Abstract
Background:
Geriatric hip fractures are common and important sentinel events regarding bone health. Although dual x-ray absorptiometry (DEXA) scans are a standard method for determining bone density, differences in use among various race/ethnic groups around the time of hip fracture may reveal disparities within the healthcare system.
Methods:
The 2014 to 2016 Medicare Standard Analytic Files PearlDiver data set was used to identify geriatric patients sustaining hip fracture. From that cohort, those who had a DEXA scan before or after their fracture were defined. For the defined cohorts, patient age, sex, Elixhauser Comorbidity Index, race/ethnicity, and income (based on zip code) were defined and compared with univariate and multivariate analyses.
Results:
Of 58,099 hip fracture patients, only 19.8% had had a DEXA scan before fracture and 3.9% of the remaining group had the DEXA scan after fracture. Of the hip fracture population, 91.0% identified as White and 9.0% as non-White (Native American, Black, Hispanic, Asian, or Other). Before hip fracture, controlling for other variables and compared with White patients, all non-White categories were at lesser odds of having had the DEXA scan. After hip fracture, Black, Hispanic, and other patients were also at lesser odds of having the DEXA scan.
Discussion:
Using a large Medicare data set, controlling for patient age, sex, Elixhauser Comorbidity Index, and income marker, this study revealed disparities in DEXA scan utilization across race/ethnic groups before and after presenting with a hip fragility fracture. Identification of such disparities highlights the needs for improved medical access and care for this at-risk hip fracture population.
Osteoporosis is a major health problem in the United States, with more than 11 million adults affected and 34 million at risk.1,2 This condition is characterized by microarchitectural bone resorption, decreased bone density, and bone fragility leading to susceptibility to fracture.3 Hip factures, in particular, are common fragility fractures that are associated with notable morbidity and mortality.4,5
Detecting osteoporosis is crucial because the condition may be both preventable and treatable.6 The current standard diagnostic criteria for osteoporosis involve bone densitometry using dual-energy x-ray absorptiometry (DEXA) scans.4 As such, the US government has attempted to limit financial barriers to osteoporosis screening by declaring that Medicare will reimburse routine bone mineral density testing every 2 years.7
Despite efforts to encourage assessing and managing bone density, osteoporosis remains underdiagnosed in the overall population, with disparities in screening rates between racial, ethnic, and socioeconomic groups.7-9 The importance of identifying patients with osteoporosis before having a hip fragility fracture is crucial because hip fractures are associated with increased risk of mortality in the year after fracture.10 Furthermore, many efforts have been initiated to ensure bone density is assessed and managed after hip fracture to mitigate the risk of other osteoporotic-related fractures.11
Although previous studies characterized disparities in overall DEXA scan utilization, no literature to our knowledge has specifically evaluated disparities in DEXA scan utilization before and after hip fragility fractures. Thus, using a large, national, administrative database, this study aimed to assess for potential race/ethnic disparities in DEXA scan utilization before and after hip fragility fractures.
Methods
Database/Study Cohort
This retrospective cohort study used the 2014 to 2016 Standard Analytic Files PearlDiver data set—a national Medicare data set containing information for more than 60 million patients. This has been used for multiple previous orthopaedic-related studies.12,13 As patient information obtained from PearlDiver data sets is output in deidentified and aggregated form, our Institutional Review Board determined the use of this data set for studies exempt from review.
Patients were isolated from the data set based on International Classification of Disease (ICD) codes for hip fracture (ICD-9-D-82000, ICD-9-D-82001, ICD-9-D-82002, ICD-9-D-82003, ICD-9-D-82009, ICD-9-D-82010, ICD-9-D-82011, ICD-9-D-82012, ICD-9-D-82013, ICD-9-D-82019, ICD-9-D-82020, ICD-9-D-82021, ICD-9-D-82022, ICD-9-D-82030, ICD-9-D-82031, ICD-9-D-82032, ICD-9-D-8208, ICD-9-D-8209, ICD-9-D-73396, ICD-9-D-82121, ICD-9-D-82131) with Current Procedural Terminology (CPT) code for fracture stabilization. Exclusion criteria included those younger than 65 years, those with high-energy trauma (whereas hip fragility fractures are considered low-energy traumas), neoplasm, and infection (based on ICD codes) within the 90 days before their procedure.
In addition, patient age, sex, Elixhauser Comorbidity Index (ECI—an estimate of patient comorbidity burden based on ICD diagnostic codes13,14,15,16), race/ethnicity (White, Black, Native American, Asian, Hispanic, and “Other”), and income (derived based on zip code and dichotomized to >$75k/yr or <$75k/yr) were identified for the study cohort.
Dual-energy X-ray Absorptiometry Studies
Patients who had undergone a DEXA scan were identified using CPT codes: CPT-77080, CPT-77085. Patients were first identified by their first hip fracture incident and whether they had received a DEXA scan at any time since the age of 65 and before their first hip fracture incident. To assess for DEXA scans after initial hip fractures, these patients were characterized by whether they received a DEXA scan any time after their initial hip fracture.
Data Analysis
Descriptive analysis was done to characterize the race/ethnic make of up of the study cohort.
Those with versus without DEXA scan before hip fracture were then characterized. Univariate analysis was done with the Pearson chi-squared test to compare categorical variables (race/ethnicity, sex, income, etc.) and the Welch t-test was to compare age and ECI variables. Multivariate analysis was done to determine independent race/ethnicity and socioeconomic differences of the two groups controlling for patient age, sex, and ECI. Odds ratios (OR) ± 95% confidence intervals were determined. Those with versus without DEXA after hip fracture were then characterized and analyzed by similar metrics. Univariate and multivariate analyses were done in a similar fashion.
Statistical analyses were conducted within the PearlDiver system with significance defined as P value <0.05. Prim9 (GraphPad Softwares) and Microsoft Excel (Microsoft) were used to generate figures.
Results
Study Cohort
A total of 58,099 patients met inclusion criteria for the study of which 52,890 (91.0%) identified as White and 5,209 (9.0%) as non-White (Figure 1). Additional breakdown of non-White individuals revealed 2,460 Black (4.2% of the overall study population), 932 Hispanic (1.6%), 875 Asian (1.5%), 375 Native American (0.6%), and 567 Other (1.0%). The rest of the demographic/socioeconomic variables are shown for the overall study population in Table 1.
Figure 1.
The pie graph displaying the breakdown of hip fracture cohort by race/ethnicity. The pie graph on the left distinguishes White from non-White patients. The pie graph on the right breaks down the non-White patients into their respective race/ethnic groups.
Table 1.
Univariate Analysis of Characteristics of Patients Who Either Had a Dual-energy X-ray Absorptiometry Scan or Did Not Have a Dual-energy X-ray Absorptiometry Scan Before Hip Fractures
| Demographic Factor | Total | No DEXA Scan Before Hip Fracture | DEXA Scan Before Hip Fracture | P |
| N | 58,099 (100%) | 46,610 (80.2%) | 11,489 (19.8%) | |
| Age | ||||
| 65-69 | 6172 (10.6%) | 5326 (11.4%) | 2117 (18.4%) | <0.0001 |
| 70-74 | 6839 (11.8%) | 5279 (11.3%) | 2758 (24%) | |
| 75-79 | 8698 (15%) | 6566 (14.1%) | 3276 (28.5%) | |
| 80-84 | 10,918 (18.8%) | 8379 (18%) | 3323 (28.9%) | |
| 85+ | 25,691 (44.2%) | 21,060 (45.2%) | 2508 (21.8%) | |
| Sex | ||||
| Female | 41,395 (71.2%) | 30,860 (66.2%) | 10,467 (91.1%) | <0.0001 |
| Male | 16,704 (28.8%) | 15,750 (33.8%) | 948 (8.3%) | |
| ECI (mean ± SD) | 7.57 ± 4.25 | 7.28 ± 4.20 | 5.27 ± 3.86 | |
| 0-3 | 10,523 (18.1%) | 9423 (20.2%) | 4423 (38.5%) | <0.0001 |
| 4-7 | 20,360 (35%) | 16,580 (35.6%) | 4183 (36.4%) | |
| 8-11 | 16,595 (28.6%) | 12,886 (27.6%) | 1980 (17.2%) | |
| 12+ | 10,616 (18.3%) | 7701 (16.5%) | 891 (7.8%) | |
| Race/ethnicity | ||||
| White | 52,890 (91%) | 42,141 (90.4%) | 10,749 (93.6%) | <0.0001 |
| Native American | 375 (0.6%) | 335 (0.7%) | 40 (0.3%) | |
| Black | 2460 (4.2%) | 2116 (4.5%) | 344 (3.0%) | |
| Hispanic | 932 (1.6%) | 798 (1.7%) | 134 (1.2%) | |
| Asian | 875 (1.5%) | 746 (1.6%) | 129 (1.1%) | |
| Other | 567 (1%) | 474 (1%) | 93 (0.8%) | |
| Income (based on zip code) | ||||
| High income (>75k/yr) | 12,842 (22.1%) | 10,393 (22.3%) | 2449 (21.3%) | <0.0001 |
| Low income (<75k/yr) | 45,257 (77.9%) | 36,217 (77.7%) | 9040 (78.7%) |
DEXA = Dual-energy X-ray Absorptiometry
Bold entries represent statistically significant P values at P < 0.05.
Dual-energy X-ray Absorptiometry Scan Before Hip Fracture
The DEXA scan was identified as having been performed before hip fracture for only 11,489 of the 58,099 patients (19.8%). By univariate analysis, there were differences in those having versus not having had DEXA before hip fracture in age, sex, ECI, race/ethnicity, and socioeconomic status (Table 1).
Multivariate analysis was then done to determine independent predictors of having had DEXA performed before hip fracture, controlling for patient age, sex, and ECI. This revealed that all non-White groups had variably lesser odds of having had DEXA (relative to White, Native American OR 0.39, Black OR 0.59, Hispanic OR 0.64, Asian 0.69, and Other OR 0.77, P < 0.001 for each other than “Other” for which P = 0.0262). Interestingly, income as derived from the zip code was not significantly predictive of having had DEXA performed before hip fracture. These results are presented in Table 2 and Figure 2.
Table 2.
Multivariate Analysis of Predictive Factors for Dual-energy X-ray Absorptiometry Scan Before Hip Fracture
| N = 58,099 | OR (95% CI) | P |
| Race/ethnicity | ||
| White (referent) | ||
| Native American | 0.39 (0.27-0.53) | <0.0001 |
| Black | 0.59 (0.52-0.66) | <0.0001 |
| Hispanic | 0.64 (0.53-0.77) | <0.0001 |
| Asian | 0.69 (0.57-0.84) | <0.0001 |
| Other | 0.77 (0.61-0.96) | 0.0262 |
| Income (based on zip code) | ||
| Income <75k/yr (referent) | ||
| Income >75k/yr | 1.0 (0.95-1.06) | 0.8889 |
CI = confidence interval, OR = odds ratio
Bold entries represent statistically significant P values at P < 0.05.
Figure 2.
Forest plot displaying predictive factors of DEXA scan utilization before hip fracture among race/ethnicity groups and income. DEXA = Dual-energy X-ray Absorptiometry.
Dual-energy X-ray Absorptiometry Scan After Hip Fracture
Of those who had not had the DEXA scan performed before hip fracture, only 1832 of 46,610 additional patients had DEXA performed after hip (3.9%). By univariate analysis, there were differences in those having versus not having DEXA after hip fracture in age, sex, ECI, and race/ethnicity (Table 3).
Table 3.
Univariate Analysis of Characteristics of Patients Who Either Had a Dual-energy X-ray Absorptiometry Scan or Did Not Have a Dual-energy X-ray Absorptiometry Scan After Having a Hip Fracture
| Demographic Factor | Total | No DEXA Scan After Hip Fracture | DEXA Scan After Hip Fracture | P |
| N | 46,610 (100%) | 44,778 (96.1%) | 1832 (3.9%) | |
| Age | ||||
| 65-69 | 5286 (11.3%) | 4864 (10.9%) | 422 (23%) | <0.0001 |
| 70-74 | 5290 (11.3%) | 4970 (11.1%) | 320 (17.5%) | |
| 75-79 | 6551 (14.1%) | 6217 (13.9%) | 334 (18.2%) | |
| 80-84 | 8382 (18%) | 8045 (18%) | 337 (18.4%) | |
| 85+ | 21,101 (45.3%) | 20,682 (46.2%) | 419 (22.9%) | |
| Sex | ||||
| Female | 30,860 (66.2%) | 29,454 (65.8%) | 1406 (76.7%) | <0.0001 |
| Male | 15,750 (33.8%) | 15,324 (34.2%) | 426 (23.3%) | |
| ECI (mean ± SD) | 7.01 ± 4.19 | 7.33 ± 4.20 | 6.69 ± 4.18 | |
| 0-3 | 9344 (20%) | 8872 (19.8%) | 472 (25.8%) | <0.0001 |
| 4-7 | 16,566 (35.5%) | 15,886 (35.5%) | 680 (37.1%) | |
| 8-11 | 12,919 (27.7%) | 12,494 (27.9%) | 425 (23.2%) | |
| 12+ | 7731 (16.6%) | 7508 (16.8%) | 223 (12.2%) | |
| Race/ethnicity | ||||
| White | 42,141 (90.4%) | 40,452 (90.3%) | 1689 (92.2%) | 0.0117 |
| Native American | 335 (0.7%) | 319 (0.7%) | 16 (0.9%) | |
| Black | 2116 (4.5%) | 2055 (4.6%) | 61 (3.3%) | |
| Hispanic | 798 (1.7%) | 781 (1.7%) | 17 (0.9%) | |
| Asian | 746 (1.6%) | 714 (1.6%) | 32 (1.7%) | |
| Other | 474 (1%) | 457 (1%) | 17 (0.9%) | |
| Income (based on zip code) | ||||
| Low income (<75k/yr) | 36,620 (78.6%) | 35,195 (78.6%) | 1425 (77.8%) | |
| High income (>75k/yr) | 9990 (21.4%) | 9583 (21.4%) | 407 (22.2%) | 0.4213 |
ECI = Elixhauser Comorbidity Index, DEXA = Dual-energy X-ray Absorptiometry
Bold entries represent statistically significant P values at P < 0.05.
Multivariate analysis was then done to determine independent predictors of having DEXA performed after hip fracture, controlling for patient age, sex, and ECI. This revealed that lesser odds of having had DEXA performed for those who were Black OR 0.66 (P = 0.0021), Hispanic OR 0.57 (P = 0.0251), and Other OR 0.75 (P = 0.0240). Again, income as derived from the zip code was not significantly predictive of having had DEXA performed after hip fracture. These results are presented in Table 4 and Figure 3.
Table 4.
Multivariate Analysis of Predictive Factors for Dual-energy X-ray Absorptiometry Scan After Having a Hip Fracture
| N = 23,953 | OR (95% CI) | P |
| Race/ethnicity | ||
| White (referent) | ||
| Native American | 1.11 (0.64-1.79) | 0.6838 |
| Black | 0.66 (0.51-0.85) | 0.0021 |
| Hispanic | 0.57 (0.34-0.9) | 0.0251 |
| Asian | 1.07 (0.73-1.52) | 0.6965 |
| Other | 0.75 (0.44-1.18) | 0.024 |
| Income (based on zip code) | ||
| Income <75k/yr (referent) | ||
| Income >75k/yr | 1.05 (0.94-1.18) | 0.3767 |
CI = confidence interval, OR = odds ratio
Bold entries represent statistically significant P values at P < 0.05.
Figure 3.
Forest plot displaying predictive factors of DEXA scan utilization after having a hip fracture among race/ethnicity groups and income. DEXA = Dual-energy X-ray Absorptiometry.
Discussion
The US Preventive Services Task Force17 recommends DEXA scans for women older than 65 years and all women with risk factors for osteoporosis. However, it has been questioned if such studies are equitably obtained for different demographic populations.7-9 Data from this study demonstrate that specific race/ethnicity groups differentially receive DEXA screening both before and after fragility hip fracture—highlighting disparities in care.
Of the 58,099 geriatric fragility hip fracture patients identified in this study, 11,489 (19.8%) had a DEXA scan performed before their fracture. Previous studies have reported varying DEXA utilization, ranging from to 19% to 44.7%.18-20 As such, the overall cohort DEXA scan utilization rate in this study aligns with previous reports. Such broad variations in DEXA scan utilization could be due to several factors, including varying clinical pathways and provider preferences. Furthermore, populations may differ in DEXA utilization secondary to differential access to care and potential referral biases.
The race/ethnicity breakdown of the overall study cohort was subsequently assessed. Ninety-one percent of geriatric fragility hip fracture patients self-identified as White, 4.2% Black, 1.6% Hispanic, 1.5% Asian, 0.6% Native American, and 1% as “Other” race/ethnicity. Previous studies analyzing the racial distribution of patients with fragility hip fractures have reported relatively similar distributions. A 2010 study analyzing 40,321 fragility hip fractures in a Medicare claims database found 95.4% of fractures to be in White patients, 3.4% in Black patients, and 1.2% in Hispanic patients.21
The utilization of DEXA scans before and after fragility hip fractures was subsequently analyzed while controlling for patient age, sex, ECI, and income. In DEXA scans before hip fracture, all minority groups analyzed were markedly less likely to have had a DEXA scan compared with White patients. In DEXA scans after hip fracture, Black, Hispanic, and “Other” patients were markedly less likely to have a scan performed. Few studies to date have characterized racial disparities in DEXA scan utilization. One study of African American and White patients with a history of geriatric fragility hip fractures found African American patients to be markedly less likely than White patients to have received a DEXA examination.19 By contrast, another study conducted at three major hospitals in Hawaii found no disparities in DEXA utilization after fragility hip fracture between White, Native Hawaiian/Pacific Islander, and Asian American patients.21 In summary, these findings suggest that although disparities in DEXA care likely exist, different cohorts may be variably affected.
Although this study has strengths associated with its large cohort size from across the United States, it also has limitations. As with any study using administrative data, the accuracy of the findings is limited to the accuracy of the billings claims data coded. For DEXA scans obtained before hip fracture, there is the possibility that a DEXA scan had been obtained before becoming active in the data set. For DEXA scans obtained after hip fracture, there is the possibility that diagnosis was made (and treatment begun) for osteoporosis simply based on the occurrence of the hip fracture.
Overall, using a large Medicare data set and controlling for patient age, sex, ECI, and income marker, this study revealed disparities in DEXA scan utilization across race/ethnic groups before and after presenting with a hip fragility fracture. Identification of such disparities highlights the needs for improved medical access and care for this at-risk hip fracture population.
Footnotes
Dr. Grauer name or an immediate family member serves as a board member, owner, officer, or committee member of serves as Editor-in-Chief of the North American Spine Society Journal (NASSJ). None of the following authors or any immediate family member has received anything of value from or has stock or stock options held in a commercial company or institution related directly or indirectly to the subject of this article: Ms. Rudisill, Mr. Ratnasamy, and Mr. Sanchez.
Contributor Information
Katelyn E. Rudisill, Email: katelyn.rudisill@temple.edu.
Philip P. Ratnasamy, Email: Philip.Ratnasamy@yale.edu.
Joshua G. Sanchez, Email: Philip.Ratnasamy@yale.edu.
References
- 1.Looker AC, Sarafrazi Isfahani N, Fan B, Shepherd JA: Trends in osteoporosis and low bone mass in older US adults, 2005-2006 through 2013-2014. Osteoporos Int 2017;28:1979-1988. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Bone Health and Osteoporosis: A Report of the Surgeon General. Rockville, MD, Office of the Surgeon General, 2004. [PubMed] [Google Scholar]
- 3.Consensus development conference: Prophylaxis and treatment of osteoporosis. Osteoporos Int 1991;1:114-117. [PubMed] [Google Scholar]
- 4.US Preventive Services Task Force: Screening for osteoporosis: U.S. Preventive Services Task Force recommendation statement. Ann Intern Med 2011;154:356-364. [DOI] [PubMed] [Google Scholar]
- 5.Munoz M, Robinson K, Shibli-Rahhal A: Bone health and osteoporosis prevention and treatment. Clin Obstet Gynecol 2020;63:770-787. [DOI] [PubMed] [Google Scholar]
- 6.LeBoff MS, Greenspan SL, Insogna KL, et al. : The clinician's guide to prevention and treatment of osteoporosis. Osteoporos Int 2022;33:2049-2102. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Gillespie CW, Morin PE: Trends and disparities in osteoporosis screening among women in the United States, 2008-2014. Am J Med 2017;130:306-316. [DOI] [PubMed] [Google Scholar]
- 8.Brennan SL, Wluka AE, Gould H, et al. : Social determinants of bone densitometry uptake for osteoporosis risk in patients aged 50yr and older: A systematic review. J Clin Densitom 2012;15:165-175. [DOI] [PubMed] [Google Scholar]
- 9.Ruiz-Esteves KN, Teysir J, Schatoff D, Yu EW, Burnett-Bowie SAM: Disparities in osteoporosis care among postmenopausal women in the United States. Maturitas 2022;156:25-29. [DOI] [PubMed] [Google Scholar]
- 10.Wehren LE, Hawkes WG, Orwig DL, Hebel JR, Zimmerman SI, Magaziner J: Gender differences in mortality after hip fracture: The role of infection. J Bone Miner Res 2003;18:2231-2237. [DOI] [PubMed] [Google Scholar]
- 11.Karmazyn B, Reher TA, Supakul N, et al. : Whole-spine MRI in children with suspected abusive head trauma. AJR Am J Roentgenol 2022;218:1074-1087. [DOI] [PubMed] [Google Scholar]
- 12.Joo PY, Wilhelm C, Adeclat G, et al. : Comparing race/ethnicity and zip code socioeconomic status for surgical versus nonsurgical management of proximal humerus fractures in a Medicare population. J Am Acad Orthop Surg Glob Res Rev 2023;7:e22.00205. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Ratnasamy PP, Oghenesume OP, Rudisill KE, Grauer JN: Racial/ethnic disparities in physical therapy utilization after total knee arthroplasty. J Am Acad Orthop Surg 2023;31:357-363. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Rudisill KE, Ratnasamy PP, Maloy GC, Grauer JN: Decline in separate incision autograft for spine surgery over the past decade: A fading gold standard. J Am Acad Orthop Surg 2023;31:938-944. [DOI] [PubMed] [Google Scholar]
- 15.Rudisill KE, Ratnasamy PP, Joo PY, Rubin LE, Grauer JN: Magnetic resonance imaging in the year prior to total knee arthroplasty: A potential overutilization of healthcare resources. J Am Acad Orthop Surg Glob Res Rev 2023;7:e22.00262. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Austin SR, Wong YN, Uzzo RG, Beck JR, Egleston BL: Why summary comorbidity measures such as the Charlson Comorbidity Index and Elixhauser score work. Med Care 2015;53:e65-e72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.United States Preventive Services Task Force: Osteoporosis to prevent fractures: Screening. Available at: https://www.uspreventiveservicestaskforce.org/uspstf/recommendation/osteoporosis-screening. Accessed September 25, 2023.
- 18.Mora AN, Blazar PE, Rogers JC, Earp BE: Patient perceptions and preferences for osteoporosis treatment. J Hand Surg Am 2020;45:153.e1-153.e5. [DOI] [PubMed] [Google Scholar]
- 19.Mikuls TR, Saag KG, George V, Mudano AS, Banerjee S: Racial disparities in the receipt of osteoporosis related healthcare among community-dwelling older women with arthritis and previous fracture. J Rheumatol 2005;32:870-875. [PubMed] [Google Scholar]
- 20.Nguyen ET, Posas-Mendoza T, Siu AM, Ahn HJ, Choi SY, Lim SY: Low rates of osteoporosis treatment after hospitalization for hip fracture in Hawaii. Osteoporos Int 2018;29:1827-1832. [DOI] [PubMed] [Google Scholar]
- 21.Nguyen-Oghalai TU, Kuo YF, Wu H, et al. : The impact of race/ethnicity on preoperative time to hip stabilization procedure after hip fracture. South Med J 2010;103:414-418. [DOI] [PMC free article] [PubMed] [Google Scholar]