The Effect of Dynamic Femoroacetabular
Impingement on Pubic Symphysis Motion
A Cadaveric Study
Patrick M. Birmingham,*yMD, Bryan T. Kelly,zMD, Robert Jacobs,yBS,
Linda McGrady,yBS, and Mei Wang,yPhD
Investigation performed at the Medical College of Wisconsin, Milwaukee, Wisconsin
Background: A link between femoroacetabular impingement and athletic pubalgia has been reported clinically. One proposed
origin of athletic pubalgia is secondary to repetitive loading of the pubic symphysis, leading to instability and parasymphyseal
tendon and ligament injury.
Hypothesis/Purpose: The purpose of this study was to investigate the effect of simulated femoral-based femoroacetabular
impingement on rotational motion at the pubic symphysis. The authors hypothesize that the presence of a cam lesion leads to
increased relative symphyseal motion.
Study Design: Controlled laboratory study.
Methods: Twelve hips from 6 fresh-frozen human cadaveric pelvises were used to simulate cam-type femoroacetabular impinge-
ment. The hips were held in a custom jig and maximally internally rotated at 90? of flexion and neutral adduction. Three-
dimensional motion of the pubic symphysis was measured by a motion-tracking system for 2 states: native and simulated
cam. Load-displacement plots were generated between the internal rotational torque applied to the hip and the responding
motion in 3 anatomic planes of the pubic symphysis.
Results: As the hip was internally rotated, the motion at the pubic symphysis increased proportionally with the degrees of the
rotation as well as the applied torque measured at the distal femur for both states. The primary rotation of the symphysis was
in the transverse plane and on average accounted for more than 60% of the total rotation. This primary motion caused the anterior
aspect of the symphyseal joint to open or widen, whereas the posterior aspect narrowed. At the torque level of 18.0 N?m, the
mean transverse rotation in degrees was 0.89? 6 0.35? for the native state and 1.20? 6 0.41? for cam state. The difference
between cam and the native groups was statistically significant (P \ .03).
Conclusion: Dynamic femoroacetabular impingement as caused by the presence of a cam lesion causes increased rotational
motion at the pubic symphysis.
Clinical Relevance: Repetitive loading of the symphysis by cam impingement is thought to lead to increased symphyseal motion,
which is one possible precursor to athletic pubalgia.
Keywords: femoroacetabular impingement; athletic pubalgia; sports hernia; pubic symphysis; cam; proximal femoral retroversion
Femoroacetabular impingement (FAI) occurs in many
forms. One form is dynamic impingement of the hip, which
has been increasingly recognized in young patients with
hip pain.3Dynamic impingement can be caused by the
presence of a cam lesion (decreased head-neck offset), a pin-
cer lesion (acetabular overcoverage), proximal femoral ret-
roversion, or coxa vara.3,14,19,23,27As defined by Mardones
et al,19the cam-type lesion is a bony increase in the diam-
eter of the femoral neck at the femoral head-neck junction.
This reduced femoral head-to-neck offset distance leads to
contact with the acetabulum early in the arc of internal
rotation of the hip. Repetitive contact through activity
can lead to labral tearing and transition zone cartilage
delamination, pain, and early onset osteoarthritis over
Normal proximal femoral version is around 15? of ante-
version, and thus any amount of version significantly less
than this represents anatomic retroversion. In patients
*Address correspondence to Patrick Birmingham, MD, Department of
Orthopaedic Surgery, Medical College of Wisconsin, 9200 West Wiscon-
sin Ave, PO Box 26099, Milwaukee, WI 53226 (e-mail: patrickbirmin
yMedical College of Wisconsin, Milwaukee, Wisconsin.
zHospital for Special Surgery, New York, New York.
Presented at the annual meeting of the AOSSM, July 2011, San
One or more of the authors has declared the following potential con-
flict of interest or source of funding: Financial support for this study was
provided by the Department of Orthopaedic Surgery, Medical College of
The American Journal of Sports Medicine, Vol. XX, No. X
? 2012 The Author(s)
AJSM PreView, published on March 5, 2012 as doi:10.1177/0363546512437723
impingement, the femoral neck contacts the acetabular rim
earlier in the arc of functional motion and causes impinge-
ment and a crush injury to the labrum, and it is associated
with the development of arthritis.8,29,31Both of these types
of FAI have been shown to decrease physiologic internal
rotation of the hip, which puts the labrum and hemipelvis
at risk for repetitive loading while participating in activi-
ties that require more functional rotation than the bony
Repetitive loading of the pubic symphysis could be a con-
tributing factor in the origin of athletic pubalgia, or sports
hernia, or osteitis pubis.9,10This is thought to occur as
a result of the symphyseal instability, leading to injury to
the parasymphyseal tendons of the rectus abdominis and
adductor longus fascial sheath.10Other suggested causes
of sports hernia include injuries to the external oblique apo-
neurosis, transversus abdominis, and posterior inguinal
floor. Sports hernia is a common clinical presentation in
athletes that typically presents as unilateral, localized groin
pain without the presence of an actual hernia.1,21,28Ahu-
mada et al1reported this as point tenderness over the pubis,
rectus abdominis origin, or adductor longus tendon origin.
The literature attributes this pain to micro-tears of the rec-
tus abdominis, the adductor longus, and their continuous
aponeurosis as they attach on the pubic tubercle.1,21,28
There has been a recent report on a subset of patients
who present with both femoroacetabular impingement and
athletic pubalgia in whom treatment of either in isolation
leads to a high rate of failure.17Hypermobility of the pubic
symphysis has also been shown to be associated with ostei-
tis pubis.33Clinically, patients with osteitis pubis and fem-
oroacetabular impingement commonly have limited internal
rotation of the hip, so this may represent another link
between these 2 pathologic entities.30
The purpose of this study was to investigate the effect of
a cam lesion on rotational motion at the pubic symphysis.
Our hypothesis is that cam impingement leads to increased
relative motion at the pubic symphysis.
Twelve hips from 6 fresh-frozen human cadaveric pelvises
were used for this study (3 male and 3 female, mean age
47.8 years, range 29-57 years). The specimens included
native femurs and all related musculature and skin but
without visceral organs. Donors with previous hip surgery,
severe hip arthritis, or cancer were excluded. Each speci-
men was kept frozen at –20?C until 36 hours before testing,
when it was taken out to thaw at room temperature.
Cam-type FAI was simulated by implanting a dome-
shaped wood button at the femoral head-neck junction
through an open surgical dislocation with trochanteric
osteotomy.7The cam lesion was simulated with a 5-mm
dome height and base diameter of 25 mm (Figure 1). The
cam lesion site was first prepared by planing flat the
head-neck junction centered at the 1:30 position in
a clock-faced representation (Figure 2). In the clock repre-
sentation used, the medial femoral head position was refer-
enced as the 3-o’clock position for both right and left hips.
The dome was fixed in place with a screw. The osteotomy of
the greater trochanter was then held rigidly with 2 bicort-
ical 3.5-mm screws, and the joint capsule was repaired
with multiple 5-0 Ethibond sutures.
A custom-designed loading jig was used for the experi-
ment. It allowed the pelvis to be suspended vertically in
an upright position by anchoring an iliac wing to the jig
frame with bolts and dental cement. The femoral epicon-
dyle contralateral to the fixed iliac wing was exposed and
potted with a special Plexiglas plate attachment that was
used to apply internal rotation to the femur. Internal rota-
tion was manually imposed while the leg was positioned in
90? of flexion and neutral adduction on a height-adjustable
supporting stance. The 90? of flexion position was chosen
because this is the position of flexion typically used to eval-
uate a clinical impingement sign. Adduction was posi-
tioned in the neutral position to remove it as another
variable to evaluate. A 6-axis load cell (Model M3; AMTI,
Figure 1. The cam lesion was simulated with a 5-mm dome
height and base diameter of 25 mm.
Figure 2. The cam lesion site was first prepared by planing
flat the head-neck junction centered at the 1:30 position in
a clock-faced representation.
2 Birmingham et alThe American Journal of Sports Medicine
Watertown, Massachusetts) was mounted onto the Plexi-
glas plate at the distal femur to measure the moments
and forces generated during the maneuver. To record the
3-dimensional motion of the pubic symphysis, we mounted
marker triads from a motion-tracking system (Optotrak
Certus Motion Tracking System; Northern Digital Inc,
Waterloo, Ontario, Canada) onto each side of the symphy-
sis. Additional markers were also attached to the base of
the load cell to track internal rotation of the femur. A sam-
pling rate of 30 Hz was used for the collection of load cell
and marker data. Euler angles were then derived from
movement of the markers based on the theory of rigid
Each hip was tested in 2 states: native and cam. The
native state represents the hip being tested in its original
state prior to trochanteric osteotomy, dislocation, and cap-
sular repair. Load-displacement plots were generated
between the internal rotational torque applied to the hip
and the responding motion in 3 anatomic planes of the
pubic symphysis. The mean ranges of symphysis rotation
were evaluated at 10 torque levels ranging from 0.0 to
18.0 N?m and compared between the 2 groups. The maxi-
mum torque of 18 N?m was selected to represent approxi-
mately 20% of the maximum hip rotation strength
reported in the literature.18The reported peak isokinetic
internal hip rotation torque in a seated position in young
healthy subjects was 62% for males and 48% for females
in ft-lbs per lb when normalized with respect to body
weight, which was equivalent to 110 N?m and 92 N?m in
our donor group. The value of 18 N?m is also at the level
of peak internal rotation torque during stair ascent for nor-
mal elderly groups.15It represents a moderate level of
physiologic loading and low enough for the specimen to
withstand repeated testing. The 10 torque levels were
evenly selected within this range as points of comparison.
The cadaveric model and software have been previously
developed and validated in our laboratory based on previ-
ous work.32Load-displacement plots of the applied torque
versus internal rotation of the hip were then evaluated.
Statistical analysis was performed using software package
Statview (SAS Institute, Cary, North Carolina). The effect
of a cam lesion on motion across the pubic symphysis was
Range of Motion of Secondary Pubic Symphysis Rotation
(in Degrees) in Response to Hip Internal Rotationa
Frontal PlaneSagittal Plane
0.22 6 0.15
0.37 6 0.20
0.45 6 0.30
0.19 6 0.12
0.37 6 0.16
0.52 6 0.15
0.21 6 0.30
0.25 6 0.13
0.39 6 0.12
0.09 6 0.06
0.19 6 0.10
0.27 6 0.18
aValues are presented as mean 6 standard deviation.
Applied Torque (N.m)
Applied Torque (N.m)
Figure 3. Mean (SE) motion of the pubic symphysis in response to the applied internal rotation torque at the hip in 2 states: native
and cam. (A) Rotation in the transverse plane and (B) equivalent total rotation. *P = .05.
Vol. XX, No. X, XXXX Effect of Cam Impingement on Pubic Symphysis Motion3
(ANOVA) and post hoc Fisher protected least significant
difference (PLSD) tests. The rate of increase in hip rotation
with respect to the applied torque was obtained from linear
region(s) of the load-displacement plot using regression.
The effect of a cam lesion on these slopes was compared
using the same ANOVA and post hoc tests. Significance
level was set at P \ .05 for all analyses.
As the hip was internally rotated, the motion at the pubic
symphysis increased proportionally with the degrees of the
rotation as well as the applied torque measured at the dis-
tal femur. The primary rotation of the symphysis was in
the transverse plane and on average accounted for more
than 60% of the total rotation. The transverse rotation
was in the direction of opening the joint anteriorly. The
secondary rotation was in the coronal plane, which
accounted for about 20% of the total rotation, and in the
direction of opening the inferior portion of the joint. The
mean total symphysis rotation and the transverse rotation
are plotted against the applied torque and presented in
Figure 3. The mean rotations of the symphysis in the
transverse plane at 3 selected levels of internal rotation
are listed in Table 1. Both testing states showed a change
in the slope of the rotation versus torque plot around
12.0 N?m. This change in slope represents a change in
the stiffness of the system, and we postulate this repre-
sents the point of bony contact.
In the symphysis rotation versus applied torque plots,
a more rapid motion increase in the primary transverse
plane was noted for the cam group once the applied torque
exceeded 12.0 N?m, further supporting that this likely repre-
sents the point of contact. The mean transverse plane rota-
tion was comparable among the groups at 12.0 N?m, valued
at 0.62? 6 0.37? for the native state and 0.68? 6 0.37? for
the cam state. At a torque level of 18.0 N?m, the mean trans-
verse rotation increased to 0.89? 6 0.35? for the native state
and 1.20? 6 0.41? for the cam state. The difference between
the cam and native groups was statistically significant (P \
.03) at the higher level of applied torque. Repeated ANOVA
and post hoc test shows a trend toward significant difference
in stiffness in the lower torque region (0-12.0 N?m), where
stiffness of the native group is 52% higher than the cam
group (P\.058), which represents the effect of the soft tissue
on stiffness. The joint stiffness became comparable at the
higher torque region (ANOVA, P = .20), which would be con-
sistent with bone-on-bone contact.
Femoroacetabular impingement and athletic pubalgia or
sports hernia have increasingly been recognized in athletes
as a source of disability and inability to perform.4,13,24A
clinical link has been reported between femoroacetabular
impingement and athletic pubalgia.17The purpose of this
study was to examine the effect of simulated cam impinge-
ment on motion of the pubic symphysis.
It has been established that both proximal femoral ret-
roversion and the presence of a cam lesion of the femoral
head-neck junction cause a loss of physiologic internal
rotation of the hip.29Mechanically, this occurs because
the bone of the retroverted neck or of the cam lesion con-
tacts the rim of the acetabulum earlier in the arc of motion
and acts as a block to further rotation of the hip. The
amount of functional internal rotation for sports is likely
about 30?.25An athlete with reduced physiologic internal
rotation secondary to cam or femoral retroversion to
around 0? to 10? would be at risk for contacting the rim
and labrum with high torque. This has been shown to
cause tears of the acetabular labrum and subsequent groin
pain and disability to the patient.2,22,26Good results with
return to full activity have been reported with arthroscopic
treatment of this injury.5,11,16,24
Loss of internal rotation of the hip has also been shown
to be highly correlated with the incidence of groin injuries
and osteitis pubis in a 2-year prospective study of Austra-
lian rules football.30Osteitis pubis in association with inju-
ries to the low abdominal musculature has been recognized
as athletic pubalgia and has been treated successfully with
high return to activity with repair of these injured
A recent study reported on a series of patients with both
femoroacetabular impingement and sports hernia. The
results of treating either entity in isolation led to signifi-
cant recurrence of symptoms, whereas treatment of both
led to 24 of 27 patients returning to full unrestricted activ-
ity.17The authors suggest that the range-of-motion restric-
tions caused by FAI could create compensatory patterns
that affect the extra-articular structures involved in ath-
letic pubalgia. There has also been a case report linking
femoroacetabular impingement to osteitis pubis.20
The current study demonstrates that simulated cam
impingement causes rotation of the pubic symphysis after
the point of bony contact, which occurred consistently
between 10.0 N?m and 12.0 N?m. The point of bony contact
was confirmed by correlating the amount of torque to the
degrees of internal rotation on a handheld level and then
repeating the same hip rotation with the capsule open
and the cam lesion and acetabular rim under direct visual-
ization. The amount of motion occurring at the pubic sym-
physis was in all planes but most significantly in the
transverse plane (60% of the total motion). This relation-
ship is present in the native state, as well as in the cam
states. However, the cam lesion caused significantly more
rotation of the symphysis (as much as 35% more) as com-
pared with the native no-cam state at every level of torque
above the point of bony contact. The rotational motion was
small, on the order of about one-third of a degree, but the
results were reproducible over all 12 specimens tested in
triplicate. It is possible that even very small amounts of
motion as a chronic repetitive loading could cause an over-
use injury to the symphysis or parasymphyseal structures.
This signifies that after the femoral neck contacts the
acetabular rim, motion occurs at the pubic symphysis,
which can be a normal occurrence. This motion becomes
4Birmingham et al The American Journal of Sports Medicine
pathologic when the bony anatomy causes this contact to
occur earlier in the arc of internal rotation than is required
for activity. Therefore, any cause of loss of internal rotation
(cam, retroversion of the proximal femur, coxa vara) could
lead to repetitive loading of the pubic symphysis for activ-
ities that require more motion.
This study was not without its limitations. The stiff-
ness, as represented by the slope of the curves or degrees
per unit of torque (N?m), was significantly higher for the
native state, where there was less motion per unit of tor-
que before bony contact. Therefore, the effect on stiffness
of the construct by cutting the capsule is further supported
by the stiffness of the native versus the cam state. Before
contact, the soft tissue acts as the major restraint to motion
and not the bone. The amount of motion at the symphysis
recorded in our results would be underestimated given this
line of reasoning.
Second, the artificial cam was not an anatomic repre-
sentation of a true loss of head-neck offset. In addition,
there could have been undetected motion of the cam sec-
ondary to inadequate fixation or subsidence into the bone
of the neck during testing. This would lead to an underes-
timation of the transfer of motion to the symphysis.
Finally, using a 6 degree-of-freedom robot would have pro-
duced a more reproducible hip internal rotation loading
path between trials. However, despite this reproducibility,
a robot model system would be more constrained than our
method and would likely represent a less natural arc of
motion than we reproduced manually.
In addition, the effect of a decreased head-neck offset
was the only type of dynamic FAI evaluated directly. The
other types of FAI not evaluated were femoral retrover-
sion, coxa vara, and the acetabular pincer. Also, the hip
was only tested in the 90? of flexion, neutral adduction
position, which may not truly reproduce range-of-motion
requirements during various activities. The motion trans-
mitted through the sacroiliac joint and lumbar spine was
not evaluated. The mobility or rigidity of these articula-
tions would affect how much motion is transmitted
through the pubic symphysis and would ideally have
been measured. Also, the 12 hips tested were right and
left hips from 6 pelvises. It is possible that the first hip
tested could have altered the mechanics of the symphysis
for the second hip tested.
The results of this study suggest that dynamic cam
impingement causes rotational motion at the pubic sym-
physis after the point of bony contact. Repetitive loading
of the symphysis is a known precursor to athletic pubal-
gia.10This could give one possible explanation for the clin-
ical observation that patients with FAI also can present
with athletic pubalgia or osteitis pubis.17,20Further stud-
ies must be carried out to support this link between pubic
symphysis motion and athletic pubalgia and osteitis pubis.
1. Ahumada LA, Ashruf S, Espinosa-de-los-Monteros A, et al. Athletic
pubalgia: definition and surgical treatment. Ann Plast Surg. 2005;
2. Beck M, Kalhor M, Leunig M, Ganz R. Hip morphology influences the
pattern of damage to the acetabular cartilage: femoroacetabular
impingement as a cause of early osteoarthritis of the hip. J Bone Joint
Surg Br. 2005;87(7):1012-1018.
3. Bedi A, Dolan M, Leunig M, Kelly BT. Static and dynamic mechanical
causes of hip pain. Arthroscopy. 2011;27(2):235-251.
4. Bizzini M, Notzli HP, Maffiuletti NA. Femoroacetabular impingement
in professional ice hockey players: a case series of 5 athletes after
open surgical decompression of the hip. Am J Sports Med.
5. Byrd JW, Jones KS. Prospective analysis of hip arthroscopy with 10-
year followup. Clin Orthop Relat Res. 2010;468(3):741-746.
6. Dorrell JH, Catterall A. The torn acetabular labrum. J Bone Joint Surg
7. Espinosa N, Beck M, Rothenfluh DA, Ganz R, Leunig M. Treatment of
femoro-acetabular impingement: preliminary results of labral refixa-
tion. Surgical technique. J Bone Joint Surg Am. 2007;89(suppl 2, pt
8. Ganz R, Parvizi J, Beck M, Leunig M, Notzli H, Siebenrock KA. Fem-
oroacetabular impingement: a cause for osteoarthritis of the hip. Clin
Orthop Relat Res. 2003;417:112-120.
9. Garras DN, Carothers JT, Olson SA. Single-leg-stance (flamingo)
radiographs to assess pelvic instability: how much motion is normal?
J Bone Joint Surg Am. 2008;90(10):2114-2118.
10. Garvey JF, Read JW, Turner A. Sportsman hernia: what can we do?
11. Ilizaliturri VM Jr, Orozco-Rodriguez L, Acosta-Rodriguez E, Cama-
cho-Galindo J. Arthroscopic treatment of cam-type femoroacetabu-
lar impingement: preliminary report at 2 years minimum follow-up. J
12. Ito K, Minka MA II, Leunig M, Werlen S, Ganz R. Femoroacetabular
impingement and the cam-effect: a MRI-based quantitative anatom-
ical study of the femoral head-neck offset. J Bone Joint Surg Br.
13. Keogh MJ, Batt ME. A review of femoroacetabular impingement in
athletes. Sports Med. 2008;38(10):863-878.
14. Klaue K, Durnin CW, Ganz R. The acetabular rim syndrome: a clinical
presentation of dysplasia of the hip. J Bone Joint Surg Br. 1991;
15. Lamontagne M, Varin D, Beaule PE. Does the anterior approach for
total hip arthroplasty better restore stair climbing gait mechanics?
J Orthop Res. 2011;29(9):1412-1417.
16. Larson CM, Giveans MR. Arthroscopic management of femoroace-
tabularimpingement:early outcomes measures.
17. Larson CM, Pierce BR, Giveans MR. Treatment of athletes with
symptomatic intra-articular hip pathology and athletic pubalgia/
sports hernia: a case series. Arthroscopy. 2011;27(6):768-775.
18. Lindsay DM, Maitland M, Lowe RC, Kane TJ. Comparison of isoki-
netic internal and external hip rotation torques using different testing
positions. J Orthop Sports Phys Ther. 1992;16(1):43-50.
19. Mardones R, Lara J, Donndorff A, et al. Surgical correction of
‘‘cam-type’’ femoroacetabular impingement: a cadaveric com-
parison of open versus arthroscopic debridement. Arthroscopy.
20. Matsuda DK. Endoscopic pubic symphysectomy for recalcitrant
osteitis pubis associated with bilateral femoroacetabular impinge-
ment [published online ahead of print March 10, 2009]. Orthopedics.
21. Meyers WC, Foley DP, Garrett WE, Lohnes JH, Mandlebaum BR. Man-
agement of severe lower abdominal or inguinal pain in high-performance
athletes. PAIN (Performing Athletes with Abdominal or Inguinal Neuro-
muscular Pain Study Group). Am J Sports Med. 2000;28(1):2-8.
22. Nepple JJ, Zebala LP, Clohisy JC. Labral disease associated with
femoroacetabular impingement: do we need to correct the structural
deformity? J Arthroplasty. 2009;24(6)(suppl):114-119.
23. Notzli HP, Wyss TF, Stoecklin CH, Schmid MR, Treiber K, Hodler J. The
contour of the femoral head-neck junction as a predictor for the risk of
anterior impingement. J Bone Joint Surg Br. 2002;84(4):556-560.
Vol. XX, No. X, XXXXEffect of Cam Impingement on Pubic Symphysis Motion5
24. Philippon MJ, Maxwell RB, Johnston TL, Schenker M, Briggs KK. Download full-text
Clinical presentation of femoroacetabular impingement. Knee Surg
Sports Traumatol Arthrosc. 2007;15(8):1041-1047.
25. Roach KE, Miles TP. Normal hip and knee active range of motion: the
relationship to age. Phys Ther. 1991;71(9):656-665.
26. Rylander L, Froelich JM, Novicoff W, Saleh K. Femoroacetabular
impingement and acetabular labral tears. Orthopedics. 2010;
27. Siebenrock KA, Schoeniger R, Ganz R. Anterior femoro-acetabular
impingement due to acetabular retroversion: treatment with periace-
tabular osteotomy. J Bone Joint Surg Am. 2003;85(2):278-286.
28. Taylor DC, Meyers WC, Moylan JA, Lohnes J, Bassett FH, Garrett
WE Jr. Abdominal musculature abnormalities as a cause of groin
pain in athletes: inguinal hernias and pubalgia. Am J Sports Med.
29. Tonnis D, Heinecke A. Acetabular and femoral anteversion: relation-
ship with osteoarthritis of the hip. J Bone Joint Surg Am. 1999;
30. Verrall GM, Slavotinek JP, Barnes PG, Esterman A, Oakeshott RD,
Spriggins AJ. Hip joint range of motion restriction precedes athletic
chronic groin injury. J Sci Med Sport. 2007;10(6):463-466.
31. Wagner S, Hofstetter W, Chiquet M, et al. Early osteoarthritic changes
of human femoral head cartilage subsequent to femoro-acetabular
impingement. Osteoarthritis Cartilage. 2003;11(7):508-518.
32. Wang M, Bryant JT, Dumas GA. A new in vitro measurement tech-
nique for small three-dimensional joint motion and its application to
the sacroiliac joint. Med Eng Phys. 1996;18(6):495-501.
33. Williams PR, Thomas DP, Downes EM. Osteitis pubis and instability
of the pubic symphysis: when nonoperative measures fail. Am J
Sports Med. 2000;28(3):350-355.
For reprints and permission queries, please visit SAGE’s Web site at http://www.sagepub.com/journalsPermissions.nav
6 Birmingham et al The American Journal of Sports Medicine