|
Cost-effectiveness of preventing hip fracture in the general female population |
Kanis J A, Dawson A, Oden A, Johnell O, de Laet C, Jonsson B |
|
|
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 Interventions to prevent fractures in the general female population were studied. These interventions were not specifically listed. Rather, four hypothetical interventions with different costs were investigated. The comparator was not specified clearly, but appears to have been a 'no intervention' strategy.
Study population The study population comprised Swedish women who were aged at least 50 years, and were at an average risk of hip fracture at any given age.
Setting The setting was primary care and secondary care. The economic analysis was carried out in Sweden and in the UK.
Dates to which data relate The effectiveness data were obtained from studies published between 1980 and 1999. The resource use data were obtained from studies published between 1997 and 1998. The price year was not specified.
Source of effectiveness data The effectiveness data were based on a review and a synthesis of completed studies. In addition, the authors made assumptions about the effectiveness.
Modelling The model used was based on the Markov model of Jonsson et al. (see Other Publications of Related Interest).
Outcomes assessed in the review The outcomes assessed were the risk reduction associated with treatment, the hip fracture rates, and the case-fatality rates associated with hip fractures.
Study designs and other criteria for inclusion in the review Sources searched to identify primary studies Criteria used to ensure the validity of primary studies Methods used to judge relevance and validity, and for extracting data Number of primary studies included Ten primary studies were included in the review.
Methods of combining primary studies The primary studies appear to have been combined in a narrative.
Investigation of differences between primary studies Results of the review A risk reduction of 35% from the onset of treatment was used as the base-case. Additional analyses used values of 15 and 50%.
The value of 35% was stated to be an approximation of the effect of vitamin D with calcium.
The value of 50% was stated to be an approximation of the average risk reduction with bisphosphonates and oestrogens.
A value of 15% was stated to accommodate "an imperfect compliance, with regard to efficacy".
The age-specific hip fracture rates were not reported.
Excess mortality after hip fracture was not reported but was included for the first year after hip fracture.
Methods used to derive estimates of effectiveness The authors made some assumptions about effectiveness.
Estimates of effectiveness and key assumptions Subsequent hip fractures were not considered, on the basis that interventions to prevent them would be at least as cost-effective, given the increase in absolute risk following the first fracture.
It was assumed that, after stopping the intervention, the risk reduction would reverse in a linear manner over a 5-year period.
The hip fracture rates were assumed not to change over the lifetime of the individuals.
Excess mortality was assumed to be the same as the general population for the second and subsequent years.
Measure of benefits used in the economic analysis The benefit measures were the number of life-years gained and the quality-adjusted life-years (QALYs). The age-specific utilities were derived from EuroQol 5D (see Other Publications of Related Interest). The health benefits were discounted at 3%.
Direct costs The costs were discounted at 3%. The quantities and costs were not analysed separately. The direct costs were defined as the difference between the cost of the intervention and the savings due to the reduction in the number of hip fractures, which would require future treatment. The intervention costs were for diagnosis, the drugs and monitoring of treatment. The cost of fractures was estimated as the difference between the costs during the years after and preceding the fracture. The additional costs in the second and subsequent years were assumed to be constant. The costs associated with increased survival, due to a reduction in the number of hip fractures, were included in a separate analysis. When a person died or had a hip fracture, the intervention costs were assumed to be zero.
The quantity/cost boundary adopted appears to have been that of the health service, although this was not clearly specified. The quantities and costs were estimated from the model. It was not specified when the quantity of resources was measured. In addition, the price year was not reported. The difference between the marginal and average costs was not discussed.
Statistical analysis of costs Indirect Costs The indirect costs were not included in the analysis.
Currency US dollars ($). The costs were derived in Swedish kroner (SEK), then converted into US dollars using an exchange rate of $1.00 = 8.00 SEK.
Sensitivity analysis The sensitivity analyses were conducted by varying the following: the cost of the intervention, the risk reduction in fractures following treatment, the cost to the health service of an increased patient lifetime, and the age when the treatment was initiated. Multi-way sensitivity analyses were used to measure the effect of this variation on to cost-effectiveness, particularly in relation to a threshold of $30,000/QALY gained.
Estimated benefits used in the economic analysis The benefits derived from the model were not reported. The duration of the benefits was lifetime. The side-effects of the treatment were not included in the analysis.
Cost results The total net costs were calculated for the four different levels of intervention cost, assuming a risk reduction in hip fracture of 35% and a 5-year treatment strategy. For an intervention costing $625/year, the total net costs were between $2,770 and $325, depending on the age when the treatment was initiated (50 to 90 years). For an intervention costing $250/year, the total net costs were between $1,035 and -$815. For an intervention costing $125/year, the total net costs were between $455 and -$1,190. Finally, for an intervention costing $63/year, the total net costs were between $175 and -$1,380.
The costs were discounted at a rate of 3%. The total costs represented the net cost of the intervention to the health service. The duration of the intervention was 5 years and the follow-up was lifetime. The cost of adverse effects was not included.
Synthesis of costs and benefits All increments were calculated as if from a two-intervention strategy. For a treatment that decreased hip fracture by 35% and cost $625/year, the cost per life-year gained varied between $8,750,000 and $5,000, depending on the age when the treatment was initiated (between 50 and 90 years). The corresponding cost per QALY gained varied between $1,095,000 and $6,000. The age at which the intervention became cost-effective (threshold of $30,000/QALY gained) was greater than 90 years for a risk reduction of 15%, and was 77 years for a risk reduction of 50%.
For a treatment costing $250/year, the cost per life-year gained varied from $3,720,000 to being cost-saving, depending on the age when the treatment was initiated (between 50 and 90 years). The corresponding cost per QALY gained varied from $410,000 to being cost-saving. The age at which the intervention became cost-effective (threshold of $30,000/QALY gained) was 86 years for a risk reduction of 15%, and 66 years for a risk reduction of 50%.
For a treatment costing $125/year, the cost per life-year gained varied from $1,440,000 to being cost-saving, depending on the age when the treatment was initiated (between 50 and 90 years). The corresponding cost per QALY gained varied from $180,000 to being cost-saving.
For a treatment costing $63/year, the cost per life-year gained varied from $530,000 to being cost-saving, depending on the age when the treatment was initiated (between 50 and 90 years). The corresponding cost per QALY gained varied from $65,000 to being cost-saving. To achieve the $30,000/QALY threshold, the age would have to be 64 years for a risk reduction of 15%, and 51 years for a risk reduction of 50%.
Authors' conclusions The cost-effectiveness demonstrated was reasonable, even with the relatively high intervention costs for those women of average risk and aged at least 84 years. For the cheapest interventions ($63/year), cost-effectiveness was shown from the age of 53 years. Variations in the effectiveness (risk reduction of 15 to 50%) had marked effects on the age at which treatment was worthwhile. Groups of apparently healthy people could benefit from treatment if the efficacy of the interventions was supported by randomised controlled trials.
CRD COMMENTARY - Selection of comparators The choice of the comparators was justified in that the interventions considered represented a range of available treatments used for the prevention of fractures. However, since any treatment cost could be combined with any risk reduction, it was difficult to attribute the consequence to a particular technology. This method was also not supported by the evidence from the review of the literature. Each intervention was also compared to 'no treatment' rather than to other treatments. This does not provide information for decision-making since a higher cost-effectiveness measured in this way could be due to higher or lower effectiveness.
Validity of estimate of measure of effectiveness The authors did not state whether they conducted a systematic review of the literature. In addition, no information was provided on how the studies were identified, or the criteria on which they were selected and assessed.
The estimates of the measure of effectiveness were derived from a published model. Consequently, the authors did not report the methods used to derive these parameters, or the values of the parameters themselves, in the current study. More detailed reporting would have improved the transparency of the validity of the effectiveness measures.
Validity of estimate of measure of benefit The benefits were derived from a Markov state transition model, which was appropriate for the problem investigated. The study had several weaknesses. Firstly, fractures other than hip fractures were excluded, which would have tended to underestimate the cost-effectiveness of the treatments. Secondly, the side-effects from treatments were not modelled. This could be a problem with HRT where there is an increased risk of disease associated with treatment.
Validity of estimate of costs The costing seemed to cover most categories, but would have been strengthened by more specific reporting. The costing demonstrated several strengths. Firstly, most categories of cost relevant to the perspective adopted were included in the analysis, with the exception of the costs of treating the side-effects of some treatments. Second, for the category of costs used, all relevant costs seem to have been included in the analysis. Third, sensitivity analyses were conducted.
The limitations in the costing were as follows:
the quantities, and unit costs were not reported separately;
the source of resource use estimates was not reported;
the cost of side-effects was omitted;
a statistical analysis of the costs and quantities was not conducted;
the price year was not reported; and
it was unclear whether charges were used to proxy prices.
The sensitivity analysis was seriously flawed by:
an illogical and unsupported correlation of the costs of technology and effectiveness (risk reduction); and
a lack of incremental analysis (the increase in cost calculated from the next most effective technology).
Other issues The authors made appropriate comparisons with the findings from other studies. The issue of generalisability to other settings was addressed, in particular the difference in the risk of hip fractures in different geographic populations. The authors acknowledged some limitations to their study. In particular, these were the inclusion of only hip fractures in the analysis, the uncertainty surrounding the offset time of many treatments, and the exclusion of side-effects arising from treatment. The authors did, however, omit useful information in terms of resource quantities and unit costs.
Implications of the study The authors suggested that global strategies for the prevention of fractures are worthy of consideration. While the cost-effectiveness was dependent on the cost of the treatment, the efficacy of intervention and the average risk of hip fracture, strategies aimed at interventions for those aged over 70 years are feasible and require testing. These views must be taken in the context of the methodological problems described.
Source of funding Supported by the Lilly Research Centre.
Bibliographic details Kanis J A, Dawson A, Oden A, Johnell O, de Laet C, Jonsson B. Cost-effectiveness of preventing hip fracture in the general female population. Osteoporosis International 2001; 12(5): 356-361 Other publications of related interest Jonsson B, Kanis JA, Dawson A, Oden A, Johnell O. Effect and offset of effect of treatments for hip fractures on health outcomes. Osteoporos Int 1999;10:193-9.
Indexing Status Subject indexing assigned by NLM MeSH Age Factors; Aged; Aged, 80 and over; Computer Simulation; Cost-Benefit Analysis; Female; Hip Fractures /economics /etiology /prevention & Humans; Middle Aged; Models, Econometric; Osteoporosis, Postmenopausal /complications /economics /therapy; Quality-Adjusted Life Years; Sweden; Value of Life; control AccessionNumber 22001001329 Date bibliographic record published 31/05/2002 Date abstract record published 31/05/2002 |
|
|
|