Updated 2017


Q1. Have the intervention thresholds been determined using health economic analyses?
A1. Health economic data were not used to set intervention thresholds. Rather, the intervention thresholds were based on previous RCP guidelines, which recommended consideration of treatment in a postmenopausal woman with a prior fracture. The reason that the threshold flattens from 70 years upwards is to ensure equality of risk for those identified with or without a previous fracture, as the latter were slightly disenfranchised by the previous thresholds (for more details, see Osteoporos Int. 2015 Aug;26(8):2091-9. doi: 10.1007/s00198-015-3176-0. Epub 2015 Jun 16). It so happens that the NOGG thresholds are cost-effective even using a willingness to pay threshold as low as £20,000/QALY.
Q2. Why are there different thresholds for BMD testing and treatment based on age, given that FRAX already includes age as a risk factor? Women aged 70 and over have to have a much higher risk of having a fracture compared to a 50 year old in order to access treatment. Does this not discriminate against the elderly?
A2. The age-dependent intervention thresholds are based on RCP and European guidelines. Up to the age of 70, the fracture probability thresholds are set at the values equivalent to a postmenopausal woman of that age with a prior fracture. Despite an apparently higher threshold, the proportion of women eligible for treatment actually increases with age. The reasons for the plateau in the threshold from age 70 upwards are explained in the answer to Q1.
Q3.Why not use a single intervention threshold for all ages?
A3. If a single intervention threshold were to be used, this would result in under-treatment of younger people and over-treatment of older people. It is important to ensure that younger patients at risk of fracture are not neglected, but their inclusion has not been at the expense of older individuals.

Treatment with and without DXA

Q4. Has the availability of DXA has been considered in setting the upper and lower assessment thresholds? Is this not a form of rationing?
A4. The availability of scanners was not considered in developing the strategy for the UK. It is true that FRAX® can be used without BMD, and indeed the performance characteristics of FRAX® without BMD are about as good as BMD alone in fracture prediction. The decision to recommend BMD in a segment of the population, rather than the whole population is that, with the exception of North America, no agency including government agencies in the UK, currently supports the use of population screening. This is because of the low sensitivity of BMD for fracture prediction. The assessment thresholds chosen by NOGG are based on minimising the likelihood of incorrectly misclassifying a high risk patient as low risk and minimising the likelihood of incorrectly misclassifying a low risk patient as high risk (for more details, see http://onlinelibrary.wiley.com/doi/10.1359/jbmr.2004.19.6.906/full). Of note, a 5 year study of population screening of women aged 70-85 years using the FRAX tool has recently been completed in the UK; the screening arm which targeted treatment to those at high risk of hip fracture showed a 28% reduction in hip fracture incidence compared to a control arm.
Q5. As this guideline will offer "best practice" advice, should not clinicians use BMD testing when they feel that this is appropriate, rather than be driven by assessment thresholds?
A5. There is nothing in the wording of our guideline that prevents a clinician from performing DXA where he or she feels that this is clinically appropriate. As with all guidelines, there has to be some flexibility and room for clinical judgement.
Q6. Should not FRAX® be used in conjunction with BMD rather than as a replacement for it?
A6. The FRAX tool incorporates BMD. The decision to do a BMD measurement depends on whether the test could influence patient management and this is more likely in individuals with probabilities that lie near or around the intervention threshold before BMD. The assessment thresholds used make incorrect classification unlikely. Having decided to treat, a physician may well order a BMD test to monitor treatment.
Q7. I am concerned about the use of treatment without a BMD test because most of the clinical trials have used treatments on the basis of BMD.
A7. It is true that the majority of the evidence base for fracture reduction with pharmacological interventions is based on women with osteoporosis or established osteoporosis. There are several reasons that indicate that responses to treatment would not be impaired in women without osteoporosis. First, there are now several trials where good treatment effects are seen with women recruited on the basis of a prior fracture, age or glucocorticoid exposure. Second, several studies have shown the expected benefits on fracture risk in women from the general population, not selected on the basis of any risk factors for fracture. Third, in the vast majority of phase III studies comparable effectiveness has been shown in subgroups with and without the clinical risk factors used in FRAX. Fourth, the vast majority of women with high fracture probability as assessed by FRAX® have low BMD. Finally, recent phase III studies have shown that the use of FRAX, even without BMD, identifies women in whom treatment is effective. The CHMP of the European Medicines Evaluation Agency have endorsed the use of fracture probability based on clinical risk factors and BMD in their revised guidelines issued in May 2007 and now demand that all new studies of pharmacological treatment enrol on the basis of fracture probability, not BMD fracture. As noted, recent phase III studies have shown efficacy of interventions for vertebral and non-vertebral fracture outcomes in patients selected on the basis of clinical risk factors.
Q8. Will treating individuals with a fracture without BMD mean that those who have severe osteoporosis with very low BMD will not be differentiated from those with normal BMD and will not undergo appropriate testing for underlying causes?
A8. A recommendation to intervene without knowledge of BMD is not the same as mandating that a BMD should not be undertaken. A clinician can still decide to request a BMD measurement where investigation is deemed appropriate or as a baseline to monitor treatment.

Centre for Metabolic Bone Diseases, University of Sheffield, UK