Q J Med 2002; 95: 305-311
© 2002 Association of Physicians
The cost utility of bisphosphonate treatment in established osteoporosis
From the Department of Health Sciences & Centre for Health Economics, University of York, York, and 1 Aventis Pharma, West Malling, UK
Received 26 November 2001 and in revised form 18 February 2002
 |
Summary |
|---|
 Top Summary IntroductionMethodsResultsDiscussionReferences |
|---|
Background: Hip fracture is an important and costly problem. Bisphosphonate therapy prevents hip and other fractures among women with established osteoporosis, but there are few published economic evaluations of this treatment.
Aim: To assess the cost-effectiveness of risedronate, a recently launched bisphosphonate for the prevention of fractures among women with established osteoporosis.
Methods: A state transition Markov model of established post-menopausal osteoporosis based upon randomized clinical trial data was developed. Uncertainty underlying model parameters and outcomes was dealt with using traditional sensitivity analysis and stochastic sensitivity analysis to produce quasi-95%CIs. We focussed on patients aged 
75 years, since this population most closely matches the randomized controlled trial, and is typical of osteoporosis patients in the UK.
Results: The baseline model of treating a cohort of 1000 75-year-old women for 3 years with risedronate and then modelling the costs and benefits over their expected lifetimes, produced net savings of £786 000 for the treatment group per 1000 treated women, (95%CI £1.55m savings to £47000 extra costs). Restricting the horizon of the analysis to only three years led to a small net cost of £138 000 per 1000 treated women (95%CI £196 000 savings to £477 000 extra costs) with a net increment in Quality Adjusted Life years (QALYs) of 16 per 1000 treated women. This resulted in a cost per QALY of £8625 per treated woman.
Conclusions: In this example, the use of risedronate therapy in 75-year-old women at high risk of hip fracture leads to an improvement in quality of life with possible cost savings. Restricting the analysis to a time horizon of only three years leads to a QALY gain at a modest net cost.
|
Introduction |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Osteoporotic fracture is a significant cause of morbidity and cost.1 For women who have established disease, that is low bone mineral density (BMD), and a prevalent vertebral fracture the risk of further fracture, particularly hip fracture, is considerable.2,3 Indeed, women with a prevalent vertebral fracture have a 1:5 chance of sustaining a further vertebral fracture within 12 months.4 These fractures are associated with a significant reduction in quality of life, and are excellent predictors of fractures in later years.3–5 However, despite the importance of these fractures, a UK study of general practice patients suggested that only 40% of women with a diagnosis of vertebral fracture, in primary care, receive pharmaceutical treatment.6
While vertebral, hip and Colles fractures are generally considered to be ‘osteoporotic’, Cauley and colleagues have shown that, in fact, most non-skull fractures are associated with low bone mass (i.e. osteoporosis) rather than just trauma.7 The combined incidence of these other fractures exceed the incidence of the ‘classical’ osteoporotic fractures. For instance, in the major bisphosphonate trials (alendronate and risedronate) and in the MORE trial of raloxifene, the combined incidence of non-hip fractures was actually greater than that of hip fractures alone (i.e. between 7 to 13 times greater).8–11 Therefore, while preventing vertebral and hip fractures is an important goal of therapy, prevention of these other fractures is also important, as osteoporosis is a disease of the whole skeleton.
Although the evidence base is strong for a number of treatments for vertebral fracture prevention, it is somewhat weaker for preventing hip fractures. Only three therapies have been shown to reduce hip fracture in randomized trials: calcium and vitamin D treatment,12 hip protectors13 and bisphosphonates.8,9,11 There is no randomized controlled trial evidence that hormone replacement therapy (HRT) prevents non-vertebral fractures among women aged >60 years.14 Similarly, while the selective oestrogen receptor modulator (SERM) raloxifene has been shown to maintain and increase bone mass, this has not translated into a significant non-vertebral fracture benefit.10
Calcium and vitamin D supplementation may prevent hip fractures among the frailest elderly, but there is still a paucity of evidence to recommend its widespread use outside of institutional care settings.12 A recent Cochrane review noted that while there was evidence that hip protectors may be effective among people in residential care, the available trials were of poor quality, and acceptance may be low among people living in the community.15 In contrast, there is good randomized evidence that bisphosphonates prevent vertebral and non-vertebral fractures among women with osteoporosis. For two of the three bisphosphonates commonly available (alendronate and risedronate), there are large trials in which their use was associated with significant reductions in hip fractures.8,9,11 Risedronate reduced hip fractures by 60% in women with osteoporosis aged 70–79 in a randomized trial that was powered to detect a reduction in hip fracture,11 while alendronate reduced hip fractures by 51% in a similar cohort of women.8,9 In the case of etidronate, a third bisphosphonate available in the UK, a recent meta-analysis suggested that it was ineffective at preventing non-vertebral fractures.16
There are few economic evaluations of non-HRT treatments for fracture prevention,17 so the cost utility of bisphosphonate treatment is unclear. To address this, we have undertaken a cost-utility analysis of risedronate for the prevention of vertebral and non-vertebral fractures among women with severe osteoporosis.
|
Methods |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
We have used cost-utility analysis (CUA) to estimate the costs and benefits, as measured in quality adjusted life-years (QALYs) of preventing fractures. Our aim was to estimate the cost-utility of risedronate therapy for preventing further fractures in older women with established osteoporosis. We define established osteoporosis in accordance with the World Health Organization (WHO) guidelines, that is with a pre-existing fragility fracture (i.e. vertebral fracture) and low bone mineral density (BMD).
Model
We adapted a Markov simulation model, which has been developed for evaluating osteoporosis treatments to a UK setting (Figure 1
).10 In the model, patients are allowed to transit between four health states (oval nodes), and are at risk of experiencing four different fracture events (polygon nodes). Transitions between nodes are represented by line connectors; double arrow connectors indicate that individuals can go back to their starting health state after the occurrence of a fracture event. Our baseline analysis was a simulation of the typical patient included in the latest clinical trial of risedronate.11 Our analysis assumed that risedronate was only effective during therapy, without any residual effect. In the baseline model we used a discount rate of 6%, and 1.5% for costs and health outcomes, respectively. This paper presents a conservative analysis, as a cohort of 1000 women were modelled to receive risedronate treatment for 3 years, but then continued in the model until they either died or reached age 100. This was because published efficacy rates come from the trial where therapy was received for three years.
|
Effectiveness
Based on randomized controlled trial data, we have assumed 60%, 41% and 39%, relative risk reductions for hip, vertebral and other fractures, respectively.11,18
Risk estimation
To estimate the fracture risk in an untreated population, we have taken the most recent epidemiological study of fracture rates in the UK.19 For the increased risk associated with a vertebral fracture, we have used a published estimates from an analysis of a Dutch cohort study. This investigation focused on predictive ability of vertebral fractures on a European patient population. The study demonstrated relative risks of appendicular fracture of 1.6 and 4.4, for a mild and severe prevalent deformities, respectively.3 As these data were published in a form that did not allow a simple multiplication of relative risk by population risk, we had to estimate the appropriate risk multipliers from data presented in the paper. We used the adjusted risk of a mild vertebral fracture (1.45) rather than the value for a severe deformity (3.0).
Hip fracture incidence
Our calculations returned a slightly higher estimated hip fracture rate of 2.57%, compared with the risedronate trial incidence of 2.48%. This slight difference was probably due to a ‘healthy cohort’ effect on trial patients.
Costs
We have used the current price of risedronate as published in MIMS (£285 per year). A bone density scan was costed at £30, which was taken from a 1993 publication and adjusted to current prices.20 This cost was included as BMD measure was a criteria of entry into the trial on which this analysis is based. We assumed that the prescription of risedronate would be associated with one extra visit to the GP in the first year of treatment, which we costed at £16.
Averted fracture costs were taken from a cost of illness study, where we adjusted prices to 1999 levels.21 Further, the cost of illness study only provided estimates of hip fracture cost for the first year of fracture. In actual clinical practice, hip fractures continue to impact the patient and health services over time. We therefore estimated what the second year of costs would be for a hip fracture, by assuming a 30% increase in mortality compared with the population for the second year and then extending the cost estimates of the survivors for another 12 months, after deducting acute hospital costs.
Utility
The major consequence of fracture for the patient is reduced quality of life. There are a number of published estimates of the disutility of hip fracture, which can be used in an economic evaluation. However, some of these estimates have problems. Evaluations of HRT have in the past used a utility estimate from a 1980 publication by Weinstein.22 More recently, the NOF guidelines from the USA convened a panel of clinicians to estimate the disutility of fractures.23 However, empirical estimates of the utility consequences of fracture have become available more recently. Dolan et al.24 published estimates of the disutility of Colles fracture, while Salkeld and colleagues,25 from Australia, published some utility values for hip fracture. In this analysis we have estimated the disutility of hip fracture directly by comparing the values for a measure of utility—the EuroQol—from patients who had sustained a fracture (Iglesias, unpublished data). These estimates proved to be very similar to the Australian data. Table 1 summarizes our assumptions for the cost utility analysis of risedronate therapy.
|
Dealing with uncertainty
As in all evaluations, estimates of costs and effects are surrounded by uncertainty. For the price of the bisphosphonate, there is zero uncertainty, as the price has no variance. However, for the other estimates of costs and effectiveness there is uncertainty. To deal with this in the analysis we undertook a Monte Carlo simulation exercise. In the simulation, we assign probability distribution functions to all the parameters which, according to a one-way sensitivity analysis, had an effect on the final results. All probability distributions were centred on the estimated means and bounded by the 95%CIs for each parameter. For example, the reduction in hip fracture was defined as a normal distribution with mean 60% and 95%CI 20–80%. The computer simulation therefore sampled randomly across this distribution, with the sampling reflecting distribution shape. We also undertook the following sensitivity analyses: we modelled the data over the length of the trial (i.e. 3 years); and used hip fracture incidence from the clinical trial of risedronate. Confidence intervals were estimated using a non-parametric approach: 2.5% and 97.5% percentiles of the probability distributions associated to parameters of interest are reported. This approach simultaneously adjusts for the uncertainties in the effectives estimates, cost estimates, and utility estimates.
|
Results |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Our 25-year follow-up baseline results are shown in Table 2
. As the table shows, treating a 75-year-old woman with a mild vertebral deformity, who has low BMD at the femur, for three years, will lead to an expected cost saving of £786 (95%CI £1550 cost saving to £47 extra cost), compared with a similar untreated patient. The expected additional benefit of this is 43 QALYs per 1000 women (i.e. 7376 and 7419 for untreated and treated women, respectively) and 24 hip fractures averted per 1000 treated women (i.e. 366 per 1000 and 342 per 1000 for untreated and treated women, respectively).
|
Sensitivity analysis
Restricting the analysis to 3 years duration (the same length as the trial) shows that treating women with a mild vertebral deformity results in a net benefit but at additional cost (Table 3
). There is an approximate cost per QALY of £1188/£8625, with/without discounting, respectively. The hip fracture incidence in the UK is somewhat greater than that observed in the clinical trial of risedronate; we therefore modelled the incidence, over 3 years, as observed in the clinical trial (Table 4). Using this estimate, the cost per QALY without discounting was £6235, and when discounting costs and benefits, risedronate was associated with cost savings of £14 per patient (95%CI £276 cost saving to £253 extra cost).
|
|
|
Discussion |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Osteoporosis has significant cost and quality-of-life consequences. Although there have been a substantial number of economic evaluations of preventative measures, these have mainly focussed on the use of HRT. The economics of such a strategy, however, have been cast into doubt by a recent meta-analysis suggesting that there is no significant effect of oestrogens when therapy is started among women aged >60 years (the age at which osteoporosis progression typically starts).14 On the other hand, there are robust data for the two bisphosphonates, alendronate and risedronate, that these are efficacious in preventing fractures. This analysis, based on the results of the latest clinical trials of the bisphosphonate, risedronate, shows that this treatment (when targeted at 75-year-old women with a vertebral fracture and low BMD) may be cost saving, when using UK hip fracture incidence data. If we restrict our analysis to only 3 years follow-up and use trial hip fracture incidence the treatment is no longer cost-saving. However, the cost per QALY is <£20 000, widely considered to be an important benchmark in the UK. Whether it is appropriate to use trial incidence data in our model is debatable. It is likely that patients enrolled in a clinical trial are less frail than the ‘average’ high-risk patient seen in normal practice, and they will have a lower absolute risk of hip fracture. On the other hand, one might argue that it is the type of patient included in the clinical trial that for whom we can be sure that risedronate is effective. Nevertheless, whichever viewpoint one takes, our analysis still suggests that bisphosphonate therapy among women with established osteoporosis produces QALYs at a reasonable cost, or is even cost-saving.
Our analysis has implicitly compared risedronate against either no treatment or calcium with vitamin D therapy, as the latter functioned as the placebo in the latest trial. We decided not to evaluate the cost-effectiveness of risedronate against other therapies for the following reasons. First, a comparison against HRT would not be sensible, given the lack of effectiveness evidence. Second, we did not attempt a comparison with etidronate, because it has not been shown to reduce non-spine fractures in randomized trials.16
Like all evaluations, this analysis has strengths and weaknesses. Firstly, although our effectiveness data came from a large randomized trial, our cost data did not. Thus, in this respect, we had to make assumptions based on the likely resource consequences of treatment rather than actual changes in resource use as observed in a clinical trial. On the other hand, the model we used avoids some of the limitations of trial-based data, in that we can extend our analysis beyond the 3-year length of the trial. Furthermore, the risedronate trials were placebo-controlled and therefore not entirely pragmatic, and had cost data been collected as part of the trial, this may not have reflected routine clinical practice. Finally, because hip fracture is the most economically important fracture, it is a strength that our evaluation is based on a trial which has used hip fracture as a principal outcome rather than surrogate markers of effect, such as changes in BMD.
In summary, risedronate treatment for women aged 
75 years, who have low BMD and a prior vertebral fracture, not only improves quality of life by reducing fractures but also leads to lower net costs, and potential cost savings.
|
Acknowledgments |
|---|
The present work was supported by Procter and Gamble Pharmaceuticals and Aventis Pharma.
|
Notes |
|---|
Address correspondence to Ms C.P. Iglesias, Department of Health Sciences & Centre for Health Economics, University of York, York YO10 5DD.
|
References |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
1. Torgerson DJ, Iglesias CP, Reid DM. The economics of Fracture Prevention. In: Barlow DH, Francis RM, Miles A, eds. UK Advance Series: Key Advances in Effective Management of Osteoporosis. London, 2000.
2.
Cummings SR, Nevitt MC, Browner WS, et al. Risk factors for hip fracture in white women. N Engl J Med1995; 332:767–73.
3.
Burger H, van Daele PLA, Algra D, Hofman A, Grobbee DE, Schutte HE, et al. Vertebral deformities as predictors of non-vertebral fractures. Br Med J1994; 309:991–2.
4.
Lindsay R, Silverman SL, Cooper C, Hanley SA, Barton I, Broy SB, et al. Risk of new vertebral fracture in the year following a fracture. JAMA2001; 285:320–3.
5. Oleksik A, Lips P, Dawson A, Minshall ME, Shen W, Cooper C, et al. Health related quality of life in postmenopausal women with low BMD with or without prevalent vertebral fractures. J Bone Miner Res2000; 15:1384–92.[Web of Science][Medline]
6.
Torgerson DJ, Dolan P. Prescribing by general practitioners after an osteoporotic fracture. Ann Rheu Dis1998; 57:378–9.
7. Cauley JA, Lucas FL, Kuller LH, Vogt MT, Browner WS, Cummings SR. Bone mineral density and risk of breast cancer in older women. JAMA1996; 17:1404–8.
8. Black DM, Cummings SR, Karpf DB, Cauley JA, Thompson DE, Nevitt MC, et al. Randomised trial of effect of alendronate on risk of fracture in women with existing vertebral fractures. Lancet1996; 348:1535–41.[Web of Science][Medline]
9.
Cummings SR, Black DM, Thompson DE, et al. Effect of alendronate on risk of fracture in women with low bone density but without fractures. JAMA1998; 280:2077–82.
10.
Ettinger B, Black DM, Mitlak BH, Knickerbocker RK, Nickelsen T, et al. Reduction of vertebral fracture risk in postmenopausal women with osteoporosis treated with raloxifene. JAMA1999; 282:637–45.
11.
McClung MR, Geusens P, Miller PD, Zipple H, Bensen WG, Roux C, et al. Effect of risedronate on the risk of hip fracture in elderly women. N Engl J Med2001; 344:333–40.
12. Chapuy MC, Arlott ME, Duboeuff F, Brun J, Crouzet B, Arnaud S, et al. Vitamin D3 and calcium to prevent hip fractures in elderly women. N Engl J Med1992; 327:1637–42.[Abstract]
13. Lauritzen JB, Petersen MM, Lund B. Effect of external hip protectors on hip fractures. Lancet1993; 341:11–13.[Web of Science][Medline]
14.
Torgerson DJ, Bell-Syer SEM. Hormone replacement therapy and prevention of non-vertebral fractures: a meta-analysis of randomized trials. JAMA2001; 285:2891–7.
15. Parker MJ, Gillespie LD, Gillespie WJ. Hip protectors for preventing hip fractures in the elderly. Cochrane Library2001; vol. 2.
16. Cranney A, Guyatt G, Krolicki N, Welch V, Griffith L, Adachi JD, et al. A meta-analysis of etidronate for the treatment of postmenopausal osteoporosis: Osteoporosis International2001; 12:140–51.
17. Torgerson DJ, Reid DM. The economics of osteoporosis and its prevention. Pharm Eco1997; 11:126–38.
18.
Harris ST, Watts NB, Genant HK, McKeever CD, Hangartner T, Keller M, et al. Effects of risedronate treatment on vertebral and non-vertebral fractures in women with postmenopausal osteoporosis. JAMA1999; 282:1344–52.
19. Johansen A, Evans RJ, Stone MD, Richmond PW, Lo SV, Woodhouse KW. Fracture incidence in England and Wales: a study based on the population of Cardiff. Injury1997; 28:655–60.[Web of Science][Medline]
20. Torgerson DJ, Garton MJ, Donaldson C, Reid DM, Russell IT. Recruitment methods for screening programmes: trial of an improved method within a regional osteoporosis study. Br Med J1993; 307:99.
21. Dolan P, Torgerson DJ. The costs of treating osteoporotic fractures in the United Kingdom female population. Osteoporosis Int1998; 8:611–17.[Web of Science][Medline]
22. Weinstein MC. Estrogen use in postmenopausal women: costs, risks, benefits. N Engl J Med1980; 303:308–16.[Abstract]
23. Eddy DM, Johnston CC, Cummings SR, Dawson-Hughes B, Lindsay R, Melton LJ, et al. Osteoporosis: cost-effectiveness analysis and review of the evidence for prevention, diagnosis and treatment. Osteoporosis Int1998; 8(Suppl. 4).
24. Dolan P, Torgerson DJ, Kakarlapudi TK. Health-related quality of life of Colles' fracture patients. Osteoporosis Int1999; 9:196–9.[Medline]
25.
Salkeld G, Cameron ID, Cumming RG, Easter S, Seymour J, Kurrle SE, et al. Quality of life related to fear of falling and hip fracture in older women: a time trade off study. Br Med J2000; 320:341–6.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  H. Liu, K. Michaud, S. Nayak, D. B. Karpf, D. K. Owens, and A. M. Garber The Cost-effectiveness of Therapy With Teriparatide and Alendronate in Women With Severe Osteoporosis. Arch Intern Med, June 12, 2006; 166(11): 1209 - 1217. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. T. Schousboe, J. A. Nyman, R. L. Kane, and K. E. Ensrud Cost-Effectiveness of Alendronate Therapy for Osteopenic Postmenopausal Women Ann Intern Med, May 3, 2005; 142(9): 734 - 741. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. R. McClung Osteopenia: To Treat or Not To Treat? Ann Intern Med, May 3, 2005; 142(9): 796 - 7. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||