Advertisement

Trends and Disparities in Osteoporosis Screening Among Women in the United States, 2008-2014

Open AccessPublished:November 21, 2016DOI:https://doi.org/10.1016/j.amjmed.2016.10.018

      Abstract

      Background

      The United States Preventive Services Task Force recommends universal osteoporosis screening among women ages 65+ and targeted screening of younger women, but historically, adherence to these evidence-based recommendations has been suboptimal.

      Methods

      To describe contemporary patterns of osteoporosis screening, we conducted a retrospective analysis using the OptumLabs Data Warehouse, a database of de-identified administrative claims, which includes medical and eligibility information for over 100 million Medicare Advantage and commercial enrollees. Study participants included 1,638,454 women ages 50+ with no prior history of osteoporosis diagnosis, osteoporosis drug use, or hip fracture. Osteoporosis screening during the most recent 2-year period of continuous enrollment was assessed via medical claims. Patient sociodemographics, comorbidities, and utilization of other services were also determined using health insurance files.

      Results

      Overall screening rates were low: 21.1%, 26.5%, and 12.8% among women ages 50-64, 65-79, and 80+ years, respectively. Secular trends differed significantly by age (P <.001). Between 2008 and 2014, utilization among women ages 50-64 years declined 31.4%, changed little among women 65-79, and increased 37.7% among women 80+ years. Even after accounting for socioeconomic status, health status, and health care utilization patterns, non-Hispanic black women were least likely to be screened, whereas non-Hispanic Asian and Hispanic women were most likely to undergo screening. Marked socioeconomic gradients in screening probabilities narrowed substantially over time, decreasing by 44.5%, 71.9%, and 59.7% among women ages 50-64, 65-79 and 80+ years, respectively.

      Conclusions

      Despite significant changes in utilization of osteoporosis screening among women ages 50-64 and 80+, in line with national recommendations, tremendous deficiencies among women 65+ remain.

      Keywords

      Clinical Significance
      • There is a persistent failure to align real-world implementation of osteoporosis screening among women ages 65+ years with evidence-based guidelines.
      • Fewer than 1 in 4 privately insured women ages 65+ utilize osteoporosis screening for primary prevention.
      • Between 2008 and 2014, screening among women ages 50-64 years declined steadily and did not improve, or improved but remained low among women ages 65-79 and 80+ years.
      • Disparities based on sociodemographic characteristics, though apparent, narrowed over time.
      The prevalence of osteoporosis—a leading cause of disability and loss of independence among older Americans—increases substantially with age,
      • Looker A.C.
      • Borrud L.G.
      • Dawson-Hughes B.
      • Shepherd J.A.
      • Wright N.C.
      Osteoporosis or Low Bone Mass at the Femur Neck or Lumbar Spine in Older Adults: United States, 2005-2008. NCHS Data Brief No. 93.
      and the medical and economic consequences of osteoporotic fracture can be severe. Half of all postmenopausal women will suffer osteoporotic fracture at some point during their lifetime; 15% will fracture a hip.
      U.S. Department of Health and Human Services
      Bone Health and Osteoporosis: A Report of the Surgeon General.
      U.S. Preventive Services Task Force
      Screening for osteoporosis: U.S. preventive services task force recommendation statement.
      Approximately 10%-20% of hip fracture patients die in the first year following the injury
      U.S. Department of Health and Human Services
      Bone Health and Osteoporosis: A Report of the Surgeon General.
      and risk of premature mortality may remain elevated for at least 10 years.
      • Haentjens P.
      • Magaziner J.
      • Colón-Emeric C.S.
      • et al.
      Meta-analysis: excess mortality after hip fracture among older women and men.
      Just 40% of patients experiencing hip fracture regain their baseline level of independence.
      U.S. Department of Health and Human Services
      Bone Health and Osteoporosis: A Report of the Surgeon General.
      Although common, osteoporosis is preventable and treatable. Once diagnosed, fracture risk can be substantially mitigated with lifestyle modification and an array of US Food and Drug Administration-approved pharmacologic therapies.
      U.S. Preventive Services Task Force
      Screening for osteoporosis: U.S. preventive services task force recommendation statement.
      • Nelson H.
      • Haney E.
      • Chou R.
      • Dana T.
      • Fu R.
      • Bougatsos C.
      Screening for Osteoporosis: Systematic Review to Update the 2002 U.S. Preventive Services Task Force Recommendation.
      National Osteoporosis Foundation
      Clinician's Guide to Prevention and Treatment of Osteoporosis.
      Given substantial evidence of the benefits of available treatments for osteoporosis among women, the United States Preventive Services Task Force (USPSTF) has recommended universal osteoporosis screening among women 65 years and older at 2-year intervals, and targeted screening of women 60-64 years based on individualized risk assessments since 2002.
      U.S. Preventive Services Task Force
      Screening for osteoporosis in postmenopausal women: recommendations and rationale.
      In 2011, the USPSTF eliminated the lower age limit to include targeted screening for all younger women with identified risk factors.
      U.S. Preventive Services Task Force
      Screening for osteoporosis: U.S. preventive services task force recommendation statement.
      In an effort to address cost barriers, Medicare has covered osteoporosis screening for qualified individuals at 2-year intervals since 1998.

      Balanced Budget Act of 1997. Public Law Number 105-33. (H.R. 2015); 1997. Available at: https://www.gpo.gov/fdsys/pkg/PLAW-105publ33/pdf/PLAW-105publ33.pdf. Accessed May 16, 2016.

      Additionally, preventive care provisions in the Affordable Care Act eliminated cost sharing for the USPSTF-recommended service among privately insured women beginning in September 2010, and for qualified Medicare beneficiaries beginning in January 2011.

      Patient Protection and Affordable Care Act. Public Law Number 111-148. (H.R. 3590); 2010. Available at: https://www.gpo.gov/fdsys/pkg/PLAW-111publ148/pdf/PLAW-111publ148.pdf. Accessed May 16, 2016.

      In the first decade following the initiation of Medicare reimbursement for this service, rates of osteoporosis screening remained consistently low. One study of early adoption of bone density testing based on administrative claims data found that fewer than 21% of nearly 36,000 female Medicare beneficiaries ages 65-89 years living in three states, who subsequently experienced hip fractures, had undergone screening during a 2-year window beginning in 1999, 2000, or 2001.
      • Neuner J.M.
      • Binkley N.
      • Sparapani R.A.
      • Laud P.W.
      • Nattinger A.B.
      Bone density testing in older women and its association with patient age.
      Another study, based on claims data from a 5% random sample of Medicare beneficiaries nationwide, found that fewer than 10% of women ages 65 years and older underwent screening during any given year from 2002 to 2009.
      • Zhang J.
      • Delzell E.
      • Zhao H.
      • et al.
      Central DXA utilization shifts from office-based to hospital-based settings among Medicare beneficiaries in the wake of reimbursement changes.
      Lastly, an analysis of claims from more than 400,000 geographically diverse women ages 65 years and older with employer-sponsored Medicare supplemental coverage showed that just 11%-13% of women were screened in any given year between 2005 and 2008 (38% were screened at least once during this 4-year interval).
      • McAdam-Marx C.
      • Unni S.
      • Ye X.
      • Nelson S.
      • Nickman N.A.
      Effect of Medicare reimbursement reduction for imaging services on osteoporosis screening rates.
      The impact of the recent expansion in coverage for consumers is unknown.
      National quality reporting efforts track disparities in self-reported lifetime screening behaviors,
      U.S. Department of Health and Human Services
      National Healthcare Disparities Report 2011.
      National Committee for Quality Assurance
      The State of Health Care Quality 2015.
      but few contemporary, population-based estimates of screening disparities based on administrative claims exist. Thus, we sought to evaluate trends in osteoporosis screening from 2008 to 2014 and to identify patient-level characteristics associated with screening among a large, nationwide cohort of privately insured women aged 50 years and older.

      Methods

      Data Source

      We analyzed data from the OptumLabs Data Warehouse, which includes de-identified medical and pharmacy claims and enrollment information from a large national health plan.
      • Wallace P.J.
      • Shah N.D.
      • Dennen T.
      • Bleicher P.A.
      • Crown W.H.
      Optum Labs: building a novel node in the learning health care system.
      The OptumLabs Data Warehouse contains longitudinal health information on more than 100 million privately insured and Medicare Advantage enrollees, representing a diverse array of ages, races/ethnicities, and geographic regions across the US. From the member files we obtained enrollee sex, birth year, race/ethnicity (a mix of self-report and public record data), estimated net worth (derived from assets and liabilities), geographic region, and dates of enrollment. The medical claims included International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9) diagnosis codes, ICD-9 procedure codes, Current Procedural Terminology, Version 4 (CPT) procedure codes, other Healthcare Common Procedure Coding System procedure codes, place of service codes, and provider specialty codes. The pharmacy claims included National Drug Codes and dates of prescription drug fills. Study data were accessed using techniques compliant with the Health Insurance Portability and Accountability Act of 1996, and because this study involved analysis of preexisting, de-identified data, it was exempt from institutional review board approval.

      Cohort Identification

      We identified 5,472,083 women ages 50 years and older with primary medical and prescription drug coverage at any point between January 1, 2008 and December 31, 2014. Only women with a minimum of 3 years of continuous enrollment (beginning as early as January 1 of the year of her 50th birthday) were considered for inclusion (n = 2,092,888). We assessed utilization of osteoporosis screening services during the most recent 2-year period of enrollment and used all claims preceding this observation period to confirm cohort eligibility. Because we were interested in screening for primary prevention, we excluded a total of 454,434 women with preexisting osteoporosis diagnoses (≥2 claims on different dates with ICD-9 733.0x listed for procedures other than bone mass measurement; n = 267,800), prior utilization of pharmacologic therapies for osteoporosis (≥1 claim for alendronate, calcitonin, denosumab, ibandronate, raloxifene, risedronate, teriparatide, or zoledronic acid; n = 333,193), prior history of qualifying hip fracture (≥1 claim with ICD-9 733.14, 820.00, 820.02, 820.03, 820.09, 820.2x, or 820.8; n = 29,838), or evidence of underlying conditions known to impact bone health (end-stage renal disease [ICD-9 585.6; n = 5115], bone metastases [ICD-9 198.5; n = 3484], Cushing syndrome [ICD-9 255.0; n = 836], Paget disease [ICD-9 731.0; n = 756], osteogenesis imperfecta [ICD-9 756.51; n = 122], and neoplastic disease [ICD-9 170.7; n = 110]). All women had a minimum of 12 months of continuous enrollment prior to the 2-year study observation period during which evidence of these exclusion criteria could be assessed.

      Measures of Utilization

      We defined utilization of bone mass measurement via CPT (76070, 76071, 76075, 76076, 76077, 76078, 76977, 77078, 77079, 77080, 77081, 77082, 77083, 78350, or 78351), Healthcare Common Procedure Coding System (G0130), or ICD-9 (V82.81) codes listed on any medical claims during the study period. We defined utilization of pharmacologic therapies for osteoporosis via National Drug Codes listed on pharmacy claims for the osteoporosis drug products listed above. We defined primary care utilization as any interaction with a family medicine, internal medicine, or obstetrics-gynecology provider, where place of service was defined as office, home, or nursing home, and any qualifying Evaluation and Management CPT code was listed on the claim (see supplemental file, available online).
      Lastly, we generated a chronic comorbidity count for each woman based on diagnosis codes listed in positions 1 through 6 on all medical claims from the 12 months prior to the study observation period. These diagnosis codes were mapped to 56 distinct chronic conditions of interest (eg, diabetes, hepatitis, hypertension), utilizing the Clinical Classification System developed by the Agency for Healthcare Research and Quality,

      Elixhauser A, Steiner C, Palmer L. Clinical Classifications Software (CCS). U.S. Agency for Healthcare Research and Quality; 2014. Available at: http://www.hcup-us.ahrq.gov/toolssoftware/ccs/ccs.jsp. Accessed November 22, 2016.

      and following the work of others.
      • Hwang W.
      • Weller W.
      • Ireys H.
      • Anderson G.
      Out-of-pocket medical spending for care of chronic conditions.
      • Naessens J.M.
      • Stroebel R.J.
      • Finnie D.M.
      • et al.
      Effect of multiple chronic conditions among working-age adults.

      Statistical Analyses

      We used simple descriptive statistics to summarize the characteristics of the cohort, overall and by age group. We assessed differences by age using analysis of variance, the nonparametric Kruskal-Wallis test of ranks, and Pearson chi-squared tests.
      We calculated osteoporosis screening rates, overall and by patient characteristics, expressed as percentages (and 95% confidence intervals [CI]) of cohort members with evidence of bone mass measurement at any time during the 2-year observation period. We used Pearson chi-squared tests to identify patient characteristics associated with screening in bivariate analyses.
      To identify factors independently associated with screening, we fit a series of multivariable logistic regression models including main effects for all measured characteristics, after stratifying the study cohort by age. Age and chronic comorbidity count were modeled continuously; all other characteristics were modeled categorically. We report model results as adjusted odds ratios (aOR) and 95% CI.
      We assessed secular trends in adjusted screening rates by incorporating interaction terms between calendar year and age group into a model including main effects for all other covariates of interest, and plotted the resultant adjusted predicted probabilities.
      Lastly, to further explore disparities in utilization by socioeconomic indicators, we fit a series of models that included interaction terms allowing the relative odds of screening for each race/ethnicity or net worth category to vary over time. We stratified these models by age group, but still controlled for residual confounding by age, and adjusted for all other observed covariates specified above. We used these interaction models to calculate and plot the adjusted marginal probabilities of screening for each sociodemographic group.
      All analyses were performed using Stata Version 14.1 (StataCorp, College Station, Texas).

      Results

      The study cohort consisted of 1,638,454 women, approximately 35% of whom were age 65 years or older (Table 1). The median duration of continuous coverage prior to the 2-year observation period was 3.1 years. The vast majority of women (97.1%) had at least one claim during the 2-year observation period, and most (85.7%) had at least one claim for primary care services (median 5.0 encounters). Overall, 2.5% initiated use of osteoporosis drugs and 0.5% experienced a qualifying hip fracture during the 2-year observation period. As expected, women ages 80+ accounted for a disproportionate share of the incident hip fractures (P <.001). Women ages 65-79 years (3.3%) were somewhat more likely than those ages 50-64 (2.0%) or 80+ (3.0%) to initiate use of osteoporosis drugs during the 2-year observation period (P <.001).
      Table 1Characteristics of the Study Cohort
      Women with ≥3 years continuous primary medical and pharmacy coverage during the most recent enrollment period (≥2 years starting on or after January 1 following her 50th birthday), and no evidence of osteoporosis drug use, osteoporosis diagnosis, hip fracture, or evidence of end-stage renal disease, bone metastases, Paget disease, Cushing syndrome, neoplastic disease, or osteogenesis imperfecta at any point prior to the 2-year study period (ie, the most recent 2-year period of continuous enrollment).
      , by Age Group
      n (%) or Median [IQR]Age at Start of Study Period, yP-Value
      P-value for differences among age groups, using analysis of variance and the nonparametric Kruskal-Wallis test of ranks to assess differences in distributions of continuous variables and Pearson's chi-squared tests to assess for differences in categorical variables.
      Overall50-6465-7980+
      N = 1,638,454n = 1,061,951n = 439,096n = 137,407
      Coverage characteristics
       Duration of continuous enrollment prior to study period
      All women had at least 12 months of continuous enrollment prior to the study period.
      3.1 [2.0-6.0]4.0 [2.0-6.5]2.4 [1.6-4.0]3.0 [2.0-6.0]<.0001
      P-value applies to both analysis of variance and Kruskal-Wallis tests.
       Type of insurance coverage at start of study period<.0001
      Commercial1,049,717 (64.1)1,003,666 (94.5)44,305 (10.1)1746 (1.3)
      Medicare Advantage588,737 (35.9)58,285 (5.5)394,791 (89.9)135,661 (98.7)
       Calendar year at start of study period<.0001
      2008147,032 (9.0)115,951 (10.9)21,568 (4.9)9513 (6.9)
      2009128,263 (7.8)92,758 (8.7)24,759 (5.6)10,746 (7.8)
      2010132,277 (8.1)103,392 (9.7)19,932 (4.5)8953 (6.5)
      2011138,277 (8.4)101,259 (9.5)26,918 (6.1)10,100 (7.4)
      2012277,139 (16.9)183,797 (17.3)72,599 (16.5)20,743 (15.1)
      2013815,466 (49.8)464,794 (43.8)273,320 (62.2)77,352 (56.3)
      Sociodemographics
       Race/ethnicity
      The race/ethnicity data in the OptumLabs Data Warehouse represent a mix of self-report and public record data. Wherever possible, missing data were imputed based on census data and member name.
      <.0001
      White, non-Hispanic1,178,543 (71.9)759,606 (71.5)317,696 (72.4)101,241 (73.7)
      Black, non-Hispanic174,756 (10.7)101,821 (9.6)54,468 (12.4)18,467 (13.4)
      Asian, non-Hispanic45,102 (2.8)32,872 (3.1)9905 (2.3)2325 (1.7)
      Hispanic112,781 (6.9)80,901 (7.6)25,182 (5.7)6698 (4.9)
      Other/not reported127,272 (7.8)86,751 (8.2)31,845 (7.3)8676 (6.3)
       Estimated net worth
      The estimated net worth data in the OptumLabs Data Warehouse are based on an analysis of member assets (eg, financial holdings such as deposit accounts, investments, and home value) and liabilities (eg, loans, mortgages, and credit card debt).
      <.0001
      <$25,000130,451 (8.0)75,993 (7.2)39,647 (9.0)14,811 (10.8)
      $25,000-$149,999286,847 (17.5)174,367 (16.4)84,358 (19.2)28,122 (20.5)
      $150,000-$249,999240,007 (14.6)146,074 (13.8)71,232 (16.2)22,701 (16.5)
      $250,000-$499,999424,700 (25.9)277,922 (26.2)112,885 (25.7)33,893 (24.7)
      $500,000+400,203 (24.4)287,598 (27.1)90,576 (20.6)22,029 (16.0)
      Not reported156,246 (9.5)99,997 (9.4)40,398 (9.2)15,851 (11.5)
       Census region<.0001
      Northeast326,895 (20.0)204,570 (19.3)88,974 (20.3)33,351 (24.3)
      South621,019 (37.9)419,030 (39.5)160,057 (36.5)41,932 (30.5)
      Midwest426,312 (26.0)245,199 (23.1)133,931 (30.5)47,182 (34.3)
      West257,505 (15.7)188,067 (17.7)54,890 (12.5)14,548 (10.6)
      Unknown6723 (0.4)5085 (0.5)1244 (0.3)394 (0.3)
      Clinical characteristics
       Chronic comorbidity count
      The chronic comorbidity count is determined using diagnosis codes on claims from the 12 months of continuous enrollment prior to the study period.
      <.0001
      0298,107 (18.2)248,622 (23.4)42,220 (9.6)7265 (5.3)
      1210,289 (12.8)170,492 (16.1)33,454 (7.6)6343 (4.6)
      2226,237 (13.8)166,530 (15.7)49,390 (11.2)10,317 (7.5)
      3227,159 (13.9)148,913 (14.0)62,922 (14.3)15,324 (11.2)
      4195,327 (11.9)114,465 (10.8)62,897 (14.3)17,965 (13.1)
      5+481,335 (29.4)212,929 (20.1)188,213 (42.9)80,193 (58.4)
       Chronic comorbidity count
      The chronic comorbidity count is determined using diagnosis codes on claims from the 12 months of continuous enrollment prior to the study period.
      3.0 [1.0-5.0]2.0 [1.0-4.0]4.0 [2.0-6.0]5.0 [3.0-8.0]<.0001
      P-value applies to both analysis of variance and Kruskal-Wallis tests.
      Indicators of health care access and utilization during the 2-y study period
       Any health care utilization
      Utilization is defined as any claim for any encounter, procedure, lab test, or prescription fill.
      1,591,320 (97.1)1,024,349 (96.5)431,100 (98.2)135,871 (98.9)<.0001
       Any primary care encounters
      Primary care utilization is defined as any interaction with a family medicine, internal medicine, or obstetrics-gynecology provider, with a qualifying Evaluation and Management (E/M) Current Procedural Terminology (CPT), occurring outside of an inpatient setting (ie, where point of service was defined as office, home, or nursing home).
      1,404,491 (85.7)906,070 (85.3)381,126 (86.8)117,295 (85.4)<.0001
       Total number of primary care encounters
      Primary care utilization is defined as any interaction with a family medicine, internal medicine, or obstetrics-gynecology provider, with a qualifying Evaluation and Management (E/M) Current Procedural Terminology (CPT), occurring outside of an inpatient setting (ie, where point of service was defined as office, home, or nursing home).
      5.0 [2.0-8.0]4.0 [2.0-8.0]6.0 [2.0-9.0]6.0 [3.0-11.0]<.0001
      P-value applies to both analysis of variance and Kruskal-Wallis tests.
       Initiated use of pharmacologic therapies for osteoporosis40,163 (2.5)21,442 (2.0)14,648 (3.3)4073 (3.0)<.0001
       Any qualifying hip fractures8493 (0.5)899 (0.1)3007 (0.7)4587 (3.3)<.0001
      Utilization of bone mass measurement
      Universal screening is recommended for women ≥65 years of age; targeted screening based on individualized risk assessments is recommended for women <65 years of age.
       Any bone mass measurement during 2-y study period357,714 (21.8)223,554 (21.1)116,534 (26.5)17,626 (12.8)<.0001
       Any bone mass measurement during the last 3 y479,776 (29.3)301,343 (28.4)155,013 (35.3)23,420 (17.0)<.0001
       Any bone mass measurement prior to 2-y study period536,088 (32.7)346,109 (32.6)157,811 (35.9)32,168 (23.4)<.0001
       Any bone mass measurement during any period of enrollment740,254 (45.2)470,879 (44.3)225,920 (51.5)43,455 (31.6)<.0001
      IQR = interquartile range.
      Women with ≥3 years continuous primary medical and pharmacy coverage during the most recent enrollment period (≥2 years starting on or after January 1 following her 50th birthday), and no evidence of osteoporosis drug use, osteoporosis diagnosis, hip fracture, or evidence of end-stage renal disease, bone metastases, Paget disease, Cushing syndrome, neoplastic disease, or osteogenesis imperfecta at any point prior to the 2-year study period (ie, the most recent 2-year period of continuous enrollment).
      P-value for differences among age groups, using analysis of variance and the nonparametric Kruskal-Wallis test of ranks to assess differences in distributions of continuous variables and Pearson's chi-squared tests to assess for differences in categorical variables.
      All women had at least 12 months of continuous enrollment prior to the study period.
      § P-value applies to both analysis of variance and Kruskal-Wallis tests.
      The race/ethnicity data in the OptumLabs Data Warehouse represent a mix of self-report and public record data. Wherever possible, missing data were imputed based on census data and member name.
      The estimated net worth data in the OptumLabs Data Warehouse are based on an analysis of member assets (eg, financial holdings such as deposit accounts, investments, and home value) and liabilities (eg, loans, mortgages, and credit card debt).
      ∗∗ The chronic comorbidity count is determined using diagnosis codes on claims from the 12 months of continuous enrollment prior to the study period.
      †† Utilization is defined as any claim for any encounter, procedure, lab test, or prescription fill.
      ‡‡ Primary care utilization is defined as any interaction with a family medicine, internal medicine, or obstetrics-gynecology provider, with a qualifying Evaluation and Management (E/M) Current Procedural Terminology (CPT), occurring outside of an inpatient setting (ie, where point of service was defined as office, home, or nursing home).
      §§ Universal screening is recommended for women ≥65 years of age; targeted screening based on individualized risk assessments is recommended for women <65 years of age.
      Contrary to recommendations, few women ages 65+ years underwent screening during the 2-year study period (26.5% and 12.8% among women ages 65-79 and 80+ years, respectively). Screening rates were higher when we considered claims from the most recent 3-year window (35.3% and 17.0%, respectively) and when we considered all periods of enrollment captured in the database (51.5% and 31.6%, respectively). Unadjusted 2-year screening rates for the 65- to 79-year age group changed little between the start and the end of the study period (26.2% vs 26.8%; Table 2). In contrast, we observed a 37.7% improvement for women aged 80+ years, increasing from 10.6% to 14.6% (P <.001). Meanwhile, 21.1% of women ages 50-64 years were screened, overall, with unadjusted screening rates declining 31.4% over time, falling from 26.7% to 18.3% (P <.001). The distinct trajectories in age-specific screening rates persisted after adjusting for potential compositional changes within each age group over time (Figure 1; P <.001 for test for interaction).
      Table 2Utilization of Bone Mass Measurement
      Universal screening is recommended for women ≥65 years of age; targeted screening based on individualized risk assessments is recommended for women <65 years of age.
      During the Most Recent 2-Year Interval of Continuous Enrollment Among Women Enrolled in Commercial or Medicare Advantage Plans, 2008-2014
      OverallAge at Start of Study Period, y
      50-6465-7980+
      N = 1,638,454n = 1,061,951n = 439,096n = 137,407
      % (95% CI)P-Value
      P-value for differences within age group, using Pearson's chi-squared tests to assess for differences in categorical variables; * indicates P <.0001.
      % (95% CI)P-Value
      P-value for differences within age group, using Pearson's chi-squared tests to assess for differences in categorical variables; * indicates P <.0001.
      % (95% CI)P-Value
      P-value for differences within age group, using Pearson's chi-squared tests to assess for differences in categorical variables; * indicates P <.0001.
      % (95% CI)P-Value
      P-value for differences within age group, using Pearson's chi-squared tests to assess for differences in categorical variables; * indicates P <.0001.
      Coverage characteristics
       Type of insurance coverage at start of study period<.0001<.0001<.0001<.0001
      Commercial21.3 (21.2-21.3)21.1 (21.1-21.2)24.4 (24.0-24.8)8.1 (6.9-9.5)
      Medicare Advantage22.9 (22.8-23.0)19.6 (19.3-19.9)26.8 (26.6-26.9)12.9 (12.7-13.1)
       Calendar year at start of study period<.0001<.0001<.0001<.0001
      200825.6 (25.3-25.8)26.7 (26.4-26.9)26.2 (25.6-26.8)10.6 (10.0-11.2)
      200924.1 (23.8-24.3)24.9 (24.6-25.2)26.7 (26.2-27.3)10.7 (10.1-11.3)
      201023.2 (22.9-23.4)24.4 (24.1-24.6)23.8 (23.2-24.4)7.8 (7.3-8.4)
      201121.2 (21.0-21.4)21.6 (21.3-21.8)24.4 (23.9-24.9)8.8 (8.2-9.3)
      201221.5 (21.3-21.6)20.3 (20.1-20.5)27.2 (26.9-27.5)12.3 (11.9-12.8)
      201320.8 (20.7-20.9)18.3 (18.2-18.4)26.8 (26.6-27.0)14.6 (14.4-14.9)
      Sociodemographics
       Race/ethnicity
      The race/ethnicity data in the OptumLabs Data Warehouse represent a mix of self-report and public record data. Wherever possible, missing data were imputed based on census data and member name.
      <.0001<.0001<.0001<.0001
      White, non-Hispanic22.3 (22.2-22.3)21.6 (21.5-21.7)26.9 (26.7-27.0)13.0 (12.8-13.2)
      Black, non-Hispanic18.2 (18.0-18.3)16.8 (16.6-17.1)22.9 (22.6-23.3)11.4 (11.0-11.9)
      Asian, non-Hispanic22.7 (22.3-23.1)21.7 (21.2-22.1)28.5 (27.6-29.4)12.4 (11.1-13.8)
      Hispanic22.3 (22.1-22.6)21.3 (21.0-21.6)28.1 (27.5-28.6)13.5 (12.7-14.3)
      Other/not reported22.0 (21.8-22.3)20.8 (20.5-21.1)27.7 (27.3-28.2)13.8 (13.1-14.5)
       Estimated net worth
      The estimated net worth data in the OptumLabs Data Warehouse are based on an analysis of member assets (eg, financial holdings such as deposit accounts, investments, and home value) and liabilities (eg, loans, mortgages, and credit card debt).
      <.0001<.0001<.0001<.0001
      <$25,00017.3 (17.1-17.5)15.9 (15.7-16.2)22.3 (21.9-22.7)10.6 (10.1-11.1)
      $25,000-$149,99919.1 (19.0-19.3)17.9 (17.7-18.1)24.2 (23.9-24.4)11.6 (11.2-12.0)
      $150,000-$249,99920.9 (20.7-21.1)19.7 (19.5-19.9)25.8 (25.5-26.2)13.3 (12.8-13.7)
      $250,000-$499,99922.9 (22.8-23.1)21.9 (21.7-22.0)27.9 (27.7-28.2)15.0 (14.6-15.4)
      $500,000+25.9 (25.7-26.0)25.2 (25.0-25.4)30.5 (30.2-30.8)15.4 (15.0-15.9)
      Not reported18.8 (18.6-19.0)18.2 (18.0-18.5)24.2 (23.8-24.6)8.2 (7.8-8.7)
       Census region<.0001<.0001<.0001<.0001
      Northeast21.3 (21.1-21.4)21.0 (20.9-21.2)25.8 (25.5-26.0)10.8 (10.5-11.2)
      South24.6 (24.5-24.7)24.0 (23.9-24.1)28.8 (28.6-29.0)14.5 (14.2-14.9)
      Midwest19.1 (18.9-19.2)17.4 (17.3-17.6)24.5 (24.2-24.7)12.0 (11.8-12.3)
      West20.6 (20.5-20.8)19.4 (19.2-19.6)26.3 (25.9-26.6)15.0 (14.4-15.6)
      Unknown14.7 (13.9-15.6)12.0 (11.2-13.0)25.4 (23.1-27.9)15.2 (12.0-19.1)
      Clinical characteristics
       Chronic comorbidity count
      The chronic comorbidity count is determined using diagnosis codes on claims from the 12 months of continuous enrollment prior to the study period.
      <.0001<.0001<.0001<.0001
      013.7 (13.6-13.9)13.7 (13.6-13.8)15.1 (14.7-15.4)6.5 (6.0-7.1)
      121.1 (21.0-21.3)20.9 (20.7-21.0)24.7 (24.2-25.2)9.6 (8.9-10.4)
      223.2 (23.0-23.3)22.7 (22.5-22.9)27.2 (26.8-27.6)11.8 (11.1-12.4)
      323.9 (23.7-24.1)23.3 (23.1-23.5)27.9 (27.6-28.3)13.4 (12.8-13.9)
      424.4 (24.2-24.6)23.9 (23.7-24.2)28.3 (28.0-28.7)13.8 (13.3-14.3)
      5+24.5 (24.4-24.6)25.4 (25.2-25.6)28.2 (28.0-28.4)13.5 (13.2-13.7)
       History of bone mass measurement prior to 2-y study period<.0001<.0001<.0001<.0001
      Yes32.9 (32.7-33.0)33.0 (32.8-33.1)34.7 (34.5-34.9)22.6 (22.1-23.1)
      No16.5 (16.4-16.5)15.3 (15.2-15.4)22.0 (21.8-22.1)9.8 (9.7-10.0)
      Indicators of health care access and utilization during the 2-y study period
       Any primary care encounters
      Primary care utilization is defined as any interaction with a family medicine, internal medicine, or obstetrics-gynecology provider, with a qualifying Evaluation and Management (E/M) Current Procedural Terminology (CPT), occurring outside of an inpatient setting (ie, where point of service was defined as office, home, or nursing home).
      <.0001<.0001<.0001<.0001
      Yes24.0 (23.9-24.1)23.5 (23.4-23.6)28.3 (28.1-28.4)13.8 (13.6-14.0)
      No8.9 (8.8-9.0)6.9 (6.8-7.0)15.1 (14.8-15.4)7.0 (6.6-7.3)
       Filled any prescriptions for pharmacologic therapies for osteoporosis<.0001<.0001<.0001<.0001
      Yes78.7 (78.3-79.1)79.9 (79.3-80.4)80.9 (80.2-81.5)65.2 (63.7-66.7)
      No20.4 (20.3-20.5)19.8 (19.8-19.9)24.7 (24.5-24.8)11.2 (11.1-11.4)
       Any qualifying hip fractures<.0001<.0001<.0001<.0001
      Yes24.2 (23.3-25.2)30.9 (28.0-34.0)32.6 (31.0-34.3)17.4 (16.3-18.5)
      No21.8 (21.8-21.9)21.0 (21.0-21.1)26.5 (26.4-26.6)12.7 (12.5-12.8)
      CI = confidence interval.
      Universal screening is recommended for women ≥65 years of age; targeted screening based on individualized risk assessments is recommended for women <65 years of age.
      P-value for differences within age group, using Pearson's chi-squared tests to assess for differences in categorical variables; * indicates P <.0001.
      The race/ethnicity data in the OptumLabs Data Warehouse represent a mix of self-report and public record data. Wherever possible, missing data were imputed based on census data and member name.
      § The estimated net worth data in the OptumLabs Data Warehouse are based on an analysis of member assets (eg, financial holdings such as deposit accounts, investments, and home value) and liabilities (eg, loans, mortgages, and credit card debt).
      The chronic comorbidity count is determined using diagnosis codes on claims from the 12 months of continuous enrollment prior to the study period.
      Primary care utilization is defined as any interaction with a family medicine, internal medicine, or obstetrics-gynecology provider, with a qualifying Evaluation and Management (E/M) Current Procedural Terminology (CPT), occurring outside of an inpatient setting (ie, where point of service was defined as office, home, or nursing home).
      Figure thumbnail gr1
      Figure 1Trends in osteoporosis screening by age among women enrolled in commercial or Medicare Advantage plans, 2008-2014.
      Osteoporosis screening rates varied significantly by all factors assessed in bivariate analyses (P <.001). Rates were markedly lower among women with no evidence of primary care during the study period. Most women who initiated osteoporosis therapies underwent testing (78.7%), but surprisingly few women experiencing incident hip fracture did so (24.2%). Non-Hispanic black women in all age groups were the least likely to undergo screening (18.2%) compared with the other racial/ethnic categories assessed (range: 22.0%-22.7%; P <.001). Screening rates were lower for this group regardless of age, insurance type, calendar year, net worth, geographic area, underlying health status (approximated via chronic comorbidity count), history of prior bone mass measurement, and primary care utilization (P <.001 for all bivariate comparisons, data not shown).
      All factors of interest remained independently associated with osteoporosis screening in the multivariable models (Table 3). The nature of the association of age with screening differed by age group: for each additional year of age, odds of utilization increased for those ages 50-64 years (aOR = 1.03; 95% CI, 1.03-1.03), but decreased for those 65-79 and 80+ years (aOR = 0.97; 95% CI, 0.97-0.97 and aOR = 0.87; 95% CI, 0.86-0.87, respectively). For all women, odds of screening increased as net worth and chronic comorbidity count increased. Utilization of primary care services, history of screening, and experiencing a hip fracture during the study period were all important independent predictors of bone mass measurement.
      Table 3Factors Independently Associated with Utilization of Bone Mass Measurement
      Universal screening is recommended for women ≥65 years of age; targeted screening based on individualized risk assessments is recommended for women <65 years of age.
      During the Most Recent 2-Year Interval of Continuous Enrollment Among Women Enrolled in Commercial or Medicare Advantage Plans, 2008-2014
      50-6465-7980+
      n = 1,061,951n = 439,096n = 137,407
      aOR (95% CI)aOR (95% CI)aOR (95% CI)
      Coverage characteristics
       Calendar year at start of study period
      2008RefRefRef
      20090.88*** (0.86-0.89)0.89*** (0.85-0.93)1.03 (0.94-1.13)
      20100.80*** (0.78-0.81)0.74*** (0.71-0.78)0.78*** (0.70-0.86)
      20110.68*** (0.66-0.69)0.76*** (0.73-0.79)0.89* (0.80-0.98)
      20120.59*** (0.58-0.60)0.80*** (0.77-0.83)1.21*** (1.12-1.31)
      20130.55*** (0.54-0.56)0.86*** (0.84-0.89)1.58*** (1.47-1.70)
      Sociodemographics
       Age at start of study period, y1.03*** (1.03-1.03)0.97*** (0.97-0.97)0.87*** (0.86-0.87)
       Race/ethnicity
      The race/ethnicity data in the OptumLabs Data Warehouse represent a mix of self-report and public record data. Wherever possible, missing data were imputed based on census data and member name.
      White, non-HispanicRefRefRef
      Black, non-Hispanic0.82*** (0.81-0.84)0.87*** (0.85-0.89)0.92** (0.87-0.97)
      Asian, non-Hispanic1.20*** (1.16-1.23)1.18*** (1.13-1.24)0.97 (0.86-1.11)
      Hispanic1.14*** (1.11-1.16)1.13*** (1.10-1.16)1.08* (1.00-1.17)
      Other/not reported1.05*** (1.03-1.07)1.14*** (1.11-1.18)1.24*** (1.16-1.33)
       Estimated net worth
      The estimated net worth data in the OptumLabs Data Warehouse are based on an analysis of member assets (eg, financial holdings such as deposit accounts, investments, and home value) and liabilities (eg, loans, mortgages, and credit card debt).
      <$25,000RefRefRef
      $25,000-$149,9991.11*** (1.08-1.13)1.08*** (1.05-1.11)1.02 (0.96-1.09)
      $150,000-$249,9991.22*** (1.19-1.25)1.16*** (1.13-1.20)1.14*** (1.07-1.22)
      $250,000-$499,9991.38*** (1.35-1.41)1.25*** (1.22-1.29)1.25*** (1.18-1.34)
      $500,000+1.63*** (1.59-1.67)1.38*** (1.34-1.42)1.30*** (1.22-1.39)
      Not reported1.20*** (1.17-1.23)1.04* (1.00-1.07)0.77*** (0.71-0.84)
       Census division
      New EnglandRefRefRef
      Mid-Atlantic0.76*** (0.74-0.78)0.70*** (0.68-0.73)0.71*** (0.66-0.76)
      South Atlantic1.01 (0.98-1.03)1.01 (0.98-1.03)1.17*** (1.10-1.24)
      East North Central0.69*** (0.67-0.71)0.79*** (0.77-0.81)0.90*** (0.84-0.96)
      East South Central0.85*** (0.82-0.88)0.83*** (0.80-0.87)1.08 (0.98-1.18)
      West North Central0.70*** (0.68-0.72)0.77*** (0.75-0.79)0.85*** (0.80-0.91)
      West South Central1.04** (1.01-1.07)0.91*** (0.87-0.94)1.20** (1.07-1.35)
      Mountain0.95*** (0.92-0.97)0.94*** (0.90-0.97)1.32*** (1.22-1.44)
      Pacific0.81*** (0.79-0.84)0.87*** (0.84-0.91)1.36*** (1.25-1.48)
      Unknown1.03 (0.94-1.13)1.20** (1.05-1.37)2.03*** (1.52-2.72)
      Clinical characteristics
       Chronic comorbidity count
      The chronic comorbidity count is determined using diagnosis codes on claims from the 12 months of continuous enrollment prior to the study period.
      1.03*** (1.03-1.03)1.02*** (1.02-1.02)1.01*** (1.01-1.02)
       History of bone mass measurement prior to 2-y study period2.18*** (2.16-2.21)1.73*** (1.71-1.76)2.24*** (2.17-2.32)
      Indicators of health care access and utilization during the 2-y study period
       Any primary care encounters
      Primary care utilization is defined as any interaction with a family medicine, internal medicine, or obstetrics-gynecology provider, with a qualifying Evaluation and Management (E/M) Current Procedural Terminology (CPT), occurring outside of an inpatient setting (ie, where point of service was defined as office, home, or nursing home).
      3.50*** (3.43-3.58)2.02*** (1.97-2.07)1.82*** (1.72-1.93)
       Any qualifying hip fractures1.56*** (1.35-1.81)1.50*** (1.39-1.63)1.81*** (1.67-1.96)
      aOR = adjusted odds ratio; CI = confidence interval.
      *P <.05; **P <.01; ***P <.001.
      Universal screening is recommended for women ≥65 years of age; targeted screening based on individualized risk assessments is recommended for women <65 years of age.
      The race/ethnicity data in the OptumLabs Data Warehouse represent a mix of self-report and public record data. Wherever possible, missing data were imputed based on census data and member name.
      § The estimated net worth data in the OptumLabs Data Warehouse are based on an analysis of member assets (eg, financial holdings such as deposit accounts, investments, and home value) and liabilities (eg, loans, mortgages, and credit card debt).
      The chronic comorbidity count is determined using diagnosis codes on claims from the 12 months of continuous enrollment prior to the study period.
      Primary care utilization is defined as any interaction with a family medicine, internal medicine, or obstetrics-gynecology provider, with a qualifying Evaluation and Management (E/M) Current Procedural Terminology (CPT), occurring outside of an inpatient setting (ie, where point of service was defined as office, home, or nursing home).
      Even after controlling for all patient characteristics described above, non-Hispanic Asian and Hispanic women in the 50-64 and 65-79-year age groups had the highest odds of screening, whereas non-Hispanic black women had the lowest odds across all age groups in our cohort. Among women 50-64 years of age, screening odds for non-Hispanic blacks were 18% lower compared with non-Hispanic whites (aOR = 0.82; 95% CI, 0.81-0.84), 28% lower compared with Hispanics (aOR = 0.72; 95% CI, 0.71-0.74), and 31% lower compared with non-Hispanic Asians (aOR = 0.69; 95% CI, 0.66-0.71). These disparities persisted but were somewhat less pronounced among women 65-79 years of age (black vs white aOR = 0.87; 95% CI, 0.85-0.89; black vs Hispanic aOR = 0.77; 95% CI, 0.74-0.79; and black vs Asian aOR = 0.73; 95% CI, 0.69-0.77). There was more parity among women 80+ (black vs white aOR = 0.92; 95% CI, 0.87-0.97; black vs Hispanic aOR = 0.85; 95% CI, 0.78-0.93; and black vs. Asian aOR = 0.94; 95% CI, 0.82-1.08), for whom screening rates were low overall.
      Lastly, we detected significant interactions between calendar year and both race/ethnicity and net worth (Figure 2; P <.001 for tests for interaction). As a result, the disparity in the adjusted marginal screening probabilities comparing non-Hispanic White and non-Hispanic Black women was reduced by 66.0% among women ages 50-64 and was virtually eliminated among women aged 65+ years. Similarly, apparent disparities in screening probabilities derived from the fully adjusted models comparing the highest and lowest net worth categories decreased by 44.5%, 71.9%, and 59.7% among women ages 50-64, 65-79 and 80+ years, respectively, over the course of the study period.
      Figure thumbnail gr2
      Figure 2Disparities in osteoporosis screening by race/ethnicity and net worth among women enrolled in commercial or Medicare Advantage plans, 2008-2014.

      Discussion

      We used medical claims data from a large, nationwide cohort to assess trends and disparities in osteoporosis screening. Between 2008 and 2014, fewer than 1 in 4 women 65 years and older underwent bone mass measurement despite recommendations for universal screening. Screening rates among women 80+ years of age increased markedly, yet remain low. These findings are largely consistent with previous claims-based studies
      • Neuner J.M.
      • Binkley N.
      • Sparapani R.A.
      • Laud P.W.
      • Nattinger A.B.
      Bone density testing in older women and its association with patient age.
      • Zhang J.
      • Delzell E.
      • Zhao H.
      • et al.
      Central DXA utilization shifts from office-based to hospital-based settings among Medicare beneficiaries in the wake of reimbursement changes.
      • McAdam-Marx C.
      • Unni S.
      • Ye X.
      • Nelson S.
      • Nickman N.A.
      Effect of Medicare reimbursement reduction for imaging services on osteoporosis screening rates.
      and demonstrate that there has been only limited improvement in osteoporosis screening rates among older women since Medicare began covering the service nearly 2 decades ago. In contrast, utilization among younger women in the general population, which was included in the 2011 Top 5 List
      Good Stewardship Working Group
      The “top 5” lists in primary care: meeting the responsibility of professionalism.
      of practices to avoid as part of an initial phase of the Choosing Wisely Campaign,

      Choosing Wisely: An Initiative of the ABIM Foundation. Available at: http://www.choosingwisely.org/. Accessed May 16, 2016.

      and is not routinely recommended due to a lack of cost-effectiveness,
      U.S. Preventive Services Task Force
      Screening for osteoporosis: U.S. preventive services task force recommendation statement.
      declined steadily.
      In 2011, the USPSTF issued an update to their recommendations, which underscored the utility of universal screening among all women ages 65+.
      U.S. Preventive Services Task Force
      Screening for osteoporosis: U.S. preventive services task force recommendation statement.
      The same year, the Centers for Medicare and Medicaid Services began tracking utilization of osteoporosis screening and treatment via the Physician Quality Reporting System, using incentivized provider reporting and peer comparisons in an attempt to improve health care delivery.

      Centers for Medicare and Medicaid Services. Physician Quality Reporting System. Available at: https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/PQRS/index.html?redirect=/pqri/. Accessed May 16, 2016.

      In addition, by 2011, the Affordable Care Act had eliminated all cost sharing for this service for privately insured women age 65 and older (and younger women, when indicated) and qualified Medicare beneficiaries.

      Patient Protection and Affordable Care Act. Public Law Number 111-148. (H.R. 3590); 2010. Available at: https://www.gpo.gov/fdsys/pkg/PLAW-111publ148/pdf/PLAW-111publ148.pdf. Accessed May 16, 2016.

      Our data suggest that the impact of this combination of external factors was modest at best.
      The large size and geographic diversity of our study cohort allowed us to assess utilization of osteoporosis screening services within a diverse cohort, including nearly 160,000 Hispanic and Asian women. We found that these women had the highest odds of screening, even higher than non-Hispanic whites, whereas non-Hispanic black women had the lowest odds of screening. Like others,
      • Zhang J.
      • Delzell E.
      • Zhao H.
      • et al.
      Central DXA utilization shifts from office-based to hospital-based settings among Medicare beneficiaries in the wake of reimbursement changes.
      we also observed a narrowing of disparities over time. Our findings are largely consistent with survey data on self-reported lifetime use of osteoporosis screening,
      U.S. Department of Health and Human Services
      National Healthcare Disparities Report 2011.
      which show lower utilization rates among non-Hispanic black women and women of low socioeconomic status. These patterns may reflect perceived risk differences, access to affordable care, and compliance with physician referrals. Although the USPSTF reports that fracture risks are highest for non-Hispanic white women, the quantity and quality of data pertaining to other racial/ethnic groups is lacking.
      U.S. Preventive Services Task Force
      Screening for osteoporosis: U.S. preventive services task force recommendation statement.
      The National Health and Nutrition Examination Survey (NHANES) oversamples non-Hispanic blacks and Mexican Americans in an effort to address such deficiencies. The age-adjusted prevalence of osteoporosis based on standardized NHANES bone mass measurements is indeed lowest among non-Hispanic black women, but about 75% higher among Mexican Americans compared with non-Hispanic White women.
      • Looker A.C.
      • Borrud L.G.
      • Dawson-Hughes B.
      • Shepherd J.A.
      • Wright N.C.
      Osteoporosis or Low Bone Mass at the Femur Neck or Lumbar Spine in Older Adults: United States, 2005-2008. NCHS Data Brief No. 93.
      Too few non-Hispanic Asian women are captured in NHANES to calculate true population-based estimates for this group.
      Despite the large size and availability of rich covariate data, our study is subject to several limitations. Although our claims were drawn from a nationwide sample, we lack information about osteoporosis screening among uninsured, Medicaid, or Medicare fee-for-service populations. Additionally, we could not account for differences in underlying risk factors that may have influenced screening behaviors, such as family history of osteoporosis, and would have allowed us to assess whether or not declines in screening rates among women ages 50-64 years were clinically indicated. Our finding that screening rates in this group fell from 26.7% to 18.3% over the study period may reflect progress toward better adherence to the targeted screening recommendations for this younger age group. A recent analysis of risk factor data collected from 5165 postmenopausal women ages 50-64 years who were participants of the Women's Health Initiative identified just 15.2% as meeting the USPSTF criteria for osteoporosis screening among women <65 years of age.
      • Crandall C.J.
      • Larson J.
      • Gourlay M.L.
      • et al.
      Osteoporosis screening in postmenopausal women 50 to 64 years old: comparison of US Preventive Services Task Force strategy and two traditional strategies in the Women's Health Initiative.
      However, the USPSTF risk assessment tool has been found to be insensitive in some populations
      • Crandall C.J.
      • Larson J.
      • Gourlay M.L.
      • et al.
      Osteoporosis screening in postmenopausal women 50 to 64 years old: comparison of US Preventive Services Task Force strategy and two traditional strategies in the Women's Health Initiative.
      • Bansal S.
      • Pecina J.L.
      • Merry S.P.
      • et al.
      US Preventative Services Task Force FRAX threshold has a low sensitivity to detect osteoporosis in women ages 50-64 years.
      • Pecina J.L.
      • Romanovsky L.
      • Merry S.P.
      • Kennel K.A.
      • Thacher T.D.
      Comparison of clinical risk tools for predicting osteoporosis in women ages 50-64.
      and mismatch between calculated risk and real-world screening practices has been documented. For example, in one regional health system, uptake of osteoporosis screening was uncommon overall and fairly similar among women ages 40-85 years with and without identified risk factors (25.1% and 19.3%, respectively).
      • Amarnath A.L.D.
      • Franks P.
      • Robbins J.A.
      • Xing G.
      • Fenton J.J.
      Underuse and overuse of osteoporosis screening in a regional health system: a retrospective cohort study.
      Finally, we interpreted our findings pertaining to women 65+ years of age in the context of the current USPSTF recommendations, which call for universal screening of this group at 2-year intervals. In reality, prior test results or other clinical factors such as remaining life expectancy may justifiably support a decision to forego screening for this group. More data on the benefits of repeat testing
      • Berry S.D.
      • Samelson E.J.
      • Pencina M.J.
      • et al.
      Repeat bone mineral density screening and prediction of hip and major osteoporotic fracture.
      and optimal screening intervals
      • Gourlay M.L.
      • Fine J.P.
      • Preisser J.S.
      • et al.
      Bone-density testing interval and transition to osteoporosis in older women.
      are needed.
      In conclusion, despite significant changes in utilization among women ages 50-64 and those 80+ years, our study underscores the large, persistent deficiencies in real-world implementation of evidence-based guidelines for osteoporosis screening. While increased awareness of the lack of cost-effectiveness among younger women may have contributed to declines in utilization among women 50-64 years of age, thus far, performance improvement strategies, including provider incentives, and efforts to eliminate cost barriers for consumers have been insufficient to ensure widespread utilization of this recommended evidenced-based service among women aged 65+ years. Osteoporosis screening and management should be an essential component of primary care for older Americans. Further work is needed to identify promising practices for increasing appropriate utilization.

      Supplementary Data

      References

        • Looker A.C.
        • Borrud L.G.
        • Dawson-Hughes B.
        • Shepherd J.A.
        • Wright N.C.
        Osteoporosis or Low Bone Mass at the Femur Neck or Lumbar Spine in Older Adults: United States, 2005-2008. NCHS Data Brief No. 93.
        National Center for Health Statistics, Hyattsville, MD2012
        • U.S. Department of Health and Human Services
        Bone Health and Osteoporosis: A Report of the Surgeon General.
        U.S. Department of Health and Human Services, Office of the Surgeon General, Rockville, MD2004
        • U.S. Preventive Services Task Force
        Screening for osteoporosis: U.S. preventive services task force recommendation statement.
        Ann Intern Med. 2011; 154: 356-364
        • Haentjens P.
        • Magaziner J.
        • Colón-Emeric C.S.
        • et al.
        Meta-analysis: excess mortality after hip fracture among older women and men.
        Ann Intern Med. 2010; 152: 380-390
        • Nelson H.
        • Haney E.
        • Chou R.
        • Dana T.
        • Fu R.
        • Bougatsos C.
        Screening for Osteoporosis: Systematic Review to Update the 2002 U.S. Preventive Services Task Force Recommendation.
        U.S. Agency for Healthcare Research and Quality, Rockville, MD2010
        • National Osteoporosis Foundation
        Clinician's Guide to Prevention and Treatment of Osteoporosis.
        National Osteoporosis Foundation, Washington, DC2010
        • U.S. Preventive Services Task Force
        Screening for osteoporosis in postmenopausal women: recommendations and rationale.
        Ann Intern Med. 2002; 137: 526-528
      1. Balanced Budget Act of 1997. Public Law Number 105-33. (H.R. 2015); 1997. Available at: https://www.gpo.gov/fdsys/pkg/PLAW-105publ33/pdf/PLAW-105publ33.pdf. Accessed May 16, 2016.

      2. Patient Protection and Affordable Care Act. Public Law Number 111-148. (H.R. 3590); 2010. Available at: https://www.gpo.gov/fdsys/pkg/PLAW-111publ148/pdf/PLAW-111publ148.pdf. Accessed May 16, 2016.

        • Neuner J.M.
        • Binkley N.
        • Sparapani R.A.
        • Laud P.W.
        • Nattinger A.B.
        Bone density testing in older women and its association with patient age.
        J Am Geriatr Soc. 2006; 54: 485-489
        • Zhang J.
        • Delzell E.
        • Zhao H.
        • et al.
        Central DXA utilization shifts from office-based to hospital-based settings among Medicare beneficiaries in the wake of reimbursement changes.
        J Bone Miner Res. 2012; 27: 858-864
        • McAdam-Marx C.
        • Unni S.
        • Ye X.
        • Nelson S.
        • Nickman N.A.
        Effect of Medicare reimbursement reduction for imaging services on osteoporosis screening rates.
        J Am Geriatr Soc. 2012; 60: 511-516
        • U.S. Department of Health and Human Services
        National Healthcare Disparities Report 2011.
        Agency for Healthcare Research and Quality, Rockville, MD2012
        • National Committee for Quality Assurance
        The State of Health Care Quality 2015.
        National Committee for Quality Assurance, Washington, DC2015
        • Wallace P.J.
        • Shah N.D.
        • Dennen T.
        • Bleicher P.A.
        • Crown W.H.
        Optum Labs: building a novel node in the learning health care system.
        Health Aff (Millwood). 2014; 33: 1187-1194
      3. Elixhauser A, Steiner C, Palmer L. Clinical Classifications Software (CCS). U.S. Agency for Healthcare Research and Quality; 2014. Available at: http://www.hcup-us.ahrq.gov/toolssoftware/ccs/ccs.jsp. Accessed November 22, 2016.

        • Hwang W.
        • Weller W.
        • Ireys H.
        • Anderson G.
        Out-of-pocket medical spending for care of chronic conditions.
        Health Aff (Millwood). 2001; 20: 267-278
        • Naessens J.M.
        • Stroebel R.J.
        • Finnie D.M.
        • et al.
        Effect of multiple chronic conditions among working-age adults.
        Am J Manag Care. 2011; 17: 118-122
        • Good Stewardship Working Group
        The “top 5” lists in primary care: meeting the responsibility of professionalism.
        Arch Intern Med. 2011; 171: 1385-1390
      4. Choosing Wisely: An Initiative of the ABIM Foundation. Available at: http://www.choosingwisely.org/. Accessed May 16, 2016.

      5. Centers for Medicare and Medicaid Services. Physician Quality Reporting System. Available at: https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/PQRS/index.html?redirect=/pqri/. Accessed May 16, 2016.

        • Crandall C.J.
        • Larson J.
        • Gourlay M.L.
        • et al.
        Osteoporosis screening in postmenopausal women 50 to 64 years old: comparison of US Preventive Services Task Force strategy and two traditional strategies in the Women's Health Initiative.
        J Bone Miner Res. 2014; 29: 1661-1666
        • Bansal S.
        • Pecina J.L.
        • Merry S.P.
        • et al.
        US Preventative Services Task Force FRAX threshold has a low sensitivity to detect osteoporosis in women ages 50-64 years.
        Osteoporos Int. 2015; 26: 1429-1433
        • Pecina J.L.
        • Romanovsky L.
        • Merry S.P.
        • Kennel K.A.
        • Thacher T.D.
        Comparison of clinical risk tools for predicting osteoporosis in women ages 50-64.
        J Am Board Fam Med. 2016; 29: 233-239
        • Amarnath A.L.D.
        • Franks P.
        • Robbins J.A.
        • Xing G.
        • Fenton J.J.
        Underuse and overuse of osteoporosis screening in a regional health system: a retrospective cohort study.
        J Gen Intern Med. 2015; 30: 1733-1740
        • Berry S.D.
        • Samelson E.J.
        • Pencina M.J.
        • et al.
        Repeat bone mineral density screening and prediction of hip and major osteoporotic fracture.
        JAMA. 2013; 310: 1256-1262
        • Gourlay M.L.
        • Fine J.P.
        • Preisser J.S.
        • et al.
        Bone-density testing interval and transition to osteoporosis in older women.
        N Engl J Med. 2012; 366: 225-233