Incremental costs of fragility fractures: a population-based matched -cohort study from Ontario, Canada

Due to its increasing prevalence and associated costs, fragility fractures are a major public health issue. Worldwide, the number of individuals with fractures due to osteoporosis has increased from 56.2 million to 157.4 million between 2000 and 2010 and this number is expected to increase at least twofold until 2040 [1, 2]. Despite variations in the incidence rates of individual fracture types across countries, the pattern of fractures appears to be similar in all developed countries examined, with major osteoporotic fractures comprising the majority of all recorded incidents [3]. Fragility fractures impose a substantial economic burden, estimated at €37.5 billion in the largest 5 countries of the European Union plus Sweden (EU6) in 2017 and $19 billion in the USA in 2005 (equivalent to $42 billion and $26 billion respectively, in 2020 USD values) [4, 5]. Cost of illness studies comparing patients with and without fractures have been conducted using claims data from the USA [6, 7], and healthcare records from the UK [8, 9], the Netherlands [10], and Belgium [11]. These studies have demonstrated that treatment and care of patients with fractures imposes a substantial burden on healthcare systems compared to patients of a similar age and sex without fractures, emphasizing the importance of primary and secondary prevention in high-risk individuals.

However, these studies were not Canadian and are not useful to support economic evaluations of osteoporosis in Canada [12]. While several studies have documented the costs associated with osteoporosis in Canada [13‐16], very few recent studies have evaluated the incremental costs associated with fragility fractures. One of the first studies, using 1995–2008 administrative data from the province of Manitoba (4% of Canadian population), estimated the incremental costs associated with fractures, with hip fracture having the highest incremental cost over a 2-year period of approximately $45,000 per patient [14]. A more recent study using data from Ontario (40% of Canadian population) reported that the first-year incremental direct healthcare costs associated with hip fracture were $282 million in Ontario or $1.1 billion in Canada in 2010 [15]. However, this study was limited to hip fractures and the current incremental burden associated with all types of fragility fractures in Canada is unknown. To better inform policy makers, payers, clinicians, and patients, the primary objective of this study was to document the incremental costs associated with fractures.

This was a retrospective population-based observational study using de-identified health services data of patients in the province of Ontario (population 14.7 million) from the ICES Data Repository [17].

Using a large population-based incident fracture cohort matched to a non-fracture cohort, our analysis showed that patients with fragility fractures incurred an excess 1-year direct healthcare cost of $26,341 across all resource utilization categories compared to those without a fracture. The main cost drivers were hospitalization and continuing care (hospital-based continuing care, home care, and LTC) across the study period. The first-year incremental healthcare costs of incident fragility fractures for the province of Ontario were estimated at $724 million, which translated into an excess $1.9 billion annually when extrapolated to Canada. Although the costs varied by fracture site, the costs generally increased with age and number of comorbidities. Costs were highest in the first year following index fracture (mean $39,089 ± SD $43,272 per patient) but remained above $12,000 in years 2–5 following the index fractures.

Comparisons with other Canadian studies are difficult due to differences in time period, databases used, types of healthcare resource utilization captured, and characteristics of patients included in the study cohorts or control groups. Nonetheless, our estimates are higher than those reported before. For example, using a similar methodology and data sources, the 1-year cost in the first year after incident hip fracture in Ontario was estimated at $52,000–$54,000 based on 2004–2008 data [15]. In contrast, we reported a first-year cost of $62,793 per patient with an incident hip fracture using 2011–2015 data from Ontario. A study from Manitoba estimated the 2-year cost of hip fractures at approximately $57,000 and the incremental costs of hip fracture at approximately $45,000 using data from 2007/2008 [14]; however, this study did not provide the data for year 1 and year 2 separately and the results were based on a small number of hip fractures (N = 751). A more recent study from Manitoba [20] used data from 1997 to 2002, but results were expressed in medians making direct comparisons difficult with our study. Although not evaluating the incremental costs of fragility fractures, a Canadian study using national administrative databases estimated the first-year direct medical costs associated with fragility fractures at approximately $25,000 per patient (compared to $39,089 in our study) [16]. While the study reported that the annual healthcare costs of fragility fractures were $4.6 billion (including $400,000 for wage loss and mobility devices), these figures were based on incident and prevalent fractures that occurred in 2011 (as opposed to incident fractures in our study). Matched cohort analyses conducted in Europe and the USA have shown the excess costs of hip fracture in the first year [6, 8‐10, 21], with incremental annual costs as high as $50,508 USD per patient in a US Medicare population [7]. Differences in study designs, healthcare systems and clinical practice patterns across countries, and the publication of these data across more than 10 years make it difficult to compare results between studies. However, regardless of the geographic region, there are substantial costs imposed on the healthcare system. While the excess costs were attributed to hospitalization costs in some countries [8, 10], they were largely due to nursing home stays, rehabilitation center stays, and home physical therapy services in others [21]. The excess costs are relevant for healthcare providers to design and implement effective strategies to prevent fractures in high-risk individuals.

Despite the differences in methodology and results across previous Canadian studies and our current analysis [13, 14, 16], the substantial economic burden that fragility fractures place on LTC and rehabilitation services is evident. The current study supports the observation that a large proportion of healthcare costs in the years following a fracture are associated with LTC, which reflects a global trend of an increase in aging populations and an expected increase in the use of LTC [16]. Cost-effective strategies for continuing care services are especially important for older patients since there is a high risk of mortality associated within the first year after experiencing a fragility fracture [22]. Additional surveys are required to monitor the effectiveness of guideline-recommended prevention strategies, especially in older residents.

This study has some limitations which should be noted before drawing any conclusions from the data. The cohort was limited to patients aged >65 with public drug coverage (ODB Program) from Ontario and a third of the cohort was older than 85 years which further limits generalizability. The study excluded patients who had experienced a fracture within 5 years prior to the index event but not beyond those 5 years; therefore, the cohort was potentially biased towards an older population. In this incidence-based analysis, patients were followed longitudinally from their index date and approximately one-third of patients had follow-up data available for the full 5-year period. Also, the excess costs of fracture beyond the first year have not been estimated as this was beyond the scope of this study. Another area of future research is to estimate the excess costs in a fixed length cohort (e.g., 5-year survivors). We acknowledge that there is some uncertainty in extrapolating the costs to Canada, as Ontario unit costs may not be representative of costs in other provinces and the population distribution of those 65 years and over varies across other provinces (e.g., from 14% in Alberta to 22% in New Brunswick; average of 18% in Canada) [23]. It should be however noted that the population distribution of those 65 years and over is the same in Ontario as Canada overall (i.e., 18%), which minimizes the potential bias in extrapolating the Ontario results to Canada based on population data. There is also potentially an underestimation of the number of fractures in the cohort studied, specifically with vertebral fractures, considering that we captured “clinical vertebral fractures” that were identified from hospital/emergency/outpatient clinics and as a result would underestimate those managed in the community (i.e., general practitioner). The cost of prescription drugs only included medications covered under ODB and may underestimate the costs of drugs dispensed at hospitals as these were not captured in the analysis. Direct costs were only included from a public payer perspective and do not include any costs covered through private insurance plans or paid out-of-pocket by the patient (for example non-OHIP funded physiotherapy, occupational therapy). Indirect costs (i.e., productivity loss, absenteeism, presenteeism) were not included and limits the contribution at a societal perspective. The healthcare costs included in this analysis may not be related to the index fragility fracture and could reflect unrelated conditions. No statistical analyses were performed to compare the differences between the matched cohorts and the differences may not be statistically significant. Furthermore, additional analyses would be required to evaluate costs and healthcare resource utilization in other provinces to attain a more holistic understanding of the economic burden of osteoporotic patients in Canada.

Despite these limitations, the strengths of the present study are the large sample size and the matched cohort analysis. The study captured all residents of Ontario older than 65 years with any type of fragility fracture and their associated healthcare resource utilization over a 5-year period in the largest province in Canada. Patients were matched from the date of index fracture by age, sex, and comorbidities. By conducting a matched analysis with a non-fracture cohort, this mitigated some general limitations observed in previous studies, such as being able to attribute subsequent healthcare resource utilization and costs to the fracture. An added value of the present study is the evaluation of costs up to 5 years following a fragility fracture, as few studies have evaluated the longer term costs beyond the first year or two [20].

Fragility fractures pose a substantial economic burden to the healthcare system. This is expected to continue to increase with an aging population in Canada. Implementation of improved fracture preventive strategies can result in overall cost savings. Further research is required to determine appropriate management of osteoporotic fracture patients and prevention methods to decrease the burden of fragility fractures on the Canadian healthcare system.