|Cost-effective osteoporosis treatment thresholds: the United States perspective
|Tosteson A N, Melton L J, Dawson-Hughes B, Baim S, Favus M J, Khosla S, Lindsay R L
This is a critical abstract of an economic evaluation that meets the criteria for inclusion on NHS EED. Each abstract contains a brief summary of the methods, the results and conclusions followed by a detailed critical assessment on the reliability of the study and the conclusions drawn.
This study aimed to identify the 10-year hip fracture probability required for osteoporosis treatment to be cost-effective for different cohorts of patients defined by their age, sex, and race or ethnicity. The osteoporosis treatment was cost-effective, at a willingness to pay of $60,000 per quality-adjusted life-year, when the 10-year hip fracture probability reached about 3%. The study was based on valid methodology, but the sources of data were not extensively presented. Caution will be required when interpreting the authors’ conclusions.
Type of economic evaluation
The primary objective was to identify the 10-year hip fracture probability required for osteoporosis treatment to be cost-effective for different cohorts of patients defined by their age, sex, and race or ethnicity.
The two strategies examined were five-year osteoporosis treatment with a bisphosphonate-like therapy and no treatment.
This economic evaluation was based on a Markov model, which simulated patient management and subsequent risk of fractures for different cohorts of patients. The time horizon of the analysis was the lifetime. The authors did not explicitly report the perspective of the study.
The clinical data appear to have been derived from a selection of known, relevant published studies supplemented by the authors’ opinions. It appears that the treatment effect in reducing fracture risk was taken from clinical trials, although this was not clearly reported. The age-, sex- and race-specific mortalities were taken from US life tables. Other details of the design and other characteristics of the primary sources of data were not given.
Monetary benefit and utility valuations:
The evaluation of the utility associated with health states was based on a published US population study which used the Euro-Qol (EQ-5D) questionnaire.
Measure of benefit:
Quality-adjusted life-years (QALYs) were used as the summary benefit measure. They were discounted at an annual rate of 3%.
The economic analysis included the costs of osteoporosis treatment (drugs, physician visits, and bone mineral density test) and the costs associated with each type of fracture. The fracture costs were derived from a previous study and were not broken down into individual items. The sources of other costs were not clearly reported. All costs were in US dollars ($) and the price year was 2005. Future costs were discounted at an annual rate of 3%.
Analysis of uncertainty:
A deterministic univariate sensitivity analysis was performed to investigate the effects of variations in drug costs, duration of fracture sequelae, and thresholds for willingness to pay per QALY gained. The ranges of values used appear to have been defined by the authors.
The incremental cost per QALY gained with treatment over no treatment was presented for all the cohorts of patients. For example, in white women, it ranged from $380,000 at age 50 to being dominant (less expensive and more effective) at age 75. In all cohorts, the cost-effectiveness of the intervention improved as the starting age increased.
At each age, the cost per QALY gained, for those at average risk, was two or more times greater for black, Asian, and Hispanic than for white patients (both for men and for women).
Assuming a threshold of $60,000 per QALY and average-risk, the intervention would be cost-effective at a starting age of over 70 years for white women, over 75 years for Hispanic and Asian women and white men, over 80 years for black women, and over 85 years for black, Hispanic and Asian men.
Despite these variations, the absolute 10-year hip fracture probability, at which the treatment cost was $60,000 per QALY gained, was quite similar across all the race and ethnicity cohorts. In women, the intervention thresholds by age ranged from 2.4% to 4.7%, while in men they ranged from 2.4% to 4.9%.
The sensitivity analysis showed that variations in drug costs affected the intervention threshold, which was also influenced by the duration of a fracture’s adverse impact on health-related quality of life. Changes in the cost-effectiveness thresholds did not substantially affect the conclusions on the cost-effectiveness of the intervention.
The authors concluded that the osteoporosis treatment was cost-effective at a willingness to pay of $60,000 per QALY, when the 10-year hip fracture probability was around 3% or more. Similar results were found across all the race or ethnicity groups, but the threshold was slightly higher for men than for women.
The comparators were appropriately selected. The authors did not consider a specific drug for the treatment strategy. They referred to a generic "bisphosphonate-like therapy".
The approach used to identify the relevant sources of data was not reported. A description of the characteristics of the published studies was not provided, and the authors did not justify their selection of specific estimates from among those found in the literature. Although it likely that the treatment effect was taken from clinical trials, this was not explicitly stated. In general, the clinical side of the study was not extensively reported. A few details of the calculation of the QALYs were provided and QALYs are a validated measure and they capture the impact of the interventions on quality of life and survival. Furthermore, they allow cross-disease comparisons.
The authors did not state the economic viewpoint of their study and did not provide a breakdown of the costs. In effect, costs were presented as macro-categories with no information on the resource consumption. This might reduce the transparency of the economic analysis. Other aspects of the study, such as the price year and the use of discounting were reported. The sensitivity analysis considered only variations in treatment costs, while changes in other economic data were not considered.
Analysis and results:
The synthesis of costs and benefits was appropriately performed and presented. The issue of uncertainty was restricted to a deterministic approach which focused on a few model inputs, but several sub-group analyses were conducted. Two approaches to simulate fracture incidence rates (with or without taking into account the first fracture) were considered and given the similar findings achieved with the two approaches, the authors used the more straightforward technique of ignoring the impact of the first fracture. The issue of generalisability of the study findings to other settings was not explicitly addressed. The data for the intervention thresholds were presented in terms of hip fracture risk, but the model also considered the impact of wrist, clinical spine, proximal humerus, pelvis, rib, and tibia or fibula fractures on the health-related quality of life.
Overall, the study was based on a valid methodology, but the sources of data were not extensively presented. Thus, caution will required when interpreting the authors’ conclusions.
Supported by the National Osteoporosis Foundation, and grants from the National Institutes of Health.
Tosteson A N, Melton L J, Dawson-Hughes B, Baim S, Favus M J, Khosla S, Lindsay R L. Cost-effective osteoporosis treatment thresholds: the United States perspective. Osteoporosis International 2008; 19(4): 437-447
Other publications of related interest
Burge R, Dawson-Hughes B, Solomon DH, et al. Incidence and economic burden of osteoporosis-related fractures in the United States, 2005-2025. J Bone Miner Res 2007;22:465-75.
Kanis JA, Borgstrom F, Zethraeus N, et al. Intervention thresholds for osteoporosis in the UK. Bone 2005;26:22-32.
Tosteson AN, Jonsson B, Grima DT, et al. Challenges for model-based economic evaluations of postmenopausal osteoporosis interventions. Osteoporos Int 2001;12:849-57.
Subject indexing assigned by NLM
Age Factors; Aged; Aged, 80 and over; Algorithms; Cost-Benefit Analysis /statistics & numerical data; Female; Fractures, Bone /economics; Health Care Costs /statistics & numerical data /trends; Humans; Male; Middle Aged; Models, Economic; Osteoporosis /economics /epidemiology /therapy; Probability; Quality-Adjusted Life Years; Time Factors; United States /epidemiology
Date bibliographic record published
Date abstract record published