SPINE Volume 37, Number 7, pp 531–542
©2012, Lippincott Williams & Wilkins
RANDOMIZED TRIAL
The Efficacy of Systematic Active Conservative
Treatment for Patients With Severe Sciatica
A Single-Blind, Randomized, Clinical, Controlled Trial
Hanne B. Albert, PT, MPH, PhD, and Claus Manniche, MD, PhD, Med Sci
Study Design. Prospective single-blind, randomized, clinical,
controlled trial.
Objective. To evaluate the efficacy of active conservative treatment
and to compare 2 active conservative treatment programs for patients
with severe sciatica.
Summary of Background Data. Reviews have demonstrated
little or no efficacy for passive conservative treatment modalities in
patients suffering from sciatica. The results for surgery are conflicting.
Cohort studies have shown excellent results for active treatment
modalities in patients with sciatica.
Methods. One hundred eighty-one consecutive patients with
radicular pain below the knee were examined at the baseline, at 8
weeks, and at 1 year after the treatment. Participants were randomized
into 2 groups: (1) symptom-guided exercises + information + advice
to stay active and (2) sham exercises + information + advice to stay
active. Symptom-guided exercises consisted of a variety of back-
related exercises given in accordance with a written algorithm in
which symptoms or response to exercises determined the exercises
given (http://www.sygehuslillebaelt.dk/wm345075, click exercises).
Sham exercises were optional, designed to increase general blood
circulation, and had no targeted effect on the back. The information
was comprehensive and included anatomy, pathogenesis, and how
discs heal without surgery. The advice included encouragement to
stay as active as possible but to reduce activity if leg pain increased.
The use of medication was optional, but only paracetamol and
nonsteroidal anti-inflammatory drugs were recommended.
From the Back Research Centre, Funen, University of Southern Denmark,
Ringe, Denmark.
Acknowledgment date: October 23, 2006. First revision date: February 10,
2009. Second revision date: January 27, 2011. Acceptance date: March 7,
2011.
The legal regulatory status of the device(s)/drug(s) that is/are the subject of this
manuscript is not applicable in my country.
Federal, institutional, and foundation funds were received in support of
this work. Although one or more of the author(s) has/have received or will
receive benefits for personal or professional use from a commercial party
related directly or indirectly to the subject of this manuscript, benefits will be
directed solely to a research fund, foundation, educational institution, or other
nonprofit organization which the author(s) has/have been associated.
Address correspondence and reprint requests to Hanne B. Albert, PT, MPH,
PhD, The Back Research Centre, Lindevej 5, DK-5750 Ringe, Denmark;
E-mail: hanne.birgit.albert@slb.regionsyddanmark.dk
DOI: 10.1097/BRS.0b013e31821ace7f
Spine
Copyright © 2012 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
BRS204492.indd 531
Results. A mean of 4.8 treatment sessions were provided.
All patients experienced statistically significant and clinically
important improvements in global assessment, functional status,
pain, vocational status, and clinical findings. The symptom-guided
exercise group improved significantly more than the sham exercise
group in most outcomes.
Conclusion. Active conservative treatment was effective for
patients who had symptoms and clinical findings that would normally
qualify them for surgery. Although participating patients had greater
faith in the sham exercises before treatment, the symptom-guided
exercises were superior for most outcomes.
Key words: herniated disc, sciatica, randomized trial, active
conservative treatment, exercise therapy. Spine 2012;37:531–542
L
ow back pain (LBP) with or without sciatica is a com-
mon problem. It has an estimated point prevalence of
around 30% and a lifetime prevalence of 80%.1 It can
be disabling and has significant direct and indirect costs to
the individual as well as to society. The prevalence of sciatica
due to nerve root compression is more difficult to estimate
because the diagnosis requires clinical examination. Never-
theless, in a large epidemiological study, Heliövaara et al2
showed a lifetime prevalence of lumbar disc herniation of 5%
for men and 4% for women. These prevalence estimates were
confirmed in a study by Manninen et al.3 Lumbar nerve root
compression is generally thought to be the cause of sciatica,
especially if the pain radiates below the knee. Even though
patients with an acute lumbar disc herniation represent only
2% of all LBP patients, they account for a disproportionately
large percentage (30%) of the annual US national expendi-
ture for the treatment of LBP.4 This is likely due to the level
of activity limitation and participation restriction associated
with acute sciatica.
Anecdotally, the natural course of sciatica due to herni-
ated discs is generally favorable, and there is evidence that
many herniations are physiologically reabsorbed after several
months.5 However, because a large proportion of patients
with herniated discs suffer from severe pain and experience
unpleasant sensory and motor disturbances, health care sys-
tems often intervene to relieve these symptoms.
Few authors have measured the natural course of sciatica.
Ellenberg et al6 monitored 18 patients with an average
www.spinejournal.com 531
13/03/12 3:52 PM
 RANDOMIZED TRIAL The Efficacy of Systematic Active Conservative Treatment • Albert and Manniche
duration of symptoms of 9.8 months. Of these patients, 14
achieved complete clinical improvement and 4 underwent sur-
gery. Weber et al7 monitored 208 LBP patients with radiating
pain and clear clinical signs of nerve root compression, who
were randomized into groups of either nonsteroidal anti-in-
flammatory drug (NSAID) treatment (piroxicam) or placebo
medicine. The results in the placebo group may be assumed
to be close to the natural course. Both groups improved sig-
nificantly within 4 weeks. At 4-week, 3-month, and 1-year
follow-up, there were no differences between the groups in
any of the outcome measures.
Since Mixter and Barr8 published results implicating rup-
tures of intervertebral discs as a direct cause of sciatica, the
presence of protruding nuclear material has encouraged many
to initiate surgery in an attempt to alleviate severe symptoms.
Currently, though, conservative treatment is generally the first
treatment option1,4 except in cases of cauda equina syndrome.
Studies have shown that accessibility to surgeons and the phy-
sician’s belief in the efficacy of surgery or conservative treat-
ment are the determinant factors for the choice of treatment.9
No general agreement exists as to the indications for surgery
or conservative treatment and as to which type of conserva-
tive treatment is the most efficacious.10,11
There is consistent evidence that the long-term efficacy of
conservative treatment and surgery are very similar.12 In one
10-year follow-up study, Weber12 found satisfactory results
in 74% of the patients who received conservative treatment
and in 71% of the patients who underwent surgery. Hentzer13
reported similar results at her 4-year follow-up, with 85%
of the patients having recovered after receiving mainly ac-
tive conservative treatment and 82% having recovered after
receiving surgery. Atlas et al14 performed a 10-year follow-
up of patients with sciatica, consisting of a cohort of patients
treated with different kinds of conservative treatments and a
cohort of surgically treated patients. At 10 years, the groups
were similar according to back pain, number of additional
surgical interventions, work, and disability status, but the
surgery group had less leg pain and activity limitation. Peul
et al15 investigated whether surgery or continued conserva-
tive treatment would be more effective for people who had
not improved after initial conservative treatment. This multi-
center study included 283 patients with 6 to 12 weeks of un-
relenting pain, aged 18 to 65 years, and with a radiologically
confirmed herniation. Patients were either randomized to a
standardized operation or returned to their general practitio-
ner (GP) with information about the generally good prognosis
and usefulness of staying active. Activity limitation, measured
with the Roland Morris Disability Questionnaire (RMDQ),
was the primary outcome measure. On average, both groups
did well, reducing their activity limitation from 16 RMDQ
points at the baseline to 4 RMDQ points at 6 months, with
no improvement thereafter. Although the surgical group had
less activity limitation and leg pain at 2 and 3 months’ follow-
ups, these differences were not present at previous or later
time points.15
Conservative care includes a large variety of treatments
such as analgesics, rest, exercises, traction, manipulation,
532 www.spinejournal.com
Copyright © 2012 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
BRS204492.indd 532
mobilization, and epidural injections. Little is known about
the efficacy and effectiveness of these treatments, with report-
ed success rates varying from 11% to 85%.10,12–14,16–19 The
most frequently used passive conservative treatment modality
is bed rest, either as a single modality or in combination with
others. In a systematic review of passive conservative treat-
ments for sciatica, which included epidural steroids, manipu-
lation, traction, and NSAIDs, Vroomen et al11 concluded that
there was insufficient evidence supporting the effectiveness of
most of these passive conservative treatments. Similarly, in a
thorough review, Rozenberg et al20 concluded that there is no
evidence that epidural steroid injections are efficacious in the
treatment of sciatica.
Perhaps because of this evidence, in recent years, there has
been a trend away from passive to active treatment of LBP.1
The distinguishing elements in active treatment are providing
patients with the necessary information and advice to stay ac-
tive and treatment modalities in which the patient is an active
participant, physically as well as taking responsibility in the
treatment process. The treatment provider has a new role as
a “coach” for the patient, instead of being the person who is
expected to provide a cure.
Cohort studies have demonstrated excellent results of in-
tensive active treatment programs for sciatica patients. Saal
and Saal17 treated 58 patients who all had indications for sur-
gery. The intensive active treatment program consists of Back
School and exercises for spinal stabilization, strength, and dy-
namic maintenance of postural control. They demonstrated
a recovery rate of good to excellent in 96% of the treated
patients. Similarly, Sang-Ho et al21 monitored 22 patients who
were treated with an intensive active conservative treatment
program, initially consisting of physical therapy, oral ste-
roids, and NSAIDs. Thereafter, they received a rehabilitation
program consisting of pelvic stabilization exercises, traction,
Back School, NSAIDs, and antidepressants. After a mean of
6.9 months (3–11), 17 patients (77%) demonstrated a clini-
cally successful result.
To our knowledge, the only RCT comparing active conser-
vative treatment with surgery was performed by Ulreich and
Kullich,22 involving patients with CT-verified herniations. A
4-week, active multidisciplinary treatment program was com-
pared with surgery. Pain was measured with the McGill ques-
tionnaire and activity limitation with the RMDQ. In compari-
son with the baseline, both pain and activity limitation were
reduced in both groups after treatment, but improvement was
only statistically significant for pain in the active, conserva-
tively treated group and for activity limitation in the surgical
group. This publication is in German and does not appear to
be well known.
Great emphasis has been invested in developing the most
efficacious operation techniques for a herniated disc. This has
resulted in an ability to thoroughly describe and standard-
ize these operations. This is very important as a prerequisite
for research to be able to identify best-practice surgical tech-
niques for clinical practice.
In contrast, despite conservative treatment being the first-line
treatment for a herniated disc, minimal research has attempted
April 2012
13/03/12 3:52 PM
 RANDOMIZED TRIAL The Efficacy of Systematic Active Conservative Treatment • Albert and Manniche
to identify the most efficacious conservative treatment. Even
in large and prestigious research projects such as SPORT,23 the
choice of conservative treatment is at the discretion of partici-
pating clinicians, resulting in diverse treatments from injection
of steroids to traction. The only similarity between these in-
terventions is that they are not surgical, and this heterogeneity
is likely to confound estimates of efficacy. Similarly, studies of
sciatica, such as those of Peul et al15 and Luijsterburg et al,24
had minimum standardization of conservative care. Beyond a
requirement to advise patients to stay active, the conservative
treatment techniques used for individual patients in both stud-
ies were at the discretion of participating clinicians.
Studies that compare the efficacy of conservative treat-
ment regimens for sciatica are rare, but Luijsterburg et al24
undertook a study and found that treatment by both a GP
and physiotherapist resulted in better patient outcomes than
the treatment by a GP alone. GP treatment consisted of mild
analgesia and advice to perform activities of daily living and
to avoid painful movements. In addition, the physiotherapy
group had active exercises that were chosen at the discre-
tion of the physiotherapist. A maximum of 9 treatments in
6 weeks were given. Both groups improved at 3, 6, and 12
weeks’ follow-up. At the 12-month follow-up, a significantly
larger proportion of the physiotherapy group had improved:
79% compared with 56% in the GP-only group. The poten-
tial heterogeneity of the active exercises received in this study
limits the conclusions that can be drawn from the findings. It
may be that variations in exercise technique, dosage, intensity,
and progression affect efficacy.
There remains a need to identify the most efficacious con-
servative treatment of sciatica caused by a herniated disc.
Therefore, the aims of this study were to describe the overall
efficacy of active conservative treatment of severe sciatica and
to compare the outcomes of 2 active conservative treatment
regimens in patients suffering from severe sciatica.
MATERIALS AND METHODS
This study was a prospective, single-blind, randomized, con-
trolled clinical trial. The setting was the Back Centre Funen,
which is a state-funded regional secondary care facility on
the island of Funen in Denmark. All patients were referred
by general practitioners, rheumatologists, and chiroprac-
tors after unsuccessful treatment in primary care. The study
was approved by the local ethical committee reference no.
VP20010134.
Inclusion and Exclusion Criteria
Patients were included if they were 18 to 65 years of age and
had radicular pain of dermatomal distribution to the knee or
below in 1 or both legs, had leg pain more than 3 on a 1- to
10-point scale at first visit to the clinic, and had a duration of
sciatica between 2 weeks and 1 year. Patients were excluded if
they had cauda equina syndrome, pending worker’s litigation,
previous back surgery, spinal tumors, pregnancy, a language
other than Danish as their first language, or an inability to
follow the rehabilitation protocol due to concomitant disease
such as depression or heart failure.
Spine
Copyright © 2012 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
BRS204492.indd 533
Treatment Programs
Patients were randomized to 1 of 2 treatment programs. Both
treatment programs contained identical information and
advice but differed in the type of exercise program that was
included.
Both the groups received thorough information regard-
ing the anatomy of the spine, the pathogenesis of a herniated
disc, and the natural healing process without surgery. It was
emphasized that sciatica usually has a positive prognosis.
Patients were encouraged to stay as active as possible but to
respect an increase in leg pain as a warning of overstepping
current physical limitations. This information was based on
behavioral theories of empowerment,25 which emphasize en-
couraging patients’ self-responsibility for their actions and
also to provide tools to enable them to control their pain.
Medication was optional, but only paracetamol and NSAIDs
were recommended.
The treatment differed between the 2 groups in their exer-
cise programs: either “symptom-guided exercises” or “sham
exercises.” Symptom-guided exercises consisted of back-related
exercises: directional end-range exercises and postural instruc-
tions guided by the individual patient’s directional preference
(based on the McKenzie method of assessing pain-related phys-
ical impairment). Furthermore, these patients were instructed
in stabilizing exercises for the transverse abdominis and mul-
tifidus muscles and dynamic exercises for the outer layers of
the abdominal wall and back extensors. The exercises were
administered in accordance with a standardized algorithm. In
the algorithm, different symptoms or responses to exercises de-
termined the choice of exercises and relevant instructions, as
well as whether the patients were ready to be progressed to
more strenuous exercises. The detailed algorithm can be down-
loaded from http://www.sygehuslillebaelt.dk/wm335360 (click
exercise program). Home exercise programs were handed out
to all patients.
Sham exercises consisted of optional exercises that were
not back related but were low-dose exercises to simulate an
increase in systemic blood circulation.
The treatment lasted for 8 weeks with a minimum of 4 and
a maximum of 8 treatments. Patients were discouraged from
receiving any additional treatment of their sciatica. At the end
of the treatment period, they were encouraged to continue
their exercises for the following year. If patients phoned the
trial coordinator, they were reassured and urged to comply
with their assigned treatment. In cases of new radiating pain,
they were referred again to the Back Centre.
Induction Procedures
Patients were included in the study after approximately
1 week after referral. They received both oral and written
information, and signed informed consent was obtained. A
blinded observer performed a thorough physical examina-
tion, and a history was obtained. The patient self-completed
all the study questionnaires.
Because blinding of the patients was not possible, it was
important that both treatment regimens were equally credible
for the patients. There was considerable emphasis on getting
www.spinejournal.com 533
13/03/12 3:52 PM
 RANDOMIZED TRIAL The Efficacy of Systematic Active Conservative Treatment • Albert and Manniche
patient understanding of the rationale for both treatments.
The rationale for the symptom-guided exercises and their
possible effect on the movement of the nucleus material were
explained. To clarify the rationale behind the sham exercises,
the metaphor of a mosquito bite was used: “A mosquito bite
is itchy and annoying, and if you scratch it, the itching will
become worse and the healing period will be extended. If you
don’t scratch the mosquito bite, it will take the time of the
natural healing process before it disappears, which is shorter
than if you scratch it.” The same applies to a herniated disc:
“Be as active as you can, but if leg pain is provoked, it is like
scratching the mosquito bite. Slowly the body’s natural heal-
ing process will heal the wound in your disc. The exercise is
meant to increase the blood circulation, and this increase will
bring more nourishment to the wound in your disc and speed
up the healing period.”
Randomization
After enrollment, nonstratified randomization was performed
by a research assistant using the random number generator
program Minimizer. The patient was allocated an appoint-
ment with a clinician, and the MRI scan was performed. This
first session lasted for 3½ to 4 hours.
Treatment Providers
The treatment was performed by 2 multidisciplinary treat-
ment teams. The symptom-guided exercise team consisted of
3 physiotherapists and 1 chiropractor, and the sham exercise
team consisted of 2 physiotherapists and 2 chiropractors. A
medical doctor and a nurse were allocated to each team for
consultancy if specific problems concerning medication or
differential diagnostic problems arose.
Blinding of the treatment providers was impossible, so it
was important to make the treatment credible and to encour-
age the clinicians’ enthusiasm so that they would promote
their own treatment. Therefore, no personnel overlap in the
treatment teams was permitted. Each treatment provider had
chosen the treatment team of his or her preference. The algo-
rithm of exercises and patient information was developed by
the treatment teams. Therefore, all treatment providers were
enthusiastic concerning their own treatment group. A pilot
study with 20 patients was performed with the aim of training
and synchronizing the treatment providers and training the
blinded examiner in the examination procedures.
Examination
The medical history and a thorough physical examination of
the spine and lower extremities were undertaken by the same
blinded observer at the baseline, 8 weeks after, at the end of
treatment, and 1 year after the end of treatment. The clini-
cal presentation of the patients is reported elsewhere.25 The
patients were repeatedly instructed to conceal their treatment
assignment from the blinded observer.
Physical Examination
A thorough examination of the spine and lower extremities
was performed. In this study, only the standardized signs of
534 www.spinejournal.com
Copyright © 2012 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
BRS204492.indd 534
nerve root compression are presented. A straight-leg-raise
(SLR) test was deemed positive if a passive lift of the extended
leg with the foot in neutral position provoked pain of ra-
dicular nature. Asymmetry of the Achilles and/or the patella
reflexes were recorded if the reflex was missing or hypo- or
hypertonic only in the involved leg. Motor deficits were clas-
sified as present if the muscle strength of the flexors/exten-
sors of the foot or the 4 lateral toes or the big toe was less
than normal on the Oxford Muscle Strength Scale. Disturbed
sensibility was present if any change in sensory perception oc-
curred in the full extent of the dermatomes of L4 or L5 or S1.
Outcome Measures
The primary outcome measures were activity limitation and
current leg pain. Activity limitation was measured using the
Danish version26 of the 23-question RMDQ,27 which includes
the assessment of back pain and back-related leg pain on ac-
tivities of daily living.28 Clinically important change in activ-
ity limitation was defined as 30% or more change from the
baseline RMDQ score, as recommended by Ostelo et al.29
Pain was measured using the Low Back Pain Rating Scale,
which measures low back and leg pain on a 0 to 10 scale. It
was originally developed in the Danish population, is exten-
sively used,31 and includes measures of current leg pain, worst
leg pain in the last 2 weeks, and average leg pain in the last
2 weeks. In this study, current leg pain was used as the pri-
mary pain outcome measure. Clinically important change in
current leg pain was defined as a change of 2 points as recom-
mended by Ostelo et al.29
Secondary outcome measures were global improvement
and number of neurological signs. Global improvement
was measured on a 5-point Likert scale, and QUALY was
measured by EuroQOL (EQ-5D) using the adjusted Danish
scores.31 A composite score was calculated, which measured
total leg pain on a 30-point scale (a sum score of current leg
pain, worst leg pain in the last 2 weeks, and average leg pain
in the last 2 weeks) as a comprehensive secondary outcome
measure of leg pain. Sick leave during the year after treat-
ment was measured by patients’ self-report on the 12-month
follow-up questionnaire. The patients’ expectations of out-
come of the 2 treatment regimens were also measured before
randomization by patients’ self-report, using the questions
“What treatment outcome will you expect to have if you re-
ceive ‘Hands on’ exercises = symptom-guided exercises” and
“What treatment outcome will you expect to have if you re-
ceive ‘Hands off’ exercises = circulation exercises?” Evalu-
ated on a 7-point Likert scale, 3 were worse, 3 were better, 1
and 1 was unchanged.
Sample Size
Improvement in current leg pain was the outcome measure
in the power analysis to determine sample size. A mean im-
provement of 2 points (range, 0–5) in the sham exercise group
and 3 points (range, 0–6) in the symptom-guided exercise
group was assumed. Using the Wilcoxon rank test, a sample
size of 180 patients, 90 per group, was required to detect this
magnitude of improvement with a power of 90% at 1000
April 2012
13/03/12 3:52 PM
 RANDOMIZED TRIAL The Efficacy of Systematic Active Conservative Treatment • Albert and Manniche
simulations. Sample size was determined before the start of
the study, and no interim analysis was performed.
Statistical Analysis
Data were described with either mean or median and quar-
tiles depending on whether they were normally distributed.
By using SPSS 13 (SPSS Inc., Chicago, IL), treatment efficacy
was evaluated with a Wilcoxon matched-pairs signed-rank
test with a P value of <0.05. Intention-to-treat analysis was
performed.
RESULTS
Participants and Dropouts
All consecutive 181 patients who met the inclusion criteria
were included and randomized to 1 of the 2 treatment groups.
An additional 6 patients had been excluded: 4 due to a spinal
tumor and 2 due to severe depression. A participant flow chart
is shown in Figure 1. During the treatment period, 6 patients
had rapidly progressing symptoms and were referred to the
neurosurgical department. One was involved in an accident
and could not continue the treatment, and 2 did not complete
the treatment for other reasons. Both at the follow-up at the
end of the treatment and 1 year after the end of the treatment,
4 patients did not show up for the clinical examination, 2 of
these because of temporary work in another part of the coun-
try, but they completed the questionnaires. The mean number
of treatments was 4.8, and, of these, 0.6 were via telephone
contact. Intention-to-treat analysis was performed but, be-
cause of the low dropout rate, did not significantly change
any result on any outcome measure.
The baseline characteristics of the 2 groups are shown in
Table 1. There were no statistically significant differences be-
tween the 2 groups on any factors. The time from onset of
sciatica until entering the study varied: 16% had sciatica for
less than 1 month, 61.3% for 1 to 3 months, 17.7% for 3 to
6 months, and 5% for 6 to 12 months.
Patients’ Expectations
A greater number of the patients thought that the sham ex-
ercise treatment would be more efficacious for them than the
symptom-guided exercise treatment (Figure 2).
TREATMENT OUTCOME
Global Assessment
Most patients (89.2%) experienced a global improvement
(better or much better) by the end of the treatment, and this
was sustained at 1 year posttreatment (91.1%). More people
in the symptom-guided exercise group experienced a global
improvement than those in the sham exercise group (P <
0.008) at both time points (Figure 3).
Activity Limitation
On average, patients experienced statistically significant (P
< 0.00001) and clinically important improvement in their
activity limitation both at the end of treatment and at 1-year
Spine
Copyright © 2012 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
BRS204492.indd 535
follow-up. The median improvement at the end of treatment
was 10.0 RMDQ points and at 1 year was 12.5 RMDQ points
relative to the baseline (Table 2). Across the whole cohort, by
the end of treatment (after 8 weeks), 75% of the patients had
improved by a clinically important number or more. There
were no statistically significant differences between the 2
treatment groups at either time point.
Leg Pain
On average, patients also experienced statistically significant
(P < 0.00001) and clinically important reductions in their
current leg pain both at the end of treatment and at 1-year
follow-up. The mean improvement at the end of treatment
was 2.6 points and at 1 year was 3.0 points relative to the
baseline (Table 3). At 1-year follow-up, there was an average
reduction of 66% of their baseline pain scores. At the end of
treatment, there appeared to be a trend (P < 0.6) toward the
symptom-guided exercise group having a greater mean reduc-
tion in current leg pain than that in the sham exercise group
(0.8 points), but the size of this difference was not clinically
important. These results were mirrored in the change in total
leg pain (Table 3), with the median score at the baseline being
18 on a 30-point scale, improving to 4 after treatment, and
improving further to 2.5 at 1-year follow-up.
Nerve Root Compression Signs
At the baseline, almost all (95%) patients had between 2
and 4 positive neurological signs, with a mean of 2.8 signs.
On average, the whole cohort of patients experienced a sta-
tistically significant (P < 0.001) reduction in the number of
positive signs at 1-year follow-up (symptom-guided exercise
group 1.9 and sham exercise group 1.3). The symptom-
guided exercise group demonstrated a greater improvement
than the sham exercise group at both time points (Table 4).
Generic Function: QUALY
The EQ-5D scores showed that all patients experienced a
mean improvement of 0.20 from the baseline at the end of
treatment and at 1-year follow-up (P < 0.001). The symptom-
guided exercise group improved from 0.62 (SD = 0.18) at the
baseline to 0.82 (SD = 0.21) at 1-year follow-up, which was
more than the sham exercise group from 0.62 (SD = 0.62)
to 0.79 (SD = 0.24), but the difference was not statistically
significant.
At the baseline, the health thermometer value for all pa-
tients was a median of 60 (40–75), at the end of treatment 82
(57–90), and at 1-year follow-up 85 (70–95). The symptom-
guided exercise group improved more than the sham exercise
group, but the difference was not statistically significant.
Sick Leave
At the 1-year follow-up, only 35.2% of the patients in the
symptom-guided exercise group and 34.1% of the patients in
the sham exercise group had had any sick leave due to back
pain in the preceding year. The patients who had had sick
leave in the symptom-guided exercise group had a mean of
73 days (SD = 79), and the patients on sick leave in the sham
www.spinejournal.com 535
13/03/12 3:52 PM
 RANDOMIZED TRIAL The Efficacy of Systematic Active Conservative Treatment • Albert and Manniche

Figure 1. Patient flow chart.

exercise group had had a mean of 107 days (SD = 107). Of Patients’ Satisfaction With Information
the people on sick leave, 23.9% of the symptom-guided ex- The information provided to the symptom-guided exercise
ercise group were on sick leave for more than 120 days as op- group was individualized, and the information provided to
posed to 43% of the sham exercise group. the sham exercise group was general. The patients in both
536 www.spinejournal.com April 2012
Copyright © 2012 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

BRS204492.indd 536 13/03/12 3:52 PM

 RANDOMIZED TRIAL The Efficacy of Systematic Active Conservative Treatment • Albert and Manniche

TABLE 1. Patient Characteristics at Baseline: Median Values (25%–75% Interquartile Range) If Not
Otherwise Stated
Variable Symptom-Guided Exercise Group Sham Exercise Group
Number included at baseline 95 96
Age, yr 46 (38–52) 44 (37–51)
Women 43% 53%
Duration of symptoms before study
0–4 weeks 25% 18%
5–12 weeks 59% 63%
12–52 weeks 16% 19%
Mean number of treatments 5 (4–6) 5 (4–6)
Roland Morris Disability 16 (11–18) 15 (12–18)
Current leg pain (0–10) 4 (3–6) 5 (3–7)
Angle where SLR is positive 30° (15°–45°) 30° (15°–45°)
Positive SLR 90 (94.7%) 81 (94.2%)
Motor deficits in leg 50 (52.6%) 43 (50%)
Disturbed sensibility in leg 74 (77.9%) 72 (83.7%)
Asymmetric reflexes in leg 53 (55.8%) 46 (53.5%)
Sum score of positive nerve root neurological signs
0 1 (1.1%) 0 (0%)
1 3 (3.2%) 5 (5.8%)
2 31 (32.6%) 24 (27.9%)
3 38 (40%) 39 (45.3%)
4 22 (23.2%) 18 (20.9%)
SLR indicates straight leg raise.

groups were generally very satisfied with the information they of the information, and, at 1-year follow-up, 90.5% felt the
were given. At the end of the treatment, 93.5% in the symp- same way (Figure 4). More patients in the symptom-guided
tom-guided exercise group had been able to use all or most exercise group had been able to use all the information.

Figure 2. The patients’ expectations

before randomization of how they
would benefit from 2 treatments. SYMIA
indicates Symptom-guided exercises +
Information + Advice to stay active
group; SHAMIA, Sham exercises + In-
formation + Advice to stay active group.
Spine www.spinejournal.com 537
Copyright © 2012 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

BRS204492.indd 537 13/03/12 3:52 PM

 RANDOMIZED TRIAL The Efficacy of Systematic Active Conservative Treatment • Albert and Manniche

Figure 3. The patients’ overall assessment of the treatment outcomes in both treatment groups. The difference between the 2 groups is statistically
significant (P < 0.008) both at the end of treatment and at 1-year follow-up. SYMIA indicates Symptom-guided exercises + Information + Advice
to stay active group; SHAMIA, Sham exercises + Information + Advice to stay active group.

DISCUSSION
In this study, 181 patients with sciatica were randomized into
2 treatment groups receiving active conservative treatment for
8 weeks. Both groups received thorough information about
the anatomy of the spine and the natural healing process of
a herniated disc, and optional paracetamol and/or NSAIDs.
The symptom-guided exercise group performed active reha-
bilitation exercises, directional end-range exercises, postural
instruction (based on the McKenzie approach), and segmen-
tal stabilization exercises for the inner unit of trunk muscles
and optional manipulation. The sham exercise group received
optional low-dose exercises to increase systemic blood circu-
lation. On average, patients experienced clinically meaningful
improvement in all outcome measures. Although the patients
had greater faith in the sham exercises before treatment, the
symptom-guided exercise treatment produced superior results
on most outcomes.
Study Design
The 2 treatment groups differed in size (86 and 95)
because no block randomization was performed, but
randomization was effective because both the groups were
similar on all measurements at the baseline. Any difference
in the outcomes of the groups was therefore not due to
discrepancies related to baseline status.
All the participants were consecutively recruited using a
standardized, pretested procedure and examined by the same
blinded examiner at all time points. Because only 3 patients

TABLE 2. The Median (25%–75% Interquartile Range) Values of Roland Morris Disability
Questionnaire (RMDQ) Measured at Baseline, End of Treatment, and at 1-Year Follow-up
Obtained 30% or More Difference Between
End of Treat- 1-year Follow- Improvement of Origi- Treatment Groups at
Baseline ment up nal Value (%)* Any Time
All patients 16 (12–18) 6 (2–12) 3.5 (1–10) 75
Symptom-guided 15.5 (11–18) 6 (2–12) 3.5 (1–10) 73 NS
exercise group
Sham exercise 15 (12–18) 6 (2–12) 3.5 (1–10) 77.5 NS
group
*Number of patients who obtained a clinically relevant improvement defined as a 30% or more improvement of the original RMDQ score.

538 www.spinejournal.com April 2012

Copyright © 2012 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

BRS204492.indd 538 13/03/12 3:52 PM

 RANDOMIZED TRIAL The Efficacy of Systematic Active Conservative Treatment • Albert and Manniche

TABLE 3. Mean (Standard Deviation) Current Leg Pain and Median (25–75% Interquartile Range)
Total Leg Pain Value at Baseline, End of Treatment, and at the Follow-up 1 Year After End of
Treatment
Symptom-Guided Exercises Sham Exercises Statistical Difference Between Groups
Baseline
Current leg pain 4.3 (SD = 2.3) 4.5 (SD = 2.5) NS
Total leg pain* 18 (15–21) 18 (12–21) NS
End of treatment
Current leg pain 1.5 (SD = 2.1) 2.3 (SD = 2.7) 0.06
Total leg pain* 4 (0–9) 4 (0–12) NS
1-year follow-up
Current leg pain 1.5 (SD = 2.1) 1.4 (SD = 2.4) NS
Total leg pain* 3 (0–10) 2 (0–8) NS
*Total leg pain (0–30) = current leg pain value (0–10) + worst leg pain last 2 weeks (0–10) + average leg pain last 2 weeks (0–10).

declined participation from a cohort of 190 potential par- self-selection pattern all indicate that patients were satisfied
ticipants, there is little possibility of selection bias. Only 2 with the treatment provided and that the treatments were
patients did not complete the treatment. Thus, the patients’ equally credible to both patients and treatment providers. The
satisfaction scores, treatment credibility scores, and clinician dropout rate was low, at 2.2% for the clinical examination

TABLE 4. The Proportion of Patients in Each Treatment Group With Neurological Signs at Baseline,
End of Treatment, and 1 Year After End of Treatment
Symptom-Guided Statistical Difference Between
Exercise Group# Sham Exercise Group# Treatment Groups
Baseline
Positive SLR 94.3% 94.0% NS
Motor deficit 53.4% 50.6% NS
Disturbed sensibility 78.4% 84.3% NS
Asymmetric reflexes 53.4% 53.0% NS
Mean sum score* 2.8 2.8 NS
End of treatment
Positive SLR 43.2% 59.0% 0.05
Motor deficit 13.6% 37.3% 0.00001
Disturbed sensibility 35.2% 54.2% 0.017
Asymmetric reflexes 22.7% 26.5% NS
Mean sum score* 1.2 1.8 0.001
1-year follow-up
Positive SLR 19.3% 33.7% 0.032
Motor deficit 13.6% 30.1% 0.009
Disturbed sensibility 31.8% 48.2% 0.027
Asymmetric reflexes 19.3% 30.1% NS
Mean sum score* 0.9 1.5 0.001
*The mean number of neurological signs per patient.
SLR indicates straight leg raise.

Spine www.spinejournal.com 539

Copyright © 2012 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

BRS204492.indd 539 13/03/12 3:52 PM

 RANDOMIZED TRIAL
Figure 4. The patients’ ability to use the given information in the 2 treatment groups, at end of treatment, and at 1-year follow-up. SYMIA indicates
Symptom-guided exercises + Information + Advice to stay active group; SHAMIA, SHAM exercises + Information + Advice to stay active group.
and 1.1% for the questionnaires. The effect of this drop-
out was negligible in that intention-to-treat analysis did not
change any result. We, therefore, believe that the patients
studied are representative and can be generalized to patients
with a similar clinical profile. To enhance external validity, all
the treatment providers were members of the existing clinical
staff at the Back Centre and were not chosen for the study
because of uncommon skills. The treatment could have been
performed by any physiotherapist with current knowledge of
the treatment of LBP and sciatica.
Conservative Treatment Versus Surgery
The patients in this study would be considered to be surgical
candidates in most hospitals because 65% of them had 3 or
4 positive root compression signs and 30% had 2 positive
signs. Despite most patients having severe symptoms, it
was still possible to improve their pain and function with
active conservative treatment. During the 8-week treat-
ment period, only 3% developed symptoms so severe that
a referral to a neurosurgeon became necessary. At the end
of the treatment, 89% of patients, and at the 1-year fol-
low-up, 91% of patients, reported being better or much
better—these numbers match or surpass the results reported
by surgical interventions.4,13–15,19,32–34
It could be argued that keeping patients with such a consid-
erable number of root compression signs away from surgery
is unsafe, but nevertheless it appears to be a safe intervention
because no patients developed cauda equina syndrome or se-
vere paresis in the intervention period, and the medication
used was mild analgesics and NSAIDs only.
A common argument for providing surgery to patients
with herniated discs is the belief that surgery brings faster
relief from pain and return to work. That argument is not
supported by the results of this study because, after 8 weeks
of treatment, 74% of the symptom-guided exercise group
and 60% of the sham exercise group were back at work, and
many had been working throughout the treatment period. In
540 www.spinejournal.com
Copyright © 2012 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
BRS204492.indd 540
The Efficacy of Systematic Active Conservative Treatment • Albert and Manniche
Denmark, surgical intervention cannot match this. Patients
are usually on sick leave for 6 weeks postsurgery and sub-
sequently on sick leave for a varying number of weeks while
undergoing rehabilitation. At least in Denmark, where this
study was performed, active conservative treatment brings the
patients back to work sooner, without the risk of surgical side
effects such as spinal scar tissue. In addition, this treatment
is cheap, uses low technology, has no side effects, is easy to
perform, and results in high patient satisfaction.
How Soon Should You Operate?
All patients in this study had received conservative treatment
of variable length, which was unsuccessful before entering this
study. Most had suffered from sciatica for 1 to 3 months; how-
ever, the results of the 18% patients who had suffered from sci-
atica for 3 to 6 months were just as favorable as the results for
more acute patients. This provides food for thought because
the most commonly used reason for surgery is “unsuccessful
conservative treatment.” However, it would appear that not
all conservative treatments are the same. Some treatments are
clearly more efficacious than others. The results of Vroomen
et al11 and Rozenberg et al20 suggest that passive conserva-
tive treatment is not efficacious, whereas the results of Saal
and Saal,17 Sang-Ho et al,21 Luijsterburg et al,24 and this study
show that active conservative treatment is efficacious. Ulreich
and Kullich23 also showed in their RCT comparing active con-
servative treatment with surgery that the active conservative
treatment was more efficacious than surgery. Surgeons can be
pressured by patients for fast action if weeks go by and con-
servative treatment has not been effective, but patients can still
get much better outcomes even after months of sciatica if they
receive appropriate active conservative treatment.
Symptom-Guided Exercises Versus Sham Exercises
The 2 treatment regimens were identical, except for the type of
exercises given. Improvements in both groups were clinically
important on all measured parameters. At the baseline, the
April 2012
13/03/12 3:52 PM
 RANDOMIZED TRIAL The Efficacy of Systematic Active Conservative Treatment • Albert and Manniche
patients had more faith in the sham exercises, but even with
this advantage, the sham exercise group did not surpass the
symptom-guided exercise group on any parameter. In fact, the
results were the opposite, with the symptom-guided exercise
group improving significantly more with respect to global
improvement, sick leave, vocational status, root compression
signs, and patients’ satisfaction with information. There was
no difference concerning back-specific, and generic, function.
With respect to current leg pain, a borderline significant differ-
ence (P < 0.06) was observed at the end of the treatment but
not at 1-year follow-up. The patients in the symptom-guided
exercise group received individualized specific information
and exercises appropriate to their current specific symptoms.
The algorithm was structured and pretested so that patients
did not get random exercises at their dinicians’ discretion.
CONCLUSION
These patients had symptoms and clinical findings that would
qualify them for surgery in most hospitals, and yet they ex-
perienced highly significant improvements in all outcome
measures. Although the patients had greater faith in the sham
exercises before treatment, the symptom-guided exercise
treatment produced better outcomes. A considerable part of
this improvement is probably due to the natural healing pro-
cess. However, it is likely that the natural healing process was
given a “helping hand” by the individualized exercises and by
information that either relieved the patients’ pain or changed
the patients’ perceptions of pain. It is possible that a consid-
erable number of operations have been avoided by this safe,
low-cost, and low-technology treatment.
➢ Key Points
‰ Active conservative treatment is an efficacious treat-
ment for patients with severe sciatica.
‰ Patients who had symptoms and clinical findings
that would qualify them for surgery in most hospitals
improved greatly with active conservative treatment.
‰ Although the patients had greater faith in the sham
exercises before treatment, the symptom-guided
exercises were superior for most outcomes.
Acknowledgments
The authors thank the following staff at the Back Center, Fu-
nen, for their enthusiasm in participating in this study as treat-
ment teams:
Lars Morsø, PT; Anders Thorup, PT; Dorte Lemvig, PT, head
physiotherapists; Hanne Møller, PT; Anette Greve, RN; Betina
Møllegaard, head secretary; Hans Kroman, PT; Rikke Becker-
mann, PT; Henrik Lauritsen, DC; Hanne Ellegaard, RN; Peter
Kryger-Baggesen, DC; Jens Jacobsen, DC, head chiropractor;
Jette Sanchez, radiographer; Susanne Petersen, radiographer;
Winnie Most, radiographer.
Joan Solgaard Sorensen, MD, leading radiologist.
Spine
Copyright © 2012 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
BRS204492.indd 541
Gert Bronfort, PhD; Alan Jordan, PhD; Charlotte Lebeuf-
Yde, PhD, for scientific and editorial advice. Bente Otte-
sen, Lise Lotte Nielsen, and Lene Hofberg for secretarial
assistance.
References
1. Danish Institute for Health Technology Assessment. Low Back
Pain. Frequency, Management and Prevention From an HTA Per-
spective. Copenhagen, Denmark: Danish Institute for Health Tech-
nology Assessment; 1999.
2. Heliövaara M, Knekt P, Aromaa A. Incidence and risk factors
of herniated lumbar intervertebral disc or sciatica leading to
hospitalization. J Chronic Dis 1987;40:251–8.
3. Manninen P, Riihimäki H, Heliövaara M, et al. Incidence and
risk factors of low-back pain in middle-aged farmers. Occup Med
(Lond) 1995;45:141–6.
4. Shvartzman L, Weingarten E, Sherry H, et al. Cost-effectiveness
analysis of extended conservative therapy versus surgical interven-
tion in the management of herniated lumbar intervertebral disc.
Spine 1992;17:176–82.
5. Jensen TS, Albert HB, Soerensen JS, et al. Natural course of disc
morphology in patients with sciatica: an MRI study using a stan-
dardized qualitative classification system. Spine 2006;31:1605–12.
6. Ellenberg MR, Ross ML, Honet JC, et al. Prospective evaluation of
the course of disc herniations in patients with proven radiculopathy.
Arch Phys Med Rehabil 1993;74:3–8.
7. Weber H, Holme I, Amlie E, et al. The natural course of acute sciat-
ica with nerve root symptoms in a double-blind placebo-controlled
trial evaluating the effect of piroxicam. Spine 1993;18:1433–8.
8. Mixter W, Barr J. Rupture of the intervertebral disc with involve-
ment of the spinal canal. N Engl J Med 1934;211:210–5.
9. Keller RB, Atlas SJ, Singer DE, et al. The Maine lumbar spine study,
part I. Background and concepts. Spine 1996;21:1769–76.
10. Kogstad O, Brauti BP. Surgery or conservative therapy of herniated
lumbar disk? A long-term study [in Norwegian]. Tidsskr Nor Lae-
geforen 1978;98:1381–3.
11. Vroomen PCAJ, de Krom MCTFM, Slofstra PD, et al. Conser-
vative treatment of sciatica: a systematic review. J Spinal Disord
2000;13:463–9.
12. Weber H. Lumbar disc herniation. A controlled, prospective study
with ten years of observation. Spine 1983;8:131–40.
13. Hentzer LA. Conservative and operative treatment of disk pro-
lapse. A follow-up examination of 152 patients [in Danish]. Ugeskr
Laeger 1973;135:2258–62.
14. Atlas SJ, Keller RB, Wu YA, et al. Long-term outcome of surgi-
cal and nonsurgical management of sciatica secondary to a lumbar
disc herniation: 10 year results from the Maine lumbar spine study.
Spine 2005;30:927–35.
15. Peul WC, van Houwelingen HC, van den Hout WB, et al. Surgery
versus prolonged conservative treatment for sciatica. N Engl J Med
2007;356:2245–56.
16. Van den Hout WB, Peul WC, Koes BW. Prolonged conservative
care versus early surgery in patients with sciatica from lumbar disc
herniation: cost utility analysis alongside a randomised controlled
trial. BMJ 2008;336:1351–4.
17. Saal JA, Saal JS. Nonoperative treatment of herniated lumbar in-
tervertebral disc with radiculopathy. An outcome study. Spine
1989;14:431–7.
18. Atlas SJ, Deyo RA, Keller RB, et al. The Maine Lumbar Spine
Study, Part II. 1-year outcomes of surgical and nonsurgical manage-
ment of sciatica. Spine 1996;21:1777–86.
19. Balagué F, Nordin M, Sheikhzadeh A, et al. Recovery of severe sci-
atica. Spine 1999;24:2516–24.
20. Rozenberg S, Dubourg G, Khalifa P, et al. Efficacy of epidu-
ral steroids in low back pain and sciatica. Rev Rhum (Engl Ed)
1999;66:79–85.
21. Sang-Ho A, Park H-W, Byun W-M, et al. Comparison of clini-
cal outcome and natural morphologic changes between seques-
tered and large central extruded disc herniations. Yonsei Med J
2002;43:283–90.
www.spinejournal.com 541
13/03/12 3:52 PM
 RANDOMIZED TRIAL The Efficacy of Systematic Active Conservative Treatment • Albert and Manniche
22. Ulreich A, Kullich W. Results of a multidisciplinary rehabilitation
concept in patients with chronic lumbar syndromes [in German].
Wien Med Wochenschr 1999;149:564–6.
23. Weinstein JN, Lurie JD, Tosteson TD, et al. Surgical vs nonoperative
treatment for lumbar disc herniation. JAMA 2006;296:2451–9.
24. Luijsterburg PAJ, Verhagen AP, Ostelo RWJG, et al. Physical ther-
apy plus general practitioners’ care versus general practitioners’
care alone for sciatica: a randomised clinical trial with a 12-month
follow-up. Eur Spine J 2008;17:509–17.
25. Freire P. De undertryktes pädagogik [in Danish]. Copenhagen,
Denmark: Christians Ejlers Publishing; 1972.
26. Albert HB, Jensen AM, Dahl D, et al. Criteria validation of the Ro-
land Morris Questionnaire. A Danish translation of the international
scale for the assessment of functional level in patients with low back
pain and sciatica [in Danish]. Ugeskr Laeger 2003;165:1875–80.
27. Leclaire R, Blier F, Fortin L, et al. A cross-sectional study compar-
ing the Oswestry and Roland-Morris Functional Disability scales in
two populations of patients with low back pain of different levels
of severity. Spine 1997;22:68–71.
542 www.spinejournal.com
Copyright © 2012 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
BRS204492.indd 542
28. Patrick DL, Deyo RA, Atlas SJ, et al. Assessing health-re-
lated quality of life in patients with sciatica. Spine 1995;20:
1899–908.
29. Ostelo RWJG, Deyo RA, Stratford P, et al. Interpreting change scores
for pain and functional status in low back pain. Spine 2008;33:
90–4.
30. Manniche C, Asmussen K, Lauritsen B, et al. Low back pain rating
scale: validation of a tool for assessment of low back pain. Pain
1994;57:317–26.
31. Møller Pedersen K, Wittrup-Jensen K, Brooks R, et al. Værdisæt-
ning af sundhed. Odense, Denmark: University of Southern Den-
mark Publishing; 2003.
32. Gibson JN, Grant IC, Waddell G. Surgery for lumbar disc prolapse.
Cochrane Database Syst Rev. 2000;(3):CD001350.
33. Dubourg G, Valls-Bellec I, Duplan B, et al. A pilot study on
the recovery from paresis after lumbar disc herniation. Spine
2002;27:1426–32.
34. Nykvist F, Hurme M, Alaranta H, et al. Severe sciatica: a 13-year
follow-up of 342 patients. Eur Spine J 1995;4:335–8.
April 2012
13/03/12 3:52 PM