Spine (Phila Pa 1976). 2006 (Mar 15); 31 (6): 611–621 ~ FULL TEXT
Eric L. Hurwitz, DC, PhD, Hal Morgenstern, PhD, Gerald F. Kominski, PhD,
Fei Yu, PhD, and Lu-May Chiang, MS
Department of Epidemiology,
UCLA School of Public Health,
Los Angeles, CA, USA.
STUDY DESIGN: Randomized clinical trial.
OBJECTIVES: To compare the long-term effectiveness of medical and chiropractic care for low back pain in managed care and to assess the effectiveness of physical therapy and modalities among patients receiving medical or chiropractic care.
SUMMARY OF BACKGROUND DATA: Evidence comparing the long-term relative effectiveness of common treatment strategies offered to low back pain patients in managed care is lacking.
METHODS: A total of 681 low back pain patients presenting to a managed-care facility were randomized to chiropractic with or without physical modalities, or medical care with or without physical therapy, and followed for 18 months. The primary outcome variables are low back pain intensity, disability, and complete remission. The secondary outcome is participants' perception of improvement in low back symptoms.
RESULTS: Of the 681 patients, 610 (89.6%) were followed through 18 months. Among participants not assigned to receive physical therapy or modalities, the estimated improvements in pain and disability and 18-month risk of complete remission were a little greater in the chiropractic group than in the medical group (adjusted RR of remission = 1.29; 95% CI = 0.80-2.07). Among participants assigned to medical care, mean changes in pain and disability and risk of remission were larger in patients assigned to receive physical therapy (adjusted RR = 1.69; 95% CI = 1.08-2.66). Among those assigned to chiropractic care, however, assignment to methods was not associated with improvement or remission (adjusted RR = 0.98; 95% CI = 0.62-1.55). Compared with medical care only patients, chiropractic and physical therapy patients were much more likely to perceive improvement in their low back symptoms. However, less than 20% of all patients were pain-free at 18 months.
CONCLUSIONS: Differences in outcomes between medical and chiropractic care without physical therapy or modalities are not clinically meaningful, although chiropractic may result in a greater likelihood of perceived improvement, perhaps reflecting satisfaction or lack of blinding. Physical therapy may be more effective than medical care alone for some patients, while physical modalities appear to have no benefit in chiropractic care.
Key words: low back pain, randomized trial, chiropractic, physical therapy, managed care.
From the Full-Text Article:
Low back pain is one of the most common reasons for
visiting a medical physician  or physical therapist  and is
the leading complaint among persons accessing chiropractic
care in the United States.  Indeed, chiropractic is
the most frequently used alternative treatment for back
problems  and is the most commonly used unconventional
therapy in the United States.  Chiropractic care is
now included in several managed-care plans, and the
trend of managed-care organizations serving an increasing
proportion of the population has continued up to the
present. [6, 7]
Despite the high economic costs, disability, and lost
productivity associated with low back pain in the United
States and Europe, [8–11] and the development of several
treatment guidelines, [12, 13] management of low back pain
remains controversial and highly variable across professions
and geographic regions.  Although several recent
studies have compared chiropractic care with medical
care and/or physical therapy for patients with low back
pain, [15–22] ours is the only randomized clinical trial
(RCT) to have been conducted entirely within a managed-care practice setting.  We previously reported that
in the short-term (6 months), chiropractic and medical
care for low back pain are comparable in their clinical
effectiveness, but physical therapy may be marginally
more effective than medical care alone for reducing disability
in some patients. 
The purpose of this paper is to extend the analysis to
participants followed through 18 months and to contrast
and compare effect estimates from the primary (clinical)
and secondary (perceived effectiveness) outcome measures.
The three objectives are to compare the long-term
clinical and perceived effectiveness of:
1) medical and chiropractic care for low back pain patients in managed care,
2) physical therapy among medical patients, and
3) physical modalities among chiropractic patients.
Materials and Methods
Study Design and Source Population.
Low back pain patients presenting at 3 primary-care centers of a 100,000-member health-care network based in southern California were
randomized in a balanced design to 4 treatment groups:
chiropractic care with and without physical modalities, and
medical care with and without physical therapy.
were mailed to participants at 2 and 6 weeks and at 6,
12, and 18 months. All study protocols and forms were approved
by the institutional review boards (IRB) from UCLA
and the healthcare network. Key design and protocol details
are given below; additional methodologic information has been
previously reported. 
Patients were eligible for the clinical trial
1) were health maintenance organization members with the network chosen as their healthcare provider;
2) sought care at one of the 3 study sites from October 30, 1995 through November 9, 1998;
3) presented with pain in the region of the lumbosacral spine and its surrounding musculature;
4) had not received professional care for low back pain within the previous 1 month; and
5) were at least 18 years of age.
Patients were excluded if they had pain due to fracture, tumor,
infection, spondyloarthropathy, or other nonmechanical cause;
had severe comorbidity; had a pacemaker or other electrical device;
had a blood coagulation disorder or were on corticosteroids
or anticoagulants; had progressive lower limb muscle weakness;
had current signs or symptoms of cauda equina syndrome; had
plans to move out of the area; could not be contacted by telephone;
lacked the ability to read English; or had pending thirdparty
liability or Workers’ Compensation cases.
Recruitment and Enrollment.
All patients presenting with
symptoms consistent with low back pain were initially screened
by the field coordinator and asked if they would be willing to
participate in a study addressing the effectiveness of different
types of care for low back pain. Patients received a brochure
describing the study, its protocols, and the requirements of
participation. A primary-care provider conducted a detailed
history and physical examination on each eligible patient, and
ordered imaging studies or additional diagnostic tests if necessary
to rule out specific spinal pathology. Radiologic examinations
were ordered on 9.5% of enrolled patients.
Patients meeting the eligibility criteria and agreeing to participate
read and signed the IRB-approved informed consent
form, which was administered by the field coordinator and
witnessed by a third party.
Our statistician wrote a computer program to generate randomized
treatment assignments in blocks of 12, stratified by
site, and placed each assignment in a numbered security envelope.
Each of the 3 sites had its own series of sequentially
numbered sealed envelopes. On receiving consent, the field coordinator
opened the appropriate envelope, informed the participant
of the assignment, directed him or her to fill out the
baseline questionnaire, and scheduled the same-day appointment
with a medical or chiropractic provider.
Chiropractic Care Without Physical Modalities (DC Group)
Participants randomized to the DC group received spinal manipulation
or mobilization, instruction in strengthening and
flexibility exercises, and instruction in proper back care.
Chiropractic Care With Physical Modalities (DCPm Group)
addition to chiropractic care (above), participants randomized
to the DCPm group received heat or cold therapy, ultrasound,
and/or electrical muscle stimulation (EMS), at the chiropractor’s
Medical Care Without Physical Therapy (MD Group)
randomized to the MD group received one or more of
the following at the discretion of the medical provider: instruction
in proper back care and strengthening and flexibility
exercises; prescriptions for analgesics, muscle relaxants,
or anti-inflammatories; and lifestyle recommendations.
Medical Care With Physical Therapy (MDPt Group)
to medical care (above), participants randomized to the
MDPt group received instruction in proper back care and, at
the discretion of the physical therapist, one or more of the
following: heat or cold therapy, ultrasound, EMS, soft tissue
and joint mobilization, traction, supervised therapeutic exercise,
and strengthening and flexibility exercises. The interventions
most commonly used on study participants were heat or
cold therapy (71%), supervised therapeutic exercise (59.5%),
ultrasound (45%), EMS (33.6%), and joint mobilization
Data Sources and Collection
Baseline Pain and Disability Data.
The baseline questionnaire
included items on low back pain and related disability.
Intensity of most severe low back pain and average level of pain
for the past week were measured with 0 to 10 numerical rating
scales, where 0 = no pain and 10 = unbearable pain. These
scales have been shown to have excellent reliability and validity
for measuring back pain.  Current disability due to low back
pain was measured with the previously validated 24-item Roland-
Morris Low-Back Disability Questionnaire, [25–31] where 0
indicates no disability and 24 indicates severe disability due to
low back pain.
We also asked participants several questions about their
current and past episodes of low back pain, including the number
of previous episodes, length of current episode, frequency
of pain, disability days, and history of trauma.
Other Baseline Data.
Sociodemographic and other possible
prognostic factors were also queried at baseline. Sociodemographic
variables included age, sex, race/ethnicity, education,
household income, marital status, and current employment status.
General health status was measured with 5 of 8 subscales
of the previously validated Medical Outcomes Study 36-Item
Short-Form Health Survey (SF-36):
1) limitations in physical activities because of physical or emotional problems;
2) limitations in usual role activities because of physical health problems;
3) limitations in usual role activities because of emotional health problems;
4) general health perceptions; and
5) general mental health. [32, 33]
Follow-up questionnaires mailed to participants
at 2, 6, 26, 52, and 78 weeks after their baseline visits
addressed low back pain intensity, related disability, and perceived
improvement. Pain status was measured with repeat 0 to
10 pain scales, and low back-related disability was measured
with repeat Roland-Morris Low-Back Disability Questionnaires.
Perceived improvement was assessed with scales of perception
of change in low back symptoms during the preceding
follow-up interval, with the following response options: “a lot
worse,” “a little worse,” “about the same,” “a little better,” or
“a lot better.” At each follow-up assessment, we also asked
participants about the frequency of pain in the past week, number
of disability days, and use of over-the-counter (OTC) and
prescription medications to deal with their low back pain.
Four primary outcome variables were
used to evaluate the effectiveness of assigned treatment group:
average level of low back pain in the past week (0–10 scale),
most severe low back pain in the past week (0–10 scale), current
back-related disability score (0–24 Roland-Morris Questionnaire),
and complete remission, defined as the first observation
during follow-up in which the above outcome variables
were zero (i.e., no low back pain in the past week and no related
disability). Cutpoints of 2 or more points (vs.<2) on the 0 to
10 scales and 3 or more points (vs.<3) on the Roland-Morris
scale were used as dichotomous outcomes. [8, 29] Differences between
groups of 2+ and 3+ points on the pain and disability
scales, respectively, were considered clinically relevant.
The secondary outcome was perceived improvement as
measured by patients’ self-reports of their perceptions of
change in low back pain symptoms during the preceding follow-
up interval. Because of the small number of respondents
reporting worsening of their symptoms, the response options
“a lot worse” and “a little worse” were collapsed into one
category; “a lot better” and “a little better” were left intact.
To estimate treatment effects, we used intention-
to-treat analyses and made 3 contrasts between treatment
groups: DC versus MD, MDPt versus MD, and DCPm versus
DC. Two statistical methods were used to estimate these 3
effects on the primary outcomes: The first involved treating
each outcome variable as a continuous variable, whereas the
second involved combining all 3 outcomes into a composite
dichotomous measure, called complete remission. A third statistical
method, polytomous logistic regression using a Markov
(transition) model, was used to estimate treatment effects on
perception of improvement. SAS 8.1 was used for data management
and analysis. 
In the first method, applied separately to each outcome at
all 6 observations, a mixed-effects linear model was fit to the
data. [35, 36] To allow for heteroscedasticity in the variancecovariance
matrix, we treated outcome values between individual
subjects as independent, while outcome values within each
subject were treated as correlated with a common unspecified
variance-covariance structure. The other covariates in the
model were follow-up time, age, sex, and 3 characteristics of
the subject’s low back pain at baseline: duration of the current
episode, number of previous episodes, and frequency of pain in
the past week. Since the effects of treatment group were represented
by the outcome changes (slopes) over follow-up time
and it is unrealistic to assume linear treatment effects (constant
slopes) for the entire 18-month follow-up period, linear splines
for treatment effects (piece-wise linear slopes) at each follow-up
time were used to allow outcome changes (slopes) to differ
across follow-up intervals within each treatment group and to
differ across treatment groups within each follow-up interval.
To measure treatment effects for each outcome, we used the
fitted model to estimate treatment-group differences (and 95%
confidence intervals [CIs]) in mean outcome change between
baseline and 6 months, baseline and 12 months, and baseline
and 18 months. In addition, the adjusted means in each outcome
by treatment group and follow-up time were graphed for
the entire 18-month follow-up period. To reflect the adjusted
mean in the population, each covariate was set equal to its
mean value in the total sample.
In the second method, we observed the time to first complete
remission after baseline. First complete remission for a given
subject was defined as the first observation during follow-up in
which all 3 outcome variables were zero (i.e., no low back pain
in the past week and no related disability). Unconditional logistic
regression was used to compare the 18-month probability
(risk) of complete remission between treatment groups. Results
of the fitted model were used to estimate risk ratios (and 95%
CIs) for the same treatment-group contrasts described above.
In addition to treatment group, other covariates in this model
were age, sex, the 3 low back pain variables described above,
and the baseline value of each outcome variable from which
composite measure was derived. To estimate risk ratios for the
treatment contrasts, each covariate was set equal to its mean
value in the total sample. Because this method excluded 10% of
all subjects lost to follow-up, we also used proportional hazards
(survival) analysis to compare the remission rates between
treatment groups. Since the results of these 2 methods were
similar, we report here only the logistic regression results.
In the third method, we fit a first-order Markov transition
model to estimate treatment effects on patient’s perception of
low back pain improvement, recorded as “a lot better,” “a little
better,” “about the same,” and “worse,” allowing for multiple
categories by relating transition probabilities to covariates and
previous outcomes through a polytomous logistic regression
model with Markov structure.  In addition to treatment group
and past outcomes, the other covariates were age, sex, duration
of current episode at baseline, and 4 indicator variables reflecting
follow-up times with 2 weeks as the reference category. All
models used “about the same” as the reference response level.
Estimated treatment effects are presented as odds ratios (ORs)
with 95% CIs. The PHREG and GENMOD procedures in SAS
were used for fitting the polytomous logistic regression models.  Additional details about the application of these models
may be found elsewhere. 
Recruitment, Enrollment, and Follow-up
Figure 1 shows the flow of patients from initial screening
through end of follow-up. A total of 886 (37.6%) of the
2,355 screened patients were excluded for the following
reasons (number in parentheses):
recent treatment for low back pain (270),
lack of health maintenance organization insurance coverage (199),
symptoms inconsistent with low back pain (144),
third party liability or Workers’ Compensation case (55),
inability to read English (46),
less than 18 years old (43),
plans to move out of the area (18),
and inaccessible by telephone (4).
Patients were also excluded for the following medical reasons:
low back pain due to fracture, tumor, or infection, ankylosing spondylitis, or other rheumatic disease (47),
severe comorbidity (37),
use of anticoagulants (13),
treatment with electrical device (5),
progressive lower limb muscle weakness (2),
abdominal aortic aneurysm (1),
blood coagulation disorder (1), and
signs or symptoms of possible cauda equina syndrome (1).
Of the 1,469 eligible patients, 779 (53.0%) declined participation because of:
lack of interest (345),
preference for medical care (116),
preference for chiropractic care (105),
desire for multiple treatment options (45), and
inability to afford copayments (31).
Because of the apparent lack of understanding of
the informed consent form, 9 additional otherwise eligible
and willing prospective participants were not enrolled.
We therefore enrolled 681 patients for an acceptance
rate [number of enrollees/number of eligible patients]
Two- and 6-week follow-up questionnaires were
completed by 679 (99.7%) and 675 (99.1%) participants,
respectively. Questionnaires at 6, 12, and 18
months were completed by 652 (95.7%), 629 (92.4%),
and 610 (89.6%) participants, respectively.
Characteristics of Study Population
Table 1 shows the baseline distributions of sociodemographic
and health-status variables by treatment group.
Fifty-two percent of the participants are female, half are
under age 50, 40% are nonwhite, and two thirds are
currently employed. The SF-36 scores are lower than the
U.S. general population norms,  although compatible
with other back pain populations. [18, 40, 41]
Table 2 shows
the baseline distributions of low back-related variables.
Almost half of the participants had been in pain for
longer than 1 year, and more than 80% reported prior
episodes of low back pain. The median low back disability
score of 11 reflects moderate disability, whereas the
pain intensity scores indicate appreciable levels of pain
perception. There are no clinically meaningful differences
between treatment groups in the distributions of
possible prognostic factors measured at baseline.
Fifty-five percent of all low back-related patient visits
took place within the first 6 weeks following randomization;
85% occurred within 6 months. In the initial 6
months, patients assigned to one of the chiropractic
groups or the physical therapy group had, on average,
almost twice as many visits (5.4 vs. 2.9) as patients assigned
to medical care alone. Average visit duration was
twice as long for physical therapy patients as for medical
and chiropractic patients (31 vs. 15 minutes). Visit frequencies
after 6 months did not appreciably differ between
groups or providers. There were a total of 3 low
back surgeries (one in each of the medical groups and
one in the chiropractic care without physical modalities
group), all performed between 6 and 18 months of
Table 3 shows the estimated effects from the mixed-effects
linear models (adjusted differences in mean changes and
95% CIs) of chiropractic care only versus medical care
only, medical care with physical therapy versus medical
care only, and chiropractic care with physical modalities
versus chiropractic care only on pain intensity and disability
from baseline to 6, 12, and 18 months.
Figures 2 to 4 show the mean levels of pain and disability by follow-up assessment and treatment group. Although medical
patients assigned to physical therapy had greater
reductions in pain and disability than did medical patients
not assigned to physical therapy, the differences are
of marginal clinical significance at best. Using our a priori
definitions of clinically relevant differences in pain
and disability, none of the estimated effects would be
considered clinically meaningful, although the estimated
effects of DC versus MD and MDPt versus MD were
relatively stronger after 18 months than after 6 months.
Table 4 shows the estimated adjusted 18-month risk
ratio of clinical remission, by treatment-group contrast,
from the logistic regression analysis. Twenty-three percent
of all participants had complete remissions. Medical
patients assigned to physical therapy were more likely
than medical patients not assigned to physical therapy
to remit during the 18-month follow-up period (RR = 1.69; 95% CI = 1.08–2.66). Chiropractic patients were
also somewhat more likely than medical patients to remit
(RR = 1.29; 95% CI = 0.80–2.07). Chiropractic with
physical modalities did not result in any apparent benefit
relative to chiropractic care without modalities (RR = 0.98; 95% CI = 0.62–1.55).
Table 5 shows the estimated adjusted treatment effects on
perceived change in low back pain, from the transition
polytomous logistic regression model. The odds of perceiving
their pain as a little or a lot better (vs. perceiving no
change) since the previous assessment were greater among
patients assigned to chiropractic care without modalities (a
lot better: OR = 1.83; 95% CI = 1.34–2.49; a little better:
OR = 1.39; 95% CI = 1.08–1.79) and medical patients
assigned to physical therapy (a lot better: OR = 1.67; 95%
CI = 1.22–2.28; a little better: OR = 1.45; 95% CI = 1.13–1.87) than among medical patients not assigned to
physical therapy. Chiropractic patients assigned to physical
modalities were more likely than those not assigned to modalities
to perceive their pain as worsening (vs. staying the
same) since the previous assessment (OR = 0.63; 95% CI = 0.45–0.91).
Frequency of Pain, Disability Days, and Use of Medications
After 18 months of follow-up, 25% of all participants
reported having low back pain most or all of the time in
the past week; less than 20% reported having no pain.
Participants in the medical care group were most likely to
have pain most or all of the time (30%) and least likely to
be pain-free (15%). The frequencies of disability days
(cut-down and bed days) due to low back pain did not
appreciably vary by treatment group. At each follow-up
assessment, about 40% of all participants reported at
least one cut-down day and 10% at least one bed day in
the past month. There were no known treatment-related
adverse events requiring IRB notification during the 18-
At each follow-up assessment, participants in the
medical care groups were more likely to report prescription-pain medication use than were participants in the
chiropractic groups (32% vs. 24% at 6 months; 29% vs.
20% at 12 months; 27% vs. 19% at 18 months). The
most frequently prescribed drugs were nonsteroidal antiinflammatory
drugs (NSAIDs), muscle relaxants, and
narcotic analgesics. Patients assigned to chiropractic care
were somewhat more likely than those assigned to medical
care to report OTC-pain medication use at 6 months
(56% vs. 49%) but not at 12 or 18 months (56% vs.
58% at 12 months; 53% vs. 51% at 18 months). Assignment
to physical modalities or physical therapy among
chiropractic and medical participants, respectively, did
not appear to influence reported use of prescription or
In a managed-care population comprised of mostly subacute
and chronic low back pain patients, we observed
relatively small differences in clinical outcomes between
4 popular low back pain management strategies during
18 months of follow-up. Larger differences were observed
for patients’ perceptions of improvement. Among
participants not assigned to receive physical therapy or
modalities, there was slightly more improvement in the
chiropractic group than in the medical group, and the
estimated effects were somewhat stronger after 18
months than after 6 months. Chiropractic patients were
much more likely than medical patients to perceive their
pain as being a little or a lot better during the 18 months
of follow-up, possibly reflecting, in part, satisfaction
with care.  Among participants assigned to medical
care, there was more remission and improvement in the
physical therapy group, with relatively stronger effects
observed after 18 months than after 6 months. These
patients were also more likely to perceive their pain as
being a little or a lot better during follow-up. As was also
found in the short-term among participants assigned to
chiropractic care,  no association was detected between
physical modalities and low back pain outcome after 18
months, and patients assigned to modalities were actually
more likely than patients not assigned to modalities
to perceive their pain as getting worse.
The management approaches in our trial each include
specific components, such as medications, exercises, spinal
manipulative therapy, and advice, many of which have
been the focus of recent systematic reviews. Published findings
show that acetaminophen,  NSAIDs, [44–48] muscle relaxants, [45, 47, 49] conditioning exercises and certain aerobic
activities,  and advice to stay active  may provide shortterm
benefit for patients with acute low back pain;
NSAIDs,  exercise therapy, [47, 48, 50, 52–55] massage, [56, 57] and
cognitive behavioral or multidisciplinary biopsychosocial
rehabilitation therapy [58, 59] may be of some benefit for patients
with chronic low back pain; and that current evidence
does not support bed rest [47, 60] or physical agents or modalities
for acute or chronic low back pain. [47, 61] There is little
or no support for the use of transcutaneous electrical
nerve stimulation,  traction,  and specific exercise regimens [50, 53];
and despite more than 20 RCTs of acupuncture
for acute and chronic back pain, evidence supporting
the effectiveness of its use is unclear.  Guidelines for
the treatment of low back pain of less than 3 months’
duration are consistent with the above findings, recommending
the use of acetaminophen, NSAIDs, and aerobic
activities, [12, 13] and recommending against bed rest,
physical agents, and transcutaneous electrical nerve
stimulation and other modalities. [12, 13]
Several RCTs and systematic reviews comparing chiropractic’s
core procedure, spinal manipulation, with
other interventions have reported favorable outcomes
relative to both sham and active treatments in acute [47, 64]
and chronic  low back pain, and guidelines support its
use. [12, 13] Nevertheless, more recent findings suggest that
spinal manipulation may not be more effective than conventional
treatments for acute or chronic low back
pain. [57, 65] In a meta-regression analysis of 39 RCTs, spinal
manipulation was found to be superior only when
compared with sham or therapies known to be ineffective,
such as traction or bed rest.  However, the investigators
of a systematic review and best evidence synthesis
of 46 RCTs of spinal manipulative therapy for low
back pain concluded that spinal manipulation is as effective
as NSAIDs and may result in marginally better clinical outcomes relative to general practitioner care in the
short-term and physical therapy in the long-term for patients
with chronic pain.  The literature also indicates
that analgesics and other medications given during the
course of standard medical care are much more likely to
be associated with adverse reactions than are comparator
treatments, [12, 45, 46, 49, 66] and may delay recovery, 
and that complications following spinal manipulation of
the lumbar spine are very rare. [64, 66, 68, 69]
In addition to our trial, chiropractic care or spinal
manipulative therapy has been compared with physical
therapy and/or medical care in a few recent, well-conducted
RCTs and observational studies. No clinically meaningful
treatment-group differences in outcomes were found
in the total study populations of the RCTs conducted by
Meade et al, [16, 17]and Skargren et al, [20, 21] although certain
subgroup differences were detected. Cherkin et al 
also reported no meaningful differences between chiropractic
and physical therapy outcomes, and their pain
and disability outcomes were only slightly better than
those observed in a group receiving only an educational
booklet. An additional RCT comparing subacute-pain
patients receiving 12 weeks of either osteopathic or routine
medical care found comparable clinical outcomes
but greater prescription-medication use in the medical
Three observational studies are worth noting. Pain and
disability outcomes were similar at 6 months among acute
and subacute low back pain patients of primary care physicians,
orthopedic surgeons, and chiropractors.  Acute
and chronic low back pain patients of chiropractors enrolled
in a large observational, practice-based study had
better pain and disability outcomes after 1 year than did
patients of medical physicians.  Among respondents of a
national survey in the United States, persons visiting chiropractors
for back or neck pain treatment in the past year
were more than twice as likely as those visiting conventional
providers to rate their care as “very helpful” (61%
vs. 27%).  However, given the lack of randomization and
evidence indicating that patients of chiropractors and medical
physicians differ with respect to health status and other
prognostic factors, [72, 73] findings from these observational
studies are not easily interpretable.
Regardless of treatment, a large proportion of persons
with low back pain do not become pain- and disability-free.
In our trial, only about 20% of participants had
absolutely no symptoms after 12 and 18 months, and
more than 25% reported having pain most or all of the
time. These figures are consistent with the findings from
a review of low back pain in general patient populations,
showing that up to 75% of patients still have pain 12
months later, and 60% of those who became pain-free
relapsed within 12 months.  Similar findings have been
reported in more recent long-term clinic-  and population-
based studies.  Given this natural history of often
persistent and recurring pain in a large segment of low
back pain sufferers, even after one or more episodes of
care, and the high prevalence of significant comorbidity, 
it is apparent that focusing on the low back may not be
the optimal strategy for long-term relief. Interventions
that address behavior modification aiming for long-term
benefits in general health and well-being, rather than
back-specific treatment, may be more effective in ameliorating
back pain and its many personal, social, and economic
consequences. For example, promoting exercise
and leisure-time physical activity may be an inexpensive,
safe, and efficacious approach to reduce not only low
back pain and associated work-loss days, [54, 77] but also psychologic
distress and pain-related fear and anxiety. [55, 78]
Consistent with findings from other recent RCTs showing
the benefits of supervised exercise for patients with subacute
and chronic low back pain, [79–81] the generally
more favorable outcomes among physical therapy participants
may be due to the therapists’ greater emphasis
on active care. 
Although randomization and an 18-month follow-up
rate of almost 90% likely preclude appreciable bias due
to confounding or selective follow-up loss, findings from
our clinical trial should be considered in light of the following
limitations. First, because of the nature of the
interventions, there could be no masking of participants
or providers. Estimates of treatment effects could be biased
by differential error in outcome measurement or by
differential behavior of providers or participants during
follow-up that might affect the outcome. The estimated
effects of treatment on perceived changes in symptoms
may be especially affected by lack of patient blinding.
Second, if the eligible patients who declined to participate
differed from participants on factors that modify
treatment effects, then our effect estimates may not accurately
reflect the results that would have been obtained
from the total source population. Although this does not
impact the study’s internal validity, generalizing the findings
may be problematic. Third, generalizing the findings
to patients in other settings or to those who receive their
health care through different reimbursement mechanisms
may not be appropriate. Healthcare practitioners
in other environments may manage their low back pain
patients in ways that affect treatment results, and fee-forservice
and other reimbursement systems (e.g.,Workers’
Compensation, personal injury) may involve incentives
that affect utilization and outcomes. Nevertheless, our
study population is clinically similar to other ambulatory
low back pain populations, [19, 25, 83–87] and we have no
evidence that nonparticipants were systematically different
from participants, or that treatment effects appreciably
vary by baseline episode duration or other clinical
Small differences in clinical outcomes were observed
among low back pain patients randomized to chiropractic
care with and without physical modalities, and medical
care with and without physical therapy during 18 months
of follow-up. The differences in pain and disability outcomes
between medical and chiropractic care without
physical therapy or modalities are not clinically meaningful,
although patients of chiropractors are more likely
to perceive improvement. Changes in reported pain and
disability scores may reflect clinical outcomes, whereas
perceived improvement may also reflect satisfaction with
care or lack of blinding. Physical therapy may be more
effective than medical care alone for some patients,
whereas physical modalities used by chiropractors appear
to have no effect. Regardless of treatment, a majority
of each group had some degree of low back pain at the
end of follow-up. Other promising multi-intervention
strategies, such as coupling manual and behavioral therapies
with the promotion of physical activity, may more
favorably influence long-term prognosis and should be
the focus of subsequent research.
A randomized clinical trial was conducted among low back pain patients to compare the long-term effectiveness of medical and chiropractic care for low back pain and to assess the effectiveness of physical therapy and modalities in a managed-care practice setting.
A total of 681 patients presenting to a managedcare facility were randomized to medical care with and without physical therapy, and chiropractic care with and without physical modalities; 610 (89.6%) were followed up through 18 months.
The differences in pain and disability outcomes between medical and chiropractic care without physical therapy or modalities are not clinically meaningful, although patients of chiropractors are more likely to perceive improvement.
Physical therapy may be more effective than medical care alone for some patients, whereas physical modalities used by chiropractors appear to have no effect. Regardless of treatment, a majority of each group had some degree of low back pain at the end of follow-up.
The authors thank Friendly Hills HealthCare Network’ s
executive officers for their initial interest, support, and commitment
of resources; Gary Pirnat, former chairman of chiropractic
services and manager of rehabilitation services,
and the medical, chiropractic, and physical therapy providers
for their cooperation during the study; and the clinic
management teams from La Habra, Brea, and Buena Park
and their front-office and health information personnel for
their help with patient enrollment. The authors also thank
UCLA personnel Karen Hemmerling and Stan Ewald for
coordinating patient recruitment and follow-up efforts,
Emerlinda Gonzalez and Silvia Sanz for assistance with patient
enrollment and tracking, and He-Jing Wang for providing
data management services at UCLA.
Hing E, Middleton K.
National Hospital Ambulatory Medical Care Survey: 2002 outpatient department summary.
Adv Data Vital Health Stat 2004; 345.
Feine JS, Lund JP.
An assessment of the efficacy of physical therapy and physical modalities for the control of chronic musculoskeletal pain.
Barnes PM, Powell-Griner E, McFann K, et al.
Complementary and alternative medicine use among adults: United States, 2002.
Adv Data Vital Health Stat 2004;343.
Eisenberg DM, Davis RB, Ettner SL, et al.
Trends in Alternative Medicine Use in the United States, from 1990 to 1997:
Results of a Follow-up National Survey
JAMA 1998 (Nov 11); 280 (18): 1569–1575
Druss BG, Rosenheck RA.
Association between use of unconventional therapies and conventional medical services.
Yelin E, Herrndorf A, Trupin L, et al.
A national study of medical care expenditures for musculoskeletal conditions: the impact of health insurance
and managed care.
Arthritis Rheum 2001;44:1160–1169.
Cherry DK, Burt CW, Woodwell DA.
National Ambulatory Medical Care Survey 2001 summary.
Adv Data Vital Health Stat 2003;337.
Hashemi L, Webster BS, Clancy EA.
Trends in disability duration and cost of workers’ compensation low back pain claims (1988–1996).
J Occup Environ Med 1998;40:1110–9.
Murphy PL, Volinn E.
Is occupational low back pain on the rise?
Low back pain: a twentieth century health care enigma.
Pai S, Sundarum LJ.
Low back pain: an economic assessment in the United States.
Orthop Clin North Am 2004;35:1–5.
Bigos S, Bowyer O, Braen G, et al.
Acute Lower Back Problems in Adults
Clinical Practice Guideline No. 14 [AHCPR Publication No. 95-0642].
Rockville, MD: Agency for Health Care Policy and Research,
Public Health Service, U.S. Department of Health and Human Services, 1994.
Waddell G, McIntosh A, Hutchinson A, et al.
Clinical Guidelines for the Management of
Acute Low Back Pain
London: Royal College of General Practitioners, 1999.
The Back Pain Revolution.
Edinburgh: Churchill Livingston, 1998.
The relative impact of chiropractic vs. medical management of low-back pain on health status in a multispecialty group practice.
J Manipulative Physiol Ther 1994;17:74–82.
Meade TW, Dyer S, Browne W, et al.
Low Back Pain of Mechanical Origin: Randomised Comparison of Chiropractic and
Hospital Outpatient Treatment
British Medical Journal 1990 (Jun 2); 300 (6737): 1431–1437
Meade TW, Dyer S, Browne W, et al.
Randomised Comparison of Chiropractic and Hospital Outpatient Management for
Low Back Pain: Results from Extended Follow up
British Medical Journal 1995 (Aug 5); 311 (7001): 349–351
Carey TS, Garrett J, Jackman A, et al.
The Outcomes and Costs of Care for Acute Low Back Pain Among Patients
Seen by Primary Care Practitioners, Chiropractors, and Orthopedic Surgeons
New England J Medicine 1995 (Oct 5); 333 (14): 913–917
Cherkin DC, Deyo RA, Battie M, et al.
A Comparison of Physical Therapy, Chiropractic Manipulation, and Provision of an
Educational Booklet for the Treatment of Patients with Low Back Pain
New England Journal of Medicine 1998 (Oct 8); 339 (15): 1021-1029
Skargren EI, Oberg BE, Carlsson PG, et al.
Cost and effectiveness analysis of chiropractic and physiotherapy treatment for low back and neck pain: sixmonth follow-up.
Skargren EI, Carlsson PG, Oberg BE.
One-year follow-up comparison of the cost and effectiveness of chiropractic and physiotherapy as primary management for back pain: subgroup analysis, recurrence, and additional health care utilization.
Andersson GBJ, Lucente T, Davis AM, et al.
A comparison of osteopathic spinal manipulation with standard care for patients with low back pain.
N Engl J Med 1999;341:1426–31.
Hurwitz EL, Morgenstern H, Harber P, et al.
A Randomized Trial of Medical Care with and without Physical Therapy and Chiropractic
Care with and without Physical Modalities for Patients with Low Back Pain: 6-month
Follow-up Outcomes From the UCLA Low Back Pain Study
Spine (Phila Pa 1976) 2002 (Oct 15); 27 (20): 2193–2204
Strong J, Ashton R, Chant D.
Pain intensity measurement in chronic low back pain.
Clin J Pain 1991;7:209–18.
Roland M, Morris R.
A study of the natural history of back pain: I. Development of a reliable and sensitive measure of disability in low-back pain.
Comparative validity of Sickness Impact Profile and shorter scales for functional assessment in low-back pain.
Deyo RA, Centor RM.
Assessing responsiveness of functional scales to clinical change: analogy to diagnostic test performance.
J Chronic Dis 1986;39: 897–906.
Jensen MP, Strom SE, Turner JA, et al.
Validity of the Sickness Impact Profile Roland scale as a measure of dysfunction in chronic pain patients.
Hsieh CJ, Phillips RB, Adams AH, et al.
Functional outcomes of low back pain: comparison of four treatment groups in a randomized controlled trial.
J Manipulative Physiol Ther 1992;15:4–9.
Bouter LM, van Tulder MW, Koes BW.
Methodologic issues in low back pain research in primary care.
Ostelo RW, de Vet HC, Knol DL, et al.
24-item Roland-Morris Disability Questionnaire was preferred out of six functional status questionnaires for
post-lumbar disc surgery.
J Clin Epidemiol 2004;57:268–76.
Ware JE, Sherbourne CD.
The MOS 36-Item Short-Form Health Survey (SF-36): I. Conceptual framework and item selection.
Med Care 1992;30: 473–83.
McHorney CA, Ware JE.
Construction and validation of an alternate form general mental health scale for the Medical Outcomes Study Short-Form Health Survey.
Med Care 1995;33:15–28.
SAS Institute Inc.
The SAS System for Windows 8.1.
Cary, NC: SAS Institute, 2000.
Zeger SL, Liang K-Y.
Longitudinal data analysis for discrete and continuous outcomes.
Liang K-Y, Zeger SL.
Longitudinal data analysis using generalized linear models.
Diggle PJ, Liang K-Y, Zeger SL.
Analysis of Longitudinal Data.
New York: Oxford University Press, 1994.
SAS Institute Inc.
SAS/STAT Software, version 8.
Cary, NC: SAS Institute, 2000.
Yu F, Morgenstern H, Hurwitz EL, et al.
Use of a Markov transition model to analyse longitudinal low-back pain data.
Stat Methods Med Res 2003;12: 321–31.
Hays RD, Brown JA, Spritzer KL, et al.
Member ratings of health care provided by 48 physician groups.
Arch Intern Med 1998;158:785–90.
Coulter ID, Hurwitz EL, Adams AA, Genovese BJ, Hays R, Shekelle PG.
Patients Using Chiropractors in North America:
Who Are They, and Why Are They in Chiropractic Care?
Spine (Phila Pa 1976) 2002; 27 (3) Feb 1: 291–298
Hertzman-Miller RP, Morgenstern H, Hurwitz EL, et al.
Comparing the Satisfaction of Low Back Pain Patients
Randomized to Receive Medical or
Chiropractic Care: Results From the UCLA Low-back Pain Study
Am J Public Health 2002 (Oct); 92 (10): 1628–1633
Hurwitz EL, Morgenstern H, Harber P, et al.
The effectiveness of physical modalities among low-back pain patients randomized to chiropractic care:
findings from the UCLA Low-Back Pain Study.
J Manipulative Physiol Ther 2002;25:10–20.
Drug therapy for back pain: which drugs help which patients?
Cherkin DC, Wheeler KJ, Barlow W, et al.
Medication use for low back pain in primary care.
Koes BW, Scholten RJPM, Mens JMA, et al.
Efficacy of non-steroidal antiinflammatory drugs for low back pain: a systematic review of randomised clinical trials.
Ann Rheum Dis 1997;56:214–23.
van Tulder MW, Koes BW, Bouter LM.
Conservative treatment of acute and chronic nonspecific low back pain: a systematic review of randomized controlled trials of the most common interventions.
van Tulder MW, Scholten RJ, Koes BW, et al.
Nonsteroidal anti-inflammatory drugs for low back pain: a systematic review within the framework of the
Cochrane Collaboration Back Review Group.
van Tulder MW, Touray T, Furlan AD, et al.
Muscle relaxants for nonspecific low back pain: a systematic review within the framework of the Cochrane Collaboration.
Spine 2003;28:1978 –92.
Abenhaim L, Rossignol M, Valat JP, et al.
The role of activity in the therapeutic management of back pain: report of the International Paris Task Force
on Back Pain.
Spine 2000;25(suppl 4):1–33.
Hagen KB, Hilde G, Jamtvedt G, et al.
The Cochrane review of advice to stay active as a single treatment for low back pain and sciatica.
Spine 2002;27: 1736–41.
Exercises: which ones are worth trying, for which patients, and when?
van Tulder M, Malmivaara A, Esmail R, et al.
Exercise therapy for low back pain: a systematic review within the framework of the Cochrane Collaboration
Back Review Group.
Kool J, de Bie R, Oesch P, et al.
Exercise reduces sick leave in patients with non-acute non-specific low back pain: a meta-analysis.
J Rehabil Med 2004; 36:49–62.
Rainville J, Hartigan C, Martinez E, et al.
Exercise as a treatment for chronic low back pain.
Spine J 2004;4:106–15.
Furlan AD, Brosseau L, Imamura M, et al.
Massage for low-back pain: a systematic review within the framework of the Cochrane Collaboration Back Review Group.
Cherkin DC, Sherman KJ, Deyo RA, et al.
A review of the evidence for the effectiveness, safety, and cost of acupuncture, massage therapy, and spinal
manipulation for back pain.
Ann Intern Med 2003;138:898–906.
van Tulder MW, Ostelo R, Vlaeyen JW, et al.
Behavioral treatment for chronic low back pain: a systematic review within the framework of the Cochrane Back Review Group.
Karlalainen K, Malmivaara A, van Tulder M, et al.
Multidisciplinary biopsychosocial rehabilitation for subacute low back pain in working-age adults: a systematic review within the framework of the Cochrane Collaboration Back Review Group.
Hagen KB, Hilde G, Jamtvedt G, et al.
The Cochrane review of bed rest for acute low back pain and sciatica.
Nordin M, Campello M.
Physical therapy. Exercises and the modalities: when, what and why?
Neurol Clin North Am 1999;17:75–89.
Brosseau L, Milne S, Robinson V, et al.
Efficacy of the transcutaneous electrical nerve stimulation for the treatment of chronic low back pain: a metaanalysis.
Harte AA, Baxter GD, Gracey JH.
The efficacy of traction for back pain: a systematic review of randomized controlled trials.
Arch Phys Med Rehabil 2003;84:1542–53.
Shekelle PG, Adams AH, Chassin MR, et al.
Spinal manipulation for lowback pain.
Ann Intern Med 1992;117:590–8.
Assendelft WJ, Morton SC, Yu EI, et al.
Spinal manipulative therapy for low back pain: a meta-analysis of effectiveness to other therapies.
Ann Intern Med 2003;138:871–81.
Bronfort, G, Haas, M, Evans, RL, and Bouter, LM.
Efficacy of Spinal Manipulation and Mobilization
for Low Back Pain and Neck Pain:
A Systematic Review and Best Evidence Synthesis
Spine J (N American Spine Soc) 2004 (May); 4 (3): 335–356
Bernstein E, Carey TS, Garrett JM.
The use of muscle relaxant medications in acute low back pain.
Assendelft WJJ, Bouter LM, Knipschild PG.
Complications of spinal manipulation: a comprehensive review of the literature.
J Fam Pract 1996;42: 475–80.
Assendelft WJJ, Koes BW, van der Heijden GJMG, et al.
The effectiveness of chiropractic for treatment of low back pain: an update and attempt at statistical
J Manipulative Physiol Ther 1996;19:499–507.
Haas M, Goldberg B, Aickin M, Ganger B, Attwood M.
A Practice-Based Study of Patients With Acute and Chronic Low Back Pain
Attending Primary Care and Chiropractic Physicians: Two-Week to 48-Month Follow-up
J Manipulative Physiol Ther 2004 (Mar); 27 (3): 160–169
Wolsko PM, Eisenberg DM, Davis RB, et al.
Patterns and perceptions of care for treatment of back and neck pain: results of a national survey.
Hurwitz EL, Morgenstern H.
The effects of comorbidity and other factors on medical versus chiropractic care for back problems.
Spine 1997;22: 2254– 64.
Cote P, Cassidy JD, Carroll L.
The Treatment of Neck and Low Back Pain:
Who Seeks Care? Who Goes Where?
Med Care. 2001 (Sep); 39 (9): 956–967
Hestbaek L, Leboeuf-Yde C, Manniche C.
Low Back Pain: What Is The Long-term Course?
A Review of Studies of General Patient Populations
European Spine Journal 2003 (Apr); 12 (2): 149–165
Jacob T, Baras M, Zeev A, et al.
A longitudinal, community-based study of low back pain outcomes.
Hurwitz EL, Morgenstern H.
Correlates of back problems and back-related disability in the United States.
J Clin Epidemiol 1997;50:669–81.
Macfarlane GJ, Thomas E, Croft PR, et al.
Predictors of early improvement in low back pain amongst consulters to general practice: the influence of
pre-morbid and episode-related factors.
Hurwitz EL, Morgenstern H, Chiao C.
Effects of Recreational Physical Activity and Back Exercises
on Low Back Pain
and Psychological Distress: Findings from the UCLA Low Back Pain Study
Am J Public Health. 2005 (Oct); 95 (10): 1817–1824
Frost H, Lamb SE, Klaber Moffett JA, et al.
A fitness programme for patients with chronic low back pain: 2-year follow-up of a randomised controlled
Moffett JK, Torgerson D, Bell-Syer S, et al.
Randomised controlled trial of exercise for low back pain: clinical outcomes, costs, and preferences.
Torstensen TA, Ljunggren AE, Meen HD, et al.
Efficiency and costs of medical exercise therapy, conventional physiotherapy, and self-exercise in patients
with chronic low back pain: a pragmatic, randomized, single-blinded, controlled trial with 1-year follow-up.
Physical therapy for musculoskeletal syndromes.
Rheum Dis Clin North Am 1996;22:551–62.
Burton AK, Tillotson KM, Main CJ, et al.
Psychosocial predictors of outcome in acute and subchronic low back trouble.
Leclaire R, Esdaile JM, Suissa S, et al.
Back school in a first episode of compensated acute low back pain: a clinical trial to assess efficacy and prevent relapse.
Arch Phys Med Rehabil 1996;77:673–9.
Malmivaara A, Hakkinen U, Aro T, et al.
The treatment of acute low back pain: bed rest, exercises, or ordinary activity?
N Engl J Med 1995;332: 351–5.
Stratford PW, Binkley J, Solomon P, et al.
Defining the minimum level of detectable change for the Roland-Morris questionnaire.
Phys Ther 1996;76: 359–65.
Stucki G, Liang MH, Fossel AH, et al.
Relative responsiveness of conditionspecific and general health status measures in degenerative lumbar spinal
J Clin Epidemiol 1995;48:1369–78
Return to the LOW BACK PAIN Section