Current practice of BHS members in the treatment of osteonecrosis of the femoral head in adults

Abstract

Introduction:
The management of osteonecrosis (ON) of the femoral head remains controversial. It is unclear the extent to which non-arthroplasty procedures are used and there has been no previous report of the trends in operative management of ON in the UK. Our objective is to report current trends in management of ON of the femoral head amongst specialist hip surgeons in the UK.

Methods:
A single-stage internet-based survey was e-mailed to 352 eligible members of the British Hip Society (BHS). This consisted of 10 question stems including 16 hypothetical clinical scenarios with imaging.

Results:
115 active Consultant members of the BHS completed the survey. For symptomatic pre-collapse ON we found core decompression (CD) was the most common operative intervention and for post-collapse ON we found that total hip arthroplasty (THA) was the most common operative intervention. We found no difference in the rate of operative intervention between 24 and 48-year-old patients at any stage of ON but joint preserving procedures were more often selected for the younger patient and arthroplasty for the older patient. Surgeons were more likely to offer arthroplasty to a 48-year-old patient at an earlier stage of disease.

Conclusions:
Our respondents would offer different operative interventions dependent on stage of ON and patient age. Core decompression (CD) and arthroplasty were common but variation in treatment options offered suggests a lack of consensus amongst UK hip surgeons. We suggest that further research such as a prospective RCT is needed to gain consensus on management of this condition.

Full text

HIP
ISSN 1120-7000
HIP International
2018, Vol. 28(1) 90 –95
© The Author(s) 2017
Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
DOI: 10.5301/hipint.5000535
journals.sagepub.com/home/hpi
ORIGINAL RESEARCH ARTICLE
Several factors inuence disease progression, including size
of lesion, age, aeological risk factors, and there are varying
levels of evidence for pharmacological and other nonopera-
ve treatments to modify risk factors (3). Several classicaon
systems are in use but no accepted way of relang the dis-
ease stages between systems (1, 4). Of 76,448 primary hip
replacements undertaken in the UK in 2012, the vast major-
ity (92%) were performed for osteoarthris (5). Only 2% were
performed for ON but this accounted for 24% of all paents
undergoing total hip arthroplasty (THA) under 30 and 19% of
paents under 40 years old (5).
In 2005, McGrory et al (6) conducted a survey of mem-
bers of the American Associaon of Hip and Knee Surgeons
(AAHKS) regarding their current pracce in management of
ON of the femoral head. Of the 403 (54%) acve members
who responded core decompression (CD) was the most com-
mon intervenon for pre-collapse ON, and THA was the most
common for post-collapse ON. The authors recognised a wide
variaon in reported pracce and advocated conducng of
mulcentre prospecve randomised controlled trials to de-
termine best pracce.
To our knowledge, no similar study has been performed in
the UK, and it would be interesng to compare variaons in
Current pracce of BHS members in the
treatment of osteonecrosis of the femoral
head in adults
Henry B. Colaço1, Jerome A. Davidson2, Dominic Davenport3, Mark. C. Norris4, Marcus J.K. Bankes3, Zameer Shah3
1 Trauma and Orthopaedics Department, Royal Free Hospital, London - UK
2 Royal Naonal Orthopaedic Hospital, Stanmore - UK
3 Department of Orthopaedics, Guy’s and St. Thomas’ NHS Foundaon Trust, London - UK
4 Department of Orthopaedics, Darent Valley Hospital, Darord - UK
Introducon
Avascular necrosis of the femoral head (ON) presents
a unique set of challenges to the orthopaedic surgeon. It
is most prevalent in younger individuals in their 3rd and 4th
decades (1). The natural history of the condion is failure of
bone remodeling progressing to subchondral fracture, col-
lapse, and ulmately hip arthris (2). Management of ON of
the femoral head is controversial, and there remains a wide
variaon in pracce worldwide. Marker et al (1) suggested a
treatment algorithm in 2008, but no established guidelines in
the UK exist for this condion.
ABSTRACT
Introducon: The management of osteonecrosis (ON) of the femoral head remains controversial. It is unclear
the extent to which non-arthroplasty procedures are used and there has been no previous report of the trends
in operave management of ON in the UK. Our objecve is to report current trends in management of ON of the
femoral head amongst specialist hip surgeons in the UK.
Methods: A single-stage internet-based survey was e-mailed to 352 eligible members of the Brish Hip Society
(BHS). This consisted of 10 queson stems including 16 hypothecal clinical scenarios with imaging.
Results: 115 acve Consultant members of the BHS completed the survey. For symptomac pre-collapse ON we
found core decompression (CD) was the most common operave intervenon and for post-collapse ON we found
that total hip arthroplasty (THA) was the most common operave intervenon. We found no dierence in the
rate of operave intervenon between 24 and 48-year-old paents at any stage of ON but joint preserving pro-
cedures were more oen selected for the younger paent and arthroplasty for the older paent. Surgeons were
more likely to oer arthroplasty to a 48-year-old paent at an earlier stage of disease.
Conclusions: Our respondents would oer dierent operave intervenons dependent on stage of ON and pa-
ent age. Core decompression (CD) and arthroplasty were common but variaon in treatment opons oered
suggests a lack of consensus amongst UK hip surgeons. We suggest that further research such as a prospecve
RCT is needed to gain consensus on management of this condion.
Keywords: Arthroplasty, Avascular necrosis, Conservave hip surgery, Core decompression, Osteonecrosis, Young
adult hip
Accepted: May 10, 2017
Published online: September 5, 2017
Corresponding author:
Dominic Davenport
Department of Orthopaedics
Guy’s Hospital
Great Maze Pond
SE1 9RT, London, UK
dominic.davenport@nhs.net

Colaço et al 91
pracce including use of novel treatments and threshold for
arthroplasty. We hypothesised that despite the development
of novel treatment opons most surgeons would oer CD for
pre-collapse ON and arthroplasty for post-collapse ON. We
hypothesised that joint preserving surgery would be more
commonly used in the younger paent group (24 years), and
the threshold for arthroplasty would be higher than for the
older paent group (48 years).
Methods
An invitaon to complete an on-line electronic survey
was e-mailed by the Brish Hip Society (BHS) to all known
acve Consultant members (n = 352) from the total number
of 447 members (remainder listed as non-Consultant grade).
We sent a single reminder e-mail aer 4 weeks, and the
survey was open for 8 weeks. Demographic data collected
included years of experience, fellowship training, geographi-
cal region and annual arthroplasty volume. The survey was
incenvised to minimise responder bias. All respondents’
details were anonymised to enhance condenality and en-
courage responses.
Our survey is comprised of 10 stems with dierent ques-
on styles including ranking, mulple choice, matrix of choic-
es, and matrix of drop-down menus (Appendix 1, available
online as Supplementary material at www.hip-int.com). The
survey was carefully designed to reduce total survey error,
by minimising the 4 major potenal sources of error: sam-
pling, noncoverage, measurement and nonresponse. The
ease of use of internet surveys can potenally be exploited to
increase the response rate, which is the generally accepted
indicator of nonresponse error. The survey will refer to pro-
gressive stages consistent with the Steinberg classicaon for
the clinical scenario quesons (7). 16 clinical scenarios are
presented: 8 dierent combinaons of Steinberg stage, level
of symptoms, and Kerboul angle in 2 dierent age groups
(24 and 48 years old). Surgeons were asked to select their
preferred treatment opon for 16 hypothecal clinical sce-
narios of paents with ON of the hip.
There were no paent idenable cases used in the sce-
narios and therefore ethical approval was not required. The
quesons have been carefully designed to discriminate be-
tween parcipants’ individual approaches to management.
Data were analysed using Microso® Excel. Descripve sta-
scs and presented, and comparisons analysed using chi-
square test.
Results
Responder demographics
We received responses from 115/352 (33%) BHS mem-
bers. 90 responders chose to indicate their experience, annual
THA operaon volume and fellowship status. 33.3% had more
than 20 years of hip surgery and the majority (83.3%) of the
responders had at least 10 years’ experience. Only 2 (2.2%) of
responders had <5 years’ experience. Parcipants performed
a high volume of THA with the majority (55.4%) performing
over 100 per year while only 14.13% of respondents per-
formed <50 THA per year.
Of our responders 54.4% had UK hip fellowship experi-
ence and 37.8% had a hip fellowship experience abroad, the
majority of which were in USA, Canada and Australia. Only
17.8% had no fellowship experience specic to hip surgery.
This gure was consistent with our nding that 79.1% of re-
spondents described themselves as either ‘hip specialists’
or ‘specialist lower limb arthroplasty’ surgeon. The survey
was completed by a wide geographical spread of surgeons
throughout the UK with similar numbers from NHS District
General Hospital and NHS Teaching Hospital (49.4% and
44.7% respecvely), the remaining 5 (5.9%) operated exclu-
sively in the independent sector.
The use of classicaon systems
The majority (89.1%) used Ficat and Arlet (8) or Ficat (9)
classicaon system to guide their treatment plan and deci-
sion making but 57% were also sased to use a binary (pre-
vs. post-collapse) system. A minority used other classicaons:
Steinberg et al (7), Gardeniers (10) (ARCO), Ohzono et al (11)
and Kerboul et al (12) classicaons were used by 14%, 12%,
7%, and 3%, respecvely.
Modicaon of risk factors
75% of parcipants would advocate treatment for alco-
hol dependency, and 77.8% would stop ongoing steroid use if
appropriate. 60% would oer bisphosphonate therapy to pa-
ents taking steroids, 35% would oer thromboprophylaxis
to paents with hypercoagulability or a stan to paents with
hyperlipidaemia. Other nonoperave treatments such as ex-
tra- corporeal shock wave therapy (3%) or hyperbaric oxygen
therapy (6%) were less frequently oered.
Clinical scenarios: management opons
Full results are summarised in Appendix 2A and 2B, avail-
able online as Supplementary material at www.hip-int.com.
Minimal symptoms, normal radiograph but abnormal MRI
(Steinberg Stage IB)
24-year-old: 26 of 94 (28%) responders would oer CD
with 10 (11%) oering single 8-10 mm technique and 16
(17%) a mulple drill hole technique. 66 (70%) would manage
this case nonoperavely. 1 (1%) respondent would oer THA.
48-year-old: 18 of 93 (19%) responders would oer CD
with 8 (9%) oering a single 8-10 mm technique and 10 (11%)
a mulple drill hole technique. 73 (79%) would manage non-
operavely. 1 (1%) respondent would oer THA.
Moderate symptoms, normal radiograph but abnormal MRI
(Steinberg Stage IIB)
24-year-old: 65 of 94 (69%) responders would oer CD
with 39 (42%) oering single 8-10 mm technique and 26
(28%) a mulple drill hole technique. 27 (29%) would manage
nonoperavely. 1 (1%) respondent would oer THA.
48-year-old: 58 of 93 (63%) responders would oer CD
with 38 (41%) oering a single 8-10 mm technique and 20

A focussed survey of hip specialist pracce in the UK
92
(22%) a mulple drill hole technique. 32 (34%) would manage
nonoperavely. 2 (2%) respondents would oer THA.
Minimal symptoms, radiographic evidence of sclerosis and
poroc femoral head (Steinberg Stage IIB)
24-year-old: 11 of 92 (12%) responders would oer CD
with 9 (10%) oering a single 8-10 mm technique, 2 (2%)
a mulple drill hole technique and 1 (1%) would oer an
osteotomy. 79 (86%) would manage nonoperavely. 1 (1%)
respondent would oer THA.
48-year-old: 9 of 91 (10%) responders would oer CD with
7 (8%) oering a single 8-10 mm technique, 2 (2%) a mulple
drill hole technique, 1 (1%) responder would oer CD with
tantalum rod implantaon. 79 (87%) would manage nonop-
eravely. 2 (2%) respondents would oer THA.
Moderate symptoms, radiographic evidence of sclerosis and
poroc femoral head (Steinberg Stage IIB)
24-year-old: 31 of 92 (34%) responders would oer CD
with 20 (22%) oering a single 8-10 mm technique, 11 (12%)
a mulple drill hole technique and 1 (1%) would oer nonvas-
cularised bula gra. 49 (53%) would manage nonoperave-
ly. 7 (8%) respondents would oer THA, 1 (1%) respondent
would oer resurfacing arthroplasty.
48-year-old: 23 of 91 (25%) responders would oer CD
with 14 (15%) oering a single 8-10 mm technique and 9
(10%) a mulple drill hole technique. 53 (58%) would manage
nonoperavely. 14 (15%) respondent would oer THA, 1 (1%)
responder would oer resurfacing arthroplasty.
Radiographic subchondral collapse of the femoral head and
moderate head involvement (Steinberg Stage IIIB), (25%, Ker-
boul angle 100⁰)
24-year-old: 57 of 94 (61%) responders would oer CD
with 32 (34%) oering a single 8-10 mm technique, 24 (26%)
a mulple drill hole technique and 1 (1%) would oer CD and
tantalum rod implantaon. 8 (9%) would oer a vascularised,
4 (4%) would oer nonvascularised bula gra and 2 (2%)
would oer osteotomy. 13 (14%) would manage nonopera-
vely. 6 (6%) respondent would oer THA, and 2 (2%) would
oer resurfacing arthroplasty.
48-year-old: 46 of 92 (50%) responders would oer CD
with 26 (28%) oering a single 8-10 mm technique, 18 (20%)
a mulple drill hole technique, 2 (2%) would oer CD and
tantalum rod implantaon. 1 (1%) would oer a vascular-
ised, 2 (2%) would oer nonvascularised bula gra. 19
(21%) would manage nonoperavely. 20 (22%) respondents
would oer THA, 3 (3%) would oer resurfacing arthroplasty.
Radiographic subchondral collapse of the femoral head and
large area of involvement (Steinberg Stage IIIB), (35%, Ker-
boul angle 230⁰)
24-year-old: 43 of 91 (47%) responders would oer CD
with 20 (22%) oering a single 8-10 mm technique, 21 (23%)
a mulple drill hole technique, 2 (2%) would oer CD and tan-
talum rod implantaon. 11 (12%) would oer a vascularised, 3
(3%) would oer nonvascularised bula gra and 3 (3%) would
oer osteotomy. 10 (10%) would manage nonoperavely. 16
(18%) respondents would oer THA, 1 (1%) would oer resur-
facing arthroplasty.
48-year-old: 36 of 90 (40%) responders would oer CD with
19 (21%) oering a single 8-10 mm technique, 15 (17%) a mul-
ple drill hole technique, 2 (2%) would oer CD and tantalum
rod implantaon. 2 (2%) would oer a vascularised, 2 (2%). 13
(14%) would manage nonoperavely. 33 (37%) respondents
would oer THA, 1 (1%) would oer resurfacing arthroplasty.
Severe symptoms and femoral head aening post collapse
(Steinberg Stage IVB)
24-year-old: 72 of 92 (78%) of respondents would oer
THA, while a further 4 (4%) would oer resurfacing arthro-
plasty. 8 of 92 (9%) respondents would oer an osteotomy, 1
(1%) respondent would oer hemiarthroplasty, 1 (1 %) would
oer vascularised bula gra and 2 (2%) would oer CD. 2
(2%) would manage this nonoperavely.
48-year-old: 85 of 91 (92%) respondents would oer THA,
while a further 4 (4%) would oer resurfacing arthroplasty.
1 (1%) of respondents would oer an osteotomy, and 1 (1%)
would manage this case nonoperavely.
Severe symptoms with radiographic femoral head collapse
and established osteoarthris (Steinberg Stage V)
24-year-old: 89 of 94 (95%) of respondents would oer
THA, while a further 3 (3%) would oer resurfacing arthro-
plasty. 1 (1%) would oer an osteotomy, and 1 (1%) would
oer arthrodesis. No respondents would choose to manage
this paent nonoperavely.
48-year-old: All 93 respondents (100%) would oer ar-
throplasty in this case, with 92 (99%) selecng THA and 1
(1%) respondent would oer resurfacing arthroplasty.
Comparison of pre- vs. post-collapse ON
For symptomac pre-collapse ON in a 24-year-old the
most common intervenon was CD (52%). 41% of responders
would manage this nonoperavely while only 5% would oer
arthroplasty.
For symptomac pre-collapse ON in a 48-year-old the
most common management was nonoperave (46%) fol-
lowed by CD (44%) and THA (9%).
For severe post-collapse ON in a 24-year-old the most
common intervenon was THA (49%) followed by CD (28%)
and other joint preserving surgery (13%). 7% of responders
would manage this paent nonoperavely.
For severe post-collapse ON in a 48-year-old the most
common intervenon was THA (63%) followed by CD (23%)
and other joint preserving surgery (2%). 9% of responders
would manage this paent nonoperavely.
Factors inuencing management decisions
Surgeons were asked to rank various paent factors in
order of importance in inuencing management decisions.
Stage of the disease and symptoms were the most commonly

Colaço et al 93
selected as the single-most important factor both selected
by 39%. Less frequently considered factors were age (11%),
disease progression (7%), size of the lesion (3%), and risk
factors (1%).
Eect of age on choice of intervenon
There was no signicant dierence in decision for op-
erave versus nonoperave intervenon regardless of stage
between the 24-year-old (67.9%) and 48-year-old (63.2%)
paent (p = 0.11 chi2).
When operaons were grouped into either joint preserv-
ing surgery (including CD, osteotomy and bula graing) or
arthroplasty, we found that joint preserving surgeries were
more commonly oered to the 24-year-old paent than the
48-year-old (58.5% vs. 44.2%, p<0.0001), conversely arthro-
plasty was oered more commonly to the 48-year-old (55.6%
vs. 41.3%, p<0.0001).
Eect of symptoms on decision to operate
We found symptoms had a signicant eect on decision to
oer operave intervenon in early disease across both ages
(p<0.0001) (Fig. 1).
At what stage of disease is arthroplasty oered?
We found that for both age groups decision to oer
arthroplasty increased with stage of disease and for the
48-year-old paent responders were signicantly more likely
to oer arthroplasty at an earlier stage (Stage III) of disease
when compared to 24-year paent (Fig. 2).
What type of arthroplasty is oered?
Arthroplasty was oered in 41.3% of all operave inter-
venons for the 24-year-old paent. Most (94%) were THA,
followed by resurfacing (5.4%) and hemiarthroplasty (0.5%).
Arthroplasty was selected for 55.6% of cases for the 48-year-
old paent with a similar trend in type (96.1% THA, 3.9%
resurfacing, 0 hemiarthroplasty). Further analysis of the THA
oered showed the most popular choice was uncemented
THA (58.6% in 24-year-old, 55.2% in 48-year-old), followed
by hybrid (30%, 30.7%) and cemented THA (11.4%, 14.1%).
However, we found no signicant dierence between types
of THA between age groups.
Conclusion
There has been no previous report of the trends in opera-
ve management of ON in the UK, and there are no estab-
lished guidelines for management of this condion. We used
this survey to evaluate the approach to diagnosis and man-
agement of hip ON in acve consultant hip surgeon members
of the Brish Hip Society. Of the 115 respondents, 83% had
at least 10 years of surgical pracce as a consultant, 82% had
undertaken a specialist hip fellowship and the majority (75%)
of our respondents describe themselves as either lower limb
arthroplasty or hip specialists.
Like the trends found by the AAHKS our respondents were
likely to recommend cessaon of alcohol or steroid use in pa-
ents with ON, however our respondents were more likely
to oer medical therapy compared to the AAHKS study (6).
60% in our survey would oer a bisphosphonate to paents
on steroids, 35% would oer thromboprophylaxis for paents
with coagulaon disorders and 35% would oer a stan to
paents with hyperlipidemia, compared to the 3%, 6% and
10% respecvely found by McGrory et al (6). There have been
several studies on pharmacological treatments for early ON
of the femoral head (13-15), which may account for their
popularity amongst UK surgeons.
Our survey established that 2 key criteria aid our respon-
dents’ evaluaon; the severity of symptoms and stage of dis-
ease. This was supported by the responses with signicantly
higher operave intervenon rate when symptoms were
moderate rather than minimal. Only 11% of our responders
stated that age was an important factor, which is consistent
with our nding that for all stages of ON rate of operave ver-
sus nonoperave management was not signicantly dierent
regardless of age.
Intervenon in early ON remains controversial. When spe-
cically isolang the responses to the scenario of a paent
Fig. 1 - Rate of operave intervenon in scenarios of mild vs. moder-
ate symptoms for pre-collapse osteonecrosis in a 24- and 48-year-
old paent.
Fig. 2 - Rate of arthroplasty in each stage of osteonecrosis in a 24-
and 48-year-old paent.

A focussed survey of hip specialist pracce in the UK
94
with minimal symptoms and pre-collapse ON most would
manage nonoperavely (70% and 79% for the 24- and 48-year-
old respecvely) however even in this early stage there are re-
spondents who would oer CD (27% for the 24-year-old, 18%
for the 48-year-old) which may reect the evidence that it can
delay disease progression despite lack of symptoms (16). Sim-
ilar results were also seen in the AAHKS study (2).
In the context of symptomac pre-collapse ON joint pre-
serving surgery including CD was the most common interven-
on oered to the 24-year-old paent, whereas nonoperave
management was more commonly oered to the 48-year-old
paent. Arthroplasty was infrequently oered in 4% and 9%
for either age group. This may reect the successful results of
joint preserving surgery in pre-collapse ON in the recent liter-
ature. A meta-analysis of 2025 hips found sasfactory clinical
results in 63.5% of hips managed by CD compared with only
22.7% of hips managed nonoperavely for all stages of ON
(17) and 71% versus 34.5% sasfactory clinic results in pre-
collapse ON in parcular (17). In a long-term study of CD in 34
paents (54 hips) with Ficat stage I-III, Bozic et al (18) report-
ed 48% sasfactory clinical outcomes and 37% radiographic
survival however they observed a 100% failure in 8 Stage III
hips. Consistent with this we found the greatest variaon in
response appears in the corresponding scenario of a paent
Stage III ON.
For scenarios involving post-collapse ON, arthroplasty
was the most popular management in both ages groups and
indeed THA has been shown to be a reliable opon for the
management of post-collapse hip ON (19); a systemac re-
view of 67 studies represenng 3,277 THRs in 2593 paents
reported a survivorship of 97% at a mean of 6 years follow up
(20). Interesngly, we found that while arthroplasty was of-
fered to both the 24- and 48-year-old paent, in subchondral
collapse (Steinberg stage III) it was signicantly more oen
oered to the 48-year-old paent – an outcome not observed
in other scenarios suggesng surgeons are more inclined to
oer arthroplasty at an earlier stage of disease to the 48-year-
old paent. Whilst advances in implant design and bearing
surface technology may have reduced the high incidence of
wear and rate of asepc loosening following arthroplasty in
younger paents, we believe our results indicate that many
surgeons are sll reluctant to oer arthroplasty to younger
paents with early ON.
THA was the most common choice of arthroplasty at ev-
ery stage regardless of age with very few surgeons choosing
resurfacing or hemiarthroplasty reecng a wider trend in
the wake of reported high revision rates (21) despite poten-
ally being a good opon for younger male paents (22).
Uncemented THA was the most popular, followed by hybrid
and cemented THR (54.0%, 28.5%, 12.5% respecvely). Our
gures dier from those reected in the Naonal Joint Reg-
istry (NJR) records for 2012 (5) where the proporon of un-
cemented THA is lower (44%) (5) which we believe reects
the increasing preference of metaphyseal press t stems in
younger paents amongst specialist hip surgeons.
This study has limitaons common to survey-based meth-
odology: respondents were disclosing their opinions based
on hypothecal scenarios rather than real-life paents. We
recognise that in clinical pracce other variables may inu-
ence individual paent management decisions, which are
beyond the scope of this study. However, we believe that
the hypothecal scenarios cover a broad spectrum of clinical
presentaons. The methodology used in the quesonnaire
was designed to minimise sampling, noncoverage, measure-
ment and nonresponse bias. We recognise that our respond-
ers reected a self-selected group of specialist hip surgeons,
but will not include all praccing hip surgeons in the UK. It
is possible that the management in these scenarios would
dier between BHS members and non-BHS member hip sur-
geons. Sampling error was minimised by inving all eligible
members of the BHS to parcipate, and including all respons-
es from parcipants who meet the inclusion criteria in the
analysis. Noncoverage error was minimised by approaching a
self-selected subgroup of praccing hip surgeons, but will not
include currently praccing UK-based hip surgeons who are
non-members of the BHS. Nonresponse error stems from fail-
ure to respond to the e-mailed invitaon, and can potenally
skew the result if responders share dierent characteriscs
to responders and despite targeng this specialist group we
only received responses from 33% of the total cohort. Several
strategies were used to increase response rate: providing in-
formaon about the survey, appealing to parcipant’s sense
of being part of a scienc community, and obligaon to
support the BHS, making the quesons varied and engaging,
and liming the survey to 10 queson stems in total. In ad-
dion, an incenve was oered for one parcipant selected
at random at the end of the study, which has been shown to
reduce nonresponse error. Measurement error refers to the
discrepancy of parcipants’ beliefs and the survey responses.
This is addressed by applying the ‘leverage-saliency’ theory of
survey response, and can be minimised by asking a relevant,
topical queson to encourage eecve interacon and valid
answers that reect parcipants’ true beliefs, in addion to
careful visual layout and queson design.
This study is relevant to the orthopaedic community in the
UK, as currently there are no established guidelines for man-
agement of ON of the femoral head. Our study found that
CD and other joint preserving surgeries are the most popular
operave intervenons for pre-collapse ON in both 24- and
48-year-old paents. THA, parcularly uncemented, is the most
popular intervenon for post-collapse ON in both ages group.
We found surgeons more likely to oer THA to the 48-year-old
paent at an earlier disease stage than the 24-year-old.
There remain many unanswered quesons which is
reected in the variaon of responses both in our study and
the previous survey by McGrory et al (6) for the AAHKS. The
use of registry data and further mulcentre clinical studies
are required to enhance our understanding of outcomes in
ON of the femoral head to provide guidelines and improve
decision making in the future management of the condion.
Disclosures
Financial support: None.
Conict of interest: None.
References
1. Marker DR, Seyler TM, McGrath MS, Delanois RE, Ulrich SD,
Mont MA. Treatment of early stage osteonecrosis of the femo-
ral head. J Bone Joint Surg Am. 2008;90(Suppl 4):175-187.

Colaço et al 95
2. Zalavras CG, Lieberman JR. Osteonecrosis of the femoral
head: evaluaon and treatment. J Am Acad Orthop Surg.
2014;22(7):455-464.
3. Parsons SJ, Steele N. Osteonecrosis of the femoral head: Part
2 - Opons for treatment. Curr Orthop. 2008;22(5):349-358.
4. Mont MA, Marulanda GA, Jones LC, et al. Systemac analysis
of classicaon systems for osteonecrosis of the femoral head.
J Bone Joint Surg Am. 2006;88(Suppl 3):16-26.
5. Naonal Joint Registry for England, Wales and Northern Ire-
land. 10th Annual Report 2013. hp://www.njrcentre.org.uk/
njrcentre/Portals/0/Documents/England/Reports/10th_annu-
al_report/NJR%2010th%20Annual%20Report%202013%20B.
pdf. Accessed March 28, 2017.
6. McGrory BJ, York SC, Iorio R, et al. Current pracces of AAHKS
members in the treatment of adult osteonecrosis of the femo-
ral head. J Bone Joint Surg Am. 2007;89(6):1194-1204.
7. Steinberg ME, Hayken GD, Steinberg DR. A quantave sys-
tem for staging avascular necrosis. J Bone Joint Surg Br.
1995;77(1):34-41.
8. Ficat RP, Arlet J. Forage-biopsie de la tete femorale dans
I’osteonecrose primave. Observaons histo-pathologiques
portantsur huit forages. Rev Rhum. 1964;31:257-264.
9. Ficat RP. Idiopathic bone necrosis of the femoral head. Early
diagnosis and treatment. J Bone Joint Surg Br. 1985;67(1):
3-9.
10. Gardeniers J. ARCO commiee on terminology and staging. A
new proposion of terminology and an internaonal classica-
on of osteonecrosis. ARCO Newsl. 1991;3(3):153-159.
11. Ohzono K, Saito M, Sugano N, Takaoka K, Ono K. The fate of
nontraumac avascular necrosis of the femoral head. A radio-
logic classicaon to formulate prognosis. Clin Orthop Relat
Res. 1992;(277):73-78.
12. Kerboul M, Thomine J, Postel M, Merle dAubigné R. The
conservave surgical treatment of idiopathic asepc ne-
crosis of the femoral head. J Bone Joint Surg Br. 1974;56(2):
291-296.
13. Lai KA, Shen WJ, Yang CY, Shao CJ, Hsu JT, Lin RM. The use of
alendronate to prevent early collapse of the femoral head in
paents with nontraumac osteonecrosis. A randomized clini-
cal study. J Bone Joint Surg Am. 2005;87(10):2155-2159.
14. Pritche JW. Stan therapy decreases the risk of osteone-
crosis in paents receiving steroids. Clin Orthop Relat Res.
2001;386:173-178.
15. Glueck CJ, Freiberg RA, Fontaine RN, Tracy T, Wang P. Hypo-
brinolysis, thrombophilia, osteonecrosis. Clin Orthop Relat
Res. 2001;386:19-33.
16. Hungerford DS, Jones LC. Asymptomac osteonecrosis: should
it be treated? Clin Orthop Relat Res. 2004;429:124-130.
17. Mont MA, Carbone JJ, Fairbank AC. Core decompression versus
nonoperave management for osteonecrosis of the hip. Clin
Orthop Relat Res. 1996;324:169-178.
18. Bozic KJ, Zurakowski D, Thornhill TS. Survivorship analysis of
hips treated with core decompression for nontraumac os-
teonecrosis of the femoral head. J Bone Joint Surg Am. 1999;
81(2):200-209.
19. Issa K, Pivec R, Kapadia BH, Banerjee S, Mont MA. Ostonecro-
sis of the femoral head: the THR soluon. Bone Joint J. 2013;
95-B(11)(Suppl A):46-50.
20. Johannson HR, Zywiel MG, Marker DR, et al. Osteonecrosis is
not a predictor of poor outcomes in primary THR: a systemac
literature review. Int Orthop. 2011;35:465-473.
21. Deborah A. Marshall PhD, Karen Pykerman MPH, Jason Werle
MD, Diane Lorenze MLS, Tracy Wasylak CHE, MSc, BN, Tom
Noseworthy MD, MSc, MPH, Donald A. Dick MD, Greg OConnor
MD, Aish Sundaram BMSc, Sanne Heintzbergen MSc, Cy Frank
MD. Hip Resurfacing versus THR: A Systemac Review Com-
paring Standardized Outcomes. Clin Orthop Relat Res. 2014;
472:2217-2230.
22. Börnert S, Lützner J, Beyer F, Günther KP, Hartmann A. Revision
Rate and Paent-Reported Outcome Aer Hip Resurfacing Ar-
throplasty: A Concise Follow-Up of 1064 Cases. J Arthroplasty.
2015;30(12):2190-2195.