Direct medical costs attributable to peripheral fractures in Canadian post-menopausal women
This study determined the cost of treating fractures at osteoporotic sites (except spine fractures) for the year following fracture. While the average cost of treating a hip fracture was the highest of all fractures ($46,664 CAD per fracture), treating other fractures also accounted for significant expenditures ($5,253 to $10,410 CAD per fracture).
This study aims to determine the mean direct medical cost of treating fractures at peripheral osteoporotic sites in the year post-fracture (through 2 years post-hip fracture).
Health administrative databases from the province of Quebec, Canada were used to estimate the cost of treating peripheral fractures at osteoporotic sites for the year following fracture (through 2 years for hip fractures). Included in costs analyses were physician claims, emergency and outpatient clinic costs, hospitalization costs, and subsequent costs for treatment of complications.
A total of 15,827 patients (mean age 72 years) who suffered one fracture at an osteoporotic site had data for analyses. Hip/femur fractures had the highest rate of hospital stays related to fracture (91%) and the highest rate of hospital stays associated with a post-fracture complication (8%). In the year following fracture, the mean (SD) costs (2009 Canadian dollars) of treating acute fractures and post-fracture complications were: hip/femur fracture $46,664 ($43,198), wrist fracture $5,253 ($18,982), and fractures at other peripheral sites $10,410 ($27,641). The average (SD) cost of treating post-fracture complications at the hip/femur in the second year post-fracture was $1,698 ($12,462). Hospitalizations associated with the fracture accounted for 88% of the total cost of fracture treatment.
The treatment of hip fractures accounts for a significant proportion of the costs associated with the treatment of peripheral osteoporotic fractures. Interventions to reduce the incidence of fractures, particularly hip fractures, would result in significant cost savings to the health care system and would preserve quality of life in many patients.
KeywordsCost Economic Fractures Hip Nonvertebral Treatment
Fragility fractures are associated with significant decrements in quality of life  and patient survival . Furthermore, the financial strain placed on personal resources and the greater health care system for acute and chronic care following a fragility fracture can be substantial .
Cost-models need to be developed to accurately detail the mean financial cost of fracture treatment at each skeletal site in order to fully appreciate the economic burden they represent. Information regarding the mean cost of treating fragility fractures can be used in conjunction with local incidence data to better estimate the overall financial burden of fragility fracture treatment in a given region so that proper resource allocation can be made. Furthermore, the treatment costs for specific fragility fracture sites can be entered into cost-effectiveness models, along with anti-fracture medication prescription costs, for the assessment of potential financial savings with various preventive therapies.
The purpose of this investigation was to obtain Canadian-specific data regarding the direct medical costs of peripheral fracture treatment for the year following fracture in postmenopausal women. Additionally, cost data was collected for hip fracture through 2 years from the incident fracture date since post-fracture complications from fractures at this site can commonly occur into the second year post-fracture.
The cohort of patients in this investigation was comprised of all women aged 50 years or greater who received treatment for an incident peripheral fracture between January 1, 2004 through December 31, 2005 in one of three health-administrative regions in the province of Quebec (Montreal, Quebec City, and Mauricie). This cohort of women with fractures was identified from a previous study that validated an algorithm for fracture case identification . Together, the three regions included in this study comprise approximately 42% of the total population of women 50 years and older in the province of Quebec.
In Canada, all citizens are covered by a universal health insurance plan which is administered by their province or territory of residence. In the province of Quebec, the Régie de l’Assurance Maladie du Québec (RAMQ) maintains computerized administrative databases of physicians’ fee-for-service claims containing the following information: patient unique identification number, patient demographic information (age, gender, social and material deprivation indexes), the medical services billing codes, the date and location of the clinical service provided, and, optionally, a diagnosis coded by the ninth revision of the International Classification of diseases, Clinical Modification (ICD-9-CM). The RAMQ also maintains the hospitalization discharge databases (Med-Echo) for all patients in Quebec containing information on hospital admission using 16 diagnosis codes (ICD-9-CM) before April 1, 2006 and 25 diagnosis codes (ICD-10-CA) thereafter. For this study, the RAMQ and MedEcho databases were linked through an anonymous unique identifier to identify fracture cases and to provide their respective physician fee-for-service and hospitalization services and cost data.
In these analyses, cases with a peripheral fracture at an osteoporotic site during the time of interest were identified and the cost of all treatments provided to these patients related to direct medical fracture care for the year following fracture was summed; follow-up was extended 2 years in the case of hip fracture as complications with hip fractures can persist longer than 1 year.
Identification of incident fracture cases
Using the physicians’ fee-for-service claim database (RAMQ) all incident peripheral fracture cases during the period of interest were identified with the aid of an algorithm developed, validated, and published by this research team . Based on a panel of clinical experts, the algorithm was designed to first select all medical services claims potentially associated with a fracture treatment: (a) claims with medical services billing codes definitively related to fracture care (i.e., open and closed reduction), or (b) claims with medical services billing codes not limited to fracture care (i.e., immobilization, consultation, principal or follow-up visit with an orthopedic surgeon (OS), emergency physician (EP), or general practitioner (GP)) if they are combined with ICD-9-CM diagnosis codes of fracture. Using these extracted claims, the algorithm considered an incident fracture had occurred if there was at least one claim associated with (a) fracture treatment (open reduction, closed reduction, immobilization), (b) principal visit with an OS with at least one other claim, or (c) consultation with an OS with at least one other claim. The claim allowing the identification of a fracture was referred as the “index claim.” The validity of this algorithm has been previously evaluated and published .
Fractures were grouped according to the fracture site indicated by the specific medical service code of the index claim or, if not specific to a treatment of fracture, to the ICD-9-CM code. To establish the complete temporal sequence of medical care for each fracture, the algorithm identified any other claim (emergency room visit, follow-up visit, etc.) related to the same anatomical site as the fracture. The date of fracture corresponded to the date of the first claim in the temporal sequence of medical care. Finally, a 6-month period was established as the “washout period” between two clinical sequences related to the same anatomical fracture to minimize potential misclassification of the follow-up of a fracture as a new incident fracture.
Fractures identified by algorithm were grouped into fractures of the hip/femur, wrist, or other peripheral osteoporotic site (shoulder, humerus, forearm, elbow, foot, ankle, pelvis, knee, tibia, and fibula). In this study, no information was available regarding the circumstances surrounding the incident fracture. Since fragility fractures represent the vast majority of fractures in women (approximately 81% of this population ) both low- and high-energy trauma fractures were included in these analyses. Furthermore, it has been recently proposed that both low- and high-energy trauma fractures be included as outcomes in osteoporosis trials and observational studies as they similarly are associated with decrease in BMD and result in increased risk of future fracture . Similarly, pathologic fractures were not excluded since they represent a small proportion and their exclusion may lead to an underestimation of the burden related to osteoporotic fractures .
Women that had multiple fractures, either suffered at the same or different times, during the assessment period were eliminated from the analyses as it was often impossible to accurately ascertain the resources utilized for each fracture event independently. Since fractures of the coccyx, vertebra, or sacrum were unreliably identified by the algorithm, they were also eliminated from these analyses.
Identification of health resources used
To establish the direct medical cost related to each fracture, health resources used were identified in both physicians’ fee-for-service claims and hospitalization discharge (Med-Echo) databases. The records of each fracture cases retained were followed forward in time between January 1, 2004 and December 31, 2007 and linked in order to create a longitudinal cohort of health resources utilization following the fracture.
Selection of claims
In physicians’ fee-for-service claims database, all claims were screened to ensure they were relevant to fracture treatment. In order to prevent the addition of claims linked to a subsequent or previous fracture type, claims were only retained if they were billed with a medical service billing code or ICD-9-CM diagnosis code directly related to the original fracture site or on a site considered to be affected by the original fracture (i.e., concomitant site).
Selection of hospital stays
Each fracture case requiring hospitalization related to fracture care or subsequent complications were identified in hospital discharge data. The Med-Echo databases were used to evaluate the cost related to a hospital stay.
Hospital stays definitively related to the fracture included all hospital stays with: (1) primary diagnostic codes of fracture at the same site, (2) primary diagnostic codes associated to a fracture with a secondary diagnosis code of fracture on the same sites, (3) primary diagnostic codes of fracture at a concomitant site or an unspecified site and occurring in a specified number of weeks after the fracture (≤120 days for shoulder/humerus, forearm/elbow, wrist, foot or ankle fractures, and ≤180 days for hip, pelvis, knee or tibia/fibula; the time limits were applied to avoid the inclusion of hospital stays related to a new fracture despite the probability of a re-fracture of the same site being very low), (4) primary diagnostic codes associated to a fracture and occurring in a specified number of weeks after the fracture (as above) with a secondary diagnostic code of fracture at a concomitant or unspecified site, (5) primary diagnostic codes unrelated to the fracture and occurring in a specified number of days after the fracture (up to 14 days for hip fracture and up to 10 days after fracture for other sites) with a secondary diagnostic code of fracture on the same site, a concomitant site or an unspecified site, and (6) primary diagnostic codes associated to the fracture with all the secondary diagnostic codes unrelated to the fracture.
Hospital stays probably related to a complication of the fracture included all hospital stays with: (1) primary diagnostic codes associated to a post-fracture complication with a secondary diagnostic code of fracture at the same site, (2) primary diagnostic codes associated with post-fracture complication and occurring in a specified number of weeks after the fracture (as above) with a secondary diagnostic code of fracture at a concomitant or unspecified site, and (3) primary diagnostic codes associated to complication with all secondary diagnostic codes unrelated to the fracture.
All other hospital stays were categorized into the group of stays unrelated to the fracture and were not used in the cost analysis. Appendix 2 details the selection of hospital stays.
Assessment of the direct costs of peripheral fracture
The direct costs of fracture treatment were calculated for each fracture case selected and represented the sum of the total costs related to physician claims (visit and procedure claims), the total costs related to hospital use associated with physician’s visits to the emergency and outpatient clinics (including staff and hospital support), the total costs related to hospital stay(s), and the total costs related to hospital stays for treating post-fracture complications.
To establish the total costs related to physician claims (visits and procedures), all claims considered to be associated with the fracture in the year following the fracture date were retained in the analysis. For each physician claim, the unit price per procedure was obtained from the medical specialists’ and general practitioners’ manual published by the RAMQ (October 2010 version) . Some physician claims involved the participation of an anaesthesiologist or an additional technical component (referred as R2). For those claims, the guidance manuals quantify the billable number of R2 units by type of claim, with the units multiplied by $14.20 for specialists and $13.65 for general practitioners. The cost for a claim with an R2 designation is the sum of the physician reimbursement plus the additional R2 costs. For each fracture case, the total cost related to physician claims corresponded to the sum of all unit prices related to claims made for the fracture treatment.
For each day where a claim was made for a physician’s visit to an emergency room or to a hospital outpatient clinic, a cost attributed to hospital use was estimated (staff and hospital support). If multiple physician claims were made from the same department for the same day, these costs were truncated to only one visit.
The cost of an emergency room visit was estimated at $236 per day, as calculated from a study of 357,227 emergency room visits (all-cause) of patients between 18 and 65 years of age in the Ontario Case Costing Initiative Hospitals for 2007/2008. This amount was an average cost to the hospital for the hospital space, health professionals, and administration for each patient visiting an emergency room. The mean cost to treat a fracture at an outpatient clinic visit was assumed to be $187 based on 1,298 ambulatory visits in the Ontario Case Costing Initiative Hospitals for 2007/2008 with patients 18 to 69 years of age visiting the clinic for open fixation/fusion with graft, open fixation/fusion, other closed fixation, other bone intervention, closed reduction, open reduction, removal internal fixation device, or follow-up/convalescence. Costs attributed to emergency room visits and outpatient clinic costs do not include the respective physician fees.
The cost of hospital stays and post-fracture complications was calculated by assessing the full sequence of hospital care, including transfers and readmissions, to estimate the length of hospital stays. The admission date was the first day of the sequence, except in the case of an emergency arrival in which case the date of arrival at the emergency ward was used as the initial day in the sequence of hospitalized care. Transfers were considered a continuation of hospitalization and not a new stay, and a 1-day outpatient surgery was calculated as a half day (0.5) of hospitalization. The length of stay was truncated to include only the days of hospitalization occurring in the year following fracture or in the second year following hip fracture. Finally, hospital stays were divided in two groups: hospitalizations related to the acute treatment of the fracture or to complications post-fracture.
Cost ($ CAD) of hospitalizations by fracture site based on the Ontario hospital corporate costing model (www.occp.com)
Most appropriate diagnosis used in OCCP costing analysis tool
Mean direct costa
Mean indirect costb
Mean total costc
Mean length of stay (d)
S52.0–52.4, S52.7, S52.9
S92.0–92.5, S92.7, S92.9
There were 45,585 claims associated with the treatment of hip fracture; 16,103 claims for the treatment of wrist fracture and 42,789 claims for the treatment of other peripheral fractures. A total of 10,833 hospital stays were related to acutely treating a fracture and 992 stays (9.2%) were associated with treating complications of a fracture were included in the cost analyses.
Skeletal site distribution for the 15,827 fractures retained in the analysis and the published respective positive predictive value and sensitivity from algorithm 
Number of fractures
% of total
Positive predictive value
Percentage of patients having at least one hospital stay related to a fracture or to a complication within the year following the fracture according to fracture group
Fractures at other peripheral sites
N = 15,827
N = 4,536
N = 3,157
N = 8,134
Mean (SD) patient age in years (total cohort)
Stay related to the fracture
Stay associated to a post-fracture complication
Stay related to the fracture or associated to a post-fracture complication
Mean direct medical cost of treatment for each fracture site grouping in the year post-fracture (in 2009 CAN$)
All fractures (1st year)
Hip/femur fractures (1st year)
Wrist fractures (1st year)
Fractures at other sites (1st year)
Hip/femur fractures (2nd year)
N = 15,827
N = 4,536
N = 3,157
N = 8,134
N = 4,536
Mean cost related to physician claims (SD)
Mean cost related to ER and OC (SD)
Mean cost associated to hospital stays related to the fracture in the year following fracture (SD)
Mean cost associated to hospital stays related a complication post-fracture (SD)
Total mean cost (±SD)
Total costs by fracture group and claims grouping
All fractures (1st year)
Hip fractures (1st year)
Wrist fractures (1st year)
Other osteoporotic peripheral fracture sites (1st year)
Hip fractures (2nd year)
Emergency room and outpatient clinics
Hospitalizations associated to the fracture
Hospitalizations associated to a complication
This investigation assessed the treatment costs associated with fractures at osteoporotic sites in women aged 50 years or greater from the province of Quebec. Data from almost 16,000 fractures were used to populate cost calculations to allow for estimates to be made for the mean cost of fracture treatment over a 1-year period. Unsurprisingly, the greatest costs, greatest percentage requiring hospitalization, and longest hospitalizations were observed with treatment of hip/femur fractures. However, other peripheral fractures also accounted for a significant financial burden. In an investigation from the USA, it was concluded that while the mean cost of treating hip fractures was greater than the mean cost of treating non-hip, nonvertebral fractures, the higher prevalence of the latter forms of fracture as compared to hip fractures (11:1 ratio, respectively) resulted in a greater cost to the health care system over the ages of 50–64 years of age . In the older age group (65 or older), the ratio of non-hip, non-vertebral fractures to hip fractures was 2:1, with the treatment of hip fractures then accounting for 52% of the financial burden. Similarly, in another assessment of the economic burden of fractures in the USA, while the incidence of fracture was highest in the vertebrae and wrist (27% and 19%, respectively), the cost of treating hip fractures was 72% of the total economic burden (14% incidence) . The incidence of vertebral fracture was not assessed in this investigation, but the cost of treating hip fractures was similarly the majority of the burden.
In comparison, a Belgian investigation from the year 2000 calculated the direct medical costs of treating a hip fracture in 170 women over a 1-year period and concluded that the in-hospital cost to treat hip fractures was 8,667 Belgian francs with costs attributed to later complications being 6,636 Belgian francs. In 2006, Borgstrom et al.  estimated the mean 1-year cost of treating hip, vertebral, and wrist fractures at 14,221 Euros (∼$19,000 CAD), 12,544 Euros (∼$17,000 CAD), and 2,147 Euros (∼$3,000 CAD), respectively (at $1 CAD = 0.74 Euros). An investigation of hip fracture cost in adults over 50 years of age in the USA concluded that between 2001 and 2004, the mean adjusted cost of treating a hip fracture was US $15,196–$26,545, depending on the age of the patient (65 or older or below 65 years, respectively) . However, comparisons among countries with differing medical and administrative structures and from different time periods are fraught with difficulty, further highlighting the need for regional cost data with respect to the treatment of fractures.
The total mean cost of treating all non-hip peripheral fractures (excluding wrist) over a 1-year period was $10,410 ($CAN 2009). When taken as a whole, these peripheral fractures cost the medical system about $42 million annually in these three regions of Quebec (42% of population), which extrapolated out to the total provincial population would represent a cost of more than $101 million annually (Fig. 3). If then applied to the Canadian population (Quebec assumed to be 23% of Canadian population), the annual cost of treating non-hip peripheral osteoporotic fractures could be conservatively estimated to be $435 million. In a similar study from the USA, the mean adjusted cost of treating a non-hip peripheral fractures in patients aged 50 years or greater over a 1-year period was US $6,106–$9,183 (2004 dollars), depending on the patient’s age (65 or older or below 65 years, respectively) .
Lastly, the total mean cost of treating a wrist fracture over a one-year period was $5,253 ($CAN 2009). When taken as a whole, wrist fractures cost the medical system about $8.3 million annually in these three regions of Quebec (42% of population), which extrapolated out to the total provincial population would represent a cost of about $20 million per year for postmenopausal women over the age of 50 years. The conservatively estimated costs for wrist fracture treatment in the year following fracture in all of Canada (Quebec assumed to be 23% of Canadian population) for postmenopausal women over the age of 50 years would be about $85 million annually (Fig. 3).
There are a number of limitations that need to be identified concerning the methodology used to identify incident fracture cases. First, no procedure for eliminating fractures related to motor vehicle accidents (MVA) or malignancies was used. However, studies have reported that the rate of MVA and malignancy-related fractures is low. Of note for these analyses, the ROCQ study, performed in a subset of the population used in these analyses, demonstrated that the majority of fractures (81%) occurring at an osteoporotic site are fragility fractures . However, the management of traumatic and pathologic fractures may differ significantly and result in significantly different costs as compared to the treatment of fragility fractures. Second, the algorithm used was not reliable in the identification of vertebral fractures due to the very low sensitivity calculated. Therefore, the results of this study confirm that vertebral fractures cannot be assessed by administrative claims databases, most likely due to the fact that they are not reliably diagnosed by health care professionals. Third, the extrapolation of the fracture cost to Quebec and Canada are conservative extrapolations as all incidences of multiple fractures were excluded from the calculations. Furthermore, differing health care systems from different provinces will have slightly different costs associated with different forms of post-fracture care and this would change costing models. Another limitation was the cost calculation for hospitalization after a fracture. The mean cost was calculated by dividing the total cost by the mean stay (12 days for hip fracture). However, the cost of acute care is higher at the beginning and decrease thereafter, but for this analysis, the same mean cost per day for patients was used independent of the length of stay. Therefore, patients with very long stays may have increased the mean cost estimated in the analysis. The costs associated with pharmacological treatment of fracture (analgesics, antibiotics, etc.), orthopedic implants, and rehabilitation were not included in this model; accordingly, the costs of orthopedic implants were not included in these analyses. The average costs calculated here would underestimate the true cost of the fracture, particularly those of the hip. Lastly, the data extrapolated in this investigation relates to the cost of a fracture event in women and may not represent the same cost proportion in the male population of Canada.
In summary, hip fractures place a significant financial burden on health care budgets, with an approximate yearly cost of a quarter billion dollars. However, other peripheral fractures also accounted for substantial costs with about 32% of the total expenditures being spent for the treatment of these non-hip peripheral fractures. With these estimated costs of different osteoporosis fractures, it is now possible to construct further models detailing resource allocation and cost effectiveness of preventive therapies.
Conflicts of interest
Dr. Bessette has received research grants from Abbott, Amgen, Bristol-Myers-Squibb, Eli Lilly, Merck, Pfizer, and Roche, has received consulting fees or other remuneration from Abbott, Amgen, Merck, Novartis, Pfizer, and Roche and has participated on the Speakers Bureau for Amgen, Novartis, Merck, Pfizer, Roche, and Warner Chilcott. Dr. Brown has received research grants from Abbott, Amgen, Bristol-Myers Squibb, Eli Lilly, Pfizer, and Roche, has received consulting fees or other remuneration from Abbott, Amgen, Eli Lilly, Novartis, Merck, and Warner Chilcott and has participated on the Speakers Bureau for Eli Lilly, Amgen, Novartis, Merck, and Warner Chilcott. Dr. Davison has received consulting fees or other remuneration from Amgen and Servier and has participated on the Speakers Bureau for Amgen, Merck, and Servier. Dr. Ste-Marie has received research grants from the Alliance for Better Bone Health and Novartis, has received consulting fees or other remuneration from the Alliance for Better Bone Health, Amgen, Novartis, Eli Lilly, and Servier and has participated on the Speakers Bureau for the Alliance for Better Bone Health, Amgen, Novartis, Eli Lilly, Servier, and Merck. No other authors have a conflict or interest to disclose.
- 1.Papaioannou A, Kennedy CC, Ioannidis G, Sawka A, Hopman WM, Pickard L, Brown JP, Josse RG, Kaiser S, Anastassiades T, Goltzman D, Papadimitropoulos M, Tenenhouse A, Prior JC, Olszynski WP, Adachi JD (2008) The impact of incident fractures on health-related quality of life: 5 years of data from the Canadian Multicentre Osteoporosis Study. Osteoporos Int 20:703–714PubMedCrossRefGoogle Scholar
- 2.Ioannidis G, Papaioannou A, Hopman WM, khtar-Danesh N, Anastassiades T, Pickard L, Kennedy CC, Prior JC, Olszynski WP, Davison KS, Goltzman D, Thabane L, Gafni A, Papadimitropoulos EA, Brown JP, Josse RG, Hanley DA, Adachi JD (2009) Relation between fractures and mortality: results from the Canadian Multicentre Osteoporosis Study. CMAJ 181:265–271PubMedCrossRefGoogle Scholar
- 4.Jean S, Candas B, Belzile E, Morin S, Bessette L, Dodin S, Brown JP (2011) Algorithms can be used to identify fragility fracture cases in physician-claims databases. Osteoporos Int (in press)Google Scholar
- 8.(2011) RAMQ medical specialists’ and general practitioners’ manual. http://wwwramq.gouv.qc.ca. Accessed 2 February 2011
- 11.Piscitelli P, Iolascon G, Gimigliano F, Muratore M, Camboa P, Borgia O, Forcina B, Fitto F, Robaud V, Termini G, Rini GB, Gianicolo E, Faino A, Rossini M, Adami S, Angeli A, Distante A, Gatto S, Gimigliano R, Guida G (2007) Incidence and costs of hip fractures compared to acute myocardial infarction in the Italian population: a 4-year survey. Osteoporos Int 18:211–219PubMedCrossRefGoogle Scholar