Estimating the future clinical and economic benefits of improving osteoporosis diagnosis and treatment among postmenopausal women across eight European countries

Osteoporosis is a chronic condition characterised by low bone mineral density (BMD) and deterioration of bone tissue, leading to increased susceptibility to fracture [1]. It primarily occurs in postmenopausal women [2]. It was estimated that 22% of women aged 50 years or older in 27 countries of the European Union (EU) plus the UK and Switzerland had osteoporosis in 2019 [3]. Furthermore, given that the prevalence of osteoporosis increases with age, the clinical and economic burden of osteoporosis in countries with ageing populations is expected to rise considerably over the coming decades [2, 4]. A recent study analysing the burden and management of fragility fractures in France, Germany, Italy, Spain, Sweden and the UK estimated that the total numbers of fractures for both men and women will increase by 23% in the next decade, from 2.7 million in 2017 to 3.3 million in 2030 [4]. Over the same period, annual fracture costs are expected to increase by 27%, from €37.5 billion to €47.4 billion [4].

Optimal diagnosis and treatment of osteoporosis, in line with clinical practice guidelines [5, 6], may help reduce this significant burden on healthcare systems. However, real-world data highlight that osteoporosis remains under-diagnosed and untreated [4, 7]. Results from a large, cross-sectional, observational European study showed that over two-thirds (68%) of women aged 70 years or older who had an increased risk of fracture had not been diagnosed with osteoporosis by their primary care physician, and three-quarters (75%) were not receiving any anti-osteoporosis medication [7]. These findings are supported by the results of a separate study conducted in France, Germany, Italy, Spain, Sweden and the UK, which estimated the osteoporosis treatment gap (i.e. the proportion of eligible women not receiving treatment with osteoporosis drugs) to be 73% in 2017 [4].

In addition to addressing diagnosis and treatment gaps, improving patient adherence to anti-osteoporosis medications may also help reduce the high burden of osteoporosis and fractures. An expert group of the European Society for Clinical and Economic Aspects of Osteoporosis, Osteoarthritis and Musculoskeletal Diseases and the International Osteoporosis Foundation determined that medication non-adherence is associated with an increased risk of fracture (for example, relative risk of 1.28 for hip fractures and 1.43 for vertebral fractures) [8], leading to a substantial decrease in the clinical and economic benefits of drug therapy [9]. This is an important consideration, as a systematic review of factors affecting medication adherence among patients with osteoporosis reported a wide variation in medication adherence rates, with some as low as 13% or as high as 95% [10]. Overall, the increasing clinical and economic burden of osteoporosis, combined with suboptimal diagnosis and under-utilised anti-osteoporosis medications, represents an urgent unmet need, and improvements in the management of this chronic condition are warranted.

The primary aim of this modelling analysis was to compare projected clinical and economic outcomes for osteoporosis under different hypothetical interventions. Various rates of risk assessment and adherence to anti-osteoporosis medications were used to assess whether these interventions would lead to net benefits on numbers of incident fractures and total costs of care. These interventions could then be used to recommend optimal management approaches. Eight European countries (Belgium, France, Germany, Ireland, Poland, Slovakia, Switzerland and the UK) were included and modelled in these analyses.

Using previously published data and modelling approaches [7, 11], a population-level, cross-sectional cohort model was developed to estimate the clinical and economic burden of osteoporosis (specifically, the annual incidence of osteoporotic fractures and related costs of care). The study populations were women aged 70 years or older across eight European countries. The model predicted outcomes over a 20-year period (from 2020 to 2040) under different hypothetical interventions with various sets of improvement from the status quo, which was based on existing disease management patterns reported by McCloskey et al. [7]. The three hypothetical interventions were assessed: intervention 1, an improvement in risk assessment rate (which would have a subsequent impact on the number of patients receiving anti-osteoporosis medications); intervention 2, a reduction in rates of non-adherence to anti-osteoporosis medications and intervention 3, a combination of interventions 1 and 2.

In this study, we developed a cross-sectional cohort model to project the annual clinical and economic burdens of osteoporotic fractures in women older than 70 years old across eight European countries over a 20-year period (2020–2040). This model was then used to assess the potential impact of various improved rates of risk assessment and/or adherence to anti-osteoporosis medications on these outcomes. Both the annual number of fractures and annual cost of care were predicted to increase by over 40% from 1.2 million fractures (€12.8 billion) in 2020 to 1.8 million fractures (€18.4 billion) in 2040 should the status quo remain. This is primarily due to a projected 40% increase in the number of women aged 70 years or older during this time frame, because it is reasonable to assume increased ageing populations in Europe. Results from both the main analysis and scenario analysis suggest interventions aimed at improving risk assessment and adherence to anti-osteoporosis medications would have a considerable impact in reducing the burden of osteoporotic fracture. In particular, interventions with 50% improved rates in risk assessment and adherence to anti-osteoporosis medications would lead to a 15.2% cost reduction, totalling nearly €2.8 billion for these eight European countries in 2040.

The substantial net benefits (i.e., reduction in numbers of fractures and cost savings) estimated by our model support the need for new policy measures and/or increased efforts to improve the assessment of osteoporotic fracture risk and adherence to anti-osteoporosis medications. The introduction of such interventions should ensure that more patients at high risk of osteoporosis are screened in a timely manner to diagnose osteoporosis. This would lead to initiation of appropriate anti-osteoporosis medications at the time when it is needed by the patient and advocate that patients remain adherent to treatment. In addition, high-quality data from registries of patients with osteoporosis and fragility fractures are needed to allow for vital national and local feedback. Importantly, considerations on how these data can be used to improve quality-of-care should be contemplated. Health education programmes for patients at risk of, or with, osteoporosis should also be considered to improve disease awareness, patient confidence and treatment adherence. Policy makers, payers and healthcare providers (e.g. primary care physicians) will be required to play key roles in the design and implementation of health education programmes for patients at risk of osteoporosis.

The results from our model are consistent with those of other similar analyses. However, unlike our model, the majority of previous studies only use data from individual countries. Previous studies also tend to present absolute values rather than estimate percentage changes in the number of fractures or total costs of care over time. Furthermore, these studies tend to focus on the impact of risk assessment without accounting for treatment adherence [4, 11, 14‐16]. In our model, we have estimated fracture risk and the clinical and economic burden individually for each country using country-specific data, then aggregated to produce overall results. Thus, the results from this study could be considered more generalisable to the wider real-world population of patients than previous studies which only use data from individual countries. For the status quo, our model estimates are supported by a recent analysis conducted by Borgstrom et al., which estimated a 23% increase in fractures and a 27% increase in fracture-related costs in France, Germany, Italy, Spain, Sweden and the UK over a 13-year period [4]; our model estimates an approximate 40% increase over a 20-year period. With respect to the positive impact of improvements in risk assessment, studies in the USA, South Korea, Japan and China have all reported substantial reductions in clinical and economic burden, with estimated cost savings ranging from US$3.1 billion (Japan) to US$61.7 billion (China) over approximately 20-year periods [4, 11, 14‐16]. The consistency of these findings across studies in different regions reflects the clear clinical and economic benefits associated with improvements in the management of osteoporosis.

As with any study, there are limitations that should be considered when interpreting the results. For example, while our population size projections considered dynamic changes in demographics, including age, over the model time horizon when estimating fracture risks using the FRAX tables, other risk factors such as smoking and prior fracture (which may also change over time) were not accounted for as separated model inputs in this analysis. Nevertheless, as the FRAX approach is still considered a reasonable and reliable estimation of fracture risk and is included in current osteoporosis diagnosis and management guidelines [5], the omission of some risk factor information would not be expected to have a significant impact on the final estimates. Furthermore, our model did not allow for individuals to incur multiple fractures of the same type within a single year, though it did allow an individual to experience up to three fractures in different locations (one each of hip, vertebral and non-hip, non-vertebral). This may lead to an underestimation of the potential burden of osteoporosis. Similarly, it was assumed that healthcare costs would remain constant, potentially underestimating projected spending growth. In addition, first year direct medical costs were used to estimate the annual costs of care. Substantial costs may occur after the first year of fracture with an increased probability of patients requiring nursing home care [17]. The model did not include costs related to the broader indirect economic and social burden of osteoporosis, such as lost productivity or reduced tax revenue as a result of productivity loss, paid and unpaid caregiver burden and home/environment adaptations. All these factors mean the model estimates are likely conservative. Indeed, in Europe, it has been shown that approximately one-third of costs related to osteoporosis are indirect, and considering the broader economic and social burdens of osteoporosis, this estimate of indirect costs is likely to be conservative [3]. Thus, the potential clinical and economic value of efforts to improve outcomes is likely to be greater than presented here. Another potential limitation of our analysis is the lack of consideration of adverse events associated with treatment. However, compared with the substantial burden associated with fractures, this would likely have minimal impact on the model results. An additional limitation was that vertebral fracture assessment was not included in our cost estimation. Studies have shown that vertebral fracture assessment may be cost effective when it is incorporated in routine screening for osteoporosis [18, 19]. Finally, the impact of increased treatment in men was not considered in our model. Osteoporosis also occurs in men aged 70 years or older [2]. It was estimated that 6.6% of men aged 50 years or older in 27 EU countries plus the UK and Switzerland had osteoporosis in 2019 [3]. Inclusion of men in this analysis would likely have increased the number of potential fractures prevented, and this is an area that should be explored in future research.

Regarding assumptions inherent in the model, assuming that all women aged 70 years or older had a BMD measurement when they underwent fracture risk assessment is likely a conservative approach, since less costly risk assessment methods, such as using clinical judgement or algorithms based on risks produced by FRAX, are also used in clinical practice. BMD measurement is often reserved for cases in which a treatment decision is uncertain. As such, potential cost savings based on improving risk assessment rates would likely be greater than those presented here. The model also assumed that the fracture reduction benefits of treatment in clinical practice are equivalent to those reported in the meta-analysis of randomised controlled trials conducted by Barrionuevo et al. [13]. In the real world, treatment effects may differ from those observed under controlled trial conditions, and future real-world evidence may provide further insights that could help improve the predictive power of the model.

In conclusion, our cross-sectional model highlights the substantial future clinical and economic burden of osteoporosis and osteoporotic fractures in postmenopausal women aged 70 years and over. This model also demonstrates that interventions aimed at improving osteoporosis care, specifically fracture risk assessment and adherence to anti-osteoporosis medications, would help to relieve this burden.