|
Simulation-based cost-utility analysis of population screening-based alendronate use in Switzerland |
Schwenkglenks M, Lippuner K |
|
|
Record Status 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. Health technology The study compared population-based dual-energy X-ray absorptiometry (DXA) screening followed by treatment with alendronate (FOSAMAX; Merck & Co) versus no screening or drug treatment for osteopenia and osteoporosis. The main screening ages were 65, 75 and 85 years. The three treatment options compared were treatment with alendronate for 5 years with full persistence, treatment with alendronate for 5 years with realistic persistence, and treatment with alendronate for 10 years with realistic persistence. Study population As this was a modelling study, the target population comprised men and women at the age of 50 years. Setting The setting was not explicitly reported. However, it appears to have been secondary care. The economic study was carried out in Switzerland. Dates to which data relate The effectiveness data used to populate the model were obtained from studies published between 1992 and 2005. However, the authors reported that some model parameters used in the previous version of the model (Schwenkglenks et al. 2005) were also retained in the current version, making it difficult to report the dates relating to those parameters. The price year was 2000. Modelling A Markov state transition model was used to model disease progression. The model was based on one reported by Schwenkglenks et al. (see 'Other Publications of Related Interest' below for bibliographic details). All modifications conducted by the authors in relation to the original version of the model were explicitly reported. A life-long time horizon from the age of 50 years was applied with a cycle length of 1 month. Health states, and cycle length, along with a number of modelling assumptions were presented in full. The model was analysed using individual, first-order Monte Carlo simulation. The methods previously used for model validation (Schwenkglenks et al. 2005) were not reported in the current paper. Study designs and other criteria for inclusion in the review The authors mainly reported the additional input parameters used in relation to the former version of the model (Schwenkglenks et al. 2005). These comprised prevalence of osteoporosis and osteopenia for men and women, relative fracture risk when suffering from osteoporosis or osteopenia for vertebral and non-vertebral fractures, relative risk of fracture during alendronate treatment, persistence with alendronate treatment and mortality. Sources searched to identify primary studies Data on the prevalence of osteopenia and osteoporosis were derived from the National Health and Nutrition Examination Survey (NHANES III), which is a continuous survey of a representative sample of the non-institutionalised US population, and the Rotterdam study, a prospective, population-based cohort study of men and women aged over 55 years. The prevalence of any previous fracture was derived from a published meta-analysis. The relative risk of fracture and fracture during alendronate treatment were derived from the Rotterdam study and from other recently published studies. However, the designs of the studies were not reported. Data on the persistence with alendronate treatment were obtained from a published randomised controlled trial. All adjustments made by the authors to published data were reported. Methods used to derive estimates of effectiveness The process used to identify the data was not reported. No inclusion criteria were specified for any parameters. The method used to select the estimates was neither reported nor discussed. Measure of benefits used in the economic analysis The measure of benefit used was the quality-adjusted-life-years (QALYs). Utilities were derived from published literature, and adjustments were made using correction factors obtained from published studies. It was reported that utilities and multiplication factors were valued using the time trade-off technique and the EQ-5D. The QALYs were discounted at an annual rate of 3%. Direct costs From the perspective of the health care system, the direct costs included in the analysis were the monthly cost of alendronate treatment, cost of screening (bone densitometry, medical consultation and typical services provided), daily inpatient costs for acute hospital care, rehabilitation facilities and nursing homes, and ambulatory treatment costs following fracture. The cost data were mainly derived from official Swiss sources, augmented by expert opinion in the case of non-availability of relevant data. The resource use data were derived from published sources. The costs and resource use data were reported separately only for the categories of inpatient types of care. It was reported that inpatient costs were modelled independently, but the relevant details were given in the earlier study (Schwenkglenks et al. 2005). The costs were appropriately adjusted for inflation and reported for the price year 2000. Discounting was performed at an annual rate of 3%. Statistical analysis of costs It appears that the costs were treated deterministically. Indirect Costs No productivity losses were accounted for in the economic analysis. Sensitivity analysis In the base-case, model entry was assumed at the age of 50 years. Alternative model entry at the main screening age (disregarding the patients’ previous diagnosis and treatment history) was also investigated. One-way sensitivity analyses were conducted on numerous model parameters. All parameters and the ranges over which the parameters were varied were explicitly reported. In addition, the authors conducted probabilistic sensitivity analyses. The assigned probability distributions were reported. Probabilistic sensitivity analyses were conducted using 500 different groups of model parameters and 2,000 simulated persons per group of input parameters and model arm. Estimated benefits used in the economic analysis The incremental benefits were reported per 1,000 persons.
Under the assumptions of model entry at the age of 50 years, 5-year intended duration of alendronate treatment, and realistic persistence, the screen-and-treat strategy when compared with no intervention resulted in 8.34 incremental QALYs for women and 2.05 QALYs for men when the main screening age was 65 years.
The screen-and-treat strategy yielded 9.85 extra QALYs for women and 2.04 QALYs for men when the main screening age was 75 years, and 5,44 additional QALYs for women and 1.01 QALYs for men when the main screening age was 85 years. Cost results The incremental costs were reported per 1,000 persons.
Under the assumptions of model entry at the age of 50 years, 5-year intended duration of alendronate treatment, and realistic persistence, the screen-and-treat strategy when compared with no intervention resulted in an incremental cost of CHF 591,920 for women and CHF 405,190 for men when the main screening age was 65 years.
The screen-and-treat strategy yielded an incremental cost of CHF 348,750 for women and CHF 251,570 for men when the main screening age was 75 years, and an incremental cost of CHF 153,340 for women and CHF 120,210 for men when the main screening age was 85 years. Synthesis of costs and benefits The screen-and-treat strategy was compared with no intervention and incremental cost-utility ratios (ICURs) were reported.
At model entry age of 50 years, the ICUR ranged from CHF 19,433 per QALY gained for women and CHF 93,184 for men (main screening age 85 years, 5 years' treatment and 100% persistence) to CHF 70,995 for women and CHF 197,460 for men (main screening age 65 years, 5 years' treatment and realistic persistence).
When the age at model entry was the same as the main screening age, the screen-and-treat strategy was cost-saving for all scenarios for women at the age of 85 years. For men, the ICUR ranged from CHF 30,763 (age 85 years, 5 years' treatment and 100% persistence) to CHF 176,670 (age 65 years, 5 years' treatment and realistic persistence).
The one-way sensitivity analysis demonstrated that the results were most sensitive to variation in the risk reduction due to alendronate, duration of the effect of alendronate post administration, and the cost of drug treatment.
The probabilistic sensitivity analyses demonstrated that for women, model entry at age 50 years, main screening age 75 years, 5 years' treatment and realistic persistence, a cost-effectiveness threshold of CHF 50,000 per QALY gained was achieved in 79% of cases (95% confidence interval of the ICUR: cost-saving to CHF 79,525). When the main screening age was reduced to 65 years, the cost-effectiveness threshold was only achieved in 16% of cases. Authors' conclusions The authors concluded that population-based dual-energy X-ray absorptiometry (DXA) screening followed by alendronate treatment for patients with osteoporosis, or with fracture and osteopenia, represents a cost-effective strategy in Switzerland only for postmenopausal women above the age of 75 years. CRD COMMENTARY - Selection of comparators The interventions were adequately described. The comparator appears to have been current practice in the study setting. Alternative strategies were not discussed. Validity of estimate of measure of effectiveness The parameters were derived from published research. No systematic search of the data was reported. However, from the authors' discussion it would appear that the quality of the evidence used to derive the estimates (i.e. national continuous survey, meta-analysis, randomised controlled trial) was adequate. The authors explicitly reported all adjustments conducted to published data. Validity of estimate of measure of benefit The estimation of health benefits (QALYs) was modelled using a Markov model. Although utility weights were taken from published studies, the methods used to estimate them were reported, along with the correction factors used by the authors to make appropriate adjustments. The benefits were properly discounted. Validity of estimate of costs The analysis of the costs was performed from the perspective of the health care system. Extra drug costs for patients who had permanently discontinued alendronate treatment were not included in the analysis, which might have resulted in an overestimation of the cost-effectiveness of the intervention. Appropriate adjustments for inflation and discounting were performed. Uncertainty in the cost data was investigated separately using one-way sensitivity analysis, and jointly with the effectiveness data by random sampling of observations to produce a cost-effectiveness plane. The costs and the quantities were not reported separately, which will hinder the replication of the analysis in other settings. Although it was reported that the inpatient costs were modelled independently, this analysis was not presented in the current paper. Other issues The authors compared their findings with the results from the single previous study in the same area and methodological differences were discussed in full. The authors acknowledged variation in the resource use and cost data between settings but did not explicitly evaluate its impact on the economic results. However, a balanced discussion on the limitation of the transferability of the results to other settings was given. The authors reported a number of limitations to their study that mainly related to the modelling nature of the study, which does not accurately reflect reality, and to the availability and robustness of the data used. The analysis was restricted to three fracture sites, thus failing to account for fractures at other skeletal sites. This might have resulted in an underestimation of the cost-effectiveness of the intervention. Finally, the analysis did not evaluate the cost-effectiveness of pre-selecting sub-populations at high risk of osteoporosis. Implications of the study The authors did not make explicit recommendations for changes in policy or practice. Given the non-availability of robust data, they called for further research into the co-morbidities of osteoporotic fracture patients in comparison with the general population. In addition, they acknowledged that the screen-and-treat strategy and its cost-effectiveness should be further evaluated, especially in sub-populations at high risk of osteoporosis. Source of funding Supported by a grant from Merck Sharp and Dohme-Chibret AG, Switzerland. Bibliographic details Schwenkglenks M, Lippuner K. Simulation-based cost-utility analysis of population screening-based alendronate use in Switzerland. Osteoporosis International 2007; 18(11): 1481-1491 Other publications of related interest Schwenkglenks M, Lippuner K, Hauselmann HJ, Szucs TD. A model of osteoporosis impact in Switzerland 2000-2020. Osteoporos Int 2005;16:659-71. Indexing Status Subject indexing assigned by NLM MeSH Absorptiometry, Photon /economics; Aged; Aged, 80 and over; Alendronate /economics /therapeutic use; Bone Density Conservation Agents /economics /therapeutic use; Cost-Benefit Analysis; Drug Costs /statistics & Female; Health Care Costs /statistics & Health Services Research /methods; Humans; Male; Markov Chains; Mass Screening /economics; Middle Aged; Osteoporosis /diagnosis /drug therapy /economics; Osteoporosis, Postmenopausal /diagnosis /drug therapy /economics; Quality-Adjusted Life Years; Sex Factors; Switzerland; numerical data; numerical data AccessionNumber 22007002211 Date bibliographic record published 23/10/2007 Date abstract record published 30/09/2008 |
|
|
|