Pain Medicine 2020 (Dec 25); 21 (12): 3567–3573 ~ FULL TEXT
James M. Whedon, DC, MS, Andrew W. J. Toler, MS, Louis A. Kazal, MD, Serena Bezdjian, PhD, Justin M. Goehl, DC, MS, and Jay Greenstein, DC
Southern California University of Health Sciences,
OBJECTIVE: Utilization of nonpharmacological pain management may prevent unnecessary use of opioids. Our objective was to evaluate the impact of chiropractic utilization upon use of prescription opioids among patients with spinal pain.
DESIGN AND SETTING: We employed a retrospective cohort design for analysis of health claims data from three contiguous states for the years 2012-2017.
SUBJECTS: We included adults aged 18-84 years enrolled in a health plan and with office visits to a primary care physician or chiropractor for spinal pain. We identified two cohorts of subjects: Recipients received both primary care and chiropractic care, and nonrecipients received primary care but not chiropractic care.
METHODS: We performed adjusted time-to-event analyses to compare recipients and nonrecipients with regard to the risk of filling an opioid prescription. We stratified the recipient populations as: acute (first chiropractic encounter within 30 days of diagnosis) and nonacute (all other patients).
RESULTS: The total number of subjects was 101,221. Overall, between 1.55 and 2.03 times more nonrecipients filled an opioid prescription, as compared with recipients (in Connecticut: hazard ratio [HR] = 1.55, 95% confidence interval [CI] = 1.11-2.17, P = 0.010; in New Hampshire: HR = 2.03, 95% CI = 1.92-2.14, P < 0.0001). Similar differences were observed for the acute groups.
CONCLUSIONS: Patients with spinal pain who saw a chiropractor had half the risk of filling an opioid prescription. Among those who saw a chiropractor within 30 days of diagnosis, the reduction in risk was greater as compared with those with their first visit after the acute phase.
KEYWORDS: Spine Pain; Back pain; Neck Pain; Chiropractic; Analgesics; Opioid
From the FULL TEXT Article:
Opioid Prescription Crisis
The risks associated with the use of prescription opioid
analgesics are well known. Drug overdose deaths in
2017 increased by almost 10% over 2016, with opioids
accounting for almost 48,000 cases.  Promising improvement
has been made in curtailing excessive and
inappropriate prescribing practices and in the acute
treatment of opioid use disorder.  However, to prevent
unnecessary use of opioids, there is a pressing
need to identify safe and cost-effective alternatives for
the treatment of pain. Increasing attention is being paid
to the potential of nonpharmacological pain treatment
as an upstream strategy for addressing the opioid
Chiropractic: A Nonpharmacological Alternative to Opioids
The Institute of Medicine has recommended the use of
nonpharmacological therapies as effective alternatives to
pharmacotherapy for patients with chronic musculoskeletal
pain.  Among US adults prescribed opioids, 59%
reported having back pain.  Recently published clinical
guidelines from the American College of Physicians
recommend nonpharmacological treatment as the firstline
approach to treating back pain, with consideration
of opioids only as the last treatment option or if other
options present substantial risk to the patient.  In general,
it is not known if the use of nonpharmacological
therapies for musculoskeletal pain influences opioid use , but in that regard there is some emerging evidence in
favor of chiropractic care.
Doctors of Chiropractic (DCs) provide nonpharmacologic
treatment for spinal pain, using a multimodal functional
approach that typically includes spinal
manipulation and exercise, in accordance with international
clinical practice guidelines.  A recent study of
the implementation of a conservative spine care pathway
reported that among patients treated for spinal pain, as
expenditures for manual care (chiropractic and physical
therapy) increased, expenditures for opioid therapy decreased,
as did costs for spinal surgery and spine care
A retrospective claims study of 165,569
adults diagnosed with low back pain found that utilization
of services delivered by DCs was associated with reduced
use of opioids.  The supply of DCs, as well as
spending on spinal manipulative therapy, is inversely correlated
with opioid prescriptions in disabled Medicare
beneficiaries under age 65 . In a cohort of 1,702
patients with a new episode of neck pain, those who received
DC care had decreased odds of being prescribed
opioids within one year of their first visit (odds ratio
[OR] = 0.54, 95% confidence interval [CI] = 0.39–0.76).  In a 2018 report of a study conducted among
14,025 veterans of recent wars, the percentage of patients
receiving opioid prescriptions was lower after receiving
DC care for low back pain as compared with before. 
Also, in 2018, a study of health insurance claims for the
treatment of low back pain in New Hampshire found a
55% reduction in the likelihood of opioid prescription
fill among recipients of chiropractic care as opposed to
In 2019, Kazis et al.  reported
that patients who were initially treated by a chiropractor,
acupuncturist, or physical therapist had decreased odds
of both short- and long-term opioid use as compared
with initial treatment by a primary care physician. Also
in 2019, a systematic review of six studies examining the
association between chiropractic use and opioid use
found that opioid prescriptions were significantly lower
among chiropractic users as compared with nonusers. 
Reduced use of opioids among recipients of chiropractic
services may in turn lead to lower costs and improved
safety. The objective of this investigation was to build
upon our research in New Hampshire with a larger study
population, a longer time frame, and more advanced
methods, evaluating data from three New England states
for the impact of chiropractic utilization upon use of prescription
opioids among patients with spinal pain.
We hypothesized that among patients diagnosed with
spinal pain, recipients of chiropractic services have a
lower risk of filling a prescription for an opioid analgesic
as compared with nonrecipients. To test this hypothesis,
we employed a retrospective cohort design to analyze
health insurance claims data. Our data sources were the
All Payer Claims Databases (APCDs) of the three contiguous
states of Connecticut (CT), Massachusetts (MA),
and New Hampshire (NH), which aggregate health
claims data from third-party payers. This project was
conducted subject to the terms of data user agreements
between the principal investigator and the states. In accordance
with standard rules for analysis of health
claims, cells with N<11 were suppressed to prevent
against disclosure of protected health information. The
research methods were reviewed and approved by the institutional
review board of the investigator’s university.
All statistical analyses were performed using SAS (SAS
Institute, Cary, NC, USA).
Study Population and Cohorts
Data extracted from the three APCDs were not merged
together, but analyzed separately, in accordance with the
terms of the data use agreements. We accessed health
claims data for the years 2012–2017. We only included
claims with payment amounts greater than zero. The
study population was comprised of adults aged 18–84 years, enrolled in a health plan, and with office visits
to a primary care physician and/or DC. We included only
patients with continuous pharmacy coverage and at least
two visits associated with a primary diagnosis of a spinal
pain disorder between seven and 90 days apart. Thus, the
study population included subjects with multiple office
visits for spinal pain. Patients diagnosed with cancer at
any time during the study period were excluded.
those included in the study population as defined above,
we identified two cohorts of subjects:
1) Recipients of
chiropractic services (recipients) received both primary
care and chiropractic care at any point in the study period.
2) Nonrecipients received primary care but did not
receive chiropractic care at any time during the study period.
For each subject, the first date associated with diagnosis
of a spinal pain disorder was designated as the
index date. We excluded all subjects with an opioid
prescription fill that occurred before the index date. We
accounted for immortal time bias by using first chiropractic
visit only as a cohort inclusion criterion. Patients
with an opioid prescription after their index date but before
their first chiropractic visit remained in the analysis.
We stratified the recipient population into two groups: 1)
acute—patients whose first chiropractic encounter occurred
within 30 days of the index date, and 2) nonacute — all other patients.
Following establishment of the cohorts, we modeled for
likelihood of opioid prescription fill for up to six years of
follow-up time. We controlled for patient demographics
and health status at baseline through Charlson comorbidity
scoring. Comorbidity scores were calculated from diagnoses
documented during the one-year period
preceding the index date. We performed adjusted timeto-
event analyses, generating hazard ratios and Kaplan-
Meier graphs to compare recipients and nonrecipients
with regard to the risk of filling an opioid prescription
subsequent to being diagnosed with a spinal pain disorder.
To assess the effect of receiving chiropractic care
early in an episode of care, we subanalyzed for risk of
prescription fill in the acute and nonacute groups.
Subject characteristics may be viewed in Table 1. Total
subjects, including both recipients and nonrecipients
from all three states, numbered 101,221. The population
was skewed toward younger age categories among subjects
in NH and nonrecipients in MA. The skewed distribution
was expected because Medicare claims were not
included in the data sets. Recipients dominated the population
in NH, whereas the opposite was true among subjects
in MA. The number of subjects extracted from the
CT database was low, considering the size of the state
population (see the Discussion section for an explanation
of the small population size for CT). Charlson comorbidity
scores were higher among recipients in MA and nonrecipients
Results for the time-to-event analysis are displayed in
Table 2. Overall, in the three states at any particular time
in the study period, between 1.55 and 2.03 times more
nonrecipients filled an opioid prescription, as compared
with recipients (in Connecticut: hazard ratio [HR] =
1.55, 95% CI = 1.11–2.17, P = 0.010; in New
Hampshire: HR = 2.03, 95% CI = 1.92–2.14,
P<0.0001). In other words, chiropractic recipients are
at about half the risk of seeking an opioid over the sixyear
Significant differences were also observed for the
acute group in all three states and for the nonacute group
in NH. For the nonacute group in MA, no difference was
observed between recipients and nonrecipients (HR =
0.95, 95% CI = 0.71–1.26, P=0.71). For the nonacute
group in CT, low numbers required data suppression and
precluded statistical analysis. In both MA and NH, the
protective effect against risk of prescription fill was
higher for the acute group as compared with the nonacute
group (in MA: HR = 1.68 acute vs HR = 0.95 nonacute;
in NH: HR = 1.92 acute vs HR = 1.29 nonacute)
The Kaplan-Meier charts in Figures 1–3 illustrate the
results of the time-to-event analysis. All six charts show a
consistent reduction in risk over time for filling a prescription
for opioids. The charts for CT illustrate the
stepwise pattern commonly seen in survival analyses with
smaller numbers of subjects. Also reflecting the relatively
low sample size in CT, the two curves are closer together
and P values are higher. The charts for MA and NH,
which are based in analyses of larger study populations,
show smoother time-to-event curves, greater separation
between cohorts, and more narrow confidence intervals.
For all three states, differences in risk for the acute group
as compared with the overall analysis were negligible.
A previously published analysis of NH data on patients
with low back pain represented a snapshot in time of the
risk of taking prescription opioids after chiropractic or
nonchiropractic treatment.  The time-to-event analysis
described here yields a more accurate depiction of
comparative risk over time: It portrays the temporal relationships
between the spine disorder index event and the
first opioid fill for two modes of treatment.
The Kaplan-Meier graphs (Figures 1–3) show a remarkably
consistent pattern over six years of follow-up
for three US states, illustrating a strong protective effect
of chiropractic care against risk of opioid prescription
fill. The effect was stronger for patients who saw a chiropractor
within the first 30 days of diagnosis of a spinal
pain disorder, as compared with those who first saw a
chiropractor in later, nonacute phases. Additionally, the
protective effect of chiropractic care was sustained beyond
1,200 days after the index date in Massachusetts
and beyond 1,500 days in the other states, indicating that
once chiropractic treatment has been engaged in the
acute phase, patients experience a lasting benefit that is
measurable in years.
The pattern of repetitive encounters typical of a course
of chiropractic care may account for the sustained benefit.
The most recent chronic low back pain guidelines
suggest two to three patient encounters per week for two
to four weeks as a trial of chiropractic care. 
Multiple opportunities for patient–doctor interaction
may allow the chiropractor to review clinical progress,
advise on home exercise, ergonomics, and other selfmanagement
strategies, and provide reassurance, all of
which may help improve outcomes and reduce the need
These findings echo previous reports of superior outcomes
associated with early chiropractic intervention. [7, 16, 17] Although the case mix, subject age, sex, and
comorbidity scores varied among the three states, the adjusted
negative correlation between chiropractic care and
opioid therapy remained strong across all models. It may
prove instructive in future research to examine
differences in medication use among patients who selfrefer
for chiropractic care compared with those who are
referred by a medical physician.
Heyward et al.  recently reported inconsistent coverage
for nonpharmacologic therapies for spine pain and
little integration in coverage between drug and nondrug
approaches. More consistent is the accumulating evidence
for increased utilization of chiropractic services as
an upstream strategy for reducing dependence upon prescription
opioid medications. In the context of the opioid
crisis, the imperative for patient protection mandates
aggressive pursuit of all promising strategies, including
the elimination of barriers to access chiropractic services,
and benefits designs intended to encourage early utilization
of chiropractic services.
This study was subject to certain limitations. Across
the three APCDs, we encountered variation in case mix
that is reflected in differences by state in the proportion
of recipients and nonrecipients. The variation may be attributable
to population differences in health care utilization
and insurance coverage or to differences by state
administration in data collection, reporting, and release.
Because the CT data set did not include provider
specialty codes, we cross-walked national provider identifiers
(NPIs) to taxonomy codes and cross-referenced
those codes to provider specialty codes. This linkage
resulted in significant data loss that is reflected in much
smaller cohort sizes for CT.
In designing this study, we identified four main possible
sources of bias. First, among recipients, there was potential
for misclassification of the time-to-event analysis
through immortal time bias. We eliminated this possibility
by identifying recipients independently of the timing
of their opioid prescriptions. Thus the time interval between
index date and opioid fill was similar for all subjects,
and measurement of time to event was consistent
across cohorts. Second, due to the observational study
design, there was potential for confounding by indication,
whereby patients with more severe pain may have
been more likely to use opioid medication. Diagnoses in
claims data are unreliable indicators of pain severity.
took steps to account for confounding by indication by
excluding cancer patients and by adjusting for Charlson
comorbidity score. Additionally, this study was subject
to chronological bias, which was apparent in the acute vs
nonacute subgroup analysis. Subjects in the nonacute
groups were fewer in number, and the effect sizes were
smaller. These differences may be attributable to differences
in patient behavior or insurance benefits. Finally,
the study may be subject to selection bias if recipients differed
from nonrecipients by unmeasured characteristics.
Despite these potential sources of bias and confounding,
however, the results were remarkably similar across all
Among patients with spinal pain disorders, for recipients
of chiropractic care, the risk of filling a prescription for
an opioid analgesic over a six-year period was reduced
by half, as compared with nonrecipients. Among those
who saw a chiropractor within 30 days of being diagnosed
with a spinal pain disorder, the reduction in risk
was greater as compared with those who visited a chiropractor
after the acute phase had passed.
Data for this research were supplied by the Center for
Health Information and Analysis (Boston, MA, USA),
the Connecticut Health Insurance Exchange, the New
Hampshire Insurance Department, and the NH
Department of Health and Human Services.
US Department of Health and Human Services.
Facing Addiction in America: The Surgeon General’s Spotlight on Opioids
Washington, DC: Office of the Surgeon General,
US Department of Health and Human Services; 2018.
Institute of Medicine (IOM)
Relieving Pain in America: A Blueprint for Transforming Prevention,
Care, Education, and Research
Washington, DC: The National Academies Press, 2011.
Hudson TJ, Edlund MJ, Steffick DE, Tripathi SP, Sullivan MD.
Epidemiology of regular prescribed opioid use: Results from a national, population-based survey.
J Pain Symptom Manage 2008;36(3):280–8.
Qaseem A, Wilt TJ, McLean RM, Forciea MA;
Noninvasive Treatments for Acute, Subacute, and Chronic Low Back Pain:
A Clinical Practice Guideline From the American College of Physicians
Annals of Internal Medicine 2017 (Apr 4); 166 (7): 514–530
The Role of Nonpharmacological Approaches to Pain Management:
Proceedings of a Workshop
National Academies of Sciences, Engineering, and Medicine
The National Academies Press, Washington, DC (2019)
Foster NE, Anema JR, Cherkin D, Chou R, Cohen SP, Gross DP, Ferreira PH, Fritz JM.
Prevention and Treatment of Low Back Pain: Evidence, Challenges, and Promising Directions
Lancet. 2018 (Jun 9); 391 (10137): 2368–2383
Weeks WB, Pike J, Donath J, Fiacco P, Justice BD.
Conservative Spine Care Pathway Implementation is Associated with Reduced
Health Care Expenditures in a Controlled, Before-After Observational Study
J General Internal Medicine 2019 (Aug); 34 (8): 1381–1382
Rhee Y, Taitel MS, Walker DR, Lau DT.
Narcotic drug use among patients with lower back pain in employer health plans:
A retrospective analysis of risk factors and health care services.
Clin Ther 2007;29(Suppl):2603–12.
Weeks WB, Goertz CM.
Cross-Sectional Analysis of Per Capita Supply of Doctors of Chiropractic and Opioid Use
in Younger Medicare Beneficiaries
J Manipulative Physiol Ther 2016 (May); 39 (4): 263–266
Horn ME, George SZ, Fritz JM.
Influence of Initial Provider on Health Care Utilization in Patients Seeking Care for Neck Pain
Mayo Clin Proc Innov Qual Outcomes. 2017 (Oct 19); 1 (3): 226–233
Lisi AJ, Corcoran KL, DeRycke EC, et al.
Opioid Use Among Veterans of Recent Wars Receiving
Veterans Affairs Chiropractic Care
Pain Med. 2018 (Sep 1); 19 (suppl_1): S54–S60
Whedon JM, Toler AWJ, Goehl JM, Kazal LA.
Association Between Utilization of Chiropractic Services for Treatment of Low-Back Pain
and Use of Prescription Opioids
J Altern Complement Med. 2018 (Jun); 24 (6): 552–556
Kazis LE, Ameli O, Rothendler J, et al.
Observational Retrospective Study of the Association of Initial Healthcare Provider for New-onset
Low Back Pain with Early and Long-term Opioid Use
BMJ Open. 2019 (Sep 20); 9 (9): e028633
Corcoran KL, Bastian LA, Gunderson CG, Steffens C, Brackett A, Lisi AJ.
Association Between Chiropractic Use and Opioid Receipt Among Patients with Spinal Pain:
A Systematic Review and Meta-analysis
Pain Medicine 2020 (Feb 1); 21 (2): e139–e145
Globe, G, Farabaugh, RJ, Hawk, C et al.
Clinical Practice Guideline: Chiropractic Care for Low Back Pain
J Manipulative Physiol Ther. 2016 (Jan); 39 (1): 1–22
Paskowski I, Schneider M, Stevens J, Ventura JM, Justice BD.
A Hospital-Based Standardized Spine Care Pathway: Report of a Multidisciplinary, Evidence-Based Process
J Manipulative Physiol Ther 2011 (Feb); 34 (2): 98–106
Keeney BJ, Fulton-Kehoe D, Turner JA, et al.
Early predictors of lumbar spine surgery after occupational back injury:
Results from a prospective study of workers in Washington State.
Heyward J , Jones CM , Compton WM , et al .
Coverage of Nonpharmacologic Treatments for Low Back Pain Among US Public and Private Insurer
JAMA Network Open 2018 (Oct 5); 1 (6): e183044
Return to SPINAL PAIN MANAGEMENT