Abstract
Objective: To examine direct costs of hip, vertebral, and non-hip non-vertebral (NHNV) fractures, and to estimate the rate of osteoporosis (OA) diagnosis and treatment in the fracture population.
Methods: Subjects ≥ 45 years with a new hip, vertebral, or NHNV fragility (closed) fracture between June 30, 2002 and June 30, 2006 were studied retrospectively. All-cause and fracture-specific medical costs were calculated from medical claims in the 12-month baseline and follow-up periods. Total healthcare costs included pharmacy and medical costs. Diagnosis for OA and OA treatment were identified in both the baseline and follow-up period from diagnosis codes on medical claims for OA, and from pharmacy claims for treatment. Analyses were performed separately for commercial (COM) and Medicare Advantage (MA) populations and stratified by fracture type. Generalized linear models were estimated for total follow-up healthcare cost.
Results: The study sample included 36,521 COM and 10,160 MA subjects. Hip fracture subjects had the highest follow-up medical costs in unadjusted and adjusted analyses (COM: mean $35,898; median $22,945; MA: mean $32,919; median $26,047). Follow-up costs were much higher than baseline costs. Fracture-related costs accounted for a large proportion of follow-up medical costs. Although rates of osteoporosis diagnosis and treatment increased from baseline to follow-up, the majority of both COM and MA subjects had no evidence of osteoporosis diagnosis or treatment in either period.
Conclusions: Despite limitations of this study, including conventional generalizability issues, and sensitivity and specificity of claims-based diagnoses, results are consistent with other research and provide compelling results of substantial cost burden of fractures related to osteoporosis. Low rates of osteoporosis diagnosis and treatment among patients with costly fragility fractures underscore the opportunity for managed care organizations to initiate comprehensive disease management programs in osteoporosis.
Introduction
Osteoporosis is a significant public health problem. In the United States, as many as 8 million women and 2 million men have osteoporosis. An additional estimated 34 million Americans have low bone mass and are at risk of developing osteoporosisCitation[1]. In 2005, osteoporosis was associated with approximately 2 million fractures, accounting for annual national direct healthcare costs of about $19 billionCitation[2]. With the aging of the population, by 2025, the number of fractures is projected to rise to >3 million at an expense of $25.3 billion in direct costsCitation[3]. Indirect costs (e.g., disability, lost productivity for patients and family) likely add billions in costs associated with osteoporotic fractures.
The majority of studies examining the economic burden of osteoporosis focus on hip and vertebral fractures, given the relatively higher expense and resource use associated with these fracture types. However, in terms of prevalence, non-hip non-vertebral (NHNV) fractures are more common, comprising more than 50% of the total fracture occurrence in the US osteoporosis patient populationCitation[3], and should be included as part of any empirical cost assessment on osteoporotic fractures.
Currently, there is significant variation in the disease burden literature on osteoporosis. Studies have employed a variety of study designs, methodologies, and subject selection criteria, and research has employed data with variable or limited geographic representation. These inherent differences among studies can pose challenges to managed-care health organizations as they seek to apply these results to their own patient populations and translate research into reliable, time-relevant decision making. Further, prior research has not examined the rate of osteoporosis diagnosis and treatment before and after fractures typically associated with osteoporosis.
The purpose of this study was to examine direct costs of hip, vertebral, and NHNV fractures in a national managed-care population. This research was designed to address some of the limitations of prior research by including a broader array of osteoporotic fractures, a large national health plan, and representation from both commercial and Medicare-eligible populations. A secondary objective was to identify the extent to which patients who experience a fracture are diagnosed and/or treated with osteoporosis medications in order to identify potential gaps in delivery of care for patients with fractures typically associated with osteoporosis.
Patients and methods
Data source
This was a retrospective cohort study of subjects with evidence of fracture during the period from June 30, 2002 through June 30, 2006. All subjects were drawn from a large proprietary administrative claims database containing medical claims, pharmacy claims, and enrollment information for commercial and Medicare Advantage enrollees associated with a large managed-care health plan affiliated with i3 Innovus. At the time the study was conducted, the administrative claims database included data for approximately 25.5 million commercial and 931,000 Medicare Advantage enrollees. Both types of health plans are structured as discounted fee-for-service independent practice association plans spanning the United States. All study data were maintained in a de-identified manner and were accessed following protocols compliant with the Health Insurance Portability and Accountability Act (HIPAA).
Subject identification
This study included male and female subjects 45 years of ageCitation[5–9] or older experiencing a single type of hip, vertebral or NHNV fragility (closed) fracture during the period from June 30, 2002 through June 30, 2006. Fractures were identified based on the presence of a primary diagnosis on a medical claim. The diagnosis codes used to identify fractures were those used in a study by Ohsfeldt et al.Citation[5], with the addition of codes to identify vertebral, ankle, and patella fractures. The service date on the first-appearing claim for the qualifying fracture was designated as the index date, and the type of fracture on the claim was identified as: vertebral, ankle, clavicle, femur, hip, humerus, patella, pelvis, tibia/fibula, or wrist/forearm (see ).
Table 1. Fracture diagnosis codes.
To ensure these were new fractures, subjects were removed from the sample if there was a medical claim for a fracture at the same site in the 6 months prior to the index date. Further, subjects had to have at least two fracture-related outpatient visits over the 12 months following the index date. For the subset of subjects with fracture of the hip, femur, tibia, or humerus (i.e., fractures usually requiring acute care), a claim submitted on the index date for emergency services or hospitalization was also required. Finally, all subjects had to have continuous health plan coverage for a period of at least 12 months prior to the index date (baseline period) and at least 12 months following the index date (follow-up period).
Three mutually exclusive cohorts were created based on the index fracture site: hip, vertebral, and NHNV. The NHNV fracture cohort comprised subjects with fractures of the ankle, clavicle, femur, humerus, patella, pelvis, tibia/fibula, or wrist/forearm.
Study measures
Subject demographic and clinical characteristics
Demographic variables for all study subjects were captured from health plan enrollment data and included subject age and sex. Whether subjects had a diagnosis of osteoporosis in the baseline and follow-up periods was determined from medical claims based on the presence of diagnosis code 733.0x in either a primary or secondary position on the claim. Similarly, whether subjects filled prescriptions for osteoporosis medications was determined separately for both baseline and follow-up periods. Medications considered were alendronate, risedronate, teriparatide, ibandronate, and raloxifene; they were identified based on NDC codes on retail pharmacy claims.
Cost and resource utilization measures
Medical costs were calculated from medical claims for all services and for fracture-specific services in the baseline and follow-up periods to allow for examination of change in costs in the year following fracture. Medical costs were also calculated in the first month following the fracture to determine the proportion of costs in the year following fracture incurred in the 30 days immediately following the event. Medical costs were computed from all medical claims and by service type: inpatient, outpatient, and other (e.g., emergency department, home healthcare). Costs for retail pharmacy services were calculated from retail pharmacy claims. Total healthcare costs were computed as the combined pharmacy and medical costs.
All costs were computed as combined health plan-paid and subject-paid amounts and were adjusted to 2006 dollars using the annual medical care component of the Consumer Price Index (CPI) to reflect inflation over the study periodCitation[4]. The Chained CPI was used, reflecting substitutions between goods that consumers might make due to price increases.
Fracture-related costs were calculated from claims with a fracture diagnosis in the primary position, or a procedure associated with fracture coded on the medical claim. The procedure codes were based on the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) procedure codes used in Huybrechts et al. (2006)Citation[10] as well as Current Procedural Terminology (CPT) codes used to identify fractures in a study by Weycker et al. (2007)Citation[11]. All codes from the prior studies were reviewed and updated to account for any changes in coding from the years used in these studies and to include additional codes determined to be relevant for this study. Codes added or deleted from those used in the prior studies are shown in .
Table 2. Procedure codes added or removed for study.
Table 3. Subject demographics.
Table 4. Inpatient, outpatient and pharmacy costs in the year prior to fracture.
Statistical analyses
All analyses were performed separately for subjects in the commercial population and the Medicare Advantage population. Results were stratified by fracture type. Numbers and percentages are provided for dichotomous measures, and means, standard deviations, 95% confidence intervals, and medians were produced for continuous measures. Generalized linear models were estimated for the commercial and Medicare Advantage populations to examine factors associated with total healthcare cost in the year after fracture. Total healthcare cost was modeled as a gamma distribution. Each model controlled for fracture cohort, age, gender, geographic region, Charlson comorbidity score, and baseline diagnosis and treatment of osteoporosis. Data extraction and statistical analyses for this study were performed using SAS, version 9.1 (SAS Institute, Cary, NC) and STATA, version 10.0 (StataCorp, College Station, TX).
Results
Subject characteristics
The final study sample consisted of 46,681 subjects, of which 36,521 (78%) were in the commercial population and 10,160 (22%) were in the Medicare Advantage population. depicts the sample attrition for this study.
Figure 1. Sample attrition.
Demographic characteristics for the study population by fracture are shown in . In the commercial population, there were approximately twice as many females as males (64 vs. 36%), while in the Medicare Advantage population, 77% were female and 23% were male. The average age of study subjects was 58 years for the commercial population and 76 years for the Medicare Advantage population. In the commercial population, most subjects were in the younger age brackets, with 43.6 % of the subjects aged 45–54, 36.7% aged 55–64, and 19.8% aged 65+. For the Medicare Advantage population, 5.9% of subjects were aged 45–64, 31.4% aged 65–74, and 62.7% of subjects aged 75 or older.
NHNV made up the largest cohort in each population and comprised a higher percentage of total fractures in the commercial vs. Medicare Advantage population, 79 vs. 58%, respectively. Hip fracture and vertebral fracture were more common in the Medicare Advantage population (16.8%, 25.5%) than in the commercial population (4.7%, 16.1%).
Post-fracture total medical costs
Subjects with hip fracture had the highest total mean and median medical costs in the 12 months following fracture both for the commercial ($35,898; CI $33,305–38,492; median $22,945) and Medicare Advantage ($32,919; CI $31,743–34,095; median $26,047) populations compared to subjects with vertebral or NHNV fractures (). This relationship persisted even when costs were adjusted for other factors, including demographic characteristics, comorbidity score, and osteoporosis diagnosis and treatment, using multivariate techniques. In the commercial population, compared to total cost after fracture for hip fractures, total cost was significantly lower for those with vertebral fractures (68% compared to hip fracture cost) and lowest for those with non-vertebral, non-hip fractures (43% compared to hip fracture cost, ). The relationship among fracture types was similar in the Medicare Advantage population (). Among commercial subjects, other significant covariates were gender, geographic region, and baseline comorbidity score. Total cost, on average, was 27% higher for each additional point of the comorbidity score and 16% higher for males than for females. In the Medicare Advantage population, total cost was also significantly higher for males (11%) and for each additional point of the comorbidity score (18%); however, geographic region was not significantly associated with cost, while increasing age was significantly associated with higher total cost (0.4% per year of age). Baseline diagnosis of and treatment for osteoporosis were not significantly related to total cost in either the commercial or Medicare Advantage population.
Table 5. Inpatient, outpatient and pharmacy costs in the first year post-fracture.
Table 6. Multivariate model for total follow-up cost in the commercial population.
Table 7. Multivariate model for total follow-up cost in the Medicare Advantage population.
Hip fracture costs in the year post-fracture were more than 3 times greater than mean or median costs in the year prior to fracture (, ) for both populations. Commercial subjects in the vertebral fracture cohort had median medical costs post-fracture of $8,261 (mean $23,634; CI $22,325–24,943), representing 2.5 times greater median medical costs (1.9 times greater mean) than those incurred in the baseline period. Median medical costs in the year post-NHNV fracture were 3.7 times greater compared to baseline, and mean cost was 2.3 times greater (post-fracture: mean $12,918; CI $12,606–13,230; median $5,722). Similar trends were observed in the Medicare Advantage population, where median medical costs incurred in the year following a vertebral fracture (mean $18,117; CI $17,056–19,177; median $8,985) were 2.4 times greater than those in the year prior (1.8 times greater mean), and median medical costs incurred in the year following NHNV fracture (mean $13,769; CI $13,254–14,284; median $6,489) were 2.9 times greater than those in the year prior (2.0 times greater mean).
In the commercial population, among hip fracture subjects, the majority of both mean and median total medical costs (mean $35,898; median $22,945) were composed of inpatient costs (mean $23,837, 66.4%, median $14,236, 62.0%). Subjects with vertebral fracture had mean inpatient and outpatient costs of similar magnitude ($11,286 and $10,001, respectively), while median inpatient and outpatient costs were quite different, $0 and $4,775, respectively. Outpatient cost made up the largest portion of medical cost among NHNV subjects (mean $6,619, median $3,831) while inpatient cost made up a smaller portion overall (mean $4,921) but did not contribute to medical cost for at least half of these subjects (median $0). Among subjects in the Medicare Advantage population, the pattern was similar to that observed in the commercial population for the hip fracture cohort. In the vertebral and NHNV cohorts, mean inpatient cost made up approximately half of medical costs, but median inpatient cost was $0. Outpatient costs made up a sizeable portion of the medical costs for these groups, whether measured by mean or median (see ).
For subjects with hip fracture, 65% of commercial and 73% of Medicare median medical costs were incurred during the first month post-fracture in both health plan populations. For subjects with vertebral and NHNV fracture, on average less than one-third (15–27%) of median medical costs were incurred within the first month post-fracture ().
Figure 2. Median medical costs in the first month and year post-fracture.
Fracture-related costs accounted for a large proportion of patients’ total medical costs in the year after fracture. Median fracture-related medical costs made up approximately two-thirds of total medical costs in the year following hip fracture. Subjects with hip fractures had the highest mean and median fracture-related medical cost in both the commercial ($23,671; CI $21,991–25,351; median $16,025) and Medicare Advantage groups ($24,317; CI $23,539–25,094; median $20,491). For vertebral and NHNV fractures, median fracture-related medical costs made up approximately 20–30% of median overall medical costs. Fracture-related medical costs for the first year are shown in .
Figure 3. Fracture-related medical costs in the first year post-fracture.
Rates of diagnosis of and treatment for osteoporosis
The majority of both commercial and Medicare Advantage subjects had no evidence of an osteoporosis diagnosis () or osteoporosis treatment () in the year prior to the fracture event. Less than 15% of commercial subjects and 15.5–29.7% of Medicare Advantage subjects with any fracture type had either diagnosis or treatment in the baseline year. Subjects with vertebral fracture were most likely to have a pre-existing diagnosis or treatment in both populations.
Figure 4. Percent of subjects with an osteoporosis diagnosis.
Figure 5. Percent of subjects treated for osteoporosis.
Following fracture, rates of both diagnosis and treatment increased. In the commercial population, rates of diagnosis increased to roughly 30% for subjects with hip or vertebral fracture and 11% for subjects with NHNV fracture. The percentage of commercial subjects treated with an osteoporosis medication increased in the follow-up period to no more than 25%. Similar increases in diagnosis and treatment rates were observed in the Medicare Advantage population.
Discussion
This study utilized claims data from one of the largest health plans in the country to comprehensively estimate medical costs related to fractures commonly associated with osteoporosis. Among patients with hip fractures, 12-month medical costs following the fracture were more than $30,000 on average for both commercial and Medicare Advantage health plan members. Vertebral fracture costs ranged from $18,000–23,000 in the year post-fracture, followed by non-vertebral fracture with costs ranging from $13,000 to $14,000. Among patients with fractures commonly associated with osteoporosis, less than 15% of commercial enrollees and less than 30% of Medicare Advantage enrollees had a diagnosis of osteoporosis in the year prior to the fracture. Following the fracture, less than 25% of patients had evidence of receiving therapy for osteoporosis, although diagnosis rates increased somewhat during the same year.
The findings from this study highlight the fact that the direct economic burden of hip, vertebral, and NHNV fractures is substantial. Furthermore, the broad-based cost assessment in our study indicates that hip and vertebral fracture can be more costly and resource intensive than previously reportedCitation[5],Citation[6],Citation[12–18]. The cost of hip fracture identified in this study (mean total cost of $38,699 and $35,898 for medical costs alone), in particular, was consistent with Kilgore (unadjusted mean cost $38,971)Citation[18], but higher than Ohsfeldt ($26,856)Citation[5], Roudsari ($17,744)Citation[20], and Shi ($21,942–33,999)Citation[19] (). The disproportionate share of post-fracture costs incurred in the first month immediately following fracture, along with the finding that the majority of medical costs were fracture-related, suggest that medical costs were mainly attributed to fractures and not skewed towards comorbid conditions in the post-fracture period. This is further supported by the extensive increase in medical costs in the year post-fracture compared to the year prior to the fracture eventCitation[18].
Table 8. Mean costs of fracture treatment: estimates reported in clinical studies.
Consistent with the literatureCitation[5],Citation[6], inpatient costs were the key driver of medical costs for hip fracture and to a significantly greater extent than for vertebral and NHNV fractures. Ohsfeldt et al.Citation[5] report 87% of 1-year post-fracture costs resulting from inpatient services, while the Max et al.Citation[6] paper, excluding nursing home and extended care costs, found inpatient costs representing approximately 54% of annual costs post-fracture. Our study, using more current data, and representing more current practice of shorter inpatient stays, found 62% of costs were attributable to inpatient services for commercially enrolled patients with hip fracture, and 67% of costs for Medicare enrolled hip fracture patients. This finding, supported by the multivariate cost analysis for fracture type, highlights the particularly high cost and resource-intensive burden of hip fractures as compared to the other fractures evaluated.
There have been a number of published studies that estimated costs associated with osteoporotic fracturesCitation[5],Citation[6],Citation[12–19]. The published studies differ substantially from one another in important ways, including the populations studied, the time periods analyzed, type of costs assessed, and the analytical approaches employed. This study is most similar to those of Ohsfeldt et al.Citation[5], Kilgore et al.Citation[18], and Roudsari et al.Citation[20]. In contrast to the Ohsfeldt et al.Citation[5] study, which used 2003 Medicare fee schedule payments to determine costs, this study and those by Kilgore et al.Citation[18] and Roudsari et al.Citation[20] used actual amounts paid and procedure codes related to fracture to help identify fracture-related costs. In addition to providing more updated cost estimates, the current study included vertebral, ankle, and patella fractures in addition to the fractures considered by Ohsfeldt et al.Citation[5] (clavicle, femur, hip, humerus, pelvis, tibia/fibula, and wrist/forearm).
While both Ohsfeldt et al.Citation[5] and Roudsari et al.Citation[20] found that hip fractures were among the most expensive fragility fractures, the cost estimates in those studies were lower than those reported by our study. Ohsfeldt et al.Citation[5] reported that median medical costs per patient per year for fractures of the hip were $17,012 (mean $26,856) in 2003 dollars. Kilgore et al.Citation[18], in a retrospective database study of Medicare patients, reported the mean cost of hip fractures and vertebral fractures (in 2007 dollars) to be $38,971 and $18,844, which are similar to our results of $32,919 and $18,116 in the Medicare Advantage population.
As is typical with published research, there is variation in estimates of cost found between this study and others; however, the literature is generally consistent in demonstrating a substantial burden of osteoporotic fractureCitation[6],17–Citation[20]. The notion of conducting an economic burden study hinges on the ability to identify underlying factors responsible for increased costs, so that appropriate and targeted interventions can be undertaken across the healthcare continuum. Published reports have indicated that more than 75% of women and about 90% of men with a high likelihood of osteoporosis are not tested for the conditionCitation[21]. Among subjects with vertebral fractures, the underdiagnosis of osteoporosis has been observed globallyCitation[22]. Additionally, among subjects with hip fracture, osteoporosis undertreatment rates of 11–29% have been reportedCitation[23]. In this study of adults with fractures commonly associated with osteoporosis, less than 15% of commercial and Medicare Advantage subjects had evidence of treatment for osteoporosis prevention in the year prior to fracture, even while rates of diagnosis were somewhat higher in both groups. Post-fracture, diagnosis rates increased to roughly 30% of commercial subjects and 50% of Medicare Advantage subjects. Although the current study did not involve confirmation of the presence of osteoporosis in the subjects using clinical data, the low rates of diagnosis and treatment of osteoporosis in the claims data either before or following a fracture commonly associated with osteoporosis support the findings of underdiagnosis and undertreatment of osteoporosis reported in the literature.
Despite the robust and consistent nature of these findings relative to other research, the results must be considered within the specific limitations of this study's data and design. While these limitations do not significantly reduce the strength of the study and its results, they must be considered during interpretation. First, the direct measurement method used in this study as well as in others such as the one by Ohsfeldt et al.Citation[5] may overestimate costs because both primary and secondary diagnoses were used to identify the claims from which costs were derived. Therefore, coincidental treatment for comorbid conditions during an episode of care for fracture may have been captured as post-fracture costs. Because both pre- and post-fracture costs were measured, and costs rose significantly following fracture, and further because costs were greatest in the first month post-fracture, we believe this potential over-estimate is relatively minor and should not detract from the estimates provided in the study.
Second, it is possible that the study sample may not be representative of the entire population with fractures in the United States. Thus, caution in generalizing the findings beyond this study setting is required. However, the database used for this research represents one of the largest commercial and Medicare Advantage enrollee populations, and results should be generalizable to a large segment of the US population.
Statistical testing in this study was limited to the evaluation of predictors of costs post-fracture. Results were not compared to a normative sample of non-fracture patients. However, this study was not used to argue that fractures associated with osteoporosis are more or less costly than other possible conditions, nor that enrollees with particular characteristics are more or less costly than other enrollees. The results provide evidence of trends in costs, diagnosis and treatment associated with fractures and osteoporosis and can be used in consideration of treatment policies, and resolving possible gaps in clinical practice.
Fourth, presence of a diagnosis code on a medical claim is not positive presence of disease, as the diagnosis code may be incorrectly coded. The current study applied more stringent criteria, requiring two or more fracture-related outpatient visits over the 12-month follow-up period, reducing the odds of incorrectly identifying a fracture. Similarly, attributing healthcare costs to a fracture is dependent upon correct coding of the diagnosis and procedures performed, although using both fracture diagnosis and fracture-care procedure codes reduces the possibility of attributing non-fracture-related care and costs to the fracture or failing to attribute fracture-related costs appropriately.
Finally, the lack of a diagnosis code for osteoporosis or filled prescription for an osteoporosis medication does not guarantee that osteoporosis was not diagnosed or treated. A claim could contribute up to four diagnosis codes for any particular date of service, and the database contains all pharmacy claims submitted for payment, including mail-order, retail, and specialty pharmacy. While it is conceivable that patients paid out of pocket for relatively inexpensive generic medications, or received free samples from their physicians, the gaps in diagnosis and treatment are large enough that it seems unlikely that they can be fully explained by these factors.
This study represents an advance in the cost of fracture literature by identifying sizeable underdiagnosis and undertreatment of osteoporosis among subjects with fragility fractures that can clearly be assumed to be the main precursor for the high economic burden of osteoporosis. The suboptimal pre- and post-fracture diagnosis and treatment rates in the present study lend themselves to use as reminders for managed-care entities to strive for optimal patient care in osteoporosis and formulate effective strategies to improve diagnosis and treatment across the continuum of care. The fracture reduction that may likely occur as an outcome of improved care will have significant macro-economic consequences, as it is estimated that a yearly reduction of 300,000 fractures can lead to annual savings of about $3 billion in the United States aloneCitation[24].
Conclusion
Fragility fractures represent a large cost burden to the healthcare system during the first year post-fracture. Hip fractures were associated with the highest costs, followed by vertebral and NHNV fractures. A strikingly low rate of diagnosis and treatment of osteoporosis was found in patients with fragility fractures both before and after fracture. This study provides evidence to managed-care organizations, physicians and healthcare policy makers that there may be an opportunity to improve patient care in the area of osteoporosis.
Transparency
Declaration of funding: This study was funded by Amgen, Inc.
Declaration of financial/other relationships: S.I., D.M. and E.B. have disclosed that they are employees of Amgen, Inc. L.C. and C.H. have disclosed that they are employed by i3 Innovus, a company that received funding from Amgen to conduct this study.
References
- US Department of Health and Human Services. Bone Health and Osteoporosis: A Report of the Surgeon General. Washington, DC: Office of the Surgeon General, October 14, 2004.
- World Health Organization. WHO scientific group on the assessment of osteoporosis at primary health care level. Summary Meeting Report (Brussels, Belgium, May 5–7, 2004), 2007.
- Burge R, Dawson-Hughes B, Solomon DH, Wong JB, King A, Tosteson A. Incidence and economic burden of osteoporosis-related fractures in the United States, 2005–2025. J Bone Miner Res. 2007; 22: 465–475
- US Department of Labor, Bureau of Labor Statistics. Consumer Price Index. Chained Consumer Price Index for all urban consumers (C-CPI-U) 1999–2008, Medical Care. Series ID: SUUR0000SAM. Washington, DC: U.S. Dept. of Labor, Bureau of Labor Statistics, 2008. http://data.bls.gov/cgi-bin/surveymost?su
- Ohsfeldt RL, Borisov NN, Sheer RL. Fragility fracture-related direct medical costs in the first year following a nonvertebral fracture in a managed care setting. Osteoporos Int 2006; 17: 252–258
- Max W, Sinnot P, Kao C, et al. The burden of osteoporosis in California, 1998. Osteoporos Int 2002; 13: 493–500
- Henry G, Bone SL, et al. Alendronate and estrogen effects in post menopausal women with low bone mineral density. J Clin Endocrinol Metab 2000; 85: 720–726
- Hosking D, Chilvers CE, Christiansen C, et al. Prevention of bone mass with alendronate in post menopausal women under 60 years of age. Early Postmenopausal Intervention Cohort Study Group. N Engl J Med 1998; 338: 485–492
- Chesnut 3rd CH, 3rd, McClung MR, Ensrud KE, et al. Alendronate treatment of the postmenopausal osteoporotic woman: effect of multiple dosages on bone mass and bone remodeling. Am J Med 1995; 99: 144–152
- Huybrechts KF, Ishak KJ, Caro JJ. Assessment of compliance with osteoporosis treatment and its consequences in a managed care population. Bone 2006; 38: 922–928
- Weycker D, Macarios D, Edelsberg J, et al. Compliance with osteoporosis drug therapy and risk of fracture. Osteoporos Int 2007; 18: 271–277
- Gabriel SE, Tosteson AN, Leibson CL, et al. Direct medical costs attributable to osteoporotic fractures. Osteoporos Int 2002; 13: 323–330
- Bass E, French DD, Bradham DD. A national perspective of Medicare expenditures for elderly veterans with hip fractures. J Am Med Dir Assoc 2008; 9: 114–119
- De Laet CE, van Hout BA, Burger H, et al. Incremental cost of medical care after hip fracture and first vertebral fracture: the Rotterdam study. Osteoporos Int 1999; 10: 66–72
- Haentjens P, Lamraski G, Boonen S. Cost and consequences of hip fracture occurrence in old age: an economic perspective. Disabil Rehabil 2005; 27: 1129–1141
- Melton 3rd LJ, 3rd. Adverse outcomes of osteoporotic fractures in the general population. J Bone Miner Res 2003; 18: 1139–1141
- Bouee S, Lafuma A, Fagnani F, et al. Estimation of direct unit costs associated with non-vertebral osteoporotic fractures in five European countries. Rheumatol Int 2006; 26: 1063–1072
- Kilgore ML, Morrisey MA, Becker DJ, et al. Health care expenditures associated with skeletal fractures among Medicare beneficiaries, 1999–2005. J Bone Miner Res 2009; 24: 2050–2055
- Shi N, Foley K, Lenhart G, et al. Direct healthcare costs of hip, vertebral, and non-hip, non-vertebral fractures. Bone 2009; 45: 1084–1090
- Roudsari BS, Ebel BE, Corso PS, et al. The acute medical care costs of fall-related injuries among the U.S. older adults. Injury 2005; 36: 1316–1322
- Nguyen TV, Center JR, Eisman JA. Osteoporosis: underrated, underdiagnosed and undertreated. Med J Aust 2004; 180: S18–22
- Delmas PD, van de Langerijt L, Watts NB, Eastell R, Genant H, Grauer A, Cahall DL, IMPACT Study Group. Underdiagnosis of vertebral fractures is a worldwide problem: the IMPACT study. J Bone Miner Res 2005; 4: 557–563
- Gardner MJ, Flik KR, Mooar P, Lane JM. Improvement in the undertreatment of osteoporosis following hip fracture. J Bone Joint Surg Am 2002; 84-A: 1342–1348
- Orsini LS, Rousculp MD, Long SR, et al. Health care utilization and expenditures in the United States: a study of osteoporosis-related fractures. Osteoporos Int 2005; 16: 359–371