APPENDIX B

Table B.1: Utilityvalues for osteoporosis without fracture

Study and country / Condition / Health state value / How valued / Who valued
Dhillon et al. (2005) / History of fracture (n =117), includes those who do / do not meet osteoporosis T-score / Mean (95% CI) / EQ-VAS EQ-5D UK tariff / 325 male and female patients from 18 general practices in Edinburgh
(secondary osteoporosis, not postmenopausal) / 25-34 / VAS: 79 (62,95), EQ-5D 0.86 (0.58,1.13) / Inc: Patients being at risk of osteoporosis with a referral from the general practitioner to DXA scanning
35-44 / VAS: 60 (41,78), EQ-5D 0.16 (n = 1)
45-54 / VAS: 71 (63,79), EQ-5D 0.58 (0.40, 0.76)
55-64 / VAS: 76 (70,82), EQ-5D 0.73 (0.64, 0.82)
Scotland, Edinburgh / 65-74
75+ / VAS: 69 (63,75), EQ-5D 0.68 (0.59, 0.77) / Exc: patients were excluded who did not have hip BMD measurements (due to hip replacements); patients who had a previous DXA scan or prior specialist consultation regarding osteoporosis.
VAS: 65 (57,74), EQ-5D 0.68 (0.61, 0.75)
Participants from a randomized, controlled, prospective trial to compare the impact of direct access DXA scanning on GP’s management decisions with hospital referral for specialist consultation. / Osteoporosis T-score <= -2.5, n=159 / The QoL questionnaire was filled in before DXA scan.
25-34 (n=1) / VAS: 39 (n=1) EQ-5D 0.06 (n=1)
35-44 (n=0) / VAS: (n=0) EQ-5D n= 0 / From the scale 159 patients were identified as having osteoporosis Mean age 65 (9). Of which 153 (96%) were female and 79 (50%) had prior history of low trauma fracture.
45-54 (n=21) / VAS: 69 (60,78) EQ-5D 0.64 (0.47, 0.8)
55-64 (n=50) / VAS: 71 (65,77) EQ-5D 0.67 (0.59,0.75)
65-74 (n=68) / VAS: 68 (64,73) EQ-5D 0.65 (0.59-0.71)
75+ (n=19) / VAS: 60 (50-69) EQ-5D 0.62 (0.51-0.73)
After adjusting for age, EQ-5D and EQ-VAS were significantly lower in patients with osteoporosis than those without, but the differences lessened with advancing age / Previous fracture group (n=117): Mean age 64 (1)
Age adjusted osteoporosis (n=159) / VAS: 68 (20), EQ-5D 0.65 (0.28) / No osteoporosis group (n=166): Mean age 55 (11).
Patients without osteoporosis (n=166) / VAS: 76 (16), EQ-5D: 0.76 (0.27)
After adjusting for age those with a history of fracture did not have a significantly lower EQ-5D or VAS than those without,
With history of fracture / VAS: 64 (21), EQ-5D 0.61(0.3)
Without a history of fracture / VAS: 69 (20), EQ-5D 0.66 (0.27)
Sawka et al. (2005) /

Self reported osteoporosis

/ 0.69 (0.27). / HUI-III / Randomly recruited 50 km radius around nine study centres
n = 421 / After adjusting for age and gender the mean difference was -0.08 (-0.11, -0.06) for osteoporosis. / Of the initial CaMos sample of 9,423 people (>25 years) examined data on those >= 65, who completed baseline HUI-III and provided information on medical history.
Canada (CaMos)
Participants in the Canadian Multicentre Osteoporosis Study (CaMos) / Initially
9,423 ≥ 25 years out of which
4,550 ≥ 65 years
4,495 completed the baseline HUI3
Of these 3,750 gave information on presence or absence of diseases of interest (2,716 women)
Osteoporosis group n = 421
98% response to questions pertaining to presence of respective medical conditions in the osteoporosis group
Mean age for full sample (n=4495), 73 years.
Utility values for established osteoporosis and vertebral deformities
Study and country / Condition / Health state value / How valued / Who valued
Ariza-Ariza et al (2004) / Primary osteoporosis (postmenopausal or senile) defined by bone mineral density in the lumbar column or femoral neck with a T-score of < -2.5 (av. -3.42 (0.9)). / 0.49 (sd 0.19). Median .50 (Range 0.14-1). / EQ-VAS (scored from best (0) to worst (10)) / 45 patients primary osteoporosis (postmenopausal or senile) patients
Spain.
Cross-section study designed to validate the OPTQol in Spanish. / Study uses the EuroQol health profile, but not the preference weights for scoring. / (43 women)Mean age 66.3 (6.8)
20 patients had vertebral fractures.
Badia et al (2001)
Spain.
Cross-section study to compare OQLQ and QALEFO / Stable osteoporosis (n=338) / 0.567 (0.192) / EQ-VAS
EQ-5D was used but only the levels are given in the paper with no preference weighted scores. / 400 women recruited at outpatient department of 12 Spanish hospitals. Of which 62 were excluded.
Inc: Clinically stable osteoporosis confirmed by DXA. Age >= 50.
Exc: Clinical deformity in previous month. Condition other than osteoporosis.
Mean age 68.8 (7.6)
Gabriel et al. (1999)
USA / Fracture subjects with osteoporosis (those with vertebral fracture and self-acknowledged osteoporosis)
Own health (n=75)
Non-fracture subjects. Osteoporosis vignette. (n=199).
Health state scenario constructed from clinician views and focus groups, includes reference to future risk of fractures (e.g. you are at least 5 times more likely to fracture a bone than a women your age) / 0.84 (0.29) Median 1 (IQR 0.89-1)
0.43 (0.40) Median 0.48 (IQR 0-0.83) / TTO. For fracture group, anchored by death and best imaginable for age.
For non-fracture group, anchored by death and current health then rescaled based on the TTO valuation of own health to death and best imaginable health. / Women at two clinics, Minnesota and New Hampshire, invited to participate in a study to understand how people feel about osteoporosis.
Inc: women ≥ 50 years
Group 1: Women free of osteoporotic fractures (absence of documentation corroborated with self-report of no hip or vertebral fracture at any time and no distal forearm fracture within 3 years)
Group 2: Women with documented first vertebral fracture within 5 years, most recent from minimal or moderate trauma, but no additional non-vertebral fractures within the last 3 years, and no hip fracture ever.
Group 3: Women with documented first hip fracture within 5 years, who were at least 1 year post-fracture. No distal forearm fracture within the preceding 3 years, never have had a vertebral fracture.
Fracture subjects n= 183. Mean age of all fracture subjects is 76 (sd 1).
Non fracture subjects (n=199) were women who were recent outpatient clinic attendees with no fracture in the last 2 years, mean age 68 (sd 1) (for the sample of n=199).
Jakob et al. (2006)
European (France, Germany, Greece, Portugal, Spain and UK).
Note this is the OSSO trial the same as Cooper et al 2007. / Any osteoporotic fractures
Overall (n= 2314)
Group 1 had experienced a fragility fracture despite drug therapy for at least 12 months (n=1309)
Group 2 discontinued osteoporosis drug therapy due to compliance problems and/or side effects. (n=1005) / VAS: 0.54 ± 0.193, EQ-5D 0.50 ± 0.33
VAS: 0.521 ± 0.192, EQ-5D 0.47 ± 0.34
VAS: 0.563 ± 0.191, EQ-5D 0.53 ± 0.32 / EQ-5D and EQ-VAS / Patients. Postmenopausal women diagnosed with osteoporosis based on axial and peripheral DXA bone mineral density (BMD) measurements, and confirmed by physician review of medical reports or radiographs.
Mean age full sample 70.2 (9.0)
Group 1: Mean age 71.2 (8.9)
Group 2: Mean age 69.0 (9.0)
Reported index fracture after 12 months drug therapy: median time since fracture 5 months (range 0-206 months);
Oleksik et al (2000)
Belgium, UK, Netherlands and Sweden
As part of the baseline assessment for the Multiple Outcomes of Raloxifene Evaluation study (MORE). / Vertebral fracture (VFX) identified by radiograph
no vertebral fracture n= 293
1 VFX n=130
2 VFX n= 69
3 VFX n= 36
≥ 4 VFX n= 60
Thoracic fracture only (T) (n=145)
Lumbar fracture only (L) (n=42) / Mean (SD)
0.822 (.205)
0.747 (.231)
0.739 (.249)
0.806 (.181)
0.657 (.298)
0.775 (.195)
0.678 (.343)
L versus T is a significant difference (p 0.033) / EQ-5D
UK tariff / N=751 healthy, ambulant women from European countries
n=302 no VFX
n=449 >= 1 VFX
Inc: <= 80; 2 years postmenopausal; osteoporotic, low lumbar or femoral neck BMD (T score ≤ -2,5) with and without vertebral fractures
Exc: women with a history of metabolic bone disease, malignancy, or recent antiosteoporotic treatment (with exception of calcium and vitamin D supplements and/or hormonal replacement up to 6 months before randomisation). Spinal deformities that impaired radiographic morphometry also were an exclusion criterion (e.g. scoliosis > 15°)
Postmenopausal women with osteoporosis according to WHO definition (BMD >= -2.5)
Of the full sample, mean age for non-fractures is 66.2±5.9 (n=302).
Mean age for fracture group is 68.8±6.3 (n=449) But this includes some without EQ-5D data.
Of the full VFX group 160 (35.6%) had other postmenopausal non-VFX fractures. Of the non-VFX group 75 (24.8%) had a postmenopausal non-VFX fracture.
Rajzbaum et al (2008)
8 European Countries (Germany, Greece, France, Netherlands, Sweden, Austria, Denmark, and Ireland) / Postmenopausal women entering teriparatide treatment / EQ-5D Mean 0.41 (0.38) Median 0.59 (IQR 0.08, 0.71).
EQ-VAS Mean 0.519 (0.219). Median 0.50 (IQR 0.35, 0.69) / EQ-5D UK Tariff
EQ-VAS / Baseline profile for EFOS study. Women with severe osteoporosis
Mean age 71.5 (8.4)
Mean lumbar spine BMD -3.3 (1.18)
Mean number of previous fractures (after aged 40) 2.9 (2.0)
Mean time since most recent fracture 2.4 (sd 3.5) years
van Schoor et al (2004)
Netherlands
Based on data from the Longitudinal Aging study Amsterdam (LASA), a 12 year cohort study. Current study is all persons who participated in the second follow-up cycle of LASA (1995-6). / Vertebral deformities (VFX)
0 (n=152)
1 (n=98)
2 (n=49)
≥ 3 (n=37)
Severity (most severe VFX)
Grade 0 no deformity, < 20% reduction in anterior, middle and/or posterior height (n=152)
Grade 1 mild deformity , 20-25% reduction (n=108)
Grade 2 moderate deformity, 25-30% reduction (n=44)
Grade 3 severe deformity, > 30% reduction (n=32)
Combined score
Mild /no osteoporosis (n=264)
Severe Osteoporosis (sever deformity, or 3 or more deformities, or 2 deformities of which at least one was moderate deformity) (n=72) / Median (IQ)
0.80 (0.69-1.0)
0.80 (0.73-1.0)
0.85 (0.69-1.0)
0.73 (0.69-1.0)
0.80 (0.69-1.0)
0.80 (0.69-1.0)
0.80 (0.72-1.0)
0.80 (0.69-0.85)
Median (IQ), Mean
0.80 (0.69-1.0), 0.80
0.80 (0.69-0.86), 0.76 / EQ-5D, UK tariff / People born in or before 1930 (being >= 65 years in January 1996), living in Amsterdam or its surroundings (n=698) invited to hospital for a radiograph of the spine.
539 were willing to attend (response rate of 77%)
Of which 527 had adequate radiographs.
Of these 69 deceased between 1995-6 and 1998-9, and 126 refused, were not eligible or could not be contacted for the second radiograph in 1998-9.
This left 344 willing to come to hospital in 1998-9, of which 338 had adequate radiographs.
Complete information was available on 336 participants, of which 152 (45.2%) had no VFX.
Median age of full sample (n= 336) 75.8 (IQ 71.9-82.0)
≥65 (1996)
≥71 (2001)
EQ-5D administered 3 years after the radiograph, hence number of VFX may have changed during that time.
Yoh et al. (2005)
Japan / Vertebral fractures
0 vertebral fracture (n=19)
1 vertebral fracture (n=15)
≥ 2 vertebral fracture (n=24) / 0.687 ±0.174
0.75±0.15
0.71±0.20
0.63±0.16 / EQ-5D, Japanese tariff / 58 female patients visiting outpatient clinic of the Department of Orthopedic Surgery, Hyogo College of Medicine, Japan, with chronic back pain.
Inc: chronic back pain ≥ 3 months; indicated vertebral fractures
Exc: Patients with secondary cause for osteoporosis; patients with acute pain (duration < 3 months)
First part of the study evaluates QoL in osteoporotic patients, and factors responsible for impaired QoL
BMD mean 0.680 (0.147)
Mean age 73.07±8.35
Utility values for clinical vertebral fracture over a specified time period
Study / Condition / Health state value / How valued / Who valued.
Cranney et al (2001)
Canada
Cranney et al. (2001) / HRT only (n=11).
Baseline scores for own health
Follow up (2 months) (n=11)
Recent vertebral fracture n= 10.
Baseline scores for current health (n=10)
Follow up (2 months) scores for current health
Valuation of vignette, VAS
Vertebral fracture.
State described according to 6 dimensions:
activities of daily living, self care functions, anxiety and depression, leisure activities, pain and side effects from treatment. / VAS: 0.92 (0.08)
SG: 0.90 (0.11)
HUI2: 0.80 (0.10)
VAS: 0.88 (0.12)
SG: 0.93 (0.07)
HUI2: 0.82 (0.07)
VAS: 0.76 (0.13) 0.50, 0.95
SG: 0.84 (0.20) 0.50, 1.0
HUI-II: 0.79 (0.92) CHECK 0.67,0.12
VAS: 0.83 (0.08) 0.73, 0.97
SG: 0.91 (0.10) 0.70, 1.0
HUI-II: 0.76 (0.14) 0.43, 0.92
Baseline, Follow up (VAS)
Group 1: 0.61 (0.35-0.8); 0.57 (0.27-0.85)
Group2: 0.50 (0.25-0.65); 0.54 (0.27-0.70)
Group 3: 0.56 (0.40-0.85); 0.61 (0.50-0.75)
Group 4: 0.50 (0.25-0.70); 0.54 (0.40-0.70) / VAS anchored at dead and perfect health.
SG (dead-perfect health)
HUI2.
VAS anchored at dead and perfect health.
SG (dead-perfect health)
HUI2. / Women who are osteoporoic and on HRT for osteoporosis prevention.
Median age 56.0 (range 45-69)
Patients own health state
Women (aged >= 50) identified through fracture clinics and the orthopaedic and geriatric wards.
Inc: Group 1: women who had not experienced a fracture but were osteopenic and undergoing HRT for osteoporosis prevention. Identified through the menopause and rheumatology clinics. Group 2-4: women with recent wrist, spine, or hip fracture. 10-15 patients per subgroup were interviewed. All fractures were confirmed by radiograph.
Exc: Unable/ unwilling to give consent; not fluent in English; severely depressed; deaf or visually impaired; had a pathological fracture.
Group 1: HRT (13 eligible, 11 consented)
Group 2: Wrist fracture (19 eligible, 11 consented)
Group 3: Vertebral fracture (15 eligible, 10 consented)
Group 4: Hip fracture patients (15 eligible, 10 consented)
2 individuals lost to follow-up
Group 3: Median age 75.5 (range 65-88)
Borgstrom et al. (2006)
Sweden
Prospective, observational study.
The costs and effects of osteoporosis-related fractures study (KOFOR study) (ongoing). / Vertebral fracture (n= 81)
Before fracture
After fracture (mean 6.56 days (sd 6.8)
4 months after fracture
12 months after fracture
Average annual loss of qol
Simple interpolation
Assuming the 4 month qol reached after 1 month
Population-based utility used as proxy for qol before fracture
Information on pre-fracture quality of life was collected from patients after the fracture. This could lead to recollection bias. It is therefore also compared to Swedish age-differentiated EQ-5D population norms. / Mean (95% CI)
0.73 (0.67-0.78)
0.18 (0.14-0.25)
0.47 (0.40-0.55)
0.49 (0.42-0.56)
0.30 (0.25-0.36)
0.26 (0.2-0.32)
0.34 (0.28-0.39) / EQ-5D UK tariff / Fracture patients at 7 hospitals in Sweden.
Inc: patient diagnosed with a fracture caused by low energy trauma; at least 50 years old; vertebral fractures have to be confirmed by an X-ray examination; included and interviewed within 4 weeks of the fracture event
Exc: multiple fractures; fractures cause by another disease (e.g. cancer); judged not able to complete the questionnaires due to dementia or psychological problems.
81 vertebral fractures.
Mean age 75.0 (n=81).
Cockerill et al (2004)
CzechRepublic, Hungary, Italy, Slovakia, Spain, Sweden, UK
Subjects recruited from participants of screening survey for vertebral osteoporosis – European Vertebral Osteoporosis Study (EVOS). / Cases (n=73)
Controls without baseline prevelant deformity (n=136)
Controls with baseline prevelant deformity (n=60)
Cases without baseline deformity (n=50)
Matched controls without baseline deformity (n=92)
Cases with baseline deformity (n=23)
Matched controls with baseline defomrity (n=20)
Median time from second spinal radiograph until qol survey was 1.9 years.
Adverse impact most marked among those who had a preexisting vertebral deformity, suggesting that the effect on quality of life of multiple fractures is cumulative. / EQ-5D 0.767 (.193), EQ-VAS .623 (.206)
EQ-5D 0.825 (.167), EQ-VAS .699 (.215)
EQ-5D 0.801 (.187), EQ-VAS .663 (.202)
EQ-5D 0.795 (.204), EQ-VAS .663 (.179)
EQ-5D 0.805 (.169), EQ-VAS .685 (.212)
EQ-5D 0.707 (.156), EQ-VAS .534 (.239)
EQ-5D 0.806 (.172), EQ-VAS .717 (.201) / EQ-5D, UK tariff
EQ-VAS / Men and women from recruited from 12 European centres. 3,205 had radiographs performed at baseline and at second invite (mean of 3.8 years later).
121 subjects with morphometric or qualitative incident fracture, of which QoL surveys (conducted a median of 1.9 years after second radiograph) available for 73.
Fracture group: mean age 64.8 years, 57 were women, 32% had evidence of prevalent vertebral deformity.
196 (151 women) matched controls (matched on age, gender, centre, 1 with prevalent deformity at baseline and 2 without) mean age 63.9 years. 31% evidence of prevalent vertebral deformity.
Cooper et al (2007)
France, Germany, Greece, Portugal, Spain and UK.
Observational Study of Severe Osteoporosis (OSSO) / Fragility fracture
Index Fracture Group (EQ-VAS)
- Incident Fracture (n = 100)
- no incident Fracture (n = 888)
Compliance/side effect Group
-incident fracture (n = 66)
-no incident fracture (n=831)
Index Fracture Group (EQ-5D)
- Incident Fracture (n = 100)
- no incident Fracture (n = 888)
Compliance/side effect Group
- incident fracture (n = 66)
- no incident fracture (n=831)
Prior Fracture < 12m (EQ-VAS)
- incident fracture (n = 107)
- no incident fracture (n = 838)
No prior Fracture < 12m
- incident fracture (n = 55)
-no incident fracture (n=843)
Prior Fracture < 12m (EQ-5D)
- incident fracture (n = 107)
- no incident fracture (n = 830)
No prior Fracture < 12m
-incident fracture (n = 56)
-no incident fracture (n=830)
Total (n=1823) / Baseline 6months 12months
50.2 (20.1) 52.7 (19.2) 52.0 (16.6)
51.7 (19.0) 56.3 (18.3) 58.7 (18.9)
56.0 (20.6) 50.7 (20.2) 54.2 (21.2)
55.9 (19.3) 58.2 (18.6) 61.2 (18.8)
0.39 (0.37) 0.40 (0.36) 0.39 (0.38)
0.45 (0.34) 0.55 (0.30) 0.58 (0.30)
0.53 (0.34) 0.43 (0.40) 0.40 (0.37)
0.52 (0.32) 0.60 (0.27) 0.62 (0.29)
Baseline 6months 12months
Mean 47.5 (19.4) 49.8 (18.7) 51.5 (16.8)
Median 50.0 50.0 50.0
Mean 51.6 (19.5) 56.3 (18.6) 58.9 (19.0)
Median 50.0 55.0 60.0
Mean 62.1 (19.0) 56.2 (20.7) 55.6 (21.4)
Median 60.0 55.0 53.0
Mean 55.9 (18.1) 58.1 (18.3) 60.9 (18.8)
Median 55.0 59.0 61.0
Mean 0.37 (0.37) 0.39 (0.37) 0.38 (0.37)
Median 0.52 0.52 0.52
Mean 0.45 (0.34) 0.55 (0.30) 0.57 (0.30)
Median 0.59 0.62 0.66
Mean 0.58 (0.29) 0.46 (0.38) 0.43 (0.38)
Median 0.66 0.59 0.52
Mean 0.53 (0.32) 0.59 (0.28) 0.62 (0.28)
Median 0.66 0.69 0.69
Mean 0.48 (0.34) 0.56 (0.30) 0.58 (0.31)
Median 0.62 0.66 0.69 / EQ-VAS
EQ-5D UK Tariff / Osteoporotic postmenopausal women recruited in 469 outpatient sites in 6 European countries.
Inc: Women diagnosed with osteoporosis based DXA & BMD and confirmed by physician review of medical reports or radiographs. A well-documented inadequate clinical response to osteoporosis medications defined as:
i) a past history of fragility fracture sustained despite prescription of any approved osteoporosis treatment for at least 12 months prior to fracture (index fracture group) and / or
ii) past history of discontinuation of osteoporosis drug therapy due to compliance problems and/or side effects
Exc: patients currently being treated with investigational drug or procedure.
2322 met eligibility criteria at baseline of which 1885 with incident fracture status available (at mean follow up of 351 days). Of these :
N=166 incident fracture
N=1719 no incident fracture
Mean age of the 1,885 sample: 70.6 years (9.0)
Cooper et al (2007) / Incident vertebral fracutre (n= 88)
Baseline
12 months / 0.40 Median 0.55
0.31 Median 0.52 / EQ-5D UK tariff / See above.
Gabriel et al (1999) / Vertebral fracture n= 94
VAS (own health)
TTO (own health)
HUI2
QWB
Multiple vertebral fractures (scenario)
Fracture subjects (n=24)
Those who had experienced a stage like the scenario or worse
Non-fracture subjects (n=199) / 0.76 (0.17) Median 0.80
0.81 (0.32) Median 1
0.80 (0.16) Median 0.85
0.66 (0.09) Median 0.66
0.68 (0.40) Median 0.97
0.31 (0.38) Median 0 / HUI-2
QWB (estimated from SF-36 values)
Own health state VAS and TTO
TTO Vignette: Multiple vertebral fracture.
TTO for own health anchored by death and best imaginable for their age.
TTO for hypothetical states anchored on death and current health.
These values were rescaled to perfect health (for their age) as upper anchor using the subjects’ own current health TTO value. / 94 patients from 2 clinics in Minnesota and New Hampshire.
Inclusion: women ≥ 50 years
Group 1: Women free of osteoporotic fractures (absence of documentation corroborated with self-report of no hip or vertebral fracture at any time and no distal forearm fracture within 3 years)
Group 2: Women with documented first vertebral fracture within 5 years, most recent from minimal or moderate trauma, but no additional non-vertebral fractures within the last 3 years, and no hip fracture ever.
Group 3: Women with documented first hip fracture within 5 years, who were at least 1 year post-fracture (chronic rather than acute states). Not have had a distal forearm fracture within preceding 3 years, never have had a vertebral fracture.
Fracture subjects n= 183, non fracture n=199.
Mean age 76 for the full patient sample (n=183) Those with non-traumatic vertebral fracture in last five years and at least 1 year since fracture.
No forearm fracture within preceding 3 years and never had hip fracture.
Non-fracture subjects women who were recent outpatient clinic attendees with no fracture in the last 2 years, mean age 68 (for the sample of n=199).
Hall et al (1999) / Fracture group (n=100)
Non fracture group (n=100) / 0.64 (0.08)
0.72 (0.07) / QWB index predicted from the SF-36 (using method from Fryback et al., 1997) / Women recruited from hospital-based or private practice bone clinics after referral by their primary care physicians.
Fracture group: Reviewed 145 radiographs.34 were excluded due to significant degenerative spondylosis or spondylolisthesis.
11 declined to be interviewed.
Mean age 74.4 (7.2)
Mean time since last fracture 5.1 (4.8) years. None < 4 months.
Average number of fractures 2.9
Control group; 167 controls from the community radiographed. 61 rejected due to silent vertebral fracture, degenerative joint disease or spondylolisthesis.