Around 200 million individuals worldwide have been diagnosed with osteoporosis and, each year, there are 1.7 million osteoporotic fractures involving the hip.1 In Spain it has been estimated that over the next 10years there will be 714,000 osteoporotic fractures, most of them in women over 70years of age.2 However, many cases of osteoporosis are not diagnosed and, frequently, no treatment is prescribed even after the patient has had a fracture.3,4 The impact on public health is enormous because of the high morbidity and mortality rates and the effect of these fractures on quality of life, in addition to their being accountable for increasing health costs.5

At the present time, attention is focused on identifying individuals with a high osteoporotic fracture risk, more than on detecting those with osteoporosis diagnosed exclusively by measuring bone mineral density (BMD).6 This identification is carried out with the help of tools—like the FRAX® score—which, integrating the information from different risk factors, enables the calculation of the absolute risk of fracture in coming years as the most accurate way to decide the need for ancillary diagnostic tests or of initiating a specific therapy. The intervention thresholds are established depending on local factors, such as reimbursement policies, health economic evaluation, willingness to pay for osteoporosis care and access to dual-energy X-ray absorptiometry (DXA).4,7

The recommendations issued by the National Osteoporosis Guideline Group (NOGG) of the United Kingdom, one of the guidelines with the greatest international influence, are based on this model of intervention and are supported by the European Society for Clinical and Economic Aspects of Osteoporosis and Osteoarthritis (ESCEO) and the International Osteoporosis Foundation (IOF), although many European countries have their own policies.8 In Spain, at this moment, there is no consensus on the thresholds for intervention; however, researchers working with the Risk Factors and Densitometry by Dual-energy X-ray Absorptiometry (FRIDEX) cohort, recently proposed a cost-effective calibration of the FRAX®.9

The objective of our study is to evaluate the impact of the application of the FRIDEX calibration in routine clinical practice and determine whether it translates into a reduction in the number of densitometries and treatments prescribed and, thus, increasing the efficiency of the associated costs. Moreover, it is compared with the NOGG guideline thresholds to detect possible differences.

The study population included 607 women with a median age of 59.4 (IQR=14) years; the average body mass index (BMI) was 26.3 (SD=4.5)kg/m2. According to the densitometry performed, they had a T-score of −1.4 in spine (SD=1.3), −1.4 in femoral neck (SD=1.3) and −2.2 in total hip (SD=1.2). The prevalences of the risk factors that are included in the FRAX® tool were, in descending order: secondary osteoporosis (24.1%), tobacco use (16.5%), parental hip fracture (15.2%), prior clinical fracture (12.4%), corticosteroid therapy (8.4%), rheumatoid arthritis (7.6%) and alcohol consumption (2.1%). In all, 31.4% of the women in the sample were prescribed medication by their attending physician after densitometry, whereas 36% of those who had a history of a fragility fracture were not being treated.

The application of the FRIDEX calibration in our sample results in an improvement in the efficiency of the management of osteoporosis both at the diagnostic level and at the time of indicating drug therapy. Moreover, it reduces costs in comparison with routine clinical practice in the León health area.

The cost analysis includes the direct costs of DXA and 5years of drug therapy in accordance with the cost of DXA in Castile and León and the real treatment profile in Spain.11 The direct and indirect costs related to the fractures were not included in the assessment. In short, the use of the thresholds for the FRIDEX calibration in our sample would make it possible to save €37.4/patient in DXA and €314.6/patient in medication costs. The total expense (DXA and treatment) would be reduced over 5years by 18.8% with respect to the approach adopted in our area. In contrast, the NOGG guideline would reduce overall costs by 35.0% over 5years. This would mean a savings in our sample of €39.9/patient in DXA and €685.4/patient in treatment for 5years. In a study on the FRIDEX model, the overall costs decreased by 28.7% compared to the standard strategy based on DXA to detect the same number of women who had fractures.9

When the FRIDEX cost-effectiveness proposal is compared with the NOGG guideline thresholds for a population in the United Kingdom, it is observed that both recommendations reduce the number of DXA requested and that of prescribed treatments when compared with the approach adopted in our health area. However, the level of agreement between the two is low. Thus, although the absolute number of DXA and that of treatments that are indicated by the study do not differ much, the persons to which they are directed are different. The FRIDEX calibration recommends the measurement of BMD in older individuals with a high body mass index, whereas the NOGG guidelines are aimed mainly at persons who have more severe clinical risk factors. The FRIDEX proposal helps by identifying more cases of osteoporosis in the lumbar spine as well as in hip with respect to NOGG. Likewise, when it comes to suggesting therapy, the NOGG guidelines would indicate treating young persons in whom risk factors were detected, whereas the FRIDEX calibration would focus more on elderly individuals and on patients with osteoporosis involving the lumbar spine. Clinical practice guidelines should integrate in the best possible way the information from models that predict the risk of fracture to enable the identification of the individuals who would most benefit from bone densitometry. The purpose would be to evaluate risk more accurately, which would lead to a better selection of candidates for antiresorptive therapy.

The NOGG guidelines and FRIDEX calibration are based on the FRAX® score, the only tool that takes into account life expectancy and the epidemiological differences among countries.10 The thresholds proposed by the FRIDEX calibration are of interest as they focus on the Spanish population and are corroborated by cost-effectiveness studies with national considerations. Therefore, the use of these thresholds would be more appropriate than those concerning the British population. On the other hand, the NOGG guidelines are those with the best results, but they have not been validated for the Spanish population. They are different in that the intervention threshold is flexible, and varies depending on age, a parameter that is important in cost-effectiveness studies.16 However, in several studies, it has been seen that the NOGG guidelines are associated with difficulties when it comes to preventing fragility fractures, since, for example, elderly individuals should have an elevated risk as candidates for treatment, despite the fact that studies have demonstrated that the treatment in those patients is cost-effective with smaller thresholds.17,18 To make an accurate evaluation of the cost-effectiveness of guidelines, it is necessary to assess savings both over the short term—reduction in the number of bone densitometries and of prescribed treatment—and the long term, with savings related to fracture prevention.19 Thus, guidelines like those proposed by the NOGG, which are cost-effective over the short term, may not be over the long term, since they have the inconvenience of being less precise in their identification of high-risk patients, at least at the time of indicating whether or not they should be treated.18,20 This means that it is necessary to conduct national studies for the purpose of identifying the intervention thresholds at which the benefits of treatment, over both the short and the long term are maximized, while the risks and costs that accompany treatment are limited, always taking into consideration the amount society is willing to pay. In Spain, at the present time, according to FRAX®, there are no valid thresholds for high-risk of fracture. Moreover, providing a single value to represent the risk of fracture may not be wholly suitable, when the incidence of hip fractures can be up to 2-fold higher depending on the Spanish autonomous community.21 Until we have those thresholds, the proposal of the FRIDEX calibration may continue to have a role, as it is an improvement in the efficiency of the management of osteoporosis and better identifies densitometric osteoporosis in our setting than the British NOGG guidelines.

This report reveals the implications of the utilization of the NOGG guidelines and the FRIDEX calibration in the diagnostic and therapeutic efficiency in the management of osteoporosis in routine clinical practice, demonstrating a considerable heterogeneity among the recommendations for approaching this condition, which can influence decision-making concerning our patient population. However, it has certain limitations that we consider we should mention. First, the sample may have selection and reference biases, as it was made up of individuals who were referred to undergo DXA in accordance with the criteria of a physician. This may affect the results, since these persons have been subjected to some type of prior selection and the sample cannot be totally representative of the general population, compromising the external validity. However, this question would not affect analyses comparing different guidelines. Second, the Spanish version of the FRAX® tool has weaknesses, such as its underestimation of the risk of major fracture.22 The lack of robust Spanish epidemiological data concerning the incidence of major osteoporotic fractures has been approached through the application of the Swedish ratio of hip fractures to major fractures. However, the Spanish epidemiological findings on hip fractures and on nonvertebral and nonfemoral fractures indicate that, while the incidence of hip fractures in Sweden is approximately twice that reported in Spain, the frequency of the latter is relatively similar.11,22 This underestimation of the risk was one of the reasons why we decided to employ the British FRAX® when applying the NOGG decision thresholds in our study, taking into account the associated difference in risk assessment, as well.22,23 The British FRAX® tool duplicates risk in comparison with the Spanish version.23 As a result, the use of the latter would lead to a further underestimation of the risk of fracture when the NOGG thresholds are utilized. Third, the NOGG guidelines are validated for a specific population. This means that the characteristics may be different (prevalence of risk factors, fracture incidence, etc.), possibly compromising external validation and, thus, extrapolating the results requires caution. A fourth limitation is that the presence of silent vertebral fractures was not assessed, a fact that could affect therapeutic decisions based on the guidelines.

Both the FRIDEX calibration and the NOGG guidelines have considerable heterogeneity. This signifies that there can be substantial variability in the number of individuals in which performance of DXA is recommended, as well as in those for whom the decision is made to initiate treatment. The use of either of them in our routine clinical practice would increase our efficiency in the management of osteoporosis, and we would achieve a significant reduction in costs. Nevertheless, further studies will help us to identify the intervention thresholds applicable in the Spanish population to maximize the cost-benefit.

The authors declare they have no conflicts of interest.