Respondent Demographics
Sixty-eight percent of the respondents were in a private orthopaedic
practice consisting of two to ten physicians, 82% had been in practice for
more than eleven years, and 89% had performed more than 100 total hip or total
knee arthroplasties in 2004 (see Appendix). Of the 403 surgeons, 23.8%
(ninety-six) practiced in the Northeast; 24.8% (100), in the Midwest; 33.0%
(133), in the South; and 18.4% (seventy-four), in the West (see Appendix).
The mean number of total joint arthroplasties (and standard deviation)
performed by the respondents in 2004 was 264 ± 150 (range, zero to
1400). A mean of 36.7% ± 11.9% (range, 0% to 80%) of the procedures
performed by the surgeons were primary total hip arthroplasties. The mean
number of surgical interventions for osteonecrosis of the femoral head was
fourteen (range, zero to 100) in 2004.
Survey Responses
Evaluation and Nonoperative Management
Surgeons typically screen for a number of risk factors in patients with
osteonecrosis. All (100%) of the respondents indicated that they asked about
alcohol abuse or steroid use; 91% asked about trauma; 56%, about clotting
disorders; 49%, about lupus erythematosus; 36%, about a family history; 33%,
about HIV infection; 32%, about lipid storage diseases; 32%, about cancer; 8%,
about hyperbaric exposure; and 2%, about hemoglobinopathy.
Most surgeons (82%, 326 of 398) routinely obtained a magnetic resonance
imaging scan to assess whether the disease was bilateral when patients
presented with unilateral osteonecrosis of the femoral head and had a
normal-appearing contralateral hip on plain radiographs. Seven percent
(twenty-eight of 397) did not routinely obtain a magnetic resonance imaging
scan but determined whether it was indicated on the basis of factors such as
etiology, patient age, symptoms, and results of other imaging studies.
Twenty-five percent of the responding surgeons (ninety-eight of 398)
required, after establishing a diagnosis of osteonecrosis of the femoral head
associated with alcohol abuse, that the patient stop consuming alcohol prior
to the surgery, and 35% (140 of 397) obtained professional intervention for
the patient's alcohol abuse.
Fifty-one percent of the respondents (201 of 398) advised patients to
discontinue using steroid medications when an association between
osteonecrosis and steroid use had been identified. Twenty-eight percent of the
surgeons (113 of 398) recommended an osteoporosis work-up for this subset of
patients if one had not already been performed, and only a small number of
surgeons (3%; thirteen of 398) considered the use of statin
therapy11,17
to decrease the risk of osteonecrosis in other bones.
Respondents did not commonly offer
anticoagulants6,17
(6%; twenty-five of 396) or
bisphosphonates17,32,33
(10%, forty of 397) to their patients.
Surgical Management
Surgeons were asked to comment on the importance of a number of factors in
their determination of the type of surgery to offer to a patient with
osteonecrosis of the femoral head, if surgery was indicated. The stage of
disease ranked highest in importance: 84% of the surgeons (331 of 396) thought
that the stage of disease was "critically important," and 14%
(fifty-six of 396) thought that it was "important." Their
responses are presented quantitatively in
Table III.
The survey queried surgeons specifically about whether they offered hip
fusion as an option for the surgical treatment of osteonecrosis of the femoral
head. Fifteen percent of the surgeons (sixty-two of 403) offered hip fusion to
some patients, and 11% (forty-five of 403) had performed a hip fusion in the
last five years. Nine percent (thirty-six of 403) stated that their decision
to offer fusion depended on one or more factors. The most common
considerations were a patient's age (reported by twenty-eight of the
thirty-six surgeons), occupation (reported by eleven), activity level (seven),
unilateral nature of the disease (five), gender (three), and body mass index
(two). Nevertheless, hip fusion was the least likely option to be offered for
treatment in the clinical scenarios. No respondent said that he or she would
offer fusion to the hypothetical forty-eight-year-old patient for any stage of
disease, whereas 0.3% (one of 386) said that they would offer it to the
twenty-four-year-old with symptomatic stage-IVB disease, 0.5% (two of 387)
said that they would offer it to the twenty-four-year-old with symptomatic
stage-V disease, and 0.5% (two of 385) said that they would offer it to the
twenty-four-year-old with symptomatic stage-VI disease.
Clinical Scenarios
The respondents' treatment preferences for the clinical scenarios
representative of each of the
Steinberg29 stages
of osteonecrosis of the femoral head for the two hypothetical patients,
twenty-four and forty-eight years of age, are presented graphically in Figures
1,
2,
3,
4.
Normal Radiograph with Abnormal Bone Scan and/or Magnetic Resonance
Imaging (30% of Head Affected) and Associated Symptoms
(Steinberg29
Stage IB)
Two hundred and sixty-two (68%) of 384 surgeons offered a
twenty-four-year-old patient core decompression; eighty-eight (23%) of the 384
offered it with and 174 (45%) of the 384 offered it without use of bone graft
or bone-graft substitute (Fig.
1). Twenty-six (7%) of the 384 surgeons offered such a patient
vascularized bone-grafting (Fig.
2). Eighty-four (22%) of the 384 surgeons suggested nonoperative
treatment.
For the same scenario in a forty-eight-year-old patient, 268 (69%) of 388
surgeons offered core decompression; ninety-one (23%) of the 388 offered it
with and 177 (46%) of the 388 offered it without bone graft or bone-graft
substitute (Fig. 1). Twelve
(3%) of the 388 surgeons offered vascularized bone-grafting
(Fig. 2), and ninety-one (23%)
of the 388 surgeons suggested nonoperative treatment.
Normal Radiograph with Abnormal Bone Scan and/or Magnetic Resonance
Imaging (30% of Head Affected) and No Associated Symptoms
(Steinberg29
Stage IB)
Two hundred and thirty-three (61%) of 385 surgeons offered a
twenty-four-year-old patient nonoperative treatment. Core decompression was
offered by 139 (36%) of the 385 surgeons, with forty (10%) of the 385 offering
it with and ninety-nine (26%) of the 385 offering it without bone graft or
bone-graft substitute. Four (1%) of the 385 surgeons offered vascularized
bone-grafting.
For the same scenario in a forty-eight-year-old, 249 (65%) of 388 surgeons
offered nonoperative treatment. Core decompression was offered by 129 (33%) of
the 388 respondents, with thirty-five (9%) of the 388 offering it with and
ninety-four (24%) of the 388 offering it without bone graft or bone-graft
substitute.
Moderate Lucent and Sclerotic Changes in the Femoral Head (15% to 30%
of Head Affected; Head Round without Collapse) and Associated Symptoms
(Steinberg29
Stage IIB)
Two hundred and sixty-five (69%) of 385 surgeons offered a
twenty-four-year-old patient core decompression; 115 (30%) of the 385 offered
it with and 150 (39%) of the 385 offered it without bone graft or bone-graft
substitute (Fig. 1).
Thirty-nine (10%) of the 385 surgeons offered vascularized bone-grafting
(Fig. 2), and thirty-six (9%)
of the 385 suggested nonoperative treatment. Twenty-three (6%) of the 385
surgeons offered total hip arthroplasty, with eighteen (5%) of the 385
offering traditional total hip arthroplasty and five (1%) of the 385 offering
resurfacing total hip arthroplasty when available
(Fig. 3).
For the same scenario in a forty-eight-year-old, 262 (68%) of 386 surgeons
offered core decompression, with 111 (29%) of the 386 offering it with and 151
(39%) of the 386 offering it without bone graft or bone-graft substitute
(Fig. 1). Twenty (5%) of the
386 surgeons offered vascularized bone-grafting
(Fig. 2), and thirty-nine (10%)
of the 386 suggested nonoperative treatment. Fifty-one (13%) of the 386
surgeons offered total hip arthroplasty, with forty-five (12%) of the 386
offering traditional and six (2%) of the 386 offering resurfacing total hip
arthroplasty when available (Fig.
3).
Moderate Lucent and Sclerotic Changes in the Femoral Head (15% to 30%
of Head Affected; Head Round without Collapse) and No Associated Symptoms
(Steinberg29
Stage IIB)
One hundred and eighty-three (47%) of 386 surgeons offered a
twenty-four-year-old patient core decompression; seventy-eight (20%) of the
386 offered it with and 105 (27%) of the 386 offered it without bone graft or
bone-graft substitute. Twenty-nine (8%) of the 386 surgeons offered
vascularized bone-grafting, and three (1%) of the 386 offered nonvascularized
grafting. One hundred and sixty-four (42%) of the 386 surgeons suggested
nonoperative treatment, and 1% of the surgeons offered arthroplasty.
For the same scenario in a forty-eight-year-old, 164 (42%) of 387 surgeons
offered core decompression, with sixty-three (16%) of the 387 offering it with
and 101 (26%) of the 387 offering it without bone graft or bone-graft
substitute. Fourteen (4%) of the 387 surgeons offered vascularized
bone-grafting, and three (1%) of the 387 offered nonvascularized grafting. One
hundred and ninety-six (51%) of the 387 surgeons suggested nonoperative
treatment, and 1% offered arthroplasty.
Moderate Subchondral Collapse (Crescent Sign) without Flattening (15%
to 30% of Weight-Bearing Surface) and Associated Symptoms
(Steinberg29
Stage IIIB)
Surgeons most often offered a twenty-four-year-old patient arthroplasty
surgery. One hundred and thirty-one (34%) of 386 surgeons offered total hip
arthroplasty, with 110 (28%) of the 386 offering traditional and twenty-one
(5%) of the 386 offering resurfacing total hip arthroplasty when available.
Forty-one (11%) of the 386 surgeons offered hemiarthroplasty, with thirty (8%)
of the 386 offering resurfacing and eleven (3%) of the 386 offering
traditional hemiarthroplasty (Fig.
3). Seventy-seven (20%) of the 386 surgeons offered core
decompression, with forty (10%) of the 386 offering it with and thirty-seven
(10%) of the 386 offering it without bone graft or bone-graft substitute
(Fig. 1). Fifty (13%) of the
386 surgeons offered vascularized bone-grafting
(Fig. 2). Twenty-two (6%) of
the 386 surgeons offered an osteotomy (intertrochanteric or rotational)
(Fig. 4). Forty-six (12%) of
the 386 surgeons did not offer surgery.
For the same scenario in a forty-eight-year-old, 227 (59%) of 388 surgeons
offered total hip replacement, with 216 (56%) of the 388 offering traditional
and eleven (3%) of the 388 offering resurfacing total hip arthroplasty when
available. Twenty-seven (7%) of the 388 surgeons offered hemiarthroplasty,
with sixteen (4%) of the 388 offering resurfacing and eleven (3%) of the 388
offering traditional hemiarthroplasty (Fig.
3). Fifty-eight (15%) of the 388 surgeons offered core
decompression, with thirty-one (8%) of the 388 offering it with and
twenty-seven (7%) of the 388 offering it without bone graft or bone-graft
substitute (Fig. 1). Fifteen
(4%) of the 388 surgeons offered vascularized bone-grafting
(Fig. 2), and six (2%) of the
388 surgeons offered an osteotomy (intertrochanteric or rotational)
(Fig. 4). Forty-seven (12%) of
the 388 surgeons did not offer surgery.
Moderate Subchondral Collapse (Crescent Sign) without Flattening (15%
to 30% of Weight-Bearing Surface) and No Associated Symptoms
(Steinberg29
Stage IIIB)
One hundred and eighty-one (47%) of 384 surgeons offered a
twenty-four-year-old patient nonoperative treatment. The type of nonoperative
treatment was not specified. One hundred (26%) of the 384 surgeons offered
core decompression; fifty-seven (15%) of the 384 offered it with and
forty-three (11%) of the 384 offered it without bone graft or bone-graft
substitute. Thirty-nine (10%) of the 384 surgeons offered vascularized
bone-grafting. Forty-two (11%) of the 384 surgeons offered some type of
arthroplasty.
For the same scenario in a forty-eight-year-old, 206 (53%) of 386 surgeons
offered nonoperative treatment. Ninety-three (24%) of the 386 surgeons offered
core decompression, with forty-four (11%) of the 386 offering it with and
forty-nine (13%) of the 386 offering it without bone graft or bone-graft
substitute. Ten (3%) of the 386 surgeons offered vascularized bone-grafting.
Sixty-three (16%) of the 386 surgeons offered some type of arthroplasty.
Moderate Flattening of the Femoral Head (15% to 30% of Weight-Bearing
Surface or 2 to 4-mm Depression) and Symptoms, Failed Nonoperative Treatment
(Steinberg29
Stage IVB)
Surgeons most often offered a twenty-four-year-old patient arthroplasty
surgery. Two hundred and fifteen (56%) of 386 surgeons offered total hip
arthroplasty, with 181 (47%) of the 386 offering traditional and thirty-four
(9%) of the 386 offering resurfacing total hip arthroplasty when available.
Fifty-nine (15%) of the 386 surgeons offered hemiarthroplasty, with forty-nine
(13%) of the 386 offering resurfacing and ten (3%) of the 386 offering
traditional hemiarthroplasty (Fig.
3). Thirty-two (8%) of the 386 surgeons offered an osteotomy
(intertrochanteric or rotational) (Fig.
4), and twenty-one (5%) of the 386 offered vascularized
bone-grafting (Fig. 2).
Thirty-four (9%) of the 386 surgeons did not offer surgery.
For the same scenario in a forty-eight-year-old, 331 (85%) of 391 surgeons
offered total hip replacement, with 321 (82%) of the 391 offering traditional
and ten (3%) of the 391 offering resurfacing total hip replacement when
available. Twenty-six (7%) of the 391 surgeons offered hemiarthroplasty, with
eighteen (5%) of the 391 offering resurfacing and eight (2%) of the 391
offering traditional hemiarthroplasty (Fig.
3). Thirteen (3%) of the 391 surgeons did not offer any
surgery.
Joint Narrowing and/or Acetabular Changes and Symptoms, Failed
Nonoperative Treatment
(Steinberg29
Stage V)
Surgeons most often offered a twenty-four-year-old patient arthroplasty
surgery. Two hundred and seventy-two (71%) of 385 surgeons offered total hip
arthroplasty, with 233 (61%) of the 385 offering traditional and thirty-nine
(10%) of the 385 offering resurfacing total hip arthroplasty when available.
Thirty-one (8%) of the 385 surgeons offered hemiarthroplasty, with twenty-six
(7%) of the 385 offering resurfacing and five (1%) of the 385 offering
traditional hemiarthroplasty (Fig.
3). Eighteen (5%) of the 385 surgeons offered an osteotomy
(intertrochanteric or rotational) (Fig.
4). Forty (10%) of the 385 surgeons did not offer surgery.
For the same scenario in a forty-eight-year-old, 368 (94%) of 390 surgeons
offered total hip arthroplasty, with 356 (91%) of the 390 offering traditional
and twelve (3%) of the 390 offering resurfacing total hip arthroplasty. Two
(1%) of the 390 surgeons offered hemiarthroplasty
(Fig. 3), and ten (3%) of the
390 surgeons did not offer surgery.
Definite Arthritis and Symptoms, Failed Nonoperative Treatment
(Steinberg29
Stage VI)
Surgeons most often offered a twenty-four-year-old patient arthroplasty
surgery. Three hundred and eighteen (82%) of 387 surgeons offered total hip
arthroplasty, with 277 (72%) of the 387 offering traditional and forty-one
(11%) of the 387 offering resurfacing total hip arthroplasty when available.
Twenty-eight (7%) of the 387 surgeons offered hemiarthroplasty, with
twenty-three (6%) of the 387 offering resurfacing and five (1%) of the 387
offering traditional hemiarthroplasty (Fig.
3). Eighteen (5%) of the 387 surgeons did not offer surgery.
For the same scenario in a forty-eight-year-old, 385 (98%) of 392 surgeons
offered total hip replacement; 377 (96%) of the 392 offered traditional and
eight (2%) of the 392 offered resurfacing total hip replacement. Three (1%) of
the 392 surgeons offered hemiarthroplasty
(Fig. 3). More than 99% (390 of
392) offered some surgical intervention.
Relationship Between Type of Surgical Intervention and Surgeon
Demographics
With the numbers reported, the type of practice setting, practice size, and
volume of total hip and knee arthroplasties that the surgeon performed in 2004
were not significantly related to the type of surgery offered by the surgeon.
Surgeons who had been in practice for between eleven and twenty years were
more likely than surgeons who had been in practice for fewer than eleven or
more than twenty years to offer core decompression surgery (p < 0.05).
Relationship Between Surgery Offered and Presence of Patient
Symptoms
Surgeons were more likely to offer surgery for each type of early-stage
disease (IB, IIB, and IIIB), for both the twenty-four-year-old and the
forty-eight-year-old hypothetical patient, if symptoms were present (p <
0.05).
Relationship Between Surgery Offered and Patient Age
Surgeons were more likely to offer resurfacing hemiarthroplasty,
vascularized bone-grafting, osteotomy, and "other" options,
including a hip arthrodesis, to the twenty-four-year-old hypothetical patient
than they were to the forty-eight-year-old hypothetical patient (p < 0.05).
In contrast, they were more likely to offer total hip arthroplasty to the
forty-eight-year-old patient (p = 0.008).
Type of Arthroplasty Offered
The type of arthroplasty offered by the surgeons differed according to the
stage of the disease and the age of the patient
(Fig. 3). When the specific
arthroplasty choices were combined for all stages of symptomatic osteonecrosis
of the femoral head, it was found that surgeons offered a resurfacing total
hip arthroplasty or a resurfacing hemiarthroplasty 22.5% of the time to the
twenty-four-year-old patient and 7.1% of the time to the forty-eight-year-old
patient. Hemiarthroplasty (either traditional or resurfacing) was offered
13.4% of the time to the twenty-four-year-old patient and 4.5% of the time to
the forty-eight-year-old patient. Traditional total hip replacement was the
arthroplasty type offered most frequently for each stage of the disease, with
four (1%) of 384 surgeons offering it for stage-I disease and 277 (71.6%) of
387 surgeons offering it for stage-VI disease in the twenty-eight-year-old and
seven (1.8%) of 388 surgeons offering it for symptomatic stage-I disease and
377 (96.2%) of 392 surgeons offering it for stage-VI disease in the
forty-eight-year-old.
There is currently no consensus regarding the treatment of the various
stages of osteonecrosis of the femoral head in
adults4,5,15,19,34,35.
We surveyed the active members of the AAHKS regarding their preferences for
the evaluation and treatment of this disease.
When evaluating a new patient with osteonecrosis of the femoral head, the
respondents universally inquired about alcohol and steroid usage and almost
all asked about a history of injury of the affected hip. More than one-fourth
of the surgeons obtained various blood tests, a consultation with an internal
medicine specialist, or diagnostic imaging if there was no obvious cause of
the osteonecrosis of the femoral head (idiopathic osteonecrosis of the femoral
head). Because many of these tests provide no actionable information, we think
that new studies are required to clarify which tests should or should not be
ordered to make treatment of osteonecrosis of the femoral head more
cost-effective.
The fact that 82% of the surgeons ordered screening of the unaffected hip
with magnetic resonance imaging strongly suggests that they recognize that
osteonecrosis of the femoral head is often
bilateral18,36
even though patients frequently present with unilateral symptoms. If alcohol
abuse was associated with the diagnosis of osteonecrosis of the femoral head,
one-fourth of the surgeons required the patient to stop drinking prior to
surgery and 35% obtained professional intervention to treat the alcohol abuse.
If the osteonecrosis of the femoral head was associated with steroid use, 51%
of the surgeons urged their patients to stop taking steroids and 28% asked for
an osteoporosis workup.
Nonoperative management was most commonly selected for both the young and
the middle-aged patient with asymptomatic disease (Steinberg stage IB, IIB,
and IIIB). The only exception was Steinberg stage-IIB disease in the
hypothetical twenty-four-year-old patient, for whom the respondents chose core
decompression more frequently (47% of the time) than they chose nonoperative
treatment (43% of the time). This treatment pattern is consistent with the
finding that hips without radiographic evidence of involvement (stage I)
rarely become painful or show radiographic
progression18
whereas those with collapse (stage III) are likely to have disease progression
despite any non-arthroplasty
intervention17. On
the other hand, emerging trends suggest that operative procedures to retard or
prevent progression prior to collapse should be considered for both
symptomatic and asymptomatic hips with radiographic findings and moderate head
involvement (stage
IIB)17,20.
This is reflected in the responses regarding the patients with asymptomatic
stage-IIB disease—56% of the surgeons stated that they would offer core
decompression or bone-grafting to the twenty-four-year-old and 47% said that
they would offer these interventions to the forty-eight-year-old.
It is of interest that few respondents offered
statins11,
anticoagulants6, or
bisphosphonates33
for possible prevention of progression of, or for treatment of, osteonecrosis
of the femoral head once an at-risk patient was identified. We surmised that
surgeons may not have embraced these interventions because of the possibility
of
complications37-39
or the paucity of clinical reports regarding their efficacy. Pharmacological
management of patients for whom nonoperative treatment is chosen is an
important area for further research and education.
Most (83%) of the respondents considered the stage of the osteonecrosis of
the femoral head to be the most critical determinant of surgical treatment and
offered surgery primarily on the basis of the stage of the disease. A lack of
symptoms in a patient with
Steinberg29
stage-IIB disease deterred many from offering surgery (43% and 51% offered
nonoperative treatment to the asymptomatic twenty-four-year-old and
forty-eight-year-old patients, respectively), but this trend is not fully
supported in the
literature17,20,34.
For example, Mont et
al.17 argued that
asymptomatic lesions of moderate size should be treated with core
decompression because many, if not most, will progress to symptomatic
lesions.
Current studies have provided evidence supporting the use of core
decompression only for early-stage osteonecrosis of the femoral
head15,19,40,
and the questionnaire responses reflect knowledge of those data. Core
decompression was the most commonly offered surgical treatment for symptomatic
early disease (stages IB and IIB) in both the twenty-four and the
forty-eight-year-old hypothetical patient, and it was offered by approximately
70% of the surgeons in each case.
Arthroplasty was commonly offered for late-stage osteonecrosis of the
femoral head, perhaps more so than is warranted by the long-term results of
arthroplasty for osteonecrosis of the femoral head reported in the
literature41-45.
We hypothesized that the surgeons surveyed were more confident about
arthroplasty being the best surgical choice for many young patients with
symptomatic osteonecrosis of the femoral head because of reports of a lower
rate of aseptic loosening with more contemporary arthroplasty
techniques46-49
and because of the hope that alternative bearing surfaces will lead to fewer
revisions and less
osteolysis50. It is
likely that bias was also a factor in the surgeons' choices, as the
orthopaedists surveyed in this study were arthroplasty surgeons (AAHKS members
devote =50% of their practice to hip and knee replacement surgery) and may
therefore have been more likely to offer joint replacement. Although we do not
have data on the volume of other procedures that AAHKS members perform (e.g.,
vascularized bone grafts and osteotomies), this bias must be taken into
account because it may have minimized the importance of non-arthroplasty
techniques.
Total hip replacement was more frequently offered to the older
(forty-eight-year-old) patients than to the younger (twenty-four-year-old)
patients (p = 0.008), and it was also offered more often to the patients with
more advanced arthritis in both age groups. The percentages of respondents who
offered resurfacing (22.5% for the twenty-four-year-old and 7.1% for the
forty-eight-year-old) and/or hemiarthroplasty (13.4% for the
twenty-four-year-old and 4.5% for the forty-eight-year-old) when arthroplasty
was indicated reflects the ongoing controversy about the indications for, and
results of, these
procedures17,51-56.
For example, the results of contemporary resurfacing arthroplasty for
osteonecrosis of the femoral head suggest possible problems with failures on
the femoral
side51,57.
It is also interesting to note that >10% of surgeons responded that they
would offer "resurfacing total hip replacement (when it is
available)" to a twenty-four-year-old with
Steinberg29
stage-VI osteonecrosis of the femoral head, although this prosthesis was not
approved by the Food and Drug Administration at the time that the
questionnaire responses were solicited.
In most instances, surgeon demographics did not significantly correlate
with the type of surgery offered. This somewhat surprising finding may be
explained by the similar education of AAHKS members (many have completed an
adult reconstruction fellowship, and they have a high rate of participation at
the annual meeting). We cannot conclude whether there might be a relationship
for general orthopaedic surgeons with a lower average annual arthroplasty
volume.
This study had several limitations. Opinions relating to patient care were
solicited, and no effort was made to collect data from actual patient
encounters. Hypothetical scenarios were described that reflected
representative examples, but all possible clinical scenarios were not
examined. However, openended response opportunities were offered for each
question and appropriate questionnaire methodology was utilized.
We believe that this study provides important information regarding the
treatment preferences of a select group of arthroplasty surgeons in the United
States. The variability in the practice patterns speaks to the need for better
research and education on this subject. Recommendations regarding the
treatment of osteonecrosis will likely evolve with a continued concerted
effort on the part of interested surgeons and will require not only
randomized, controlled, prospective studies of patients with each stage of the
disease but also the combination of data from multiple centers.