Abstract
Nonpharmacologic approaches are recommended as first-line treatment for chronic pain and their importance is heightened among individuals with cooccurring opioid use disorder (OUD), in whom opioid therapies may be particularly detrimental. Our objectives were to assess the receipt and trajectories of nonpharmacologic pain treatment and determine the association of OUD diagnosis with these trajectories. This retrospective cohort study used Medicare claims data from 2016–2018 and applied group-based trajectory models to identify distinct patterns of physical therapy (PT) or chiropractic care treatment over the 12 months following a new episode of chronic low back pain. We used logistic regression models to estimate the association of cooccurring OUD with group membership in PT and chiropractic trajectories. Our sample comprised 607,729 beneficiaries at least 18 years of age, of whom 11.4% had a diagnosis of OUD. The 12-month prevalence of PT and chiropractic treatment receipt was 24.7% and 27.1%; respectively, and lower among Medicare beneficiaries with cooccurring OUD (PT: 14.6%; chiropractic: 6.8%). The final models identified three distinct trajectories each for PT (no/little use [76.6% of sample], delayed and increasing use [8.2%], and early and declining use [15.2%]); and chiropractic (no/little use [75.0% of sample], early and declining use [17.3%], and early and sustained use [7.7%]). People with OUD were more likely to belong in trajectories with little/no PT or chiropractic care as compared to other trajectories. The findings indicate that people with cooccurring chronic pain and OUD often do not receive early or any nonpharmacologic pain therapies as recommended by practice guidelines.
Keywords: chronic pain, opioid use disorder, physical therapy, chiropractic, group-based trajectories
INTRODUCTION
Chronic low back pain is an important public health problem for which noninvasive nonpharmacologic therapies are recommended for its management.1, 2 Chronic low back pain is the most prevalent type of chronic pain and is a leading cause of disability in the US. An estimated 13% of US adults have chronic low back pain, with one-third experiencing moderate to high impact chronic pain that frequently limits life and work activities.3 Practice guidelines,2, 4 including the updated 2022 Centers for Disease Control and Prevention Opioid Prescribing Guideline, recommend nonpharmacologic therapies as first-line therapy for chronic pain in general, including chronic low back pain,2, 5 and regard them important in reducing reliance on opioids and other potentially high-risk medications.6–8 Recommended nonpharmacologic therapies include therapeutic exercise, spinal manipulation, acupuncture, and multidisciplinary rehabilitation which clinical trials suggest confer small to moderate sized treatment effects on pain control and functional ability although with large heterogeneity in individual responses.1, 9 Despite being widely recommended, critical knowledge gaps exist regarding the utilization of nonpharmacologic therapies.
A nontrivial share of people with chronic pain in general, including chronic low back pain, have cooccurring opioid use disorder (OUD). Epidemiological studies suggest that 8% to 12% of chronic pain patients have OUD10 and half of people with OUD have chronic pain.11 Chronic pain and OUD have a reciprocal relationship driven in part by long-term opioid prescribing to treat chronic pain, which is associated with increased risk of developing OUD.12–16 Managing chronic low back pain in people with OUD can be complex because of their increased tolerance to opioids and elevated overdose risk among other factors.17–28 Nonpharmacologic therapies are generally underutilized for a variety of reasons including lack of or insufficient insurance coverage29 and inability to self-pay, underlying general health,30 patient preferences, low supply and geographic accessibility of providers,31, 32 low referral rates, and lack of awareness among physicians about Medicare reimbursement policies.33, 34 Furthermore, although opioid prescribing has declined in the US,35, 36 opioids remain a relatively common choice for chronic pain treatment,37 particularly in Medicare.38 This is a particular concern in people with OUD because opioids are not a safe option and there are likely additional barriers to receiving nonpharmacologic therapies such as other cooccurring mental health conditions, stigma, barriers to access;20, 24–28 however, these issues have not been systematically studied.
Medicare is a suitable choice to further understand the relationship between OUD diagnosis and pain treatment because chronic low back pain increases with age and certain nonpharmacologic therapies are covered including physical therapy (PT), chiropractic care, and acupuncture (starting in 2020). A study of 2011–2014 data found that many Medicare beneficiaries who developed new low back pain received guideline non-concordant care such as opioids before other modalities like PT.39 However, much remains unknown about the uptake of and persistence with nonpharmacologic therapies among people with chronic low back pain generally and in those with comorbid OUD. Therefore, the study objectives were to identify longitudinal trajectories of PT and chiropractic care use among Medicare beneficiaries with new episodes of chronic low back pain during 2016–2018, and determine the association of OUD diagnosis with membership in these trajectories. We assumed that optimal guideline-concordant care would be characterized by trajectories early and consistent receipt of nonpharmacologic therapies in the months following a new episode of chronic low back pain. However, considering the previously described barriers to nonpharmacologic therapy use, we hypothesized that receipt of, and persistence with nonpharmacologic therapies would be low overall, and even lower among people with OUD.
METHODS
Data Sources
This study used a 20% random national sample of Medicare data from 2016 to 2018. The data used were enrollment files, Part A (inpatient data from the Medicare Provider Analysis and Review file), and Part B (outpatient) claims. We linked enrollment data to the United States Department of Agriculture 2013 Rural Urban Continuum codes by the Federal Information Processing System code, which uniquely identifies geographic areas, to determine the metropolitan status of a beneficiary’s county of residence.
Study Design and Sample
We conducted a retrospective cohort study of fee-for-service, community-dwelling beneficiaries with chronic low back pain. We first used inpatient and outpatient claims to identify beneficiaries with low back pain using International Classification of Diseases (Tenth Revision) diagnostic codes. Chronic low back pain was defined based on 2 or more back pain diagnosis codes that were at least 90 days apart but less than 1 year apart. The index date was the first instance of a low back pain diagnosis code following a period of at least 6 months with no diagnosis of low back pain. Among beneficiaries with multiple qualifying episodes of chronic low back pain, we randomly selected a single episode for inclusion in the analysis.
We applied several exclusion criteria. These were index date before July 1, 2016 (to ensure 6-month look-back window) and index date after December 31, 2017 (to ensure 12-month follow-up window), age less than 18 years, any nursing home stay, any Medicare Advantage enrollment, and discontinuous enrollment in Medicare inpatient, outpatient, and pharmacy benefits. We also excluded beneficiaries diagnosed with cancer or who received hospice care due to potential differences in the approaches and goals for pain management.
Measures
Nonpharmacologic Treatments
We analyzed PT and chiropractic care, two evidence-based treatments for chronic low back pain that were reimbursed by Medicare during the study period. Current procedural therapy codes (see eTable 1 in the supplement) for PT evaluations, modalities, therapeutic procedures were used to identify PT and chiropractic care receipt occurring on unique service dates.
OUD Definition
We used International Classification of Diseases (Tenth Revision) diagnostic codes for opioid ‘abuse’ and dependence to determine OUD status. Recognizing that OUD is potentially underreported in claims data, we used inpatient and outpatient files to capture OUD and considered both baseline and follow up periods when classifying beneficiaries as having OUD or not.
Outcomes
We examined membership in a distinct trajectory of nonpharmacologic therapy use during the first year after the index date. Group membership was determined by: (1) constructing monthly measures of any use of PT and chiropractic care, and (2) identifying distinct initiation and persistence patterns of PT and chiropractic use by applying group-based trajectory models. The output from trajectory models included estimated probabilities of group membership for each individual, estimated trajectory curves over time, and proportion of each group trajectory (see statistical analysis section below).
Covariates
Baseline demographic, enrollment, medical condition covariates were captured during the 6-month baseline window. Demographic covariates included age, sex as recorded, race and ethnicity, low-income subsidy status, urbanicity of the beneficiary’s county of residence, and geographic region. We described dual Medicare and Medicaid enrollment and reason for current Medicare entitlement but did not adjust for them due to high collinearity with age. Medical conditions included non-back musculoskeletal pain, acute pain conditions (e.g., fractures, surgery), mental illness (e.g., depression, anxiety), history of alcohol or other non-OUD substance use disorder, and Gagne Comorbidity Index. We were interested in examining the influence of prior experience with nonpharmacologic therapies; therefore, we created indicators for baseline receipt of PT and chiropractic care prior to the index date.
Statistical Analysis
We sought to assess the patterns of ongoing use of PT and chiropractic care use in addition to any use of these therapies, thus requiring trajectory modeling to identify subgroups of individuals within Medicare beneficiaries with chronic low back pain that followed distinct trajectories over time. Trajectory models were constructed using the PROC TRAJ plug-in for SAS version 9.4.40 We fit a number of models with 2 to 4 groups. There was no considerable improvement in the model parameters beyond 4 groups. In selecting the optimal number of groups, we considered the Bayesian Information Criterion, group size (each trajectory group had to constitute at least 5% of the sample), group interpretability, and Nagin’s criteria to assess final model adequacy.41–43 Nagin’s criteria specifies an average posterior probability of ≥0.7 for all groups, an odds of correct classification of ≥5.0 for all groups, and narrow confidence intervals for estimated group membership probabilities.41 After identifying the optimal number of groups, we varied the functional form of the trajectory curve (i.e., intercept, zero-order(constant), linear, quadratic, and cubic).
We examined the probability of membership in each group and assigned the trajectory with the highest probability of membership to each individual. We compared differences between groups using Chi-square tests for categorical variables and t-tests for continuous variables. Group assignments were used as dependent variables in logistic regression models, also adjusting for demographic, geographic, and medical covariates, to calculate odds of group membership taking into account state-level clustering via general estimating equations. We weighted the regression models by the probability of group membership to adjust for uncertainty in group assignments. We applied a stricter statistical significance level of 0.01 due to the large sample size. The study was reviewed and deemed exempt by the Brown University Institutional Review Board on the basis of analysis of deidentified secondary data (exemption 4) and informed consent was not required.
RESULTS
Overall Sample Characteristics
Our sample comprised 607,729 fee-for-service beneficiaries diagnosed with chronic low back pain. See eFigure 1 in the supplement for the cohort selection flow diagram. The cohort had a mean (standard deviation) age of 68.2 (12.2) years, 64.1% were female, 80.3% were non-Hispanic White, 33.4% were dually enrolled in Medicare and Medicaid, and 75.4% resided in metropolitan counties (Table 1). Overall, 24.7% and 27.1% of chronic low back pain patients utilized any PT or chiropractic care in the 12 months following the index date, respectively. Almost 1 in 10 (9.4%) utilized both PT and chiropractic care (data not shown). The prevalence of OUD was 11.4%. The OUD subsample (N=51,643) was younger, with a lower proportion of females, more racially and ethnically diverse, had higher prevalence of low-income subsidy receipt and dual enrollment than the subsample without OUD. Non-back musculoskeletal pain, acute pain, mental illness, and non-OUD SUD were more common in beneficiaries with chronic low back pain and cooccurring OUD relative to those without OUD. Among those with cooccurring OUD, 14.6% and 6.8% utilized PT and chiropractic care; over the 12-month follow-up period; respectively.
Table 1:
Characteristics of Medicare beneficiaries with new episodes of chronic low back pain, overall and by OUD status
| Overall | OUD Diagnosis | No OUD Diagnosis | |
|---|---|---|---|
| Age ≥65 years, % | 444,620 (73.2) | 19,310 (37.4) | 425,310 (76.5) |
| Mean age (SD), years | 68.2 (12.2) | 59.3 (12.6) | 69.1 (11.8) |
| Female sex, % | 389,707 (64.1) | 30,979 (60.0) | 358,728 (64.5) |
| Race, % | |||
| Non-Hispanic White | 487,717 (80.3) | 39,754 (76.9) | 447,963 (80.6) |
| Black or African-American | 56,016 (9.2) | 7,331 (14.2) | 48,685 (8.8) |
| Asian or Pacific Islander | 14,457 (2.4) | 310 (0.6) | 14,147 (2.5) |
| Hispanic (any race) | 34,296 (5.6) | 3,061 (5.9) | 31,235 (5.6) |
| American Indian or Alaska Native | 3,737 (0.6) | 577 (1.1) | 3,160 (0.6) |
| Other | 3,776 (0.6) | 250 (0.5) | 3,526 (0.6) |
| Unknown | 7,730 (1.27) | 360 (0.7) | 7,370 (1.3) |
| Had low-income subsidy, % | 222,678 (36.6) | 34,038 (65.9) | 188,640 (33.9) |
| Reason for current Medicare entitlement, % | |||
| Old age and survivors Insurance | 444,624 (73.2) | 19,314 (37.4) | 425,310 (76.5) |
| Disability insurance benefits | 161,585 (26.6) | 32,040 (62.1) | 129,545 (23.3) |
| End-stage renal disease | 868 (0.14) | 169 (0.3) | 699 (0.1) |
| Had dual enrollment, % | 203,060 (33.4) | 31,145 (60.3) | 171,915 (30.9) |
| U.S. Region, % | |||
| Midwest | 151,770 (24.9) | 9,303 (18.0) | 142,467 (25.6) |
| Northeast | 105,693 (17.4) | 8,042 (15.6) | 97,651 (17.6) |
| South | 239,249 (39.4) | 25,434 (49.3) | 213,815 (38.5) |
| West | 110,331 (18.2) | 8,853 (17.1) | 101,478 (18.3) |
| Other | 686 (0.11) | 11 (0.02) | 675 (0.1) |
| Metropolitan status, % | |||
| Metropolitan | 457,906 (75.4) | 38,753 (75.1) | 419,263 (75.4) |
| Rural | 48,075 (7.9) | 4,209 (8.2) | 43,974 (7.9) |
| Urban | 101,433 (16.7) | 8,681 (16.8) | 928,49 (16.7) |
| Medical conditions diagnosed | |||
| Non-back musculoskeletal pain | 330,196 (54.33) | 32,758 (63.4) | 297,438 (53.5) |
| Falls | 35,057 (5.8) | 4,178 (8.1) | 30,879 (5.6) |
| Fractures | 27,848 (4.6) | 3,332 (6.5) | 24,516 (4.4) |
| Surgeries | 25,718 (4.2) | 2,664 (5.2) | 23,054 (4.2) |
| Tobacco use disorder | 49,137 (8.1) | 11,526 (22.3) | 37,611 (6.8) |
| Alcohol use disorder | 8,669 (1.4) | 2,394 (4.6) | 6,275 (1.1) |
| Non-opioid drug use disorder | 14,362 (2.36) | 6,071 (11.8) | 8,291 (1.5) |
| Mental health disorder | 190,454 (31.34) | 28,289 (54.8) | 162,165 (29.2) |
| Diabetes (type 1 or 2) | 170,265 (28) | 15,021 (29.1) | 155,244 (27.9) |
| Cardiovascular disease | 127,458 (20.97) | 10,981 (21.3) | 116,477 (20.9) |
| Chronic obstructive pulmonary disease | 94,494 (15.5) | 12,879 (24.9) | 81,615 (14.7) |
| Asthma | 53,276 (8.8) | 6,103 (11.8) | 47,173 (8.5) |
| Hypertension | 387,044 (63.7) | 32,106 (62.2) | 354,943 (63.8) |
| Chronic kidney disease | 66,987 (11.1) | 5,734 (11.1) | 61,253 (11.1) |
| Drug overdose | 4,031 (0.6) | 1,758 (3.4) | 2,323 (0.4) |
| Dementia | 15,235 (2.51) | 1,121 (2.2) | 14,114 (2.5) |
| Gagne comorbidity score | |||
| <0 or 0 | 299,233 (49.24) | 18,676 (36.2) | 280,557 (50.5) |
| 1 | 124,663 (20.51) | 11,759 (22.8) | 112,904 (20.3) |
| 2 to 3 | 116,920 (19.24) | 12,362 (23.9) | 104,558 (18.8) |
| 4 or more | 66,913 (11.0) | 8,846 (17.1) | 58,067 (10.4) |
| Died during follow-up | 12,762 (2.1) | 1,859 (3.6) | 10,566 (1.9) |
| Physical therapy use | |||
| Baseline | 83,248 (13.7) | 4,687 (9.1) | 78,561 (14.1) |
| Follow-up | 149,831 (24.7) | 7,541 (14.6) | 142,290 (25.6) |
| Chiropractic care use | |||
| Baseline | 106,473 (17.5) | 2,330 (4.5) | 104,143 (18.7) |
| Follow-up | 164,370 (27.0) | 3,509 (6.8) | 160,861 (28.9) |
NOTE: OUD = opioid use disorder, SD = standard deviation
Non-back musculoskeletal pain includes neck pain, fibromyalgia, and osteoarthritis; Non-opioid drug use disorder includes cannabis, sedative, cocaine and other use disorders; Mental health disorder includes schizophrenia, bipolar disorder, mood disorder, major depressive disorder, anxiety disorder, and psychotic disorder; Cardiovascular disease includes ischemic heart disease and heart failure
All comparisons of groups with OUD versus without OUD had P-value <0.0001 with the exception of metropolitan status (P=0.046), cardiovascular disease (P=0.0901), and chronic kidney disease (P=0.5407)
Characteristics of Members in PT Trajectory Groups
We identified three distinct trajectories for PT use (no/little use [76.6% of sample], delayed and increasing use [8.2%], and early and declining use [15.2%]). See Figure 1A. The no/little use group had 2.1% of its members who received any PT during the 12-month follow-up period. Beneficiaries within the no/little PT use group had a mean (standard deviation) age of 67.6 (12.6) years, 63.1% were female, and 79.9% non-Hispanic White (Table 2). The group with no/little PT use had the highest proportion of beneficiaries under 65 years compared to both the delayed and increasing use and early and declining use groups. People with OUD were under-represented in groups indicating early and declining use (4.3%) or delayed and increasing use (6.6%) versus no/little use (9.4%).
Figure 1A:
Final model for trajectories of physical therapy use in the 12 months following a new episode of chronic low back pain
Group 1: no or little PT use, Group 2: delayed and increasing PT use, and Group 3: early and declining PT use
Table 2:
Characteristics of Medicare beneficiaries with chronic low back pain in trajectory groups of physical therapy use
| No or little PT use (Group 1) | Delayed and increasing PT use (Group 2) | Early and declining PT use (Group 3) | |
|---|---|---|---|
| Age ≥65 years, % | 339,445 (70.3) | 36,582 (80.9) | 68,593 (86.1) |
| Mean age (SD), years | 67.6 (12.6) | 69.6 (10.7) | 70.8 (10.1) |
| Female sex, % | 304,820 (63.1) | 30,834 (68.2) | 54,053 (67.9) |
| Race, % | |||
| Non-Hispanic White | 385,736 (79.9) | 37,027 (81.9) | 64,954 (81.5) |
| Black or African-American | 47,863 (9.9) | 3,338 (7.4) | 4,815 (6.0) |
| Asian or Pacific Islander | 9,730 (2.0) | 1,258 (2.8) | 3,469 (4.4) |
| Hispanic (any race) | 27,972 (5.8) | 2,447 (5.4) | 3,877 (4.9) |
| American Indian or Alaska Native | 3,274 (0.7) | 199 (0.4) | 264 (0.33) |
| Other | 2,752 (0.6) | 303 (0.7) | 721 (0.91) |
| Unknown | 5,511 (1.1) | 662 (1.5) | 1,557 (2.0) |
| Had low-income subsidy, % | 193,160 (40.0) | 11,646 (25.8) | 17,872 (22.4) |
| Reason for current Medicare entitlement, % | |||
| Old age and survivors insurance | 339,448 (70.3) | 36,583 (80.9) | 68,593 (86.1) |
| Disability insurance benefits | 141,982 (29.4) | 8,590 (19.0) | 11,013 (13.8) |
| End-stage renal disease | 799 (0.2) | 35 (0.1) | 34 (0.04) |
| Had Dual Enrollment, % | 175,607 (36.4) | 10,845 (24.0) | 16,608 (20.9) |
| U.S. Region, % | |||
| Midwest | 127,414 (26.4) | 9,695 (21.4) | 14,661 (18.4) |
| Northeast | 77,084 (16.0) | 8,993 (19.9) | 19,616 (24.6) |
| South | 198,950 (41.2) | 15,898 (35.2) | 24,401 (30.6) |
| West | 78,887 (16.3) | 10,581 (23.4) | 20,863 (26.2) |
| Other | 503 (0.1) | 67 (0.2) | 116 (0.15) |
| Metropolitan Status, % | |||
| Metropolitan | 353,005 (73.1) | 37,208 (82.3) | 67,693 (85.0) |
| Rural | 42,270 (8.8) | 2,336 (5.2) | 3,469 (4.35) |
| Urban | 87,361 (18.1) | 5,656 (12.5) | 8,416 (10.6) |
| Medical conditions diagnosed | |||
| Opioid use disorder | 45,257 (9.4) | 2,981 (6.6) | 3,405 (4.3) |
| Non-back musculoskeletal pain | 255,754 (53.0) | 26,418 (58.4) | 48,024 (60.3) |
| Falls | 27,588 (5.7) | 2,463 (5.5) | 5,006 (6.3) |
| Fractures | 21,336 (4.4) | 2,060 (4.6) | 4,452 (5.6) |
| Surgeries | 17,871 (3.7) | 2,122 (4.7) | 5,725 (7.2) |
| Tobacco use disorder | 43,755 (9.1) | 2,382 (5.3) | 3,000 (3.8) |
| Alcohol use disorder | 7,400 (1.5) | 493 (1.1) | 776 (1.0) |
| Non-opioid drug use disorder | 12,580 (2.6) | 728 (1.6) | 1,054 (1.3) |
| Mental health disorder | 156,082 (32.3) | 13,350 (29.5) | 21,022 (26.4) |
| Diabetes (type 1 or 2) | 137,362 (28.5) | 12,106 (26.8) | 20,797 (26.1) |
| Cardiovascular disease | 102,363 (21.2) | 9,297 (20.6) | 15,798 (19.83) |
| Chronic obstructive pulmonary disease | 79,932 (16.6) | 5,828 (12.9) | 87,34 (11.0) |
| Asthma | 41,787 (8.7) | 4,326 (9.6) | 7,163 (9.0) |
| Hypertension | 308,973 (64.0) | 28,813 (63.7) | 49,263 (61.8) |
| Chronic kidney disease | 54,688 (11.3) | 4,671 (10.3) | 7,628 (9.6) |
| Drug overdose | 3,463 (0.7) | 238 (0.5) | 330 (0.4) |
| Dementia | 12,775 (2.6) | 827 (1.8) | 1,633 (2.1) |
| Gagne comorbidity score | |||
| <0 or 0 | 235,023 (48.7) | 22,825 (50.4) | 41,385 (52.0) |
| 1 | 98,686 (20.4) | 9,486 (21.0) | 16,491 (20.7) |
| 2 to 3 | 93,737 (19.4) | 8,588 (19.0) | 14,595 (18.3) |
| 4 or more | 55,392 (11.5) | 4,335 (9.6) | 7,186 (9.0) |
| Died during follow-up, % | 11,105 (2.3) | 226 (0.5) | 797 (1.0) |
NOTE: PT = physical therapy, SD = standard deviation
Non-back musculoskeletal pain includes neck pain, fibromyalgia, and osteoarthritis; Non-opioid drug use disorder includes cannabis, sedative, cocaine and other use disorders; Mental health disorder includes schizophrenia, bipolar disorder, mood disorder, major depressive disorder, anxiety disorder, and psychotic disorder; Cardiovascular disease includes ischemic heart disease and heart failure
P-values for group comparisons were all <0.0001 except for falls (P=0.0041)
Characteristics of Members in Chiropractic Care Trajectory Groups
The final model for chiropractic care use trajectories had three groups indicating no/little use (75.0%), early and declining use (17.3%), and early and persistent use (7.7%). See Figure 1B. Among the no/little chiropractic use group, 3.6% received any chiropractic care during the 12-month follow-up period. Beneficiaries within the no/little use group had a mean (standard deviation) age of 67.5 (12.8) years, 64.3% were female, and 76.6% non-Hispanic White (Table 3). Beneficiaries within the early and declining use group had a mean (standard deviation) age of 70.6 (9.8) years, 62.4% were female, and 91.1% non-Hispanic White and were generally similar to beneficiaries within the early and sustained use group: mean (standard deviation) age of 70.6 (9.4) years, 66.1% female, and 92.8% non-Hispanic White. The prevalence of OUD across chiropractic trajectories was 14.2% (no/little use), 2.9% (early and declining), and 1.9% (early and persistent).
Figure 1B:
Final model for trajectories of chiropractic care use in the 12 months following a new episode of chronic low back pain
Group 1: no or little chiropractic use, Group 2: early and declining chiropractic use, and Group 3: early and sustained chiropractic use
Table 3:
Characteristics of Medicare beneficiaries with chronic low back pain in trajectory groups of chiropractic care use
| No or little chiropractic use (Group 1) | Early and declining chiropractic use (Group 2) | Early and sustained chiropractic use (Group 3) | |
|---|---|---|---|
| Age ≥65 years, % | 313,438 (68.1) | 89,783 (88.4) | 41399 (89.7) |
| Mean age (SD), years | 67.5 (12.8) | 70.6 (9.8) | 70.6 (9.4) |
| Female sex, % | 295,868 (64.3) | 63,323 (62.4) | 30516 (66.1) |
| Race, % | |||
| Non-Hispanic White | 352,373 (76.6) | 92,524 (91.1) | 42820 (92.8) |
| Black or African-American | 52,963 (11.5) | 2,292 (2.3) | 761 (1.7) |
| Asian or Pacific Islander | 12,584 (2.7) | 1,426 (1.4) | 447 (1.0) |
| Hispanic (any race) | 30,909 (6.7) | 2,544 (2.5) | 843 (1.8) |
| American Indian or Alaska Native | 3,291 (0.7) | 321 (0.3) | 125 (0.3) |
| Other | 3,041 (0.7) | 520 (0.5) | 215 (0.5) |
| Unknown | 4,898 (1.1) | 1,902 (1.9) | 930 (2.0) |
| Had low-income subsidy, % | 200,752 (43.6) | 15,858 (15.6) | 6068 (13.2) |
| Reason for current Medicare entitlement, % | |||
| Old age and survivors insurance | 313,449 (68.1) | 89,776 (88.4) | 41399 (89.7) |
| Disability insurance benefits | 145,217 (31.6) | 11,651 (11.5) | 4717 (10.2) |
| End-stage renal disease | 794 (0.2) | 63 (0.1) | 11 (0.02) |
| Had dual enrollment, % | 183,620 (39.9) | 14,104 (13.9) | 5336 (11.6) |
| U.S. Region, % | |||
| Midwest | 97,142 (21.1) | 36,665 (36.1) | 17963 (38.9) |
| Northeast | 78,987 (17.2) | 17,800 (17.5) | 8906 (19.3) |
| South | 199,469 (43.4) | 28,411 (28.0) | 11369 (24.6) |
| West | 83,892 (18.2) | 18,553 (18.3) | 7886 (17.1) |
| Other | 569 (0.1) | 100 (0.1) | 17 (0.04) |
| Metropolitan status, % | |||
| Metropolitan | 355,567 (77.3) | 69,343 (68.3) | 32996 (71.5) |
| Rural | 31,902 (6.9) | 11,826 (11.7) | 4347 (9.4) |
| Urban | 72,347 (15.7) | 20,309 (20.0) | 8777 (19.0) |
| Medical conditions diagnosed | |||
| Opioid use disorder | 48,941 (10.6) | 2,103 (2.1) | 599 (1.3) |
| Non-back musculoskeletal pain | 245,191 (53.3) | 57,063 (56.2) | 27942 (60.6) |
| Falls | 30,195 (6.6) | 3,491 (3.4) | 1371 (3.0) |
| Fractures | 24,147 (5.3) | 2,661 (2.6) | 1040 (2.3) |
| Surgeries | 21,438 (4.7) | 3,042 (3.0) | 1,238 (2.7) |
| Tobacco use disorder | 44,886 (9.8) | 3,253 (3.2) | 998 (2.2) |
| Alcohol use disorder | 7,705 (1.7) | 743 (0.7) | 221 (0.5) |
| Non-opioid drug use disorder | 13,375 (2.9) | 765 (0.8) | 222 (0.5) |
| Mental health disorder | 161,743 (35.2) | 20,290 (20.0) | 8,421 (18.3) |
| Diabetes (type 1 or 2) | 139,188 (30.3) | 21,964 (21.6) | 9,113 (19.8) |
| Cardiovascular disease | 103,099 (22.4) | 17,664 (17.4) | 6,695 (14.5) |
| Chronic obstructive pulmonary disease | 81,527 (17.7) | 9,323 (9.2) | 3,644 (7.9) |
| Asthma | 43,676 (9.5) | 6,420 (6.3) | 3,180 (6.9) |
| Hypertension | 308,846 (67.1) | 54,238 (53.4) | 23,965 (51.9) |
| Chronic kidney disease | 55,757 (12.1) | 7,949 (7.8) | 3281 (7.1) |
| Drug overdose | 3,755 (0.8) | 212 (0.2) | 64 (0.1) |
| Dementia | 13,710 (3.0) | 1,137 (1.1) | 388 (0.84) |
| Gagne comorbidity score | |||
| <0 or 0 | 207,849 (45.2) | 61,968 (61.1) | 2,9416 (63.8) |
| 1 | 97,156 (21.1) | 18,957 (18.7) | 8,550 (18.5) |
| 2 to 3 | 96,433 (21.0) | 14,534 (14.3) | 5,953 (12.9) |
| 4 or more | 58,621 (12.7) | 6,070 (6.0) | 2,222 (4.8) |
| Died during follow-up, % | 10,581 (2.3) | 1,117 (1.1) | 876 (1.1) |
NOTE: OUD = opioid use disorder, SD = standard deviation
Non-back musculoskeletal pain includes neck pain, fibromyalgia, and osteoarthritis; Non-opioid drug use disorder includes cannabis, sedative, cocaine and other use disorders; Mental health disorder includes schizophrenia, bipolar disorder, mood disorder, major depressive disorder, anxiety disorder, and psychotic disorder; Cardiovascular disease includes ischemic heart disease and heart failure
P-values for group comparisons were all <0.0001
Regression Results for PT
The reference category was the group with little/no PT use. After adjusting for demographic, geographic, and medical covariates, OUD diagnosis was associated with lower likelihood of membership in groups with delayed and increasing (adjusted odds ratio (aOR)=0.88, 99% confidence interval (CI)=0.83–0.92) or early and declining (aOR=0.61, 99%CI=0.58–0.64) PT use. Beneficiaries below age 65 (aOR=0.82, 99%CI=0.79–0.85), male (aOR=0.86, 99%CI=0.83–0.88), and residing in rural counties (aOR=0.64, 99%CI=0.60–0.67) were less likely to belong in groups with delayed and increasing PT use (Table 4). Similarly, beneficiaries below age 65 (aOR=0.63, 99%CI=0.61–0.65), male (aOR=0.91, 99%CI=0.89–0.94), and Black/African American (aOR=0.87, 99%CI=0.83–0.91) were less likely to be in groups with early and declining PT use. Beneficiaries with low-income subsidies had lower likelihood of membership in groups with delayed and increasing or early and declining PT use. Beneficiaries that had prior utilization (during 6-month baseline) of PT were more likely to be in groups with delayed and increasing (aOR=2.65,99% CI=2.56–2.75) and early and declining (aOR=9.73, 99% CI=9.49–9.97) PT use. Finally, beneficiaries that had prior utilization of chiropractic care were less likely to belong in groups with early and declining (aOR=0.89, 99% CI=0.87–0.97) PT use. Regression results comparing the group with early and declining PT use against the group with delayed and increasing use are available in eTable 2 in the supplement.
Table 4:
Multivariable regression results for the association between OUD diagnosis and membership in trajectory groups
| Physical therapy (delayed and increasing vs. no/little) use | Physical therapy (early and declining vs. no/little) use | Chiropractic care (early and declining vs. no/little) use | Chiropractic care (early and sustained vs. no/little) use | |
|---|---|---|---|---|
| OUD diagnosis (ref = none) | 0.88 (0.83–0.92) | 0.61 (0.58–0.64) | 0.37 (0.35–0.39) | 0.27 (0.23–0.31) |
| Age ≥65 years (ref=under 65 years) | 0.82 (0.79–0.85) | 0.63 (0.61–0.65) | 0.56 (0.54–0.58) | 0.69 (0.64–0.74) |
| Male (ref=female) | 0.86 (0.83–0.88) | 0.91 (0.89–0.94) | 1.15 (1.12–1.18) | 0.96 (0.92–1.01) |
| Race and ethnicity (ref = Non-Hispanic White) | ||||
| Black or African American | 0.93 (0.88–0.97) | 0.87 (0.83–0.91) | 0.40 (0.38–0.43) | 0.41 (0.37–0.47) |
| Hispanic (any race) | 0.98 (0.92–1.04) | 0.90 (0.86–0.96) | 0.62 (0.58–0.66) | 0.55 (0.48–0.62) |
| Asian or Pacific Islander | 1.20 (1.10–1.29) | 1.56 (1.47–1.66) | 0.69 (0.63–0.76) | 0.51 (0.43–0.61) |
| American Indian or Alaska Native | 0.83 (0.69–1.01) | 0.79 (0.67–0.93) | 0.67 (0.56–0.81) | 0.79 (0.54–1.14) |
| Other | 1.04 (0.89–1.22) | 1.22 (1.09–1.38) | 0.76 (0.65–0.88) | 0.79 (0.59–1.07) |
| Unknown | 1.14 (1.03–1.27) | 1.38 (1.27–1.49) | 1.27 (1.16–1.39) | 1.49 (1.24–1.79) |
| Low-income subsidy (ref=none) | 0.61 (0.59–0.63) | 0.58 (0.56–0.60) | 0.51 (0.49–0.53) | 0.57 (0.53–0.60) |
| U.S. Region (ref = South) | ||||
| Midwest | 0.94 (0.91–0.97) | 0.96 (0.93–0.98) | 2.04 (1.98–2.11) | 2.00 (1.89–2.12) |
| Northeast | 1.25 (1.21–1.30) | 1.61 (1.56–1.65) | 1.47 (1.42–1.52) | 1.56 (1.46–1.67) |
| West | 1.44 (1.39–1.48) | 1.57 (1.52–1.61) | 1.38 (1.33–1.43) | 1.33 (1.24–1.42) |
| Other | 1.16 (0.82–1.65) | 1.21 (0.88–1.65) | 1.65 (1.19–2.29) | 0.88 (0.43–1.82) |
| Metropolitan status (ref = metropolitan) | ||||
| Rural | 0.64 (0.60–0.67) | 0.62 (0.59–0.65) | 1.65 (1.59–1.72) | 1.02 (0.93–1.11) |
| Urban | 0.72 (0.69–0.74) | 0.67 (0.66–0.70) | 1.36 (1.31–1.39) | 1.05 (0.98–1.12) |
| Medical conditions (ref = absent) | ||||
| Non-back musculoskeletal pain | 1.18 (1.15–1.20) | 1.11 (1.09–1.14) | 1.09 (1.07–1.13) | 1.07 (1.02–1.12) |
| Falls | 0.98 (0.92–1.03) | 1.05 (1.00–1.11) | 0.74 (0.69–0.78) | 0.68 (0.61–0.77) |
| Fractures | 0.94 (0.89–1.01) | 0.99 (0.94–1.05) | 0.58 (0.54–0.62) | 0.52 (0.46–0.60) |
| Surgeries | 1.06 (1.00–1.13) | 1.32 (1.25–1.38) | 0.64 (0.59–0.69) | 0.49 (0.44–0.56) |
| Tobacco use disorder | 0.81 (0.76–0.85) | 0.73 (0.69–0.77) | 0.75 (0.71–0.79) | 0.59 (0.53–0.67) |
| Alcohol use disorder | 1.03 (0.91–1.17) | 1.09 (0.98–1.21) | 1.27 (1.12–1.44) | 0.97 (0.74–1.27) |
| Non-opioid drug use disorder | 0.90 (0.81–0.99) | 0.98 (0.90–1.07) | 0.95 (0.84–1.07) | 0.71 (0.56–0.91) |
| Mental health disorder | 1.02 (0.99–1.05) | 0.92 (0.90–0.95) | 0.80 (0.78–0.83) | 0.75 (0.71–0.79) |
| Diabetes (type 1 or 2) | 0.99 (0.96–1.02) | 0.99 (0.97–1.02) | 0.95 (0.93–0.98) | 0.88 (0.84–0.94) |
| Cardiovascular disease | 1.01 (0.97–1.04) | 0.99 (0.97–1.03) | 0.99 (0.96–1.03) | 0.87 (0.82–0.93) |
| Chronic obstructive pulmonary disease | 0.87 (0.83–0.90) | 0.81 (0.78–0.83) | 0.79 (0.77–0.83) | 0.78 (0.72–0.84) |
| Asthma | 1.16 (1.11–1.21) | 1.14 (1.10–1.19) | 1.08 (1.03–1.13) | 1.21 (1.11–1.32) |
| Hypertension | 0.97 (0.94–0.99) | 0.85 (0.83–0.87) | 0.61 (0.59–0.63) | 0.64 (0.61–0.67) |
| Chronic kidney disease | 0.92 (0.87–0.96) | 0.81 (0.78–0.83) | 1.01 (0.96–1.06) | 1.09 (0.99–1.19) |
| Drug overdose | 0.99 (0.83–1.18) | 0.83 (0.70–0.98) | 0.83 (0.66–1.04) | 0.76 (0.47–1.24) |
| Dementia | 0.70 (0.64–0.76) | 0.76 (0.71–0.82) | 0.62 (0.56–0.69) | 0.65 (0.54–0.79) |
| Gagne comorbidity score (ref = <0 or 0) | ||||
| 1 | 1.05 (1.01–1.09) | 0.99 (0.96–1.02) | 0.81 (0.79–0.84) | 0.78 (0.73–0.83) |
| 2–3 | 1.04 (1.00–1.08) | 0.95 (0.92–0.98) | 0.75 (0.72–0.78) | 0.70 (0.65–0.76) |
| >4 | 1.00 (0.94–1.06) | 0.93 (0.88–0.98) | 0.69 (0.66–0.74) | 0.66 (0.59–0.74) |
| Baseline physical therapy use | 2.65 (2.56–2.75) | 9.73 (9.49–9.97) | 0.67 (0.64–0.69) | 0.58 (0.55–0.62) |
| Baseline chiropractic use | 1.02 (0.99–1.05) | 0.89 (0.87–0.97) | 38.29 (37.06–39.56) | 252.17 (240.26–264.67) |
| Year, index date for back pain diagnosis (2017 vs. 2016) | 1.07 (1.04–1.09) | 1.08 (1.06–1.11) | 1.04 (1.02–1.07) | 1.52 (1.45–1.59) |
NOTE: OUD = opioid use disorder, aOR = adjusted odds ratio
Non-back musculoskeletal pain includes neck pain, fibromyalgia, and osteoarthritis; Non-opioid drug use disorder includes cannabis, sedative, cocaine and other use disorders; Mental health disorder includes schizophrenia, bipolar disorder, mood disorder, major depressive disorder, anxiety disorder, and psychotic disorder; Cardiovascular disease includes ischemic heart disease and heart failure
Regression Results for Chiropractic Care
The group with little/no chiropractic use was the reference category. After adjustment, OUD diagnosis was associated with significantly diminished likelihood of group membership in trajectories of early and declining (aOR=0.37, 99% CI=0.34–0.39) and early and sustained (aOR=0.28, 99% CI=0.25–0.32) chiropractic use (Table 3). Beneficiaries below age 65 (aOR=0.56, 99%CI=0.54–0.58) and Black/African American (aOR=0.40, 99%CI=0.38–0.43) were less likely to be in groups with early and declining chiropractic use. Beneficiaries with low-income subsidies were less likely to be in trajectories of early chiropractic use regardless of the subsequent trajectory: declining (aOR=0.50, 99%CI=0.49–0.53) or sustained (aOR=0.60, 99%CI=0.58–0.62). The characteristics strongly associated with greater likelihood of membership in groups with early and sustained chiropractic use included residence in the Midwest region, musculoskeletal pain diagnosis other than back pain, secular time, and history of receiving of chiropractic care. Prior utilization of chiropractic care at baseline appeared almost necessary for membership in groups with early and declining (aOR=38.3,99% CI=37.1–39.6) and early and sustained (aOR=252.2, 99% CI=240.3–264.7) chiropractic use. Beneficiaries that had baseline utilization of PT had lower likelihood of membership in groups with early and declining (OR=0.67, 99% CI=0.64–0.69) and early and sustained (OR=0.58, 99% CI=0.55–0.62) chiropractic use. Regression results comparing the group with early and sustained chiropractic use against the group with early and declining use are available in eTable 2.
DISCUSSION
We investigated the longitudinal use of nonpharmacologic therapies among fee-for-service Medicare beneficiaries diagnosed with chronic low back pain and examined differences by OUD status. Our study yielded several important findings. First, considering guideline recommendations, we expected early initiation to represent an optimal trajectory; however, people with chronic low back pain often did not receive guideline-recommended pain management. The vast majority of beneficiaries belonged in trajectory groups representing limited or no receipt of PT (76.6%) or chiropractic care (75.0%). Second, OUD diagnosis was associated with decreased likelihood of membership in potentially optimal trajectories characterized by early and consistent PT or chiropractic care use over time. Third, the trajectory curves for PT use differed noticeably from those for chiropractic care. Unlike PT, we found that the timing of chiropractic care receipt predominantly coincided with the beginning of the new episode of chronic pain and remained consistent (regardless of level) during the 12-month follow-up period. We found that chiropractic services were initiated earlier and those who received chiropractic care persisted with the services more than the recipients of PT. Potential explanations for these observations are that chiropractic services do not require referrals which may reduce barriers to accessing the services compared with PT which requires a referral. Provider practices such as watchful waiting before issuing a referral for PT may also explain delayed receipt of PT.44
Our finding that beneficiaries with early PT initiation were on average older, Non-Hispanic White, more financially secure (as indicated by lower prevalence of low-income subsidy and dual enrollment) and predominantly resided in metropolitan or urban locations compared with beneficiaries in groups with delayed or little/no PT use is consistent with prior research on social, economic, and geographic facilitators of access to and use of PT.34, 45, 46 The factors associated with early uptake of chiropractic care were similar to those observed for PT; however, the magnitude of the associations was larger indicating more pronounced differences when compared to the characteristics of beneficiaries with little/no chiropractic use. For example, racial and ethnic disparities were greater for chiropractic care than for PT receipt. This may be explained by relatively narrow Medicare coverage for chiropractic care which is limited to spinal manipulation to correct subluxation; whereas, PT services are coverable for broader reasons. Cultural variations in attitudes toward chiropractic versus PT services are further potential reasons for the observed disparities.47 Our findings also point to prior experience with chiropractic or PT as a strong predictor of future receipt of these services during a new episode of chronic low back pain. This association was stronger for chiropractic use where we saw that the vast majority of individuals who belonged in trajectory groups with early and moderate/high utilization of chiropractic had received chiropractic in the year preceding the new pain episode. An observation that reinforces the idea that individuals who actively engage in chiropractic pain management have potentially longstanding positive attitudes and experiences with this care.
We found that beneficiaries with cooccurring chronic low back pain and OUD were significantly less likely to receive PT or chiropractic care. This is concerning for several reasons. This finding suggests that people with OUD experience disparities in pain management and/or that they may choose not to pursue nonpharmacologic therapy when it is recommended to them by their providers.48 Potential factors driving disparities include stigma, active substance use preventing people with OUD from seeking care, low referral rates due to provider expectations that people with OUD will not follow through with nonpharmacologic therapies, and psychiatric and medical comorbidities. Prior studies indicate that early initiation of PT and increasing the availability of PT providers per capita are associated with reductions in future opioid use among individuals with chronic musculoskeletal pain.49–51 Therefore, people who are not started on or do not receive nonpharmacologic therapies maybe more likely to be prescribed opioids, which are neither recommended nor effective for chronic pain. Improving access to and use of nonpharmacologic therapies is a critical step to advance safe and equitable pain management especially in people with OUD who have been reported to frequently receive opioids at high doses and for long periods of time38, 52 but also are vulnerable to tapering and discontinuation of opioid therapy.53 Furthermore, since it is much easier to write a script for an opioid, even (or perhaps especially) for someone with OUD, than to engage in discussion about options for nonpharmacologic pain therapies, there is a need to understand and optimize the decision-making process for nonpharmacologic approaches to pain treatment. The reasons why individuals may choose to forgo nonpharmacologic therapies despite being provided with education and referral also warrant exploration in efforts to advance pain management consistent with practice guidelines.
The study has limitations. First, the administrative claims data used for the analysis preclude the ability to capture over-the-counter pain medication, self-management of pain, and self-paid care. Second, our definition for receipt of PT excluded procedure codes for prosthetic services, wound care and cardiorespiratory rehabilitation. While more limited, our definition for PT allowed us to better identify services that are more relevant for back pain management. Third, our analysis which included individuals with dual enrollment in Medicare and Medicaid may underestimate nonpharmacologic therapy use to the extent that these individuals receive these services through Medicaid for which data were unavailable. Fourth, our pre-specified criteria for identifying the best trajectory models ruled out groups which represented less than 5% of the cohort. Therefore, our efforts to aid the interpretation and presentation of the findings may have led to distinct trajectories representing small subgroups being missed. This is particularly relevant for chiropractic care were one of the models we considered and did not select had 2% of the cohort showing delayed initiation (data not shown). Lastly, information about referrals and patient preferences for pain treatment was unavailable.
A strength of this study is that it is among the first to describe the trajectories of nonpharmacologic therapy use. Our findings extend the existing literature by demonstrating longitudinal variability in the receipt of PT and chiropractic care, two key nonpharmacologic pain therapies that were Medicare-coverable during the study period. Our study identified unanswered questions and opportunities for future research which are especially relevant amid updated clinical practice guidelines which continue to call for expanded uptake of nonpharmacologic therapies to manage chronic pain.54, 55 For example, are clinicians even more likely to prescribe an opioid to a person with OUD compared with someone without OUD rather than engage in a discussion as to why nonpharmacologic approaches may be the best next step? Studies are needed to explore the impact of adding nonpharmacologic therapies on opioid dosage and duration as well as important patient-centered outcomes including pain control, function, and quality of life. Future research is also needed to examine the uptake of acupuncture services, following new Medicare coverage that began in 2020, and whether acupuncture shifted PT and chiropractic care receipt.
In conclusion, this study demonstrates that only a minority of Medicare beneficiaries with new episodes of chronic low back pain receive and persist with PT or chiropractic care in the subsequent months. People with OUD have increased risk of not receiving nonpharmacologic pain management at all and deserve particular consideration to ensure their access to adequate and guideline-concordant care for chronic pain. Evaluations of barriers and facilitators to nonpharmacologic pain treatment accompanied by systematic interventions to disseminate and implement multimodal and guideline-concordant management of chronic pain are critically needed, especially in subgroups for whom opioid therapy carries increased risk of harm.
Supplementary Material
HIGHLIGHTS.
The prevalence of physical therapy use within one year of pain diagnosis was 24.7%
The prevalence of chiropractic care use over the same period was 27.1%
Those with OUD were more likely to have little/no PT compared to other trajectories
Among recipients, the timing of chiropractic care use was often earlier than for PT
People with OUD may face disparities in accessing nonpharmacologic pain therapies
Perspective:
Physical therapy and chiropractic care use was low overall and even lower among Medicare beneficiaries with co-occurring OUD compared with those without OUD. As updated guidelines on pain management are promulgated, targeted interventions (e.g., insurance policy, provider and patient education) are needed to ensure equitable access to guideline-recommended pain therapies.
Funding sources:
This work was funded by the National Institutes of Health under grant P20GM130414 from the National Institute of General Medical Sciences (NIGMS).
Role of the funder:
The NIGMS had no role in the design and conduct of the study, analysis and interpretation, and preparation of this manuscript.
Conflicts of interest:
Dr. Moyo reported receiving grants from the National Institutes of Health during the conduct of the study, serving as a technical expert panelist for an Abt Associates study focused on opioid use and misuse in older adults, and being a member of a National Academies of Sciences, Engineering, and Medicine ad hoc committee on evaluating the effects of opioids and benzodiazepines on all-cause mortality in Veterans. Dr. Merlin reported receiving grant funding from the Cambia Health Foundation. Dr. Marshall reported receiving grants from the National Institutes of Health, Arnold Ventures, and Open Society Foundations during the conduct of the study. No other disclosures were reported.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
REFERENCES
- 1.Chou R, Deyo R, Friedly J, et al. Nonpharmacologic Therapies for Low Back Pain: A Systematic Review for an American College of Physicians Clinical Practice Guideline. Ann Intern Med. 2017;166(7):493–505. doi: 10.7326/M16-2459 [DOI] [PubMed] [Google Scholar]
- 2.Qaseem A, Wilt TJ, McLean RM, et al. Noninvasive Treatments for Acute, Subacute, and Chronic Low Back Pain: A Clinical Practice Guideline From the American College of Physicians. Ann Intern Med. 2017;166(7):514–530. doi: 10.7326/m16-2367 [DOI] [PubMed] [Google Scholar]
- 3.Stevans JM, Delitto A, Khoja SS, et al. Risk Factors Associated With Transition From Acute to Chronic Low Back Pain in US Patients Seeking Primary Care. JAMA Netw Open. 2021;4(2):e2037371. doi: 10.1001/jamanetworkopen.2020.37371 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Sandbrink F, Murphy JL, Johansson M, et al. The Use of Opioids in the Management of Chronic Pain: Synopsis of the 2022 Updated U.S. Department of Veterans Affairs and U.S. Department of Defense Clinical Practice Guideline. Ann Intern Med. 2023;176(3):388–397. doi: 10.7326/M22-2917 [DOI] [PubMed] [Google Scholar]
- 5.Dowell D, Haegerich TM, Chou R. CDC Guideline for Prescribing Opioids for Chronic Pain--United States, 2016. JAMA. 2016;315(15):1624–1645. doi: 10.1001/jama.2016.1464 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Miller GF, Guy GP Jr., Zhang K, Mikosz CA, Xu L. Prevalence of Nonopioid and Opioid Prescriptions Among Commercially Insured Patients with Chronic Pain. Pain Med. 2019;20(10):1948–1954. doi: 10.1093/pm/pny247 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Cicero TJ, Wong G, Tian Y, Lynskey M, Todorov A, Isenberg K. Co-morbidity and utilization of medical services by pain patients receiving opioid medications: data from an insurance claims database. Pain. 2009;144(1–2):20–27. doi: 10.1016/j.pain.2009.01.026 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Reuben DB, Alvanzo AA, Ashikaga T, et al. National Institutes of Health Pathways to Prevention Workshop: the role of opioids in the treatment of chronic pain. Ann Intern Med. 2015;162(4):295–300. doi: 10.7326/M14-2775 [DOI] [PubMed] [Google Scholar]
- 9.Meroni R, Piscitelli D, Ravasio C, et al. Evidence for managing chronic low back pain in primary care: a review of recommendations from high-quality clinical practice guidelines. Disabil Rehabil. 2021;43(7):1029–1043. doi: 10.1080/09638288.2019.1645888 [DOI] [PubMed] [Google Scholar]
- 10.Vowles KE, McEntee ML, Julnes PS, Frohe T, Ney JP, van der Goes DN. Rates of opioid misuse, abuse, and addiction in chronic pain: a systematic review and data synthesis. Pain. 2015;156(4):569–576. doi: 10.1097/01.j.pain.0000460357.01998.f1 [DOI] [PubMed] [Google Scholar]
- 11.Uebelacker LA, Van Noppen D, Tremont G, Bailey G, Abrantes A, Stein M. A pilot study assessing acceptability and feasibility of hatha yoga for chronic pain in people receiving opioid agonist therapy for opioid use disorder. J Subst Abuse Treat. 2019;105:19–27. doi: 10.1016/j.jsat.2019.07.015 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Burgess HJ, Siddiqui A, Burgess FW. Long-term opioid therapy for chronic pain and the risk of opioid addiction. R I Med J (2013). 2014;97(10):25–28 [PubMed] [Google Scholar]
- 13.Chou R, Turner JA, Devine EB, et al. The effectiveness and risks of long-term opioid therapy for chronic pain: a systematic review for a National Institutes of Health Pathways to Prevention Workshop. Ann Intern Med. 2015;162(4):276–286. doi: 10.7326/M14-2559 [DOI] [PubMed] [Google Scholar]
- 14.Darnall BD, Stacey BR, Chou R. Medical and psychological risks and consequences of long-term opioid therapy in women. Pain Med. 2012;13(9):1181–1211. doi: 10.1111/j.1526-4637.2012.01467.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Harned M, Sloan P. Safety concerns with long-term opioid use. Expert Opin Drug Saf. 2016;15(7):955–962. doi: 10.1080/14740338.2016.1177509 [DOI] [PubMed] [Google Scholar]
- 16.Nazarian A, Negus SS, Martin TJ. Factors mediating pain-related risk for opioid use disorder. Neuropharmacology. 2021;186:108476. doi: 10.1016/j.neuropharm.2021.108476 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Speed TJ, Parekh V, Coe W, Antoine D. Comorbid chronic pain and opioid use disorder: literature review and potential treatment innovations. Int Rev Psychiatry. 2018;30(5):136–146. doi: 10.1080/09540261.2018.1514369 [DOI] [PubMed] [Google Scholar]
- 18.Gorfinkel L, Voon P, Wood E, Klimas J. Diagnosing opioid addiction in people with chronic pain. BMJ. 2018;362:k3949. doi: 10.1136/bmj.k3949 [DOI] [PubMed] [Google Scholar]
- 19.Becker WC, Ganoczy D, Fiellin DA, Bohnert AS. Buprenorphine/Naloxone Dose and Pain Intensity Among Individuals Initiating Treatment for Opioid Use Disorder. J Subst Abuse Treat. 2015;48(1):128–131. doi: 10.1016/j.jsat.2014.09.007 [DOI] [PubMed] [Google Scholar]
- 20.Szalavitz M No One Should Have to Prove Their Worth to Get Medical Care, Regardless of Addiction or Pain. Narrat Inq Bioeth. 2018;8(3):233–237. doi: 10.1353/nib.2018.0072 [DOI] [PubMed] [Google Scholar]
- 21.Collins ED, Streltzer J. Should opioid analgesics be used in the management of chronic pain in opiate addicts? Am J Addict. 2003;12(2):93–100 [PubMed] [Google Scholar]
- 22.Miotto K, Kaufman A, Kong A, Jun G, Schwartz J. Managing co-occurring substance use and pain disorders. Psychiatr Clin North Am. 2012;35(2):393–409. doi: 10.1016/j.psc.2012.03.006 [DOI] [PubMed] [Google Scholar]
- 23.Voon P, Karamouzian M, Kerr T. Chronic pain and opioid misuse: a review of reviews. Subst Abuse Treat Prev Policy. 2017;12(1):36. doi: 10.1186/s13011-017-0120-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Bonar EE, Ilgen MA, Walton M, Bohnert AS. Associations among pain, non-medical prescription opioid use, and drug overdose history. Am J Addict. 2014;23(1):41–47. doi: 10.1111/j.1521-0391.2013.12055.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Samet JH, Kertesz SG. Suggested Paths to Fixing the Opioid Crisis: Directions and Misdirections. JAMA Netw Open. 2018;1(2):e180218. doi: 10.1001/jamanetworkopen.2018.0218 [DOI] [PubMed] [Google Scholar]
- 26.Darnall BD, Juurlink D, Kerns RD, et al. International Stakeholder Community of Pain Experts and Leaders Call for an Urgent Action on Forced Opioid Tapering. Pain Med. 2019;20(3):429–433. doi: 10.1093/pm/pny228 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Pergolizzi JV Jr., Rosenblatt M, LeQuang JA. Three Years Down the Road: The Aftermath of the CDC Guideline for Prescribing Opioids for Chronic Pain. Adv Ther. 2019;36(6):1235–1240. doi: 10.1007/s12325-019-00954-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Stuart GL, Shorey RC, France CR, et al. Empirical Studies Addressing the Opioid Epidemic: An Urgent Call for Research. Subst Abuse. 2018;12:1178221818784294. doi: 10.1177/1178221818784294 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Astin JA, Pelletier KR, Marie A, Haskell WL. Complementary and alternative medicine use among elderly persons: one-year analysis of a Blue Shield Medicare supplement. J Gerontol A Biol Sci Med Sci. 2000;55(1):M4–9. doi: 10.1093/gerona/55.1.m4 [DOI] [PubMed] [Google Scholar]
- 30.Ngo L, Latham NK, Jette AM, Soukup J, Iezzoni LI. Use of physical and occupational therapy by Medicare beneficiaries within five conditions: 1994–2001. Am J Phys Med Rehabil. 2009;88(4):308–321. doi: 10.1097/PHM.0b013e318198a791 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Davis MA, Yakusheva O, Gottlieb DJ, Bynum JP. Regional Supply of Chiropractic Care and Visits to Primary Care Physicians for Back and Neck Pain. J Am Board Fam Med. 2015;28(4):481–490. doi: 10.3122/jabfm.2015.04.150005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Davis MA, Yakusheva O, Liu H, Tootoo J, Titler MG, Bynum JPW. Access to chiropractic care and the cost of spine conditions among older adults. Am J Manag Care. 2019;25(8):e230–e236 [PMC free article] [PubMed] [Google Scholar]
- 33.Weigel PA, Hockenberry JM, Bentler SE, Kaskie B, Wolinsky FD. Chiropractic episodes and the co-occurrence of chiropractic and health services use among older Medicare beneficiaries. J Manipulative Physiol Ther. 2012;35(3):168–175. doi: 10.1016/j.jmpt.2012.01.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Mayer-Oakes SA, Hoenig H, Atchison KA, Lubben JE, De Jong F, Schweitzer SO. Patient-related predictors of rehabilitation use for community-dwelling older Americans. J Am Geriatr Soc. 1992;40(4):336–342. doi: 10.1111/j.1532-5415.1992.tb02131.x [DOI] [PubMed] [Google Scholar]
- 35.Bohnert ASB, Guy GP Jr., Losby JL. Opioid Prescribing in the United States Before and After the Centers for Disease Control and Prevention’s 2016 Opioid Guideline. Ann Intern Med. 2018;169(6):367–375. doi: 10.7326/M18-1243 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Khouja T, Tadrous M, Matusiak L, Suda K. Opioid Prescribing in United States Health Systems, 2015 to 2019. Value Health. 2021;24(9):1279–1284. doi: 10.1016/j.jval.2021.04.1274 [DOI] [PubMed] [Google Scholar]
- 37.Groenewald CB, Murray CB, Battaglia M, Scaini S, Quinn PD. Prevalence of Pain Management Techniques Among Adults With Chronic Pain in the United States, 2019. JAMA Netw Open. 2022;5(2):e2146697. doi: 10.1001/jamanetworkopen.2021.46697 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Moyo P, Vaillant J, Girard A, et al. Prevalence of opioid and nonopioid pain management therapies among Medicare beneficiaries with musculoskeletal pain conditions from 2016 to 2019. Drug Alcohol Depend. 2023;248:109930. doi: 10.1016/j.drugalcdep.2023.109930 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Ly DP. Evaluation and Treatment Patterns of New Low Back Pain Episodes for Elderly Adults in the United States, 2011–2014. Med Care. 2020;58(2):108–113. doi: 10.1097/mlr.0000000000001244 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Jones BL, Nagin DS. Advances in Group-Based Trajectory Modeling and an SAS Procedure for Estimating Them. Sociological Methods Research. 2007;35(4):542–571. [Google Scholar]
- 41.Nagin DS, Odgers CL. Group-based trajectory modeling in clinical research. Annu Rev Clin Psychol. 2010;6:109–138. doi: 10.1146/annurev.clinpsy.121208.131413 [DOI] [PubMed] [Google Scholar]
- 42.Zhou L, Bhattacharjee S, Kwoh CK, et al. Dual-trajectories of opioid and gabapentinoid use and risk of subsequent drug overdose among Medicare beneficiaries in the United States: a retrospective cohort study. Addiction. 2021;116(4):819–830. doi: 10.1111/add.15189 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Pines HA, Gorbach PM, Weiss RE, et al. Sexual risk trajectories among MSM in the United States: implications for pre-exposure prophylaxis delivery. J Acquir Immune Defic Syndr. 2014;65(5):579–586. doi: 10.1097/QAI.0000000000000101 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Pradhan BB. Evidence-informed management of chronic low back pain with watchful waiting. Spine J. 2008;8(1):253–257. doi: 10.1016/j.spinee.2007.10.028 [DOI] [PubMed] [Google Scholar]
- 45.Gell NM, Mroz TM, Patel KV. Rehabilitation Services Use and Patient-Reported Outcomes Among Older Adults in the United States. Arch Phys Med Rehabil. 2017;98(11):2221–2227.e2223. doi: 10.1016/j.apmr.2017.02.027 [DOI] [PubMed] [Google Scholar]
- 46.Weigel P, Hockenberry JM, Bentler SE, et al. A longitudinal study of chiropractic use among older adults in the United States. Chiropr Osteopat. 2010;18:34. doi: 10.1186/1746-1340-18-34 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Whedon JM, Song Y. Racial disparities in the use of chiropractic care under Medicare. Altern Ther Health Med. 2012;18(6):20–26 [PMC free article] [PubMed] [Google Scholar]
- 48.Bhondoekhan F, Marshall BDL, Shireman TI, Trivedi AN, Merlin JS, Moyo P. Racial and Ethnic Differences in Receipt of Nonpharmacologic Care for Chronic Low Back Pain Among Medicare Beneficiaries With OUD. JAMA Netw Open. Sep 5 2023;6(9):e2333251. doi: 10.1001/jamanetworkopen.2023.33251 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Karmali RN, Skinner AC, Trogdon JG, Weinberger M, George SZ, Hassmiller Lich K. The association between the supply of select nonpharmacologic providers for pain and use of nonpharmacologic pain management services and initial opioid prescribing patterns for Medicare beneficiaries with persistent musculoskeletal pain. Health Serv Res. 2021;56(2):275–288. doi: 10.1111/1475-6773.13561 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Thackeray A, Hess R, Dorius J, Brodke D, Fritz J. Relationship of Opioid Prescriptions to Physical Therapy Referral and Participation for Medicaid Patients with New-Onset Low Back Pain. The Journal of the American Board of Family Medicine. 2017;30(6):784–794. doi: 10.3122/jabfm.2017.06.170064 [DOI] [PubMed] [Google Scholar]
- 51.Feldman DE, Carlesso LC, Nahin RL. Management of Patients with a Musculoskeletal Pain Condition that is Likely Chronic: Results from a National Cross Sectional Survey. J Pain. 2020;21(7–8):869–880. doi: 10.1016/j.jpain.2019.11.014 [DOI] [PubMed] [Google Scholar]
- 52.Krashin D, Sullivan M, Ballantyne J. What are we treating with chronic opioid therapy? Curr Rheumatol Rep. 2013;15(3):311. doi: 10.1007/s11926-012-0311-1 [DOI] [PubMed] [Google Scholar]
- 53.Pergolizzi JV, Varrassi G, Paladini A, LeQuang J. Stopping or Decreasing Opioid Therapy in Patients on Chronic Opioid Therapy. Pain Ther. 2019;8(2):163–176. doi: 10.1007/s40122-019-00135-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54.Dowell D, Ragan KR, Jones CM, Baldwin GT, Chou R. Prescribing Opioids for Pain - The New CDC Clinical Practice Guideline. N Engl J Med. 2022;387(22):2011–2013. doi: 10.1056/NEJMp2211040 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55.Sandbrink F, Murphy JL, Johansson M, et al. The Use of Opioids in the Management of Chronic Pain: Synopsis of the 2022 Updated U.S. Department of Veterans Affairs and U.S. Department of Defense Clinical Practice Guideline. Ann Intern Med. 2023;176(3):388–397. doi: 10.7326/m22-2917 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.