Skip to main content
NCBI home page
VA Author Manuscripts logoLink to VA Author Manuscripts
. Author manuscript; available in PMC: 2019 Jul 15.
Published in final edited form as: Spine (Phila Pa 1976). 2018 Jul 15;43(14):1007–1017. doi: 10.1097/BRS.0000000000002499

Associations of Race and Ethnicity with Patient-Reported Outcomes and Health Care Utilization among Older Adults Initiating a New Episode of Care for Back Pain

Carlo J Milani 1, Sean Rundell 1,2,12, Jeffrey G Jarvik 2,3, Janna Friedly 1,2, Patrick J Heagerty 2,3,4,5, Andy Avins 6, David Nerenz 7, Laura S Gold 2, Judith A Turner 1,2,11, Thiru Annaswamy 9, Srdjan S Nedeljkovic 10, Pradeep Suri 1,2,4
PMCID: PMC5972040  NIHMSID: NIHMS920466  PMID: 29189640

Abstract

Study Design

Secondary analysis of the Back Pain Outcomes using Longitudinal Data (BOLD) cohort study.

Objective

To characterize associations of self-reported race/ethnicity with back pain (BP) patient-reported outcomes (PROs) and health care utilization among older adults with a new episode of care for BP.

Summary of Background Data

No prior longitudinal studies have characterized associations between multiple race/ethnicity groups, and BP-related PROs and health care utilization in the US.

Methods

This study included 5,117 participants ≥65 years from three US health care systems. The primary BP-related PROs were BP intensity and back-related functional limitations over 24 months. Health care utilization measures included common diagnostic tests and treatments related to BP (spine imaging, spine-related relative value units (RVUs), and total RVUs) over 24 months. Analyses were adjusted for multiple potential confounders including sociodemographics, clinical characteristics, and study site.

Results

Baseline BP ratings were significantly higher for Blacks vs. Whites (5.8 vs. 5.0; p<0.001). Participants in all race/ethnicity groups showed statistically significant, but modest improvements in BP over 24 months. Blacks and Hispanics did not have statistically significant improvement in BP-related functional limitations over time unlike Whites, Asians, and non-Hispanics; however, the magnitude of differences in improvement between groups was small. Blacks had less spine-related health care utilization over 24 months than Whites (spine-related RVU ratio of means 0.66, 95% confidence interval[CI] 0.51–0.86). Hispanics had less spine-related health care utilization than non-Hispanics (spine-related RVU ratio of means 0.60; 95% CI 0.40–0.90).

Conclusions

Blacks and Hispanics had slightly less improvement in BP-related functional limitations over time and less spine-related health care utilization as compared to Whites and non-Hispanics, respectively. Residual confounding may explain some of the association between race/ethnicity and health outcomes. Further studies are needed to understand the factors underlying these differences and which differences reflect disparities.

Keywords: older adults, race, ethnicity, low back pain, functional limitation, disability, utilization, disparity, Black, Asian, Hispanic

INTRODUCTION

Back pain (BP) is a leading cause of disability worldwide1 and results in an economic burden exceeding $100 billion annually in the United States (US).2,3 Prior studies have reported that Blacks in the US receive less diagnostic workup and treatment for BP compared to Whites4,5 despite reporting greater pain intensity and functional limitations.4,6–8 Few studies have evaluated back-related outcomes comparing other races or ethnicities; however, two reports suggest a similar pattern of less treatment despite greater reported BP intensity among Hispanics compared to non-Hispanics.5,9 Differences in back-related outcomes and health care utilization associated with race and ethnicity may be explained by patient-, provider-, and/or system-level factors.10 An accurate understanding of how the typical course and management of BP vary for different racial and ethnic groups in the US is a necessary step toward identifying possible disparities in medical care. The importance of this goal is highlighted by current projections of US demographics which indicate that by 2050, >50% of the US population will be non-White or Hispanic.11–13

We used data from the Back pain Outcomes using Longitudinal Data (BOLD) cohort to examine associations of self-reported race and ethnicity with back-related outcomes and health care utilization for older adults with new episodes of primary care for BP. Given that race and ethnicity are different social constructs that may be distinct or overlap we refer to them jointly here as ‘race/ethnicity’.14 The primary aim of this study was to examine longitudinal associations between race/ethnicity and patient-reported outcomes over 2-year follow-up after adjustment for potential confounders. The secondary aim was to examine associations between race/ethnicity and health care utilization.

METHODS

Definitions

Differences in back-related PROs or health care utilization between race/ethnicity groups may exist for various reasons. The term "disparity" in health services and policy research often pertains specifically to differences between groups that might signal injustice or inequity.14 The Healthy People 2020 Initiative written by the US Office of Disease Prevention and Health Promotion considers a “disparity” to be a difference between groups that is closely linked with social, economic, and/or environmental disadvantage.15 In this article, we apply the term “difference” to compare race/ethnicity groups without inferences as to underlying causes and use the term “disparities” for the more specific instance where differences are thought to be explained by socioeconomic and/or environmental disadvantage.

Study Design and Participant Sample

The BOLD cohort has been described previously.16 In brief, 5,239 participants ≥ 65 years who presented for a new (‘index’) primary care or emergency department visit for BP were enrolled between 2010 and 2013. Participants had no back-related visits within the six months prior to the index visit, and received care from one of three US health care systems: Henry Ford Health System, Kaiser Permanente Northern California, and Harvard Vanguard Medical Associates. For the current study, we excluded participants without available electronic medical record (EMR) data during the study period or who had unknown clinic locations for their index visit. The institutional review boards at the University of Washington and all sites approved the BOLD study methods and all partcipants provided informed consent.

Patient-Reported Race and Ethnicity

The US Office of Management and Budget (OMB) uses the 1997 Revisions to the Standards for the Classification of Federal Data on Race and Ethnicity to define race and ethnicity categories for official data collection.17 These revised standards include five race categories (American Indian or Alaska Native, Asian, Black or African American, Native Hawaiian or Other Pacific Islander, and White) and two ethnicity categories (Hispanic or Latino and non-Hispanic or non-Latino).18,19 This classification scheme provides standard categories for self-identification, collection, and description of racial and ethnic data.18 In this study, individuals could report none, one, or more than one race/ethnicity group. Race and ethnicity were ascertained in two separate questions and thus were not mutually exclusive categories.

Patient-Reported Outcomes (PROs)

PROs were collected via phone interview or mailed questionnaires at baseline and at 3, 6, 12, and 24 months later. The baseline assessment occurred within 3 weeks of a participant’s index visit for BP. The primary PROs were BP intensity and back-related functional limitations. Participants used a 0 to 10 numerical rating scale (NRS) to rate their average BP intensity during the past 7 days. We also examined proportions with ≥30% improvement in BP ratings at 24 months.20 A modified version of the Roland-Morris Disability Questionnaire (RMDQ) was used to measure functional limitations related to back and leg pain (sciatica) and is described elsewhere.21 We also examined proportions with ≥30% improvement in RMDQ scores at 24 months.22,23 A secondary outcome was BP resolution at 24-month follow-up defined as ≥30 days with no BP. BP resolution questions were added after study enrollment began and were obtained in a subset of participants (n=1943).24

Health Care Utilization Outcomes

Health care utilization measures including diagnostic imaging and common treatments used for BP were obtained from the EMR. EMR data were collected from 12 months before the index visit through 24 months after the index visit or until a participant withdrew or died. EMR data included Current Procedural Terminology 4 (CPT) codes and International Classification of Diseases Clinical Modification (ICD-9-CM) codes. We used CPT and ICD-9-CM codes to identify health care utilization potentially related to BP including lumbar spine imaging (radiographs, computed tomography [CT], magnetic resonance imaging [MRI]), physical therapy (PT) visits, opioid prescriptions written, psychotherapy visits, and spine surgeries for each participant from the index visit through 24 months after the index visit. Opioid prescriptions, PT, and psychotherapy visits were not specifically linked to BP and may have reflected some treatment for other conditions. Since guidelines recommend a selective approach to obtaining imaging for BP25, we also examined the subcategories of spine radiographs (‘x-rays’) or MRI/CT ordered ‘early’ (<6 weeks from the date of the index visit). We mapped each CPT code to its calendar year-specific relative value unit (RVU), a metric which reflects the relative amount of work attributed to each encounter used by the Centers for Medicare and Medicaid Services and most other payers.26–28 For each individual, we determined cumulative health care utilization by summing RVUs (total RVUs and “spine-specific RVUs”) accumulated from the date of the index visit through 24 months after the index visit using methods described elsewhere.29–31

Covariates

Covariates were chosen for this analysis based on conceptual importance as possible confounders. Sociodemographic covariates included patient-reported age, sex, marital status, educational attainment, employment status, and ongoing litigation related to current pain. Administrative/EMR data based on study-site included private/commercial health insurance or Medicaid coverage and location of the index visit (primary care vs. emergency department vs. other). All participants were Medicare-eligible due to age. We used US census data to define tertiles of mean per capita income for the cities in which the index visits took place because more specific location data (address or zip code) were not available.32 Participants reported their duration of BP at baseline and confidence in BP recovery rated on a 0 to 10 scale where ‘0’ reflected ‘not at all confident’ and ‘10’ reflected ‘extremely confident’. Depressive and anxiety symptoms were assessed with the Patient Health Questionnaire-2(PHQ-2) and Generalized Anxiety Disorder scale-2 (GAD-2).33,34 ICD-9-CM codes over 12 months prior to the index visit were used to generate the Quan comorbidity index.35 Participants reported whether they smoked currently, had quit within the past year, or never smoked. For each participant, we summed all RVUs accumulated in the 12 months prior to the index visit as a reflection of baseline overall health care utilization.

Statistical Analysis

We used descriptive statistics to examine participant baseline characteristics according to race/ethnicity groups. We used linear regression to compare baseline pain intensity and RMDQ scores between race/ethnicity groups. We used a longitudinal generalized estimating equation (GEE) approach for regression models36 to examine changes over time within each race/ethnicity group and to compare changes over time between race/ethnicity groups (characterized by race/ethnicity × time interactions). We used an analogous approach to examine associations between race/ethnicity and other back-related outcomes over 24 months using logistic regression for binary outcomes (BP resolution and health care utilization categories) and GEE with a log link and gamma distribution for RVU outcomes. All models were adjusted for baseline values of the outcome, study site, and the covariates described above and accounted for non-response using inverse probability weighting.37 Race analyses were performed separately from ethnicity analyses to provide parallel evaluations. Analyses of race were restricted to comparisons between Whites, Blacks, and Asians, excluding race groups where the sample size was insufficient to permit reliable statistical inferences (Native American/Alaskan Natives and Hawaiian/Pacific Islanders) or where race designations did not permit interpretable comparisons (‘other’, ‘unknown’, or multiple races). Whites were considered the reference group for between-race statistical comparisons, and non-Hispanics were the reference group for between-ethnicity comparisons, since these represented the largest groups in our study. Two-sided P-values < 0.05 were considered statistically significant. Analyses were performed using Stata IC, version 14 (Stata Corp., College Station, TX).

RESULTS

Patient-reported Outcomes

A participant flow diagram is presented in Figure 1. Baseline characteristics of the cohort (n=5117) are presented in Table 1. The majority of participants were female (65%). There were other differences in baseline characteristics including, but not limited to, educational attainment (8% of Blacks completed graduate school vs. 19% of Whites), per capita income of the city where the index visit occurred (65% of Blacks in the lowest income tertile vs. 24% of Whites), and clinical location of index visit (13% Blacks in emergency department/urgent care vs. 3% Whites). Race/ethnicity groups were concentrated at certain sites; for example, 61% of Blacks were from the Detroit area and 64% of Whites, 84% of Asians, and 89% of Hispanics were from Northern California.

Figure 1. Flowchart of study participants*.

Figure 1

Open in a new tab

*missing follow-up data accounted for using non-response weighting

Table 1.

Baseline characteristics by race and ethnicity (unadjusted)*

Characteristic Whitea
(n = 3746)		 Blacka
(n = 781)		 Asiana
(n = 195)		 non-Hispanica
(n = 4791)		 Hispanica
(n = 302)
Sociodemographics
Age (years) 74.1 (7.0) 73.1 (6.5) 73.6 (6.8) 73.9 (6.9) 72.9 (6.0)
Female sex 2397 (64) 561 (72) 106 (54) 3093 (65) 202 (67)
Education
  Less than high school 147 (4) 114 (15) 5 (3) 268 (6) 37 (12)
  High school graduate/GED/voc/tech 943 (25) 303 (39) 34 (17) 1299 (27) 119 (39)
  Some college 888 (24) 217 (28) 39 (20) 1154 (24) 93 (31)
  Four-year college graduate 1056 (28) 85 (11) 66 (34) 1224 (26) 24 (8)
  Professional/graduate degree 708 (19) 61 (8) 50 (26) 839 (18) 27 (9)
Marital
  Married/living with partner 2421 (65) 331 (42) 135 (69) 2930 (61) 164 (54)
  Separated/divorced 379 (10) 138 (18) 13 (7) 537 (11) 52 (17)
  Single, never married 169 (5) 57 (7) 11 (6) 239 (5) 12 (4)
  Widowed 770 (21) 254 (33) 36 (18) 1076 (22) 73 (24)
Work
  Full-time/part-time 445 (12) 70 (9) 15 (8) 536 (11) 35 (12)
  Retired (not due to ill health) 3047 (81) 639 (82) 166 (85) 3904 (81) 244 (81)
  Retired/disabled (due to health) 77 (2) 45 (6) 7 (4) 135 (3) 7 (2)
  Other 165 (4) 25 (3) 7 (4) 201 (4) 15 (5)
Ongoing litigation related to current pain 15 (0.4) 7 (0.9) 2 (1.0) 28 (0.6) 2 (0.7)
Health insuranceb
  Any private insurance 2498 (67) 579 (74) 131 (67) 3266 (68) 177 (59)
  AnyMedicaid 70 (1.9) 88 (11.3) 9 (4.6) 172 (3.6) 5 (1.7)
Study sitec
  Northern California 2413 (64) 225 (29) 171 (88) 2853 (60) 269 (89)
  Detroit, Michigan metropolitan area 414 (11) 474 (61) 16 (8) 913 (19) 20 (7)
  Boston, Massachusetts metropolitan area 919 (25) 82 (10) 8 (4) 1025 (21) 13 (4)
Per capita income of city where index visit occurredd
  Lowest income tertile 907 (24) 510 (65) 55 (28) 1487 (31) 114 (38)
  Medium income tertile 1504 (40) 189 (24) 61 (31) 1795 (37) 110 (36)
  High income tertile 1335 (36) 82 (10) 79 (41) 1509 (31) 78 (26)
Index visit clinical location
  ED or urgent care visit 95 (3) 103 (13) 4 (2) 204 (4) 9 (3)
  Primary care 3651 (97) 678 (87) 191 (98) 4587 (96) 293 (97)
Back Pain Characteristics and Comorbidities
Duration of back pain
  < 1 month 1262 (34) 224 (29) 78 (40) 1582 (33) 115 (38)
  1–3 months 752 (20) 121 (15) 48 (25) 937 (20) 51 (17)
  3–6 months 268 (7) 37 (5) 13 (7) 318 (7) 16 (5)
  6–12 months 209 (6) 70 (9) 10 (5) 286 (6) 21 (7)
  1–5 years 511 (14) 148 (19) 26 (13) 706 (15) 51 (17)
  >5 years 740 (20) 181 (23) 20 (10) 958 (20) 48 (16)
Back pain intensity ratings (0–10)e 4.8 (2.7) 6.2 (2.8) 4.9 (2.7) 5.0 (2.8) 5.4 (2.8)
Back-related functional limitations (RMDQ; 0–24)f 8.9 (6.3) 12.1 (6.4) 9.7 (6.2) 9.5 (6.4) 10.7 (6.1)
Confidence in back pain recovery (0–10)g 5.4 (3.7) 5.3 (3.7) 6.7 (3.3) 5.4 (3.7) 6.0 (3.6)
Quan comorbidity indexh 0.8 (1.4) 1.2 (1.7) 0.8 (1.5) 0.9 (1.4) 0.8 (1.4)
PHQ-2 scorei 0.6 (1.2) 0.8 (1.3) 0.6 (1.3) 0.7 (1.2) 0.9 (1.3)
GAD-2 scorej 0.8 (1.4) 1.0 (1.6) 1.2 (1.7) 0.9 (1.5) 1.3 (1.5)
Current smoking 197 (5) 78 (10) 7 (4) 295 (6) 16 (5)
Health care utilization (in 12 months prior to index visit)
Total Relative Value Units 15.8 (7.2, 35.6) 20.8 (8.4, 45.4) 11.1 (4.7, 31.2) 16.2 (7.3, 37.1) 17.0 (8.0, 36.9)
Any opioid prescriptionk 526 (14) 98 (13) 25 (13) 666 (14) 51 (17)
Open in a new tab

SD = standard deviation; IQR = interquartile range

*

Values presented include means (standard deviations) or medians (interquartile range: bottom quartile, top quartile) for continuous variables, and numbers (%) for categorical variables. All percentages represent a proportion within each respective race or ethnicity group.

a

Individuals could choose none, one, or more than one race/ethnicity group. Analysis was restricted to participants who self-reported only one race group, or one ethnicity group.

b

Data on primary and secondary insurance carriers were grouped into the two non-mutually exclusive categories of ‘any private insurance’ or ‘any Medicaid’.

c

Northern California site is Kaiser Permanente health system; Detroit site is Henry Ford health system; Boston site is Harvard Vanguard health system.

d

Per capita income data obtained from www.census.gov/quickfacts, accessed 8/1/2015. All values are per capita income in 2013 dollars(2009–2013).

e

Numerical rating scale; higher scores represent more pain.

f

Roland-Morris Disability Questionnaire (RMDQ); higher scores represent greater functional limitations.

g

Expectation that back and/or leg pain would be completely gone or much better 3 months; rated on a 0 to 10 scale; 0 = ‘not at all confident’ and 10 = ‘extremely confident’.

h

ICD-9-CM codes over the 12 months prior to the index visit were used to generate the Quan comorbidity index.

i

PHQ-2 (Patient Health Questionnaire-2): higher scores represent greater depressive symptoms.

j

GAD-2 (Generalized Anxiety Disorder-2): higher scores represent greater anxiety symptoms.

k

Number of participants who had at least one opioid prescription written during the 12 months prior to the index visit.

Figure 2 and Table S1 (supplemental digital content) present adjusted means for BP intensity ratings (NRS) stratified by race and ethnicity. Adjusted baseline BP ratings were significantly higher for Blacks vs. Whites (mean 5.8 vs. 5.0; p<0.001), but not for Hispanics vs. non-Hispanics (5.3 vs. 5.2; p=0.29). There were statistically significant improvements over time in adjusted BP ratings within all race/ethnicity groups (all p-values <0.001). There were no significant differences in improvements in BP ratings between groups over time comparing Blacks vs. Whites (mean NRS improvement between baseline and 24 months 1.4 vs. 1.5; p=0.20), Asians vs. Whites (mean NRS improvement 1.8 vs. 1.5; p=0.12), or Hispanics vs. non-Hispanics (mean NRS improvement 1.4 vs. 1.5; p=0.06).

Figure 2. Back pain intensity over time by race and ethnicity*.

Figure 2

Open in a new tab

A: Multivariate-adjusted mean scores for back pain intensity by race.

B: Multivariate-adjusted mean scores for back pain intensity by ethnicity.

NRS: numerical rating scale

*Means are adjusted for all covariates listed in Table 1. Numeric values of adjusted means and 95% confidence intervals are provided as digital supplemental content (Table S1)

Figure 3 and Table S2 (supplemental digital content) present adjusted RMDQ means stratified by race and ethnicity. Adjusted baseline RMDQ scores were not significantly different for Blacks vs. Whites (mean 10.6 vs. 9.7; p=0.06) or Hispanics vs. non-Hispanics (10.2 vs. 10.0; p=0.86). Adjusted RMDQ scores indicated no significant functional improvement over time within race/ethnicity groups for Blacks (mean RMDQ improvement 0.6 points; p=0.10) or Hispanics (0.5; p=0.12), but there was statistically significant functional improvement over time for Whites (1.3; p<0.001), Asians (1.9; p=0.001), and non-Hispanics (1.2; p<0.001). Blacks had significantly less functional improvement over time as compared to Whites (adjusted mean RMDQ improvement between baseline and 24 months 0.6 vs. 1.3; p=0.03), but differences in functional improvement between other race/ethnicity groups were not statistically significant.

Figure 3. Back-related functional limitations over time by race and ethnicity*.

Figure 3

Open in a new tab

A: Multivariate-adjusted mean scores for Roland-Morris Disability Questionnaire by race.

B: Multivariate-adjusted mean scores for Roland-Morris Disability Questionnaire by ethnicity.

RMDQ: Roland-Morris Disability Questionnaire

*Means are adjusted for all covariates listed in Table 1. Numeric values of adjusted means and 95% confidence intervals are provided as digital supplemental content (Table S2)

The proportion of individuals with ≥30% improvement in BP intensity at 24 months was 47.3% for Whites, 35.5% for Blacks, 54.6% for Asians, 45.9% for non-Hispanics, and 43.5% for Hispanics. The adjusted odds of ≥30% improvement in BP intensity at 24 months was significantly lower in Blacks vs. Whites (odds ratio [OR] 0.76, 95% confidence interval [CI] 0.60–0.96, p=0.02), but there were no statistically significant differences for Asians compared to Whites, or Hispanics compared to non-Hispanics (Table 2). The proportion of individuals with ≥30% improvement in back-related functional limitations at 24 months was 39.3% for Whites, 28.5% for Blacks, 46.7% for Asians, 38.0% for non-Hispanics, and 31.7% for Hispanics. The adjusted odds of ≥30% improvement in back-related functional limitations at 24 months was significantly lower in Blacks vs. Whites (odds ratio [OR] 0.72, 95% confidence interval [CI] 0.56–0.92, p=0.008), but there were no statistically significant differences for Asians compared to Whites, or Hispanics compared to non-Hispanics (Table 2). The proportion of individuals with BP resolution at 24 months was 26.6% for Whites, 16.1% for Blacks, 42.4% for Asians, 25.7% for non-Hispanics, and 23.3% for Hispanics. The adjusted odds of BP resolution at 24 months (Table 2) were significantly lower in Blacks (OR 0.70, 95% CI 0.52–0.94, p=0.02) and higher in Asians (OR 2.38, 95% CI 1.60–3.52, p<0.001) compared to Whites, but were similar between Hispanics and non-Hispanics.

Table 2.

Improvement in back pain or back-related functional limitations at 24 months

Race/Ethnicityb ≥30% improvement in back pain ≥30% improvement in back-related
functional limitations		 Participant-reported complete
resolution of back paina
n (%)* aOR(95% CI) p-value n (%)* aOR (95% CI) p-value n (%)* aOR (95% CI) p-value
White (n=2935) 1387 (47.3) reference NA 1154 (39.3) reference NA 769 (26.6) reference NA
Black (n=593) 210 (35.5) 0.76 (0.60–0.96) 0.02 169 (28.5) 0.72 (0.56–0.92) 0.008 95 (16.1) 0.70 (0.52–0.94) 0.02
Asian (n=152) 83 (54.6) 1.11 (0.77–1.58) 0.58 71 (46.7) 1.11 (0.77–1.59) 0.57 61 (42.4) 2.38 (1.60–3.52) <0.001
non-Hispanic (n=3724) 1708 (45.9) reference NA 1414 (38.0) reference NA 942 (25.7) reference NA
Hispanic (n=230) 100 (43.5) 0.83 (0.57–1.23) 0.36 73 (31.7) 0.82 (0.55–1.22) 0.33 52 (23.3) 0.97 (0.63–1.51) 0.90
Open in a new tab

All odds ratios and p-values are adjusted for the variables presented in Table 1.

aOR=adjusted odds ratio, CI=confidence interval

*

Values represent numbers and proportions of participants within each race or ethnicity group who achieved the given level of improvement.

a

Participant-reported complete resolution of back pain defined as ≥30 days with no pain.

b

Participants completing RMDQ at 24 months.

Health Care Utilization

The proportion in each group receiving various treatments and diagnostic studies is shown in Table 3. In adjusted analyses (Table 3), Blacks were not significantly different from Whites with respect to receiving PT, psychotherapy, or opioid prescriptions over 24 months, but they were significantly less likely to receive early (<6 weeks from index visit) spine radiographs (OR 0.67, 95% CI 0.53–0.84) or spine MRI (OR 0.71, 95% CI 0.55–0.93) at any time. Conversely, Blacks were more likely to receive spine CT at any time compared to Whites (OR 1.62, 95% CI 1.06–2.45). Post hoc analyses indicated that the greater odds of receiving a spine CT for Blacks was driven almost entirely by the Henry Ford site where most Black participants were recruited and where CT utilization was higher than at other sites (data not shown).

Table 3.

Associations between race/ethnicity and health care utilization over 24 months

Health care resource White
(n=3746)		 Blacka
(n=781)		 Asiana
(n=195)		 non-Hispanic
(n=4791)		 Hispanicb
(n=302)
n (%)* n (%)* aOR (95% CI)d n (%)* aOR (95% CI)d n (%)* n (%)* aOR(95% CI)d
Diagnostic Spine Imaging
  Early spine radiographsc 929 (24.8) 182 (23.3) 0.67 (0.53–0.84) 55 (28.2) 0.98 (0.69–1.38) 1177 (24.6) 78 (25.8) 0.94 (0.65–1.36)
  Early spine MRI/CTc 355 (9.5) 84 (10.8) 0.87 (0.65–1.17) 16 (8.2) 0.69 (0.40–1.21) 462 (9.6) 25 (8.3) 0.65 (0.34–1.21)
  Any spine radiographs 1329 (35.5) 277 (35.5) 0.83 (0.68–1.02) 77 (39.5) 0.95 (0.69–1.29) 1701 (35.5) 117 (38.7) 0.94 (0.68–1.30)
  Any spine CT 76 (2.0) 66 (8.5) 1.62 (1.06–2.45) 4 (2.1) 0.88 (0.27–2.88) 148 (3.1) 2 (0.7) 0.46 (0.12–1.75)
  Any spine MRI 742 (19.8) 104 (13.3) 0.71 (0.55–0.93) 38 (19.5) 0.82 (0.56–1.21) 899 (18.8) 62 (20.5) 0.91 (0.60–1.37)
Non-Surgical Care
  Any PT 1486 (39.7) 266 (34.1) 1.18 (0.96–1.45) 86 (44.1) 0.78 (0.57–1.07) 1869 (39.0) 130 (43.1) 0.91 (0.65–1.26)
  Any psychotherapy 157 (4.2) 21 (2.7) 0.77 (0.44–1.32) 7 (3.6) 0.54 (0.23–1.28) 193 (4.0) 14 (4.6) 0.78 (0.43–1.42)
  Any opioid prescription 746 (19.9) 88 (11.3) 0.78 (0.58–1.04) 40 (20.5) 0.72 (0.50–1.05) 900 (18.8) 63 (20.9) 0.68 (0.45–1.01)
Surgical care
Any fusion 45 (1.2) 6 (0.8) 0.60 (0.25–1.45) 3 (1.5) 1.21 (0.35–4.21) 56 (1.2) 2 (0.7) 0.88 (0.22–3.58)
Any non-fusion 79 (2.1) 10 (1.3) 0.55 (0.26–1.14) 6 (3.1) 1.28 (0.55–2.95) 98 (2.1) 5 (1.7) 0.77 (0.25–2.37)
RVUs(ratio of means)
Total RVUs NA NA 0.91 (0.82–1.02) NA 0.77 (0.62–0.96) NA NA 0.77 (0.64–0.92)
Spine-specific RVUs NA NA 0.66 (0.51–0.86) NA 0.92 (0.55–1.54) NA NA 0.60 (0.40–0.90)
Open in a new tab

All odds ratios are adjusted for the variables presented in Table 1.

aOR=adjusted odds ratio, CI=confidence interval, MRI=magnetic resonance imaging, CT=computed tomography, PT = physical therapy, RVU=Relative Value Unit, NA=not applicable

*

Values represent numbers and proportions of participants within each race or ethnicity group who received a specified health care resource.

a

Whites are the reference group.

b

non-Hispanics are the reference group.

c

’Early’ imaging was obtained before 6 weeks post-index visit.

d

Values represent odds (95% confidence interval) of having received a specified health care resource compared to the reference race/ethnicity group.

Unadjusted characteristics of RVUs over 24 months are presented in supplemental digital content Table S3. Adjusted mean spine-related RVUs for Blacks were one-third fewer than those for Whites over 24-month follow-up (ratio of means 0.66, 95% CI 0.51–0.86)(Table 3). In absolute terms, adjusted mean spine-related RVUs were 14.0 for Blacks and 21.2 for Whites (Table S4, supplemental digital content). Post hoc analysis indicated that this difference in spine-related RVUs was driven by fewer ‘late’ RVUs in months 6–24 following the index visit (ratio of means 0.44, 95% CI 0.28–0.68) suggesting relatively less spine-related health care utilization due to ongoing treatment in Blacks as compared to Whites (data not shown). Although Hispanics were comparable to non-Hispanics with respect to specific diagnostic imaging use and non-surgical/surgical treatments, adjusted mean total RVUs for Hispanics were fewer than those for non-Hispanics over 24-month follow-up (ratio of means 0.77, 95% CI 0.64–0.92), as were spine-specific RVUs (ratio of means 0.60, 95% CI 0.40–0.90)(Table 3). In absolute terms, adjusted mean total RVUs were 87.0 for Hispanics and 113.5 for non-Hispanics and adjusted mean spine-related RVUs were 12.4 for Hispanics and 20.6 for non-Hispanics (Table S4, supplemental digital content). There were no material differences with respect to ‘late’ RVUs (data not shown).

DISCUSSION

This study of older adults with BP recruited from three US health care systems found that Blacks had a small but statistically significant greater baseline BP intensity (adjusted mean NRS 0.8). All race/ethnicity groups demonstrated only modest within-group improvements in BP intensity over 24 months and there were no statistically significant differences between race/ethnicity groups in NRS improvement over time. Blacks and Hispanics did not have statistically significant within-group improvements in back-related functional limitations over time in contrast to Whites, Asians, and non-Hispanics; however, the magnitude of differences in improvement between groups was quite small (e.g., the difference between Whites and Blacks in adjusted mean RMDQ improvement over 24 months was 0.6 points on the 0–24 scale). Despite these small differences in adjusted mean scores, Blacks had about three-quarters the odds of Whites of attaining ≥30% improvement in BP intensity or functional limitations, or BP resolution at 24 months.

To our knowledge, this study is the first to compare longitudinal changes in back-related PROs and health care utilization between multiple race/ethnicity groups in the US. Our finding of greater baseline BP intensity and slightly less back-related functional improvement in Blacks compared to Whites is generally consistent with results from clinical studies of BP6–9,38 and experimental pain studies.39 The reasons for differences in BP reporting between Blacks and Whites are unclear. One explanation for such differences is confounding due to psychological, socioeconomic, or cultural factors.40–42 Although our study adjusted for a range of possible confounders, these are expected to have provided limited adjustment for several key factors such as pain beliefs/attitudes, social roles and experiences, socioeconomic status, and other factors related to pain and the reporting of pain. Residual confounding may explain the small differences between Blacks and Whites observed in back-related PROs. Another explanation for these differences could be actual biological differences through mechanisms of endogenous pain modulation43,44, endocrine factors43,45, and genetic factors.39,46,47

Differences in the diagnosis and management of various pain conditions according to race/ethnicity have been reported previously with Whites generally being prescribed higher doses of opioids for pain,48–55 having earlier and more comprehensive diagnostic workup,5,51 and receiving earlier surgical intervention compared to Blacks.56–58 Our findings in the specific context of BP are similar. In this study, Blacks were less likely to receive early spine X-rays or MRI of the spine at any time compared to Whites. Interestingly, Blacks were more likely to receive spine CT at any time. This finding was driven primarily by data from one study site where spine CT orders for back pain were common in the emergency department (ED) and may have been due to Blacks receiving a higher proportion of overall care services through the ED at that site. Blacks had less total healthcare utilization over 24 months and specifically had fewer spine-related RVUs than Whites. Given that Blacks also had fewer spine-related RVUs later in follow-up (6–24 months), when utilization was less likely due to initial diagnostic work-up, such differences in RVUs appear not to be explained by diagnostic imaging utilization alone. The magnitude of difference in RVUs between Blacks and Whites (7.2 RVUs) is of possible clinical relevance when considering RVU amounts for a lumbar spine radiograph (0.22) and a transforaminal lumbar epidural injection (1.9) as of January 1st, 2017 (Table 4). Our study found no statistically significant differences in opioid prescriptions for Blacks vs. Whites. This contrasts with some studies in other contexts54,55 which have found less opioid prescribing for Blacks. A novel finding of our study is the magnitude of difference in RVUs between Hispanics and non-Hispanics which is also of possible clinical relevance, demonstrating a difference of 26.4 total RVUs and 8.2 spine-related RVUs. Since these differences were not explained by our analysis of those receiving diagnostic imaging or specific treatments, possible explanations include other spine-related diagnostics/treatments we did not evaluate or differences in volume of services provided which our analyses did not address.

Table 4.

Work RVUs for selected healthcare resourcesa

CPT code Health care
resource		 Associated
Work RVUsb
72100 Lumbosacral spine (X-ray) 2 or 3 view 0.22
97163 PT evaluation (complex) 1.2
72148 MRI lumbar spine without contrast 1.48
64483 Transforaminal lumbar injection with guidance 1.9
90837 Psychotherapy (60 min) 3.0
Open in a new tab

CPT=Current Procedural Terminology, RVU=Relative Value Unit, PT=physical therapy, MRI=Magnetic Resonance Imaging

a

Resources shown are examples commonly used in the diagnosis and treatment of back pain.

b

Work RVUs according to the 2017 CMS update.

Unlike many other health conditions, less treatment and diagnostic workup is often considered better-quality care for new episodes of BP in the absence of historical or clinical red-flags.59 Therefore, less spine-related health care utilization for Blacks and Hispanics could paradoxically represent better BP care. It is unclear whether these differences in PROs and health care utilization by race/ethnicity constitute actual disparities per se (explained by socioeconomic and/or environmental disadvantage) since our analytic approach did not isolate mediation through individual covariates linked with disadvantage and several potentially important confounders linked with disadvantage were not measured. In theory, the observed differences in health care utilization by race/ethnicity may have explanations unrelated to disadvantage such as culturally-motivated patient preferences. Nevertheless, given the historical legacy of health care disparities affecting ethnic minorities in the United States,60 it is important to consider whether utilization differences in our study may reflect inequity in the provision of care. Our work demonstrates several important differences in back-related outcomes associated with race/ethnicity in a large US sample, even after adjustment for specific patient-level characteristics and some system-level factors. Future work in the specific context of BP is needed to further examine these differences by race/ethnicity, to account for more provider-level and system-level factors to clarify drivers underlying the differences observed in this study, and to identify which factors reflect disparities that are modifiable.

Strengths of the current study include the longitudinal design and large sample size drawn from different regions in the US. Nevertheless, further studies are needed to confirm whether our findings hold true for other regions of the country. A limitation of our study is the general classification of race/ethnicity using the OMB definitions. For example, the novel finding that US Asians had a greater likelihood of BP resolution, but otherwise similar trajectories of BP and functional improvements as compared to Whites raises questions. The categorization of ‘Asian’ participants in this study combines individuals originating from disparate regions of the Middle East, East Asia, and South Asia, possibly obscuring important differences between individuals from different regions. This concern applies to all race/ethnicity designations examined in this article. To address this issue, future studies using self-report to examine differences in race/ethnicity, in samples with large subgroups of various races/ethnicities, may benefit from creating further subdivisions within the OMB classifications or potentially moving beyond self-reporting entirely by accounting for specific genetic markers.61

CONCLUSIONS

Identifying differences in back-related outcomes associated with race/ethnicity is a necessary step toward improving health outcomes in the US and creating an equitable health care system. These findings warrant replication in other regions of the country and further critical examination to determine how racial/ethnic health disparities where present for patients with BP may be mitigated.

Supplementary Material

Supplemental Data File Revised _doc._ pdf._ xls._ etc._
Tables - Revised

Acknowledgments

We wish to thank the BOLD study participants. We also wish to thank the BOLD study staff and research teams.

The manuscript submitted does not contain information about medical device(s)/drug(s). The Agency for Healthcare Research and Quality(AHRQ) (grants 1R01HS01922201 and 1R01HS022972-01) funds were received in support of this work. Dr. Suri’s participation in this study was funded by VA Puget Sound Health Care System. Dr. Suri is funded by Career Development Award #1IK2RX001515 from VA Rehabilitation Research and Development. No relevant financial activities outside the submitted work.

References

  • 1.Hoy D, March L, Brooks P, et al. The global burden of low back pain: estimates from the Global Burden of Disease 2010 study. Annals of the rheumatic diseases. 2014;73:968–74. doi: 10.1136/annrheumdis-2013-204428. [DOI] [PubMed] [Google Scholar]
  • 2.Davis MA, Onega T, Weeks WB, et al. Where the United States spends its spine dollars: expenditures on different ambulatory services for the management of back and neck conditions. Spine (Phila Pa 1976) 2012;37:1693–701. doi: 10.1097/BRS.0b013e3182541f45. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Katz JN. Lumbar disc disorders and low-back pain: socioeconomic factors and consequences. J. Bone Joint Surg. Am. 2006;88(Suppl 2):21–4. doi: 10.2106/JBJS.E.01273. [DOI] [PubMed] [Google Scholar]
  • 4.Carey TS, Garrett JM. The relation of race to outcomes and the use of health care services for acute low back pain. Spine. 2003;28:390–4. doi: 10.1097/01.BRS.0000048499.25275.51. [DOI] [PubMed] [Google Scholar]
  • 5.Taylor BA, Casas-Ganem J, Vaccaro AR, et al. Differences in the work-up and treatment of conditions associated with low back pain by patient gender and ethnic background. Spine. 2005;30:359–64. doi: 10.1097/01.brs.0000152115.79236.6e. [DOI] [PubMed] [Google Scholar]
  • 6.Tait RC, Chibnall JT. Work injury management of refractory low back pain: relations with ethnicity, legal representation and diagnosis. Pain. 2001;91:47–56. doi: 10.1016/s0304-3959(00)00418-8. [DOI] [PubMed] [Google Scholar]
  • 7.Chibnall JT, Tait RC. Disparities in occupational low back injuries: predicting pain-related disability from satisfaction with case management in African Americans and Caucasians. Pain medicine. 2005;6:39–48. doi: 10.1111/j.1526-4637.2005.05003.x. [DOI] [PubMed] [Google Scholar]
  • 8.Chibnall JT, Tait RC. Long-term adjustment to work-related low back pain: associations with socio-demographics, claim processes, and post-settlement adjustment. Pain medicine. 2009;10:1378–88. doi: 10.1111/j.1526-4637.2009.00738.x. [DOI] [PubMed] [Google Scholar]
  • 9.Carey TS, Freburger JK, Holmes GM, et al. Race, care seeking, and utilization for chronic back and neck pain: population perspectives. J. Pain. 2010;11:343–50. doi: 10.1016/j.jpain.2009.08.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Green CR, Anderson KO, Baker TA, et al. The unequal burden of pain: confronting racial and ethnic disparities in pain. Pain Med. 2003;4:277–94. doi: 10.1046/j.1526-4637.2003.03034.x. [DOI] [PubMed] [Google Scholar]
  • 11.Pallin DJ, Espinola JA, Camargo CA., Jr US population aging and demand for inpatient services. J. Hosp. Med. 2014;9:193–6. doi: 10.1002/jhm.2145. [DOI] [PubMed] [Google Scholar]
  • 12.Dall TM, Gallo PD, Chakrabarti R, et al. An aging population and growing disease burden will require a large and specialized health care workforce by 2025. Health Aff. (Millwood) 2013;32:2013–20. doi: 10.1377/hlthaff.2013.0714. [DOI] [PubMed] [Google Scholar]
  • 13.Colby SaO, Jennifer . Projections of the Size and Compsition of the U.S. Population: 2014 to 2060. Current Population Reports. Washington, DC: 2014. [Google Scholar]
  • 14.Hebert PL, Sisk JE, Howell EA. When does a difference become a disparity? Conceptualizing racial and ethnic disparities in health. Health Aff (Millwood) 2008;27:374–82. doi: 10.1377/hlthaff.27.2.374. [DOI] [PubMed] [Google Scholar]
  • 15.U.S. Department of Health and Human Services. The Secretary’s Advisory Committee on National Health Promotion and Disease Prevention Objectives for 2020. Phase I report: Recommendations for the framework and format of Healthy People 2020 [Internet] [cited 2010 January 6];Section IV: Advisory Committee findings and recommendations. Available from: http://www.healthypeople.gov/sites/default/files/PhaseI_0.pdf.
  • 16.Jarvik JG, Comstock BA, Bresnahan BW, et al. Study protocol: the Back Pain Outcomes using Longitudinal Data (BOLD) registry. BMC musculoskeletal disorders. 2012;13:64. doi: 10.1186/1471-2474-13-64. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17. [Accessed October 30];Revisions to the Standards for the Classification of ederal Data on Race and Ethnicity. [serial online] 1997 [Google Scholar]
  • 18.Health NIo. Racial and Ethnic Categories and Definitions for NIH Diversity Programs and for Other Reporting Purposes. 2015 [Google Scholar]
  • 19.Services DoHaH. U.S. Department of Health and Human Services Implementation Guidance on Data Collection Standards for Race, Ethnicity, Sex, Primary Language, and Disability Status. 2011 [Google Scholar]
  • 20.Farrar JT, Young JP, Jr, LaMoreaux L, et al. Clinical importance of changes in chronic pain intensity measured on an 11-point numerical pain rating scale. Pain. 2001;94:149–58. doi: 10.1016/S0304-3959(01)00349-9. [DOI] [PubMed] [Google Scholar]
  • 21.Jarvik JG, Comstock BA, Bresnahan BW, et al. Study protocol: the Back Pain Outcomes using Longitudinal Data (BOLD) registry. BMC Musculoskelet. Disord. 2012;13:64. doi: 10.1186/1471-2474-13-64. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Jordan K, Dunn KM, Lewis M, et al. A minimal clinically important difference was derived for the Roland-Morris Disability Questionnaire for low back pain. J. Clin. Epidemiol. 2006;59:45–52. doi: 10.1016/j.jclinepi.2005.03.018. [DOI] [PubMed] [Google Scholar]
  • 23.Dworkin RH, Turk DC, McDermott MP, et al. Interpreting the clinical importance of group differences in chronic pain clinical trials: IMMPACT recommendations. Pain. 2009;146:238–44. doi: 10.1016/j.pain.2009.08.019. [DOI] [PubMed] [Google Scholar]
  • 24.Rundell SD, Sherman KJ, Heagerty PJ, et al. The clinical course of pain and function in older adults with a new primary care visit for back pain. J. Am. Geriatr. Soc. 2015;63:524–30. doi: 10.1111/jgs.13241. [DOI] [PubMed] [Google Scholar]
  • 25.Chou R, Qaseem A, Owens DK, et al. Diagnostic imaging for low back pain: advice for high-value health care from the American College of Physicians. Ann. Intern. Med. 2011;154:181–9. doi: 10.7326/0003-4819-154-3-201102010-00008. [DOI] [PubMed] [Google Scholar]
  • 26.Medicare program; payment policies under the physician fee schedule and other revisions to Part B for CY 2010. Final rule with comment period. Federal register. 2009;74:61737–2188. [PubMed] [Google Scholar]
  • 27.Medicare program; payment policies under the physician fee schedule, five-year review of work relative value units, clinical laboratory fee schedule: signature on requisition, and other revisions to part B for CY 2012. Final rule with comment period. Federal register. 2011;76:73026–474. [PubMed] [Google Scholar]
  • 28.Medicare Program; Revisions to Payment Policies Under the Physician Fee Schedule, DME Face-to-Face Encounters, Elimination of the Requirement for Termination of Non-Random Prepayment Complex Medical Review and Other Revisions to Part B for CY 2013. Federal register. 2012 [PubMed] [Google Scholar]
  • 29.Jarvik JG, Gold LS, Comstock BA, et al. Association of early imaging for back pain with clinical outcomes in older adults. JAMA. 2015;313:1143–53. doi: 10.1001/jama.2015.1871. [DOI] [PubMed] [Google Scholar]
  • 30.Karvelas DA, Rundell SD, Friedly JL, et al. Subsequent health-care utilization associated with early physical therapy for new episodes of low back pain in older adults. Spine J. 2017;17:380–9. doi: 10.1016/j.spinee.2016.10.007. [DOI] [PubMed] [Google Scholar]
  • 31.Gold LS, Bryan M, Comstock BA, et al. Associations Between Relative Value Units and Patient-Reported Back Pain and Disability. Gerontol Geriatr Med. 2017;3 doi: 10.1177/2333721416686019. 2333721416686019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Commerce UDo. [Accessed August 2015];All values are per capita income in past 12 months (in 2013 dollars), 2009–2013. Available at: www.census.gov/quickfacts.
  • 33.Kroenke K, Spitzer RL, Williams JB. The Patient Health Questionnaire-2: validity of a two-item depression screener. Med. Care. 2003;41:1284–92. doi: 10.1097/01.MLR.0000093487.78664.3C. [DOI] [PubMed] [Google Scholar]
  • 34.Donker T, van Straten A, Marks I, et al. Quick and easy self-rating of Generalized Anxiety Disorder: validity of the Dutch web-based GAD-7, GAD-2 and GAD-SI. Psychiatry Res. 2011;188:58–64. doi: 10.1016/j.psychres.2011.01.016. [DOI] [PubMed] [Google Scholar]
  • 35.Quan H, Parsons GA, Ghali WA. Validity of information on comorbidity derived from ICD-9-CCM administrative data. Med. Care. 2002;40:675–85. doi: 10.1097/00005650-200208000-00007. [DOI] [PubMed] [Google Scholar]
  • 36.Diggle P, Heagerty P, Liang K, Zeger S. Analysis of longitudinal data. 2. Oxford; New York: Oxford University Press; 2002. [Google Scholar]
  • 37.Little RJA, Rubin DB. Statistical analysis with missing data. 2. New York: Wiley; 2002. [Google Scholar]
  • 38.Carey TS, Garrett JM. The relation of race to outcomes and the use of health care services for acute low back pain. Spine (Phila Pa 1976) 2003;28:390–4. doi: 10.1097/01.BRS.0000048499.25275.51. [DOI] [PubMed] [Google Scholar]
  • 39.Rahim-Williams B, Riley JL, 3rd, Williams AK, et al. A quantitative review of ethnic group differences in experimental pain response: do biology, psychology, and culture matter? Pain Med. 2012;13:522–40. doi: 10.1111/j.1526-4637.2012.01336.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Campbell P, Foster NE, Thomas E, et al. Prognostic indicators of low back pain in primary care: five-year prospective study. J. Pain. 2013;14:873–83. doi: 10.1016/j.jpain.2013.03.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Dunn KM, Jordan KP, Croft PR. Contributions of prognostic factors for poor outcome in primary care low back pain patients. Eur. J. Pain. 2011;15:313–9. doi: 10.1016/j.ejpain.2010.07.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Jones GT, Johnson RE, Wiles NJ, et al. Predicting persistent disabling low back pain in general practice: a prospective cohort study. Br. J. Gen. Pract. 2006;56:334–41. [PMC free article] [PubMed] [Google Scholar]
  • 43.Mechlin MB, Maixner W, Light KC, et al. African Americans show alterations in endogenous pain regulatory mechanisms and reduced pain tolerance to experimental pain procedures. Psychosom. Med. 2005;67:948–56. doi: 10.1097/01.psy.0000188466.14546.68. [DOI] [PubMed] [Google Scholar]
  • 44.Wang H, Papoiu AD, Coghill RC, et al. Ethnic differences in pain, itch and thermal detection in response to topical capsaicin: African Americans display a notably limited hyperalgesia and neurogenic inflammation. Br J Dermatol. 2010;162:1023–9. doi: 10.1111/j.1365-2133.2009.09628.x. [DOI] [PubMed] [Google Scholar]
  • 45.Grewen KM, Light KC, Mechlin B, et al. Ethnicity is associated with alterations in oxytocin relationships to pain sensitivity in women. Ethn. Health. 2008;13:219–41. doi: 10.1080/13557850701837310. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Nielsen CS, Stubhaug A, Price DD, et al. Individual differences in pain sensitivity: genetic and environmental contributions. Pain. 2008;136:21–9. doi: 10.1016/j.pain.2007.06.008. [DOI] [PubMed] [Google Scholar]
  • 47.Fillingim RB, Kaplan L, Staud R, et al. The A118G single nucleotide polymorphism of the mu-opioid receptor gene (OPRM1) is associated with pressure pain sensitivity in humans. J. Pain. 2005;6:159–67. doi: 10.1016/j.jpain.2004.11.008. [DOI] [PubMed] [Google Scholar]
  • 48.Cintron A, Morrison RS. Pain and ethnicity in the United States: A systematic review. J. Palliat. Med. 2006;9:1454–73. doi: 10.1089/jpm.2006.9.1454. [DOI] [PubMed] [Google Scholar]
  • 49.Anderson KO, Green CR, Payne R. Racial and ethnic disparities in pain: causes and consequences of unequal care. J. Pain. 2009;10:1187–204. doi: 10.1016/j.jpain.2009.10.002. [DOI] [PubMed] [Google Scholar]
  • 50.Burgess DJ, Nelson DB, Gravely AA, et al. Racial differences in prescription of opioid analgesics for chronic noncancer pain in a national sample of veterans. J. Pain. 2014;15:447–55. doi: 10.1016/j.jpain.2013.12.010. [DOI] [PubMed] [Google Scholar]
  • 51.Chibnall JT, Tait RC, Andresen EM, et al. Race differences in diagnosis and surgery for occupational low back injuries. Spine (Phila Pa 1976) 2006;31:1272–5. doi: 10.1097/01.brs.0000217584.79528.9b. [DOI] [PubMed] [Google Scholar]
  • 52.Stover BD, Turner JA, Franklin G, et al. Factors associated with early opioid prescription among workers with low back injuries. J. Pain. 2006;7:718–25. doi: 10.1016/j.jpain.2006.03.004. [DOI] [PubMed] [Google Scholar]
  • 53.Tamayo-Sarver JH, Hinze SW, Cydulka RK, et al. Racial and ethnic disparities in emergency department analgesic prescription. Am. J. Public Health. 2003;93:2067–73. doi: 10.2105/ajph.93.12.2067. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Chen I, Kurz J, Pasanen M, et al. Racial differences in opioid use for chronic nonmalignant pain. J. Gen. Intern. Med. 2005;20:593–8. doi: 10.1111/j.1525-1497.2005.0106.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Pletcher MJ, Kertesz SG, Kohn MA, et al. Trends in opioid prescribing by race/ethnicity for patients seeking care in US emergency departments. JAMA. 2008;299:70–8. doi: 10.1001/jama.2007.64. [DOI] [PubMed] [Google Scholar]
  • 56.Skinner J, Zhou W, Weinstein J. The influence of income and race on total knee arthroplasty in the United States. The Journal of bone and joint surgery. 2006;88:2159–66. doi: 10.2106/JBJS.E.00271. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57.Steel N, Clark A, Lang IA, et al. Racial disparities in receipt of hip and knee joint replacements are not explained by need: the Health and Retirement Study 1998–2004. J. Gerontol. A Biol. Sci. Med. Sci. 2008;63:629–34. doi: 10.1093/gerona/63.6.629. [DOI] [PubMed] [Google Scholar]
  • 58.Neuman MD, Fleisher LA, Even-Shoshan O, et al. Nonoperative care for hip fracture in the elderly: the influence of race, income, and comorbidities. Med. Care. 2010;48:314–20. doi: 10.1097/mlr.0b013e3181ca4126. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59.Deyo RA, Mirza SK, Turner JA, et al. Overtreating chronic back pain: time to back off? J Am Board Fam Med. 2009;22:62–8. doi: 10.3122/jabfm.2009.01.080102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 60.Institute of Medicine. Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care. Washington (DC): 2003. [Google Scholar]
  • 61.Patterson N, Price AL, Reich D. Population structure and eigenanalysis. PLoS Genet. 2006;2:e190. doi: 10.1371/journal.pgen.0020190. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Data File Revised _doc._ pdf._ xls._ etc._
NIHMS920466-supplement-Supplemental_Data_File_Revised__doc___pdf___xls___etc__.docx (39.4KB, docx)
Tables - Revised
NIHMS920466-supplement-Tables_-_Revised.docx (49KB, docx)

RESOURCES