

23.

Journal of the American Academy of Orthopaedic Surgeons

Robert L. Barrack, MD

24.

25.

26.

27.

28.

29.

30.

Lubinus SP-2 hip prosthesis. Acta

Orthop Scand 1996;67:204-205.
Fackler CD, Poss R: Dislocation in
total hip arthroplasties. Clin Orthop
1980;151:169-178.
Amstutz HC, Kody MH: Dislocation
and subluxation, in Amstutz HC (ed):
Hip Arthroplasty. New York, NY:
Churchill Livingstone, 1991, pp 429-447.
Barrack RL, Butler RA, Laster DR,
Andrews P: Stem design and dislocation after revision total hip arthroplasty:
Clinical results and computer modeling.
J Arthroplasty 2001;16(8 suppl 1):8-12.
Letournel E, Lagrange J: LagrangeLetournel hip prosthesis: Results of
1355 cases, in The Hip: Proceedings of the
Third Open Scientific Meeting of the Hip
Society, 1975. St. Louis, MO: CV Mosby,
1975, pp 278-299.
Brien WW, Salvati EA, Wright TM,
Burstein AH: Dislocation following
THA: Comparison of two acetabular
component designs. Orthopedics 1993;
16:869-872.
Cobb TK, Morrey BF, Ilstrup DM: The
elevated-rim acetabular liner in total
hip arthroplasty: Relationship to postoperative dislocation. J Bone Joint Surg
Am 1996;78:80-86.
Cobb TK, Morrey BF, Ilstrup DM:

Vol 11, No 2, March/April 2003

31.

32.

33.

34.

35.

36.

Effect of the elevated-rim acetabular

liner on loosening after total hip arthroplasty. J Bone Joint Surg Am 1997;79:
1361-1364.
Maxian TA, Brown TD, Pedersen DR,
Callaghan JJ: Finite element modeling
of dislocation propensity in total hip
arthroplasty. Orthop Trans 1996-97;
20:647-648.
Barrack RL, Lavernia C, Ries M,
Thornberry R, Tozakoglou E: Virtual
reality computer animation of the effect
of component position and design on
stability after total hip arthroplasty.
Orthop Clin North Am 2001;32:569-577.
Lewinnek GE, Lewis JL, Tarr R, Compere CL, Zimmerman JR: Dislocations
after total hip-replacement arthroplasties. J Bone Joint Surg Am 1978;60:
217-220.
Coventry MB, Beckenbaugh RD,
Nolan DR, Ilstrup DM: 2,012 total hip
arthroplasties: A study of postoperative course and early complications.
J Bone Joint Surg Am 1974;56:273-284.
McCollum DE, Gray WJ: Dislocation
after total hip arthroplasty: Causes
and prevention. Clin Orthop 1990;261:
159-170.
Pollard JA, Daum WJ, Uchida T: Can
simple radiographs be predictive of

37.

38.

39.

40.

41.

total hip dislocation? J Arthroplasty

1995;10:800-804.
Pierchon F, Pasquier G, Cotten A,
Fontaine C, Clarisse J, Duquennoy A:
Causes of dislocation of total hip arthroplasty: CT study of component alignment. J Bone Joint Surg Br 1994;76:45-48.
DLima DD, Urquhart AG, Buehler
KO, Walker RH, Colwell CW Jr: The
effect of the orientation of the acetabular and femoral components on the
range of motion of the hip at different
head-neck ratios. J Bone Joint Surg Am
2000;82:315-321.
Robinson RP, Simonian PT, Gradisar
IM, Ching RP: Joint motion and surface
contact area related to component position in total hip arthroplasty. J Bone
Joint Surg Br 1997;79:140-146.
Seki M, Yuasa N, Ohkuni K: Analysis
of optimal range of socket orientations
in total hip arthroplasty with use of
computer-aided design simulation.
J Orthop Res 1998;16:513-517.
Schmalzried TP, Guttmann D, Grecula
M, Amstutz HC: The relationship
between the design, position, and articular wear of acetabular components
inserted without cement and the development of pelvic osteolysis. J Bone Joint
Surg Am 1994;76:677-688.

99