Musculoskeletal disorders

..................................................

Herniated lumbar disc

Search date June 2010
Jo Jordan, Kika Konstantinou, and John O'Dowd

ABSTRACT
INTRODUCTION: Herniated lumbar disc is a displacement of disc material (nucleus pulposus or annulus fibrosis) beyond the intervertebral
disc space. The highest prevalence is among people aged 30 to 50 years, with a male to female ratio of 2:1. There is little evidence to suggest
that drug treatments are effective in treating herniated disc. METHODS AND OUTCOMES: We conducted a systematic review and aimed
to answer the following clinical questions: What are the effects of drug treatments, non-drug treatments, and surgery for herniated lumbar
disc? We searched: Medline, Embase, The Cochrane Library, and other important databases up to June 2010 (Clinical Evidence reviews
are updated periodically; please check our website for the most up-to-date version of this review). We included harms alerts from relevant
organisations such as the US Food and Drug Administration (FDA) and the UK Medicines and Healthcare products Regulatory Agency
(MHRA). RESULTS: We found 37 systematic reviews, RCTs, or observational studies that met our inclusion criteria. We performed a
GRADE evaluation of the quality of evidence for interventions. CONCLUSIONS: In this systematic review, we present information relating
to the effectiveness and safety of the following interventions: acupuncture, advice to stay active, analgesics, antidepressants, bed rest,
corticosteroids (epidural injections), cytokine inhibitors (infliximab), discectomy (automated percutaneous, laser, microdiscectomy, standard),
exercise therapy, heat, ice, massage, muscle relaxants, non-steroidal anti-inflammatory drugs (NSAIDs), percutaneous disc decompression,
spinal manipulation, and traction.

QUESTIONS
What are the effects of drug treatments for herniated lumbar disc?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
What are the effects of non-drug treatments for herniated lumbar disc?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
What are the effects of surgery for herniated lumbar disc?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

INTERVENTIONS
DRUG TREATMENTS Massage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Unknown effectiveness
Unlikely to be beneficial
Analgesics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Bed rest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Antidepressants . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Traction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
Corticosteroids (epidural injections) . . . . . . . . . . . . . 4
Cytokine inhibitors . . . . . . . . . . . . . . . . . . . . . . . . . 13
SURGERY
Muscle relaxants . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Likely to be beneficial
Microdiscectomy (as effective as standard discectomy)
Unlikely to be beneficial
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
NSAIDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Standard discectomy (short-term benefit) . . . . . . . 51

NON-DRUG TREATMENTS
Unknown effectiveness
Likely to be beneficial
Automated percutaneous discectomy . . . . . . . . . . 56
Spinal manipulation . . . . . . . . . . . . . . . . . . . . . . . . 20
Laser discectomy . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Percutaneous disc decompression . . . . . . . . . . . . . 58
Unknown effectiveness
Acupuncture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Covered elsewhere in Clinical Evidence
Advice to stay active . . . . . . . . . . . . . . . . . . . . . . . . 28 Chronic low back pain
Exercise therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Non-specific acute low back pain
Heat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
Ice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

Key points

• Herniated lumbar disc is a displacement of disc material (nucleus pulposus or annulus fibrosis) beyond the inter-
vertebral disc space.
The highest prevalence is among people aged 30 to 50 years, with a male to female ratio of 2:1.
• There is little high-quality evidence to suggest that drug treatments are effective in treating herniated disc.
NSAIDs and cytokine inhibitors do not seem to improve symptoms of sciatica caused by disc herniation.
We found no RCT evidence examining the effects of analgesics, antidepressants, or muscle relaxants in people
with herniated disc.

© BMJ Publishing Group Ltd 2011. All rights reserved. .................... 1 .................... Clinical Evidence 2011;06:1118
 Musculoskeletal disorders
Herniated lumbar disc
We found several RCTs that assessed a range of different measures of symptom improvement and found incon-
sistent results, so we are unable to draw conclusions on effects of epidural injections of corticosteroids.
• With regard to non-drug treatments, spinal manipulation seems more effective at relieving local or radiating pain
in people with acute back pain and sciatica with disc protrusion compared with sham manipulation, although concerns
exist regarding possible further herniation from spinal manipulation in people who are surgical candidates.
Neither bed rest nor traction seem effective in treating people with sciatica caused by disc herniation.
We found insufficient RCT evidence about advice to stay active, acupuncture, massage, exercise, heat, or ice
to judge their efficacy in treating people with herniated disc.
• About 10% of people have sufficient pain after 6 weeks for surgery to become a consideration.
Standard discectomy and microdiscectomy seem to increase self-reported improvement to a similar extent.
We found insufficient evidence judging the effects of automated percutaneous discectomy, laser discectomy, or
percutaneous disc decompression.

DEFINITION Herniated lumbar disc is a displacement of disc material (nucleus pulposus or annulus fibrosis)
[1]
beyond the intervertebral disc space. The diagnosis can be confirmed by radiological examination.
[2] [3]
However, MRI findings of herniated disc are not always accompanied by clinical symptoms.
This review covers treatment of people with clinical symptoms relating to confirmed or suspected
disc herniation. It does not include treatment of people with spinal cord compression, or people
with cauda equina syndrome, which require emergency intervention. The management of non-
specific acute low back pain and chronic low back pain are covered elsewhere in Clinical Evidence.

INCIDENCE/ The prevalence of symptomatic herniated lumbar disc is about 1% to 3% in Finland and Italy, de-
[4] [5]
PREVALENCE pending on age and sex. The highest prevalence is among people aged 30 to 50 years, with
[6]
a male to female ratio of 2:1. In people aged 25 to 55 years, about 95% of herniated discs occur
at the lower lumbar spine (L4/5 and L5/S1 level); disc herniation above this level is more common
[7] [8]
in people aged over 55 years.

AETIOLOGY/ Radiographical evidence of disc herniation does not reliably predict low back pain in the future, or
RISK FACTORS correlate with symptoms; 19% to 27% of people without symptoms have disc herniation on imaging.
[2] [9]
Risk factors for disc herniation include smoking (OR 1.7, 95% CI 1.0 to 2.5), weight-bearing
sports (e.g., weight lifting, hammer throw), and certain work activities, such as repeated lifting.
Driving a motor vehicle has been suggested to be a risk factor for disc herniation, although evidence
[6] [10] [11]
is inconclusive (OR 1.7, 95% CI 0.2 to 2.7).

PROGNOSIS The natural history of disc herniation is difficult to determine, because most people take some form
[6]
of treatment for their back pain, and a formal diagnosis is not always made. Clinical improvement
is usual in most people, and only about 10% of people still have sufficient pain after 6 weeks to
consider surgery. Sequential MRIs have shown that the herniated portion of the disc tends to
[12]
regress over time, with partial to complete resolution after 6 months in two-thirds of people.

AIMS OF To relieve pain; increase mobility and function; improve quality of life; and minimise adverse effects
INTERVENTION of treatments.

OUTCOMES Primary outcomes: pain, including global symptom relief; functional improvement; patient perception
of improvement; quality of life; and adverse effects of treatment. Secondary outcomes: return to
work; use of analgesia; and duration of hospital admission.

METHODS Clinical Evidence search and appraisal June 2010. The following databases were used to identify
studies for this systematic review: Medline 1966 to June 2010, Embase 1980 to June 2010, and
The Cochrane Database of Systematic Reviews, May 2010 (online; 1966 to date of issue). An ad-
ditional search within The Cochrane Library was carried out for the Database of Abstracts of Reviews
of Effects (DARE) and Health Technology Assessment (HTA). We also searched for retractions of
studies included in the review. Abstracts of the studies retrieved from the initial search were assessed
by an information specialist. Selected studies were then sent to the contributor for additional as-
sessment, using predetermined criteria to identify relevant studies. Study design criteria for inclusion
in this review were: published systematic reviews of RCTs and RCTs in any language, at least
single blinded, and containing >20 people of whom >80% were followed up. There was no minimum
length of follow-up required to include trials. We excluded all trials described as "open", "open label",
or not blinded unless blinding was impossible. We included systematic reviews of RCTs and RCTs
where harms of an included intervention were studied applying the same study design criteria for
inclusion as we did for benefits. In addition we use a regular surveillance protocol to capture harms
alerts from organisations such as the FDA and the MHRA, which are added to the reviews as re-

© BMJ Publishing Group Ltd 2011. All rights reserved. ........................................................... 2

 Musculoskeletal disorders
Herniated lumbar disc
[13]
quired. The contributors used confidence interval analysis and chi-square test analysis from
[14]
PEPI version 4.0 in their own calculations, which are presented in the review. To aid readability
of the numerical data in our reviews, we round many percentages to the nearest whole number.
Readers should be aware of this when relating percentages to summary statistics such as relative
risks (RRs) and odds ratios (ORs). We have performed a GRADE evaluation of the quality of evi-
dence for interventions included in this review (see table, p 62 ). The categorisation of the quality
of the evidence (high, moderate, low, or very low) reflects the quality of evidence available for our
chosen outcomes in our defined populations of interest. These categorisations are not necessarily
a reflection of the overall methodological quality of any individual study, because the Clinical Evi-
dence population and outcome of choice may represent only a small subset of the total outcomes
reported, and population included, in any individual trial. For further details of how we perform the
GRADE evaluation and the scoring system we use, please see our website (www.clinicalevi-
dence.com).

QUESTION What are the effects of drug treatments for herniated lumbar disc?

OPTION ANALGESICS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

• For GRADE evaluation of interventions for Herniated lumbar disc, see table, p 62 .
• We found no direct information from RCTs about analgesics in the treatment of people with symptomatic herni-
ated lumbar disc.

Benefits and harms

Analgesics:
We found no systematic review or RCTs on the use of analgesics for treatment of people with symptomatic herniated
lumbar disc.

-
-
-
Further information on studies

-
-
Comment: None.

OPTION ANTIDEPRESSANTS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

• For GRADE evaluation of interventions for Herniated lumbar disc, see table, p 62 .
• We found no direct information from RCTs about antidepressants in the treatment of people with symptomatic
herniated lumbar disc.

Benefits and harms

Antidepressants:
We found no systematic review or RCTs on the use of antidepressants for treatment of people with symptomatic
herniated lumbar disc.

-
-
-
Further information on studies

-
-
Comment: None.

© BMJ Publishing Group Ltd 2011. All rights reserved. ........................................................... 3

 Musculoskeletal disorders
Herniated lumbar disc
OPTION CORTICOSTEROIDS (EPIDURAL INJECTIONS). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

• For GRADE evaluation of interventions for Herniated lumbar disc, see table, p 62 .
• We found several RCTs, which assessed a range of different measures of symptom improvement and found in-
consistent results, so we are unable to draw conclusions on the effects of epidural injections of corticosteroids.

Benefits and harms

Epidural corticosteroid injections versus no epidural corticosteroid injection:
We found 5 systematic reviews assessing epidural corticosteroid injections in people with radicular pain caused by
[15] [16] [17] [18] [19]
disc herniation. The first review (search date 1998, 4 RCTs, 332 people) performed a meta-
[15]
analysis assessing patient perception of improvement, which we report below. The second systematic review
(search date 2003, 3 RCTs, none included in the first review, 264 people) did not perform a meta-analysis because
[16]
of heterogeneity among trial parameters, so we report results from each RCT it identified separately. The third
systematic review (search date 2008, 2 RCTs, 80 people) of caudal epidural injections identified one additional RCT
[17]
not included in previous reviews and did not include a meta-analysis, so we also report this RCT separately. The
fourth systematic review (search date 2008, 2 RCTs, 215 people) of transforaminal epidural injections did not find
[18]
any additional RCTs and did not include a meta-analysis, so we do not report it further. The fifth systematic review
(search date 2008, 3 RCTs, 437 people) of lumbar interlaminar epidural injections also did not include a meta-anal-
ysis. It included two RCTs identified by the first review but reported on different outcomes and included one further
[19]
RCT not identified by any of the other reviews, so we report all three RCTs separately. We found one additional
[20] [21]
RCT not included by any of the reviews and one subsequent RCT, which we also report below.

-
Pain
Compared with no epidural corticosteroid Epidural corticosteroids may be more effective at improving limb pain at
2 weeks, but may be no more effective after more than 2 weeks in people with disc herniation (low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Pain
[16]
49 people with radi- Proportion of people with Significance not assessed
ologically con- symptom relief , 3 months
Systematic Randomisation method not report-
firmed disc hernia-
review 54% with triamcinolone interlami- ed
tion
nar perineural injection
Data from 1 RCT
40% with placebo (saline) inter-
laminar perineural injection plus
intramuscular triamcinolone
Absolute numbers not reported
Placebo group received triamci-
nolone 10 mg intramuscularly

[16]
160 people with Proportion of people with Reported as not significant
lower-limb pain symptom relief , 12 months
Systematic P value not reported
caused by con-
review 65% with corticosteroid injections Not significant
firmed disc hernia-
tion 65% with saline placebo injection
Data from 1 RCT Absolute numbers not reported

[17]
23 people with Proportion of people with im- Reported as significant in favour
nerve root compro- provement in back and leg pain of corticosteroid injection
Systematic
mise (unspecified) , 4 weeks
review No further data reported
Data from 1 RCT with caudal corticosteroid injec-
tion of 25 mL triamcinolone ace-
tonide 80 mg with or without 0.5%
procaine hydrochloride
with placebo (25 mL saline injec-
tion)
Absolute results not reported
2 caudal injections were given,
the first after admission to the tri-
al, and the second after 2 weeks

© BMJ Publishing Group Ltd 2011. All rights reserved. ........................................................... 4

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
[17]
23 people with Proportion of people with im- Reported as no significant differ-
nerve root compro- provement in back and leg pain ence between groups at 12
Systematic
mise (unspecified) , 12 months months
review
Data from 1 RCT with caudal corticosteroid injec- No further data reported
tion of 25 mL triamcinolone ace-
tonide 80 mg with or without 0.5%
procaine hydrochloride
Not significant
with placebo (25 mL saline injec-
tion)
Absolute results not reported
2 caudal injections were given,
the first after admission to the tri-
al, and the second after 2 weeks

[19]
228 people with Proportion of people with im- Reported as no significant differ-
unilateral sciatica, provement in leg pain (unspec- ence between groups
Systematic
possibly caused by ified) measured by visual ana-
review No further data reported by re-
disc herniation logue scale (VAS) , 3 weeks
view
Data from 1 RCT with triamcinolone 80 mg plus
Not significant
10 mL bupivacaine 0.25%
with 2 mL normal saline
Absolute results not reported
Interlaminar epidural injection

[19]
228 people with Proportion of people with im- P <0.01
unilateral sciatica, provement in leg pain (unspec-
Systematic
possibly caused by ified) measured by Likert scale
review
disc herniation , 3 weeks
triamcinolone
Data from 1 RCT 61% with triamcinolone 80 mg
80 mg plus 10 mL
plus 10 mL bupivacaine 0.25%
bupivacaine 0.25%
40% with 2 mL normal saline
Absolute numbers not reported
Interlaminar epidural injection

[19]
228 people with Proportion of people with im- Reported as no significant differ-
unilateral sciatica, provement in leg pain (unspec- ence between groups
Systematic
possibly caused by ified) measured by VAS , 6
review No further data reported by re-
disc herniation weeks
view
Data from 1 RCT with triamcinolone 80 mg plus Not significant
10 mL bupivacaine 0.25%
with 2 mL normal saline
Absolute results not reported

[19]
228 people with Proportion of people with im- Reported as no significant differ-
unilateral sciatica, provement in leg pain (unspec- ence between groups
Systematic
possibly caused by ified) measured by Likert scale
review No further data reported by re-
disc herniation , 6 weeks
view
Data from 1 RCT with triamcinolone 80 mg plus Not significant
10 mL bupivacaine 0.25%
with 2 mL normal saline
Absolute results not reported

[19]
158 people with Improvement in leg pain (un- P = 0.03
sciatica caused by specified) , 6 weeks
Systematic
herniated nucleus
review with methylprednisolone acetate
pulposus methylpred-
(80 mg and 8 mL of isotonic
nisolone acetate
Data from 1 RCT saline)
(80 mg and 8 mL
with 1 mL isotonic saline of isotonic saline)
Absolute results not reported
Interlaminar epidural injection

© BMJ Publishing Group Ltd 2011. All rights reserved. ........................................................... 5

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
Greater improvement with
methylprednisolone acetate
(80 mg and 8 mL of isotonic
saline) than with 1 mL isotonic
saline

[19]
158 people with Improvement in leg pain (un- Reported as no significant differ-
sciatica caused by specified) , 3 months ences between groups
Systematic
herniated nucleus
review with methylprednisolone acetate No further data reported
pulposus
(80 mg and 8 mL of isotonic
Data from 1 RCT saline) Not significant
with 1 mL isotonic saline
Absolute results not reported
Interlaminar epidural injection

[19]
51 people with Pain (unspecified) , 3 months Reported as no significant differ-
lumbar root com- ences between groups
Systematic with 80 mg methylprednisolone
pression document-
review (2 mL) No further data reported
ed by neurological
deficit and abnor- with 2 mL normal saline Not significant
mality noted on
myelography Absolute results not reported

Data from 1 RCT Interlaminar epidural injection

[19]
151 people with Pain (unspecified) , 14 months Reported as no significant differ-
lumbar root com- ence between groups
Systematic with 80 mg methylprednisolone
pression document-
review (2 mL) No further data reported
ed by neurological
deficit and abnor- with 2 mL normal saline Not significant
mality noted on
myelography Absolute results not reported

Data from 1 RCT

[20]
85 people with sci- Mean change in pain scores Mean difference –5.1
atica caused by from baseline measured by
RCT 95% CI –18.7 to +8.4
herniated disc unspecified VAS , 35 days
–30.3 mm with epidural corticos-
teroid injections (2 mL pred- Not significant
nisolone acetate at 2-day inter-
vals for a total of 3 injections)
–25.2 mm with placebo (2 mL
isotonic saline injection)

[21]
76 people with leg Improvement in leg pain mea- Significance not assessed
and back pain sured by VAS score , 3 months
RCT
caused by herniat-
mean change of 27.4 with
ed disc
methylprednisolone 40 mg plus
local anaesthetic
mean change of 24.3 with local
anaesthetic alone
The local anaesthetic used was
2 mL bupivacaine 0.25%

[21]
124 people with Improvement in back pain P = 0.57
leg and back pain measured by VAS score , 3
RCT
caused by herniat- months
ed disc (76 people)
mean change of 6.9 with methyl-
or spinal stenosis
prednisolone 40 mg plus local
(48 people)
anaesthetic
Not significant
mean change of 9.9 with local
anaesthetic alone
Baseline range 34.4 to 38.1
The local anaesthetic used was
2 mL bupivacaine 0.25%

© BMJ Publishing Group Ltd 2011. All rights reserved. ........................................................... 6

 Musculoskeletal disorders
Herniated lumbar disc
-
[15]
No data from the following reference on this outcome.

-
Functional improvement
Compared with no epidural corticosteroid Epidural corticosteroids may be no more effective in the longer term at
improving disability, as measured by the Roland Morris Disability Questionnaire and Oswestry Disability Index scores,
or functional outcomes such as straight leg raising and lumbar flexion, in people with disc herniation (moderate-
quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Functional improvement
[20]
85 people with sci- Roland Morris Disability Ques- ARR –2.1
atica caused by tionnaire score (mean change
RCT 95% CI –5.0 to +0.8
herniated disc from baseline) , 35 days
–5.3 with epidural corticosteroid
injections (2 mL prednisolone Not significant
acetate at 2-day intervals for a
total of 3 injections)
–3.2 with placebo (2 mL isotonic
saline injection)

[19]
228 people with Oswestry Disability Index , 3 Reported as significant differ-
unilateral sciatica, weeks ence; see further information on
Systematic
possibly caused by studies
review with triamcinolone 80 mg plus
disc herniation
10 mL bupivacaine 0.25% P value not reported
Data from 1 RCT
with 2 mL normal saline triamcinolone
Absolute results not reported 80 mg plus 10 mL
bupivacaine 0.25%
Interlaminar epidural injection
Greater improvement with triam-
cinolone 80 mg plus 10 mL bupi-
vacaine 0.25% than with 2 mL
normal saline

[19]
228 people with Oswestry Disability Index , 6 Reported as no significant differ-
unilateral sciatica, weeks ence; see further information on
Systematic
possibly caused by studies
review with triamcinolone 80 mg plus
disc herniation
10 mL bupivacaine 0.25% P value not reported
Not significant
Data from 1 RCT
with 2 mL normal saline
Absolute results not reported
Interlaminar epidural injection

[19]
158 people with Oswestry Disability Index , 3 Significance not assessed
sciatica due to her- weeks
Systematic
niated nucleus pul-
review with methylprednisolone acetate
posus
(80 mg and 8 mL of isotonic
Data from 1 RCT saline)
with 1 mL isotonic saline
Absolute results not reported
Interlaminar epidural injection
Slightly greater improvement with
methylprednisolone acetate
(80 mg and 8 mL of isotonic
saline) than with isotonic saline
1 mL

[19]
158 people with Oswestry Disability Index , 3 Reported as not significant
sciatica due to her- months
Systematic No further data reported
niated nucleus pul-
review with methylprednisolone acetate
posus Not significant
(80 mg and 8 mL of isotonic
Data from 1 RCT saline)
with 1 mL isotonic saline

© BMJ Publishing Group Ltd 2011. All rights reserved. ........................................................... 7

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
Absolute results not reported

[19]
228 people with Oswestry Disability Index 75% Significance not assessed; see
unilateral sciatica, improvement in scores , 52 further information on studies
Systematic
possibly caused by weeks
review
disc herniation
32.5% with triamcinolone 80 mg
Data from 1 RCT plus 10 mL bupivacaine 0.25%
29.6% with 2 mL normal saline
Interlaminar epidural injection

[21]
76 people with leg Mean change in Oswestry Dis- Significance not assessed
and back pain ability Index , 3 months
RCT
caused by herniat-
13.6 with methylprednisolone
ed disc
40 mg plus local anaesthetic
3.8 with local anaesthetic alone
Baseline values were 43.4 (in-
terquartile range [IQR] 32–54) for
methylprednisolone plus local
anaesthetic and 46.6 (IQR
34–58) for local anaesthetic alone
The local anaesthetic used was
2 mL bupivacaine 0.25%

-
[15] [16] [17]
No data from the following reference on this outcome.

-
Patient perception of improvement
Compared with no epidural corticosteroid Epidural corticosteroids may be more effective at increasing subjective
global improvement and patient satisfaction in the short term only (2 weeks), but may be no more effective in the
longer term (after 2 weeks) in people with disc herniation (low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Patient perception of improvement
[15]
332 people Proportion of people with self- OR 2.2
perceived global improvement
Systematic 4 RCTs in this 95% CI 1.0 to 4.7
(which was not defined) , 2 to
review analysis
30 days
73/160 (46%) with epidural corti-
costeroid injections
56/172 (33%) with placebo Not significant

Corticosteroids used were 8 mL

methylprednisolone 80 mg; 2 mL
methylprednisolone 80 mg;
10 mL methylprednisolone
80 mg; and 2 mL methylpred-
nisolone acetate 80 mg

[20]
85 people with sci- People rating improvement as P = 0.91
atica caused by "recovery" or "marked improve-
RCT
herniated disc ment" , 35 days
21/43 (49%) with epidural corti-
costeroid injections (2 mL pred- Not significant
nisolone acetate at 2-day inter-
vals for a total of 3 injections)
20/42 (48%) with placebo (2 mL
isotonic saline injection)

-
[16] [19] [17] [21]
No data from the following reference on this outcome.

-
© BMJ Publishing Group Ltd 2011. All rights reserved. ........................................................... 8
 Musculoskeletal disorders
Herniated lumbar disc
Need for surgery
Compared with no epidural corticosteroid We don't know if epidural corticosteroid injection is more effective at reducing
the need for surgery in the short term (very low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Need for surgery
[19]
158 people with Proportion having back Reported as not significant
sciatica caused by surgery , 12 months
Systematic No further data reported
herniated nucleus
review 26% with methylprednisolone
pulposus
acetate (80 mg and 8 mL of iso-
Data from 1 RCT tonic saline) Not significant
25% with 1 mL isotonic saline
Absolute numbers not reported
Interlaminar epidural injection

[16]
55 people for Proportion of people having RR 0.43
whom 6 weeks of surgery , end of treatment peri-
RCT 95% CI 0.23 to 0.82
physiotherapy (un- od
defined), oral use NNT 3
8/28 (29%) with transforaminal
of NSAIDs, and
corticosteroid plus anaesthetic 95% CI 2 to 6 transforaminal cor-
bracing had failed
18/27 (67%) with injections of Contributors' own calculations ticosteroid plus
anaesthetic alone anaesthetic

The corticosteroid group received

up to 4 injections of 1 mL be-
tamethasone (6 mg/mL) plus
1 mL bupivacaine 0.25%

-
[15] [17] [20] [21]
No data from the following reference on this outcome.

-
Adverse effects

-
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
Adverse effects
[15]
332 people Adverse effects , 2 to 30 days
Systematic 4 RCTs in this with epidural corticosteroid injec-
review analysis tions
with placebo
Absolute results not reported
No serious adverse effects were
reported in the RCTs identified
by the first systematic review, al-
though 26 people complained of
transient headache or transient
increase in sciatic pain

[16]
264 people Adverse effects
RCT 3 RCTs in this with epidural corticosteroid injec-
analysis tion
with placebo injection
The review noted a 1.9% inci-
dence of headache with epidural
injections in one RCT, and a
retroperitoneal haematoma in one
person having anticoagulation
treatment in another RCT

© BMJ Publishing Group Ltd 2011. All rights reserved. ........................................................... 9

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
[20]
85 people with sci- Clinically important adverse P = 0.68
atica caused by effects , 35 days
RCT
herniated disc
2/43 (5%) with epidural corticos-
teroid injections (2 mL pred-
nisolone acetate at 2-day inter-
vals for a total of 3 injections)
Not significant
3/42 (7%) with placebo (2 mL
isotonic saline injection)
The RCT reported that headache
occurred in two people in each
group, and thoracic pain in one
person with control

-
[17] [19] [21]
No data from the following reference on this outcome.

-
-
Epidural corticosteroid plus conservative non-operative treatment versus conservative treatment alone:
[22]
We found one RCT.

-
Pain
Epidural corticosteroids plus conservative non-operative treatment compared with conservative treatment only
Epidural corticosteroids plus conservative non-operative treatment may be no more effective at 6 weeks and 6 months
at improving pain scores in people with disc herniation (low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Pain
[22]
36 people with disc Pain scores (visual analogue P = 0.18
herniation con- scale: 0 = no pain, 100 = most
RCT The RCT also found no signifi-
firmed by MRI pain possible) , 6 months
cant difference at 6 weeks
32.9 (range 0–85) with epidural
corticosteroid plus conservative
non-operative treatment
39.2 (range 0–100) with conser-
vative treatment alone
The corticosteroid group received
three injections of methylpred- Not significant
nisolone 100 mg in 10 mL bupiva-
caine 0.25% during the first 14
days in hospital
Conservative treatment involved
initial bed rest and analgesia fol-
lowed by graded rehabilitation
(including hydrotherapy, elec-
troanalgesia, and postural exer-
cise classes) followed by physio-
therapy

-
Functional improvement
Epidural corticosteroids plus conservative non-operative treatment compared with conservative treatment only
Epidural corticosteroids plus conservative non-operative treatment may be no more effective at 6 months at improving
mobility scores in people with disc herniation (low-quality evidence).

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 10

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
Functional improvement
[22]
36 people with disc Hannover Functional Ability P = 0.15
herniation con- Questionnaire from 0% (lowest
RCT
firmed by MRI mobility) to 100% (highest mo-
bility) , 6 months
61.8 (range 25–83) with epidural
corticosteroid plus conservative
non-operative treatment
57.2 (range 17–83) with conser-
vative treatment alone
The corticosteroid group received
Not significant
three injections of methylpred-
nisolone 100 mg in 10 mL bupiva-
caine 0.25% during the first 14
days in hospital
Conservative treatment involved
initial bed rest and analgesia fol-
lowed by graded rehabilitation
(including hydrotherapy, elec-
troanalgesia, and postural exer-
cise classes) followed by physio-
therapy

[22]
36 people with disc People returning to work , 6 RR 1.19
herniation con- months
RCT 95% CI 0.75 to 1.33
firmed by MRI
15/17 (88%) with epidural corti-
costeroid plus conservative non-
operative treatment
14/19 (74%) with conservative
treatment alone
The corticosteroid group received
three injections of methylpred-
Not significant
nisolone 100 mg in 10 mL bupiva-
caine 0.25% during the first 14
days in hospital
Conservative treatment involved
initial bed rest and analgesia fol-
lowed by graded rehabilitation
(including hydrotherapy, elec-
troanalgesia, and postural exer-
cise classes) followed by physio-
therapy

-
Need for surgery
Epidural corticosteroids plus conservative non-operative treatment compared with conservative treatment only
Epidural corticosteroids plus conservative non-operative treatment may be no more effective at 6 months at reducing
the need for surgery (low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Need for surgery
[22]
36 people with disc Proportion of people needing RR 0.56
herniation con- back surgery , 6 months
RCT 95% CI 0.09 to 2.17
firmed by MRI
2/17 (12%) with epidural corticos-
Contributors' own calculations
teroid plus conservative non-op-
erative treatment Reported as not significant by
original RCT
4/19 (21%) with conservative Not significant
treatment alone
The corticosteroid group received
three injections of methylpred-
nisolone 100 mg in 10 mL bupiva-
caine 0.25% during the first 14
days in hospital

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 11

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
Conservative treatment involved
initial bed rest and analgesia fol-
lowed by graded rehabilitation
(including hydrotherapy, elec-
troanalgesia, and postural exer-
cise classes) followed by physio-
therapy

-
Adverse effects

-
-
[22]
No data from the following reference on this outcome.

-
-
Epidural corticosteroid injection versus discectomy:
[23] [24]
We found one systematic review (search date 2007, 1 RCT ) comparing epidural injections versus surgery.

-
Pain
Compared with standard discectomy Epidural corticosteroid injections may be less effective at 1 to 3 months at im-
proving leg pain in people with lumbar disc herniation (very low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Pain
[24]
100 people with Difference in pain on 11-point P = 0.001
lumbar disc hernia- visual analogue scale , 1 to 3
RCT The difference between treat-
tion >25% of cross- months
ments was not sustained at 2 to
sectional area of
with epidural corticosteroid injec- 3 years' follow-up (results present-
spinal canal, who
tions (betamethasone 10–15 mg, ed graphically; see further infor-
had 6 weeks of un-
1 week apart up to 3 times until mation on studies below)
successful non-in-
successful) discectomy
vasive treatment
(physiotherapy, with discectomy (no further de-
chiropractic treat- tails reported)
ment, rest, analge-
sia, or a combina- Absolute results reported graphi-
tion) cally
[23]
In review

-
Functional improvement
Compared with standard discectomy Epidural corticosteroid injections may be less effective at 1 to 3 months at im-
proving Oswestry Disability Index scores in people with lumbar disc herniation (very low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Functional improvement
[24]
100 people with Oswestry Disability Index P = 0.015
lumbar disc hernia- score , 1 to 3 months
RCT The difference between treat-
tion >25% cross-
with epidural corticosteroid injec- ments was not sustained at 2 to
sectional area of
tions (betamethasone 10–15 mg, 3 years' follow-up (results present-
spinal canal, who
1 week apart up to 3 times until ed graphically; see further infor-
had 6 weeks of un- discectomy
successful) mation on studies below)
successful non-in-
vasive treatment with discectomy (no further de-
(physiotherapy, tails reported)
chiropractic treat-
ment, rest, analge- Absolute results reported graphi-
cally

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 12

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
sia, or a combina-
tion)
[23]
In review

-
Adverse effects

-
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
Adverse effects
[24]
100 people with Adverse effects , 1 to 3 months The RCT found that 2/50 (4%)
lumbar disc hernia- people in the epidural group had
RCT with epidural corticosteroid injec-
tion >25% cross- an incidental dural puncture, and
tions (betamethasone 10–15 mg,
sectional area of 3/50 (6%) people had recurrent
1 week apart up to 3 times until
spinal canal, who disc herniation for 2 to 3 years'
successful)
had 6 weeks of un- follow-up period
successful non-in- with discectomy (no further de-
vasive treatment tails reported) discectomy
(physiotherapy,
chiropractic treat- Absolute results reported graphi-
ment, rest, analge- cally
sia, or a combina-
tion)
[23]
In review

-
-
-
Further information on studies
[16]
The RCT also reported that corticosteroid injections significantly improved subjective limb pain, straight leg
raising, lumbar flexion, and patient satisfaction in the short term at 2 weeks, but not after 2 weeks (data not re-
ported).
[17]
The additional RCT also reported a significant improvement in straight leg raise at both 4 weeks and 12 months.
[19]
This systematic review reports on a double-blinded RCT with 228 participants in which the treatment group re-
ceived an epidural injection of triamcinolone 80 mg plus 10 mL bupivacaine 0.25% and the placebo group received
an epidural injection of normal saline. The RCT found that by 6 weeks the benefits of epidural corticosteroids
were lost, and at 52 weeks, improvement in symptoms was 33% in the treatment group and 30% in the placebo
group, an improvement that the authors of the systematic review conclude was probably related to the natural
course of the disease.
[24]
The RCT allowed the 27 people in whom the epidural had failed to improve their symptoms (self-assessment)
to receive discectomy. This group was analysed as failures for the epidural corticosteroid injections, and also
as a separate subgroup.Two further people in each group who completely crossed over to receive other treatment
were analysed according to the intervention they received.There seemed to be multiple hypothesis tests without
mention of adjusting the analysis to account for this. Also, no attempt was made to blind the measurement of
outcomes. These results should therefore be interpreted with caution.

-
-
Comment: None.

OPTION CYTOKINE INHIBITORS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

• For GRADE evaluation of interventions for Herniated lumbar disc, see table, p 62 .
• Cytokine inhibitors do not seem to improve symptoms of sciatica caused by disc herniation.
• A drug safety alert has been issued by the FDA on the risk of clinically significant liver injury associated with
natalizumab.

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 13

 Musculoskeletal disorders
Herniated lumbar disc
Benefits and harms
Infliximab versus placebo:
[25]
We found one RCT comparing a cytokine inhibitor (infliximab) versus placebo (saline infusion over 2 hours).

-
Pain
Compared with placebo Infliximab seems no more effective at 12 weeks or 12 months at improving leg or back pain
scores in people with sciatic pain caused by herniated disc (moderate-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Pain
[25]
41 people with Median reduction in leg pain Mean difference –7 mm
acute or subacute score (rated on a 100-mm visu-
RCT 95% CI −21 mm to +31 mm
(2–12 weeks' dura- al analogue scale [VAS], details
tion) sciatic pain, not reported) , 12 weeks P = 0.77
caused by herniat-
43 mm with infliximab (single iv
ed disc confirmed
infusion of 5 mg/kg over 2 hours)
by MRI Not significant
50 mm with placebo (saline infu-
Participants had to
sion over 2 hours)
be eligible for
surgery, and were
screened for tuber-
culosis and other
infections

[25]
41 people with Median reduction in leg pain Mean difference –6 mm
acute or subacute score (rated on a 100-mm VAS,
RCT 95% CI −30 mm to +32 mm
(2–12 weeks' dura- details not reported) , 1 year
tion) sciatic pain, P = 0.98
38 mm with infliximab (single iv
caused by herniat-
infusion of 5 mg/kg over 2 hours)
ed disc confirmed
by MRI 44 mm with placebo (saline infu- Not significant
sion over 2 hours)
Participants had to
be eligible for
surgery, and were
screened for tuber-
culosis and other
infections

[25]
41 people with Median reduction in back pain Mean difference +8 mm
acute or subacute score (rated on a 100-mm VAS,
RCT 95% CI –19 mm to +16 mm
(2–12 weeks' dura- details not reported) , 12 weeks
tion) sciatic pain, P = 0.93
12 mm with infliximab (single iv
caused by herniat-
infusion of 5 mg/kg over 2 hours)
ed disc confirmed
by MRI 4 mm with placebo (saline infu- Not significant
sion over 2 hours)
Participants had to
be eligible for
surgery, and were
screened for tuber-
culosis and other
infections

[25]
41 people with Median reduction in back pain Mean difference –4 mm
acute or subacute score (rated on a 100-mm VAS,
RCT 95% CI −38 mm to +18 mm
(2–12 weeks' dura- details not reported) , 12
tion) sciatic pain, months P = 0.48
caused by herniat-
13 mm with infliximab (single iv
ed disc confirmed
infusion of 5 mg/kg over 2 hours)
by MRI Not significant
17 mm with placebo (saline infu-
Participants had to
sion over 2 hours)
be eligible for
surgery, and were
screened for tuber-
culosis and other
infections

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 14

 Musculoskeletal disorders
Herniated lumbar disc
Functional improvement
Compared with placebo Infliximab may be no more effective at 12 weeks or 12 months at reducing disability index
scores in people with sciatic pain caused by herniated disc (low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Functional improvement
[25]
41 people with Oswestry Disability Index P = 0.37
acute or subacute scores , 12 weeks
RCT
(2–12 weeks' dura-
with infliximab (single iv infusion
tion) sciatic pain,
of 5 mg/kg over 2 hours)
caused by herniat-
ed disc confirmed with placebo (saline infusion over
by MRI 2 hours) Not significant
Participants had to Absolute results reported graphi-
be eligible for cally
surgery, and were
screened for tuber-
culosis and other
infections

[25]
41 people with Oswestry Disability Index P = 0.48
acute or subacute scores , 1 year
RCT
(2–12 weeks' dura-
28 with infliximab (single iv infu-
tion) sciatic pain,
sion of 5 mg/kg over 2 hours)
caused by herniat-
ed disc confirmed 23 with placebo (saline infusion
by MRI over 2 hours) Not significant
Participants had to
be eligible for
surgery, and were
screened for tuber-
culosis and other
infections

[25]
41 people with Median cumulative sick leave P = 0.91
acute or subacute , 12 weeks
RCT
(2–12 weeks' dura-
28 days with infliximab (single iv
tion) sciatic pain,
infusion of 5 mg/kg over 2 hours)
caused by herniat-
ed disc confirmed 25 days with placebo (saline infu-
by MRI sion over 2 hours) Not significant
Participants had to
be eligible for
surgery, and were
screened for tuber-
culosis and other
infections

[25]
41 people with Median cumulative sick leave P = 0.60
acute or subacute , 1 year
RCT
(2–12 weeks' dura-
42 days with infliximab (single iv
tion) sciatic pain,
infusion of 5 mg/kg over 2 hours)
caused by herniat-
ed disc confirmed 25 days with placebo (saline infu-
by MRI sion over 2 hours) Not significant
Participants had to
be eligible for
surgery, and were
screened for tuber-
culosis and other
infections

-
Need for surgery
Compared with placebo Infliximab seems no more effective at 12 weeks or 12 months at reducing the requirement
for surgery in people with sciatic pain caused by herniated disc (moderate-quality evidence).

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 15

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
Need for surgery
[25]
41 people with Proportion of people having P = 0.60
acute or subacute discectomy , 12 weeks
RCT
(2–12 weeks' dura-
7/21 (33%) with infliximab (single
tion) sciatic pain,
iv infusion of 5 mg/kg over 2
caused by herniat-
hours)
ed disc confirmed
by MRI 7/19 (37%) with placebo (saline Not significant
infusion over 2 hours)
Participants had to
be eligible for
surgery, and were
screened for tuber-
culosis and other
infections

[25]
41 people with Proportion of people having P = 1.0
acute or subacute discectomy , 1 year
RCT
(2–12 weeks' dura-
8/21 (38%) with infliximab (single
tion) sciatic pain,
iv infusion of 5 mg/kg over 2
caused by herniat-
hours)
ed disc confirmed
by MRI 8/19 (42%) with placebo (saline Not significant
infusion over 2 hours)
Participants had to
be eligible for
surgery, and were
screened for tuber-
culosis and other
infections

-
Adverse effects

-
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
Adverse effects
[25]
41 people with Adverse effects , 12 weeks P = 0.23
acute or subacute
RCT 3/21 (14%) with infliximab (single
(2–12 weeks' dura-
iv infusion of 5 mg/kg over 2
tion) sciatic pain,
hours)
caused by herniat-
ed disc confirmed 0/19 (0%) with placebo (saline
by MRI infusion over 2 hours) Not significant
Participants had to Described as non-serious: rhini-
be eligible for tis, diarrhoea, otitis media with
surgery, and were sinusitis maxillaris
screened for tuber-
culosis and other
infections

-
-
Other cytokine inhibitors (adalimumab, etanercept, or natalizumab):
A drug safety alert has been issued by the FDA on the risk of clinically significant liver injury associated with natal-
izumab (www.fda.gov).

-
-
-
Further information on studies

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 16

 Musculoskeletal disorders
Herniated lumbar disc
-
Comment: One RCT comparing adalimumab versus placebo in people with acute and severe radicular leg
pain and imaging-confirmed lumbar disc herniation has been published subsequent to the search
[26]
date of this Clinical Evidence review. We will assess this RCT for inclusion at the next update
of this review.

OPTION MUSCLE RELAXANTS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

• For GRADE evaluation of interventions for Herniated lumbar disc, see table, p 62 .
• We found no direct information from RCTs about muscle relaxants in the treatment of people with symptomatic
herniated lumbar disc.

Benefits and harms

Muscle relaxants:
We found no systematic review or RCTs on the use of muscle relaxants for treatment of people with symptomatic
herniated lumbar disc.

-
-
-
Further information on studies

-
-
Comment: None.

OPTION NSAIDS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

• For GRADE evaluation of interventions for Herniated lumbar disc, see table, p 62 .
• NSAIDs do not seem to improve symptoms of sciatica caused by disc herniation.
• A drug safety alert has been issued by the European Medicines Agency (EMEA) on the increased risk of GI adverse
effects and serious skin reactions associated with piroxicam.

Benefits and harms

NSAIDs versus placebo:
[15]
We found one systematic review (search date 1998, 3 RCTs, 321 people).

-
Pain
Compared with placebo NSAIDs may be no more effective at improving global pain at 5 to 30 days in people with
sciatic pain caused by disc herniation (low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Pain
[15]
321 people Proportion of people with im- OR for global improvement 0.99
proved pain , 5 to 30 days
Systematic 3 RCTs in this 95% CI 0.60 to 1.70
review analysis 80/172 (47%) with NSAIDs
57/149 (38%) with placebo
The NSAIDs used were piroxicam Not significant
40 mg daily for 2 days or 20 mg
daily for 12 days; indometacin
(indomethacin) 75 mg to 100 mg
three times daily; phenylbutazone
1200 mg daily for 3 days or
600 mg daily for 2 days

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 17

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
Relevance of outcomes assessed
unclear — see further information
on studies

-
Adverse effects

-
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
Adverse effects
[15]
321 people Adverse effects
Systematic 3 RCTs in this with NSAIDs
review analysis
with placebo
The review reported no adverse
effects with NSAIDs; however,
NSAIDs are associated with well-
documented adverse effects. See
comment below for further details

-
-
NSAIDs versus electroacupuncture:
We found one small RCT (40 people with sciatica for >2 years caused by disc herniation; verified by MRI, CT scan,
or x-ray; see comment below) comparing an NSAID (diclofenac 50 mg 3 times/day) versus electroacupuncture
[27]
(electrical stimulator [G6805-II] for 25 minutes/day for 7 days).

-
Pain
Compared with electroacupuncture We don't know how NSAIDs compare with electroacupuncture at improving pain
(very low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Pain after treatment
[27]
40 people with sci- Mean angle of Lasègue's sign Mean difference 5.8°
atica for >2 years during straight leg raising test
RCT 95% CI 4.6° to 7.0°
caused by disc , end of treatment
herniation; verified P <0.05
70.8° with diclofenac 50 mg three
by MRI, CT, or x- electroacupuncture
times daily
ray
76.7° with electroacupuncture
Weak methods,
(electrical stimulator [G6805-II]
see further informa-
for 25 minutes/day for 7 days)
tion on studies

[27]
40 people with sci- Buttock tenderness visual Mean difference –7.6
atica for >2 years analogue scale (VAS) (0 = no
RCT 95% CI –9.3 to –6.0
caused by disc tenderness to 10 = extreme
herniation; verified tenderness, converted to a P <0.05
by MRI, CT, or x- scale of 0–100) , end of treat-
ray ment
electroacupuncture
Weak methods, 33.3 with diclofenac 50 mg three
see further informa- times daily
tion on studies
25.7 with electroacupuncture
(electrical stimulator [G6805-II]
for 25 minutes/day for 7 days)

[27]
40 people with sci- Leg tenderness VAS (0 = no P >0.05
atica for >2 years tenderness to 10 = extreme Not significant
RCT
caused by disc tenderness, converted to a

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 18

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
herniation; verified scale of 0–100) , end of treat-
by MRI, CT, or x- ment
ray
25.3 with diclofenac 50 mg three
Weak methods, times daily
see further informa-
21.0 with electroacupuncture
tion on studies
(electrical stimulator [G6805-II]
for 25 minutes/day for 7 days)

[27]
40 people with sci- Tenderness in posterior side P >0.05
atica for >2 years of the thigh VAS (0 = no tender-
RCT
caused by disc ness to 10 = extreme tender-
herniation; verified ness, converted to a scale of
by MRI, CT, or x- 0–100) , at end of treatment
ray Not significant
28.6 with diclofenac 50 mg three
Weak methods, times daily
see further informa-
21.2 with electroacupuncture
tion on studies
(electrical stimulator [G6805-II]
for 25 minutes/day for 7 days)

-
Functional improvement
Compared with electroacupuncture NSAIDs may be less effective at improving straight leg raising in people with
sciatica caused by disc herniation (low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Functional improvement
[27]
40 people with sci- Mean angle of Lasègue's sign Mean difference 5.8°
atica for >2 years during straight leg raising test
RCT 95% CI 4.6° to 7.0°
caused by disc , at end of treatment
herniation; verified P <0.05
70.8° with diclofenac 50 mg three
by MRI, CT, or x- electroacupuncture
times daily
ray
76.7° with electroacupuncture
Weak methods,
(electrical stimulator [G6805-II]
see further informa-
for 25 minutes/day for 7 days)
tion on studies

-
Adverse effects

-
-
[27]
No data from the following reference on this outcome.

-
-
-
Further information on studies
[15]
The absolute data in the RCTs relate to the outcomes of improvement in pain (3 RCTs) and return to work (1
RCT). However, the meta-analysis used the outcome measure of global improvement. The relationship between
these measures is unclear.
[27]
The RCT comparing diclofenac versus electroacupuncture may have included people without a conclusive di-
agnosis of disc herniation, as x-ray was used for diagnosis in some cases. The outcome measures used in this
RCT, such as buttock tenderness, may not be comparable to more commonly reported pain measures. The
method of randomisation was not reported.

-
-

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 19

 Musculoskeletal disorders
Herniated lumbar disc
Comment: Adverse effects of NSAIDs:
NSAIDs may cause GI, cardiovascular, and other complications (see review on NSAIDs). COX-2
inhibitors have been particularly associated with an increased risk of cardiovascular events, leading
[28] [29]
to the withdrawal of rofecoxib in September 2004. A drug safety alert has been issued by
the European Medicines Agency (EMEA) on the increased risk of GI adverse effects and serious
skin reactions associated with piroxicam (www.emea.europa.eu).

QUESTION What are the effects of non-drug treatments for herniated lumbar disc?

OPTION SPINAL MANIPULATION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

• For GRADE evaluation of interventions for Herniated lumbar disc, see table, p 62 .
• With regard to non-drug treatments, spinal manipulation seems more effective at relieving local or radiating pain
in people with acute back pain and sciatica with disc protrusion compared with sham manipulation, although
concerns exist regarding possible further herniation from spinal manipulation in people who are surgical candidates.

Benefits and harms

Spinal manipulation versus placebo or sham treatment:
[30] [31]
We found one systematic review (search date 2006) and one subsequent RCT. The review identified no
RCTs comparing spinal manipulation versus placebo. The subsequent RCT compared active spinal manipulation
(assessment of range of motion, soft tissue manipulations, and brisk rotational thrusting) versus sham manipulation
[31]
(soft muscle pressing and no rapid thrusts). We also found three subsequent systematic reviews evaluating adverse
[32] [33] [34]
effects.

-
Pain
Compared with sham manipulation Active spinal manipulation is more effective at 6 months at relieving local or radi-
ating pain in people with acute back pain and sciatica with disc protrusion (moderate-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Pain
[31]
102 people with Proportion of people who were P <0.005
acute back pain free of local or radiating pain
RCT
(pain <10 days and (visual analogue scale [VAS]
pain-free for the score = 0 on scale where 0 = no
previous 3 months) pain to 10 = unbearable pain) ,
and sciatica with 6 months active manipulation
disc protrusion
15/53 (28%) with active manipu-
lation
3/49 (6%) with sham manipula-
tion

[31]
102 people with Proportion of people who were P <0.0001
acute back pain free of radiating pain (VAS
RCT
(pain <10 days and score = 0 on scale where 0 = no
pain-free for the pain to 10 = unbearable pain) ,
previous 3 months) 6 months
active manipulation
and sciatica with
29/53 (55%) with active manipu-
disc protrusion
lation
10/49 (20%) with sham manipula-
tion

[31]
102 people with Treatment failure (defined as P value and significance not re-
acute back pain stopping of treatment because ported
RCT
(pain <10 days and of no pain reduction) , 6
pain-free for the months
previous 3 months)
1/53 (1.9%) with active manipula-
and sciatica with
tion
disc protrusion
1/49 (2.0%) with sham manipula-
tion

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 20

 Musculoskeletal disorders
Herniated lumbar disc
Functional improvement
Compared with sham manipulation We don't know whether microdiscectomy is more effective at improving physical
function (low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Functional improvement
[31]
102 people with Mean score for Short Form P value not reported
acute back pain (SF)-36 Health Survey, physical
RCT Reported as not significant
(pain <10 days and functioning domain , 6 months
pain-free for the Not significant
67.4 with active manipulation
previous 3 months)
and sciatica with 60.5 with sham manipulation
disc protrusion

-
Adverse effects

-
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
Adverse effects
[30]
People with herniat- Worsening of neurological
ed lumbar disc symptoms
Systematic
(number not report-
review with people receiving spinal ma-
ed)
nipulation
Small medical
with baseline
records review
identified by sys- Absolute results not reported
tematic review
The small review of people with
significant worsening of neurolog-
ical symptoms after spinal manip-
ulation found that some were lat-
er given a different diagnosis af-
ter an MRI scan. See further infor-
mation on studies for full details

[32]
135 cases of seri- Serious complications
ous complications
Systematic with people receiving spinal ma-
after spinal manipu-
review nipulation
lation; published
between 1950 and with baseline
1980
Absolute results not reported
Review of case re-
ports identified by The frequency of complications
systematic review was not certain. The case review
attributed these complications to
cervical manipulation, misdiagno-
sis, presence of coagulation
dyscrasias, presence of herniated
nucleus pulposus, or improper
techniques

[33]
4712 treatments in Adverse effects
1058 people hav-
Systematic with people receiving spinal ma-
ing both cervical
review nipulation
and lumbar spinal
manipulations with baseline
Results from Absolute results not reported
largest prospective
observational study The most common serious effects
found by the re- were cerebrovascular accidents,
view and other adverse effects includ-
ed local discomfort, headache,
tiredness, radiating discomfort,
dizziness, nausea, and hot skin.
However, the authors of the re-
view advise interpreting the re-
sults with caution because of un-
reliable assumptions made. See

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 21

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
further information on studies for
full details

[34]
People with herniat- Further disc herniation or cau-
ed disc (number da equina syndrome
Systematic
not reported)
review with people receiving spinal ma-
Data from 8 re- nipulation
views, 9 prospec-
with baseline
tive/retrospective
studies, and 2 Absolute results not reported
cross-sectional
surveys identified The review estimated that the risk
by the systematic of causing further disc herniation
review or cauda equina syndrome by
spinal manipulation in people in
the US is 1 in 3.7 million manipu-
lations. However, this estimate is
prone to error. See further infor-
mation on studies for full details

-
[31]
No data from the following reference on this outcome.

-
-
Spinal manipulation versus heat treatment:
[30]
We found one systematic review (search date 2006, 1 RCT).

-
Patient perception of improvement
Compared with heat treatment Spinal manipulation may be more effective than three sessions of infrared heat
treatment a week at increasing overall self-perceived improvement at 2 weeks in people with herniated lumbar disc
(very low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Patient perception of improvement
[30]
233 people, 132 Self-perceived improvement , P value not reported
people randomised 2 weeks
RCT Reported as significant
to manipulation
98/123 (80%) with spinal manipu-
and 101 people The RCT provided weak evi-
lation (by a physiotherapist, every
randomised to heat dence that manipulation may be spinal manipulation
day if necessary; total number of
effective in the short term be-
Data from 1 RCT sessions not reported)
cause of methodological limita-
56/84 (67%) with infrared heat (3 tions (see further information on
times weekly) studies below)

-
Adverse effects

-
-
[30]
No data from the following reference on this outcome.

-
-
Spinal manipulation versus exercise therapy:
We identified one systematic review (search date 2006, see comment below) that identified one methodologically
[30]
weak RCT.

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 22

 Musculoskeletal disorders
Herniated lumbar disc
Pain
Compared with exercise therapy We don't know whether spinal manipulation is more effective at 1 month or at 3 to
4 months at improving pain scores in people with herniated lumbar disc (very low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Pain
[30]
322 people Pain scores , 28 days and at 3 Reported no significant difference
to 4 months among groups (interventions
Systematic Data from 1 RCT
compared using a factorial de-
review with spinal manipulation
sign)
4-armed with exercise therapy Not significant
P value not reported
trial
with manual traction
The RCT had weak methods; see
with corsets further information on studies

Absolute results not reported

-
Patient perception of improvement
Compared with exercise therapy We don't know whether spinal manipulation is more effective at 1 month or at 3 to
4 months at increasing overall self-perceived improvement in people with herniated lumbar disc (very low-quality
evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Patient perception of improvement
[30]
322 people Self-perceived improvement , Reported no significant difference
28 days and at 3 to 4 months among groups (interventions
Systematic Data from 1 RCT
compared using a factorial de-
review with spinal manipulation
sign)
4-armed with exercise therapy Not significant
P value not reported
trial
with manual traction
The RCT had weak methods; see
with corsets further information on studies

Absolute results not reported

-
Adverse effects

-
-
[30]
No data from the following reference on this outcome.

-
-
Spinal manipulation versus traction:
[30]
We identified one systematic review (search date 2006, 2 RCTs).

-
Patient perception of improvement
Compared with traction We don't know whether spinal manipulation is more effective at 1 month at increasing overall
self-perceived improvement in people with herniated lumbar disc (very low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Patient perception of improvement
[30]
322 people Self-perceived improvement , Reported no significant difference
28 days between spinal manipulation and
Systematic Data from 1 RCT
manual traction (interventions
review with spinal manipulation
The remaining compared using a factorial de- Not significant
4-armed arms evaluated ex- with manual traction sign)
trial ercise therapy and
Absolute results not reported P value not reported
corsets

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 23

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
The RCT had weak methods; see
further information on studies

-
Functional improvement
Compared with traction Spinal manipulation may be more effective at improving lumbar function and straight leg
raising in people with herniated lumbar disc (low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Functional improvement
[35]
112 people with Proportion of people "im- P <0.05
symptomatic herni- proved" or "cured" , timescale
RCT
ated lumbar disc not reported
[30]
In review 54/62 (87%) with pulling and
turning manipulation
33/50 (66%) with traction
spinal manipulation
"Improved" was defined as ab-
sence of lumbar pain, improve-
ment in lumbar functional move-
ment; "cured" was defined as
absence of lumbar pain, straight
leg raising of >70°, ability to re-
turn to work

-
Adverse effects

-
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
Adverse effects
[35]
112 people with Syncope , timescale not report-
symptomatic herni- ed
RCT
ated lumbar disc
with pulling and turning manipula-
[30]
In review tion
with traction
The RCT found that 2/60 (3%)
people receiving traction had
syncope; no adverse effects were
reported in people receiving ma-
nipulation

-
-
-
Further information on studies
[31]
Spinal manipulation versus placebo or sham treatment: Both groups were treated according to a pre-planned
30-day protocol of up to 20 sessions lasting 5 minutes on 5 days a week by experienced chiropractors with the
same formal training. Pain scores were assessed using a 10-cm visual analogue scale (VAS; 0 = no pain to
[30] [34]
10 = unbearable pain). The review identified one systematic review of adverse effects, and a small ret-
rospective medical record review of 18 people reporting significant worsening of neurological symptoms imme-
[36]
diately after spinal manipulation by different chiropractors in New York State. Although people were not
scanned before treatment, 12 people had disc herniation (8 of whom had lumbar disc herniation) when scanned
by MRI or CT after the adverse event occurred. Two people had symptoms at the site of the manipulation who
had originally presented symptoms elsewhere. The author of the review suggested that imaging should be
carried out before manipulation to avoid worsening any existing significant disc herniation. However, this was
a small medical record review, and does not state how many people in total received spinal manipulation.

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 24

 Musculoskeletal disorders
Herniated lumbar disc
[33]
The largest study identified by the review (4712 treatments in 1058 people having both cervical and lumbar
spinal manipulations) found that the most common reaction was local discomfort (53%), followed by headache
(12%), tiredness (11%), radiating discomfort (10%), dizziness (5%), nausea (4%), hot skin (2%), and other
complaints (2%). The incidence of serious adverse effects is reported as rare, and is estimated from published
case series and reports to occur in one in 1–2 million treatments. The most common serious effects were
cerebrovascular accidents (total proportion of people having manipulations not reported, rate of adverse effects
cannot be estimated). However, it is difficult to assess whether such events are directly related to treatment.
The percentages included both cervical and lumbar spinal manipulations, which may overestimate the effect
of lumbar spinal manipulations. The authors of the review advise caution in interpreting these results, as they
are speculative and based on assumptions about the number of manipulations performed and of unreported
cases.
[34]
The estimates calculated were based on rough estimates in the literature (best available) using what the author
thought to be the most accurate, recent, or conservative values. This estimate is also prone to error because
of the possible lack of reporting of many cases of disc herniation or cauda equina syndrome. Mild symptoms
after spinal manipulation are not included in these calculations. More reliable data are needed on the incidence
of specific risks of spinal manipulation. It is unclear whether the populations studied in the RCTs cited included
people who were surgical candidates for disc herniation. Concerns exist regarding possible further herniation
from spinal manipulation in people who are surgical candidates.
[30]
Spinal manipulation versus heat treatment: The review commented that the identified RCT provided weak
evidence, because it did not report method of randomisation, group baseline characteristics, whether the control
group received the same number of treatments as the other group, what happened to those lost to follow-up at
2 weeks (9/132 [7%] with spinal manipulation v 22/123 [18%] with heat), or whether it used intention-to-treat
analysis.
[30]
Spinal manipulation versus exercise therapy or traction: The review commented on the methodological
weaknesses of the 4-armed RCT, which did not describe the method of randomisation, and was not single-
blinded. It gave insufficient detail about baseline characteristics for groups at baseline, and may have included
people without herniated disc.

-
-
[37]
Comment: We found one further trial on manipulative reduction that was written in Chinese. We are cur-
rently awaiting full text translation and we will assess this for inclusion in our next update.

OPTION ACUPUNCTURE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

• For GRADE evaluation of interventions for Herniated lumbar disc, see table, p 62 .
• We found insufficient evidence about acupuncture to judge its efficacy in treating people with herniated disc.

Benefits and harms

Acupuncture versus sham acupuncture:
[38]
We found one systematic review (search date 1998) in people with back and neck pain, which identified one
small RCT of acupuncture in people with sciatica.

-
Pain
Compared with sham acupuncture We don't know whether acupuncture is more effective at reducing pain intensity
at rest in people with acute sciatica caused by disc herniation (very low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Pain
[38]
30 people with Pain intensity at rest , 5 days The RCT found that acupuncture
acute sciatica significantly improved three out-
Systematic with acupuncture at electronically
comes compared with sham
review Data from 1 RCT detected non-traditional points
acupuncture, and that there was
with sham acupuncture an overall benefit of acupuncture.
However, the review disagreed Not significant
with the overall beneficial conclu-
sion of the RCT, only finding a
significant difference between
groups in 3/12 (25%) outcome
measures, and no significant dif-

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 25

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
ference between acupuncture
and sham acupuncture in pain
intensity at rest — the most clini-
cally relevant outcome — after 5
days (absolute numbers and P
value not reported)

-
Adverse effects

-
-
[38]
No data from the following reference on this outcome.

-
-
Laser acupuncture versus sham laser acupuncture:
[38]
We found one systematic review (search date 1998) in people with back and neck pain, which identified one
small crossover RCT of laser acupuncture at traditional points versus sham laser acupuncture.

-
Pain
Compared with sham laser acupuncture We don't know whether laser acupuncture is more effective at reducing pain
intensity in people with radicular and pseudo-radicular cervical and lumbar pain caused by stenosis, herniated disc,
or both (very low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Pain
[38]
42 people, radicu- Reduction of pain intensity , The review found no significant
lar and pseudo- after 24 hours difference between groups in re-
Systematic
radicular cervical duction of pain intensity after 24
review with laser acupuncture at tradition-
and lumbar pain hours, although pain was signifi-
al points
Crossover caused by steno- cantly improved in the laser
design sis, herniated disc, with sham laser acupuncture acupuncture group at 15 minutes,
or both 1 hour, and 6 hours compared
with sham laser acupuncture Not significant
Data from 1 RCT
The sample size
was small, and it is
unclear whether
the data are gener-
alisable to herniat-
ed disc

-
Adverse effects

-
-
[38]
No data from the following reference on this outcome.

-
-
Electroacupuncture versus NSAIDs:
See option on NSAIDs, p 17 .

-
-

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 26

 Musculoskeletal disorders
Herniated lumbar disc
Adding acupuncture to manipulation compared with manipulation alone:
[39]
We found one RCT comparing acupuncture plus manipulation versus manipulation alone.

-
Pain
Adding acupuncture to manipulation compared with manipulation alone Adding acupuncture to manipulation may
be more effective at improving pain in people with herniated lumbar disc (very low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Pain
[39]
58 people with di- Change in pain from baseline P <0.01
agnosed herniated (visual analogue scale: 0 = no
RCT
lumbar disc con- pain, 10 = unbearable severe
firmed by imaging pain) , evaluated after 20 ses-
(details not report- sions (time not reported)
ed); duration of ill-
from 4.98 to 0.83 with acupunc-
ness 24 days to 10
ture plus manipulation
years acupuncture plus
from 4.77 to 2.85 with manipula- manipulation
tion alone
See further information on studies
for full details of the interventions
used
The randomisation procedure
used in this study was not clear

[39]
58 people with di- Recovery rate (the proportion P <0.05
agnosed herniated of people with 100% improve-
RCT
lumbar disc con- ment according to the
firmed by imaging Japanese Orthopaedic Associ-
(details not report- ation Lumbar Vertebral Disease
ed); duration of ill- Therapy Scale) , evaluated after
ness 24 days to 10 20 sessions (time not reported)
years
7/30 (23%) with acupuncture plus
acupuncture plus
manipulation
manipulation
3/28 (11%) with manipulation
alone
See further information on studies
for full details of the interventions
used
The randomisation procedure
used in this study was not clear

[39]
58 people with di- Overall effectiveness (the pro- P <0.05
agnosed herniated portion of people with improve-
RCT
lumbar disc con- ments of >25% according to
firmed by imaging the Japanese Orthopaedic As-
(details not report- sociation Lumbar Vertebral
ed); duration of ill- Disease Therapy Scale) , evalu-
ness 24 days to 10 ated after 20 sessions (time not
years reported)
7/30 (23%) with acupuncture plus acupuncture plus
manipulation manipulation
3/28 (11%) with manipulation
alone
See further information on studies
for full details of the interventions
used
The randomisation procedure
used in this study was not clear

-
Adverse effects

-
-

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 27

 Musculoskeletal disorders
Herniated lumbar disc
[39]
No data from the following reference on this outcome.

-
-
-
Further information on studies
[39]
Acupoints and technique of acupuncture were selected depending on the location of pain, level of pain, and
duration of symptoms, and involved 30 minutes' treatment daily for 2 courses of 10 sessions, with 3 to 5 days'
gap between courses. Manipulation involved 20 minutes each session of forcible thrusting, pinching, grasping,
rolling, and pulling of the lower back and legs, pressing acupoints, relaxing muscles, followed by passive exer-
cises of low back and legs and oblique pulling of the low back.

-
-
Comment: None.

OPTION ADVICE TO STAY ACTIVE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

• For GRADE evaluation of interventions for Herniated lumbar disc, see table, p 62 .
• We found no direct information from RCTs about advice to stay active in the treatment of people with sciatica
caused by lumbar disc herniation.

Benefits and harms

Advice to stay active:
We found one systematic review (search date 1998) of conservative treatments for sciatica caused by disc herniation,
[15]
which found no RCTs of advice to stay active. We found no subsequent RCTs.

-
-
-
Further information on studies

-
-
Comment: None.

OPTION EXERCISE THERAPY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

• For GRADE evaluation of interventions for Herniated lumbar disc, see table, p 62 .
• We found insufficient evidence about exercise to judge its efficacy in treating people with herniated disc.

Benefits and harms

Exercise therapy versus placebo or no treatment:
We found one systematic review (search date 1998) of conservative treatments for sciatica caused by disc herniation.
[15]
It found no RCTs comparing exercise therapy versus no treatment or placebo. We found no subsequent RCTs.

-
-
Exercise therapy versus spinal manipulation:
See option on spinal manipulation, p 20 .

-
-
Exercise therapy versus traction:
[40] [30]
We found two systematic reviews (search dates 1998 and 2006 ), each of which identified a different RCT.
© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 28
 Musculoskeletal disorders
Herniated lumbar disc
-
Pain
Compared with traction We don't know whether exercise therapy is more effective than isometric exercises at
achieving global improvement in pain at 1 month in people with herniated lumbar disc (very low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Global improvement scores
[40]
50 people admitted Pain free or improved Reported as not significant
for possible
Systematic 10/26 (38%) with isometric exer-
surgery for herniat-
review cise
ed lumbar disc,
verified by myelo- 10/24 (42%) with manual traction Not significant
gram
See further information on studies
Data from 1 RCT for full details of interventions and
outcomes

[30]
322 people Overall self-perceived improve- Reported as not significant
ment, pain scores or return to
Systematic Data from 1 RCT P value not reported
work , after 28 days and at 3 to
review
4 months
4-armed
with exercise therapy
trial
with manual traction
Not significant
with spinal manipulation
with corsets
Absolute results not reported
Weak methods; see further infor-
mation on studies for full details

-
-
Adding exercise plus education to conventional non-surgical treatment versus conventional non-surgical
treatment alone:
We found one RCT (40 people with invertebral disc herniation) comparing exercise plus education plus conventional
[41]
non-surgical treatment versus conventional non-surgical treatment alone.

-
Functional improvement
Adding exercise plus education to conventional non-surgical treatment compared with conventional non-surgical
treatment alone We don't know whether adding exercise and education to conventional non-surgical treatment is
more effective at 6 months to 3 years at improving lumbodorsal function or decreasing recurrences in people with
invertebral disc herniation (low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Functional improvement
[41]
40 people with in- Proportion of people in both P <0.01
vertebral disc herni- groups with improvement in
RCT Weak methods; see further infor-
ation lumbodorsal function , 6
mation on studies
months
exercise plus edu-
with exercise plus education plus
cation
conventional non-surgical treat-
ment
with conventional non-surgical
treatment alone

[41]
40 people with in- People with "excellent" or P <0.01
vertebral disc herni- "good" efficacy (assessed us- exercise plus edu-
RCT Weak methods; see further infor-
ation ing the modified Macnab crite- cation
mation on studies
ria) , 3 years

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 29

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
17/20 (85%) with exercise plus
education plus conventional non-
surgical treatment
11/20 (55%) with conventional
non-surgical treatment alone

[41]
40 people with in- Recurrence , 3 years P <0.01
vertebral disc herni-
RCT 4/20 (20%) with exercise plus Weak methods; see further infor-
ation
education plus conventional non- mation on studies exercise plus edu-
surgical treatment cation
11/20 (55%) with conventional
non-surgical treatment alone

-
Adverse effects

-
-
[41]
No data from the following reference on this outcome.

-
-
-
Further information on studies
[40]
Isometric exercises were done for 20 minutes daily for 5 to 7 days; abdominal, back, hip, and thigh muscle
contractions held for 6 to 8 seconds, repeated 5 to 10 times for each muscle group in crook and side-lying, and
supine positions. Manual traction involved 10 minutes of static traction daily for 5 to 7 days at a force of 300 N.
The global measure of improvement used in the RCT comparing exercise versus traction was assessed by a
neurologist (blind to intervention received), based on a 4-point scale that ranged from "symptom free" to "un-
changed". An improvement was considered as: 15 cm or greater increase in straight leg raising test; 2 cm or
greater increase in range of movement of lumbar spine in sagittal plane; 25% or greater reduction in pain
measured by pain intensity (visual analogue score 0–10 cm) and pain distribution (pain drawing); or an improve-
ment in activities of daily living (interview graded according to Roland Morris Disability Questionnaire). Only
short-term outcomes were measured — long-term effectiveness was not evaluated.
[30]
The review commented on the methodological weaknesses of the 4-arm RCT, which did not describe the method
of randomisation, and was not single blinded. It gave insufficient detail about baseline characteristics for groups
at baseline, and may have included people without herniated disc.
[41]
The authors of the RCT reported a significant difference between the groups in self-assessed function at 6
months, but when these differences were recalculated by the contributor for this Clinical Evidence review, they
were not significant. Exercise involved dorsal muscle strengthening with self-massage of the lumbar region and
hands (frequency not reported). Education involved rehabilitation education (knowledge and understanding
about the condition, psychological rehabilitation (dispelling adverse moods, adjusting patient's psychology, and
strengthening their resolve and confidence in recovery), and education on preventive methods (advice on posture
and activities). Conventional non-surgical treatment was not defined.

-
-
Comment: None.

OPTION HEAT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

• For GRADE evaluation of interventions for Herniated lumbar disc, see table, p 62 .
• We found insufficient RCT evidence about heat to judge its efficacy in treating people with herniated disc.

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 30

 Musculoskeletal disorders
Herniated lumbar disc
Benefits and harms
Heat versus placebo or no treatment:
We found one systematic review (search date 1998) of conservative treatments for sciatica caused by disc herniation,
[15]
which identified no RCTs on the use of heat for herniated lumbar disc. We found no subsequent RCTs.

-
-
Heat versus spinal manipulation:
See option on spinal manipulation, p 20 .

-
-
-
Further information on studies

-
-
Comment: None.

OPTION ICE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

• For GRADE evaluation of interventions for Herniated lumbar disc, see table, p 62 .
• We found no direct information from RCTs about ice in the treatment of people with sciatica caused by lumbar
disc herniation.

Benefits and harms

Ice compared with no ice:
We found one systematic review (search date 1998) of conservative treatments for sciatica caused by disc herniation,
[15]
which identified no RCTs on the use of ice for herniated lumbar disc. We found no subsequent RCTs.

-
-
-
Further information on studies

-
-
Comment: None.

OPTION MASSAGE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

• For GRADE evaluation of interventions for Herniated lumbar disc, see table, p 62 .
• We found insufficient information from RCTs to assess the effects of massage in people with herniated lumbar
disc.

Benefits and harms

Massage versus no massage:
We found one systematic review (search date 1998) of conservative treatments for sciatica caused by disc herniation,
[15]
which found no RCTs of massage.

-
-

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 31

 Musculoskeletal disorders
Herniated lumbar disc
Massage/manipulation versus massage/manipulation plus functional training exercises versus traction:
We found one RCT that was a three-arm trial comparing massage/manipulation versus massage/manipulation plus
[42]
functional training exercises versus traction.

-
Pain
Massage/manipulation compared with massage/manipulation plus functional training exercises We don't know
whether massage/manipulation is more effective at improving lumbar pain in people with herniated lumbar disc (very
low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Pain
[42]
110 people "Significant efficacy" (defined Reported as not significant for
as cure or >60% improvement massage/manipulation v mas-
RCT The remaining arm
from baseline in lumbar pain sage/manipulation plus functional
evaluated traction
3-armed and function) training exercises
trial
39/55 (71%) with massage/manip- Not significant
ulation
39/55 (71%) with massage/manip-
ulation plus functional training
exercises

-
Adverse effects

-
-
[42]
No data from the following reference on this outcome.

-
-
Massage/manipulation versus traction:
We found one RCT that was a three-arm trial comparing massage/manipulation versus massage/manipulation plus
[42]
functional training exercises versus traction.

-
Pain
Massage/manipulation compared with traction Massage/manipulation may be more effective at improving outcomes
in people with herniated lumbar disc (very low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Pain
[42]
110 people "Significant efficacy" (defined P >0.05 for massage/manipula-
as cure or >60% improvement tion v traction
RCT The remaining arm
from baseline in lumbar pain
evaluated mas-
3-armed and function)
sage/manipulation Not significant
trial
plus functional 39/55 (71%) with massage/manip-
training exercises ulation
24/55 (44%) with traction

-
Adverse effects

-
-
[42]
No data from the following reference on this outcome.

-
-

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 32

 Musculoskeletal disorders
Herniated lumbar disc
-
Further information on studies
[42]
Massage/manipulation involved 20-minute sessions, three times weekly, for a total of 20 sessions of waist-
rolling massage and passive backward stretching, lumbar manual vertebral mobilisation, rotational manipulation,
passive hip extension while lying prone, pressure correction, improved lumbar vertebrae inclined turning, prone
lying and active backward stretching, forced leg raising, and remedial manipulation. Massage/manipulation plus
functional training was as above, plus exercises of the lumbar and abdominal muscles, including stretching and
strengthening exercises for the back and legs, for 20 to 30 minutes, three times weekly before going to sleep.
People receiving traction had 20 minutes daily for a total of 20 treatments using a TF-4 computerised traction
bed, starting at half of body weight and increasing to full body weight.

-
-
Comment: Although the intervention used in the RCT was called massage, it included spinal manipulation
[42]
techniques. Therefore, the results may not be comparable with other massage-only interventions.

OPTION BED REST. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

• For GRADE evaluation of interventions for Herniated lumbar disc, see table, p 62 .
• Bed rest does not seem effective in treating people with sciatica caused by disc herniation.

Benefits and harms

Bed rest versus no treatment (watchful waiting):
[15] [43]
We found one systematic review and one subsequent RCT. The systematic review (search date 1998)
[15]
identified no RCTs of bed rest for treatment of people with symptomatic herniated disc.

-
Pain
Compared with no treatment Bed rest may be no more effective than watchful waiting at improving pain scores at
12 weeks in people with sciatica (low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Pain
[43]
183 people with Mean pain scores (McGill Pain Difference –0.6
sciatica, intensity Questionnaire) , 12 weeks
RCT 95% CI –3.3 to +2.1
sufficient to justify
8 with bed rest at home (instruct-
2 weeks of bed Based on regression analysis
ed to stay in the supine or lateral
rest as treatment
recumbent position with 1 pillow
Most people had under the head) Not significant
nerve root compres-
7 with watchful waiting
sion on MRI
(109/161 [68%]
people who had
MRI performed)

-
Functional improvement
Compared with no treatment Bed rest may be no more effective than watchful waiting at improving disability scores
at 12 weeks in people with sciatica (low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Functional improvement
[43]
183 people with Revised Roland Morris Disabil- Difference –0.5
sciatica, intensity ity Questionnaire , 12 weeks
RCT 95% CI –2.6 to +1.6
sufficient to justify
15.2 with bed rest at home (in-
2 weeks of bed Not significant
structed to stay in the supine or
rest as treatment
lateral recumbent position with 1
Most people had pillow under the head)
nerve root compres-

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 33

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
sion on MRI 15.7 with watchful waiting
(109/161 [68%]
people who had
MRI performed)

-
Patient perception of improvement
Compared with no treatment Bed rest may be no more effective than watchful waiting at improving people's perception
of improvement at 12 weeks in people with sciatica (low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Patient perception of improvement
[43]
183 people with Mean satisfaction scores , 12 Difference –0.1
sciatica, intensity weeks
RCT 95% CI –0.6 to +0.3
sufficient to justify
7 with bed rest at home (instruct-
2 weeks of bed Based on regression analysis
ed to stay in the supine or lateral
rest as treatment
recumbent position with 1 pillow
Most people had under the head) Not significant
nerve root compres-
8 with watchful waiting
sion on MRI
(109/161 [68%]
people who had
MRI performed)

-
Adverse effects

-
-
[15] [43]
No data from the following reference on this outcome.

-
-
-
Further information on studies
[43]
The regression analysis in the RCT adjusted odds ratios and differences between treatments for several variables
including baseline differences in age, sex, presence or absence of paresis, disease duration, and people's
history with respect to sciatica.

-
-
Comment: We found one further systematic review (search date 1996) of bed rest and advice to stay active
in people with acute low back pain, which found three RCTs including people with sciatica or radi-
[44]
ating pain. However, no further details were given on the proportion of people in these RCTs
with herniated disc. The review concluded that there was little evidence on bed rest specifically for
herniated lumbar disc, although the RCTs identified questioned the efficacy of bed rest for sciatica.
[44]

OPTION TRACTION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

• For GRADE evaluation of interventions for Herniated lumbar disc, see table, p 62 .
• Traction does not seem effective in treating people with sciatica caused by disc herniation.

Benefits and harms

Traction versus no traction or sham traction:
[15] [45]
We found one systematic review (search date 1998) and one subsequent RCT.

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 34

 Musculoskeletal disorders
Herniated lumbar disc
-
Pain
Traction compared with no traction or sham traction Traction may be no more effective at achieving overall global
improvement or pain intensity in people with sciatica caused by lumbar disc herniation (very low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Pain
[15]
329 people with Global improvement OR 1.2
sciatica who may
Systematic with traction 95% CI 0.7 to 2.0
or may not have
review
had disc herniation with no traction or sham traction
4 RCTs in this Absolute results not reported
analysis Not significant
See further information on studies
for full details of interventions
Global improvement included
pain intensity, mobility of lumbar
spine, straight leg raising test,
and function

-
[45]
No data from the following reference on this outcome.

-
Functional improvement
Manual traction compared with no traction or sham traction We don't know whether manual traction is more effective
at increasing Oswestry Disability Index scores in people with herniated disc (low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Functional improvement
[45]
102 people with Mean changes from baseline Mean difference +6.00
herniated disc diag- Oswestry Disability Index
RCT 95% CI –0.42 to +12.43
nosed by clinical scores
examination or P = 0.067
19.25 with manual traction Not significant
MRI
25.25 with sham traction
See further information on studies
for details of interventions used

-
[15]
No data from the following reference on this outcome.

-
Patient perception of improvement
Manual traction compared with no traction or sham traction We don't know whether manual traction is more effective
at increasing the number of people reporting complete recovery or much improvement in people with herniated disc
(low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Patient perception of improvement
[45]
102 people with Proportion of people reporting P = 0.889
herniated disc diag- a complete recovery or much
nosed by clinical improvement
examination or
38/54 (70%) with manual traction Not significant
MRI
34/48 (71%) with sham traction
See further information on studies
for details of interventions used

-
-
© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 35
 Musculoskeletal disorders
Herniated lumbar disc
Traction versus exercise therapy:
See exercise therapy, p 28 .

-
-
Traction versus spinal manipulation:
See spinal manipulation, p 20 .

-
-
Traction versus massage:
See massage, p 31 .

-
-
Autotraction versus passive traction:
[15] [46] [47]
The review identified two RCTs comparing autotraction versus passive traction.

-
Functional improvement
Autotraction compared with passive traction We don't know whether autotraction is more effective at achieving
overall global improvement (based on Lasègue's sign, functional ability, and patient's opinion) or at increasing response
rates immediately after treatment in people with herniated lumbar disc (very low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Functional improvement
[46]
49 hospitalised Global assessment by neurolo- Results at 3 months were the
people with con- gist: AR for "no effect" , 2 same as for 2 weeks
RCT
firmed herniated weeks
P values and CIs not reported
disc
21/26 (81%) with autotraction
[15]
In review
16/23 (70%) with manual traction Not significant
See further information on studies
for details of interventions used.
Global assessment based on
Lasègue's sign, functional ability,
and patient's opinion

[47]
44 people with her- Proportion of people who P <0.001
niated disc verified classified themselves as re-
RCT
by CT scan or MRI sponders , immediately after
[15] treatment
In review
17/22 (77%) with 3 sessions of
autotraction
4/22 (18%) with 5 sessions of
passive traction
autotraction
See further information on studies
for details of interventions used.
It was only possible to determine
results immediately after treat-
ment, as non-responders in both
groups were given the interven-
tion from the other group, and no
intention-to-treat analysis was
presented

-
Adverse effects

-
-
[15] [46] [47]
No data from the following reference on this outcome.

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 36

 Musculoskeletal disorders
Herniated lumbar disc
-
-
-
Further information on studies
[15]
The RCTs identified by the review comparing traction versus placebo used a variety of traction techniques and
placebo treatments (comparisons: continuous traction, about 45 kg for 30 minutes/day for up to 3 weeks v infrared
heat three times/week; intermittent motorised traction force of a third of body weight for 20 minutes/day for 5–7
days v simulated traction of 7 kg; motorised traction force of 40–70 kg for 20 minutes/day for 5–7 days v simu-
lated traction [force not reported]; autotraction with a force of a third to full body weight in sessions lasting 1
hour plus hyperextension orthosis v orthosis only). The review included RCTs in people with sciatica, who may
[48]
not have had lumbar disc herniation. An earlier systematic review (search date 1992) identified all 4 placebo-
[15]
controlled RCTs identified in the later review, but considered two of these RCTs in acute low back pain
rather than herniated lumbar disc. Neither of the RCTs considered to be in people with lumbar disc herniation
by both systematic reviews found any significant differences between traction and placebo.
[45]
The RCT compared manual traction (20 minutes, 3 times weekly: intermittent hold for 45 seconds, rest for 30
seconds, 90° hip flexion and 90° knee flexion, therapist applied force of 35–50% of body weight) versus sham
traction (same as manual traction, but therapist applying <20% of body weight). People in both groups also re-
ceived NSAIDs, an advice booklet on appropriate activities for back protection and back exercises, and appli-
cation of superficial heat to the back at home.
[46]
The RCT compared autotraction (using the Lind technique; held from a few seconds up to a couple of minutes
with force between a third to full body weight, session lasting 1 hour) versus manual traction (static traction held
by therapist weight up to 30 kg twice, each pull lasting 5 minutes).
[47]
The RCT compared three sessions of autotraction (Natchev technique with specially designed traction table)
versus 5 sessions of passive traction (static traction held by chain to table of 35% of body weight; sessions of
45 minutes every day for 5 days). In the RCT, people classified their condition as "responsive" (fully recovered
or improved), "unchanged", or "worsened".

-
-
Comment: We also found a study on electroacupuncture under continuous traction, which was written in Chi-
[49]
nese. We are currently awaiting full text translation and we will assess this for inclusion in our
next update.

QUESTION What are the effects of surgery for herniated lumbar disc?

OPTION MICRODISCECTOMY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

• For GRADE evaluation of interventions for Herniated lumbar disc, see table, p 62 .
• Microdiscectomy and standard discectomy seem to increase self-reported improvement to a similar extent.

Benefits and harms

Microdiscectomy versus conservative treatment:
[50] [51]
We found two RCTs comparing microdiscectomy with conservative treatment.

-
Pain
Compared with conservative treatment Microdiscectomy may be more effective at reducing leg pain intensity at 8
weeks, but may be no more effective at reducing leg or back pain after 6 months to 2 years (very low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Leg pain
[50]
56 people Leg pain, measured on a 100- Reported as not significant
mm visual analogue scale
RCT
(VAS) from 0 = no pain to Not significant
100 = worst possible pain ,
baseline

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 37

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
61 with microdiscectomy plus
physiotherapeutic instructions
57 with conservative treatment
(physiotherapeutic instructions
plus continued isometric exercis-
es)

[50]
56 people Leg pain, measured on a 100- P <0.01
mm VAS from 0 = no pain to
RCT
100 = worst possible pain , 6
weeks
12 with microdiscectomy plus
microdiscectomy
physiotherapeutic instructions
25 with conservative treatment
(physiotherapeutic instructions
plus continued isometric exercis-
es)

[50]
56 people Leg pain, measured on a 100- Reported as not significant
mm VAS from 0 = no pain to
RCT
100 = worst possible pain , 3
months
9 with microdiscectomy plus
Not significant
physiotherapeutic instructions
16 with conservative treatment
(physiotherapeutic instructions
plus continued isometric exercis-
es)

[50]
56 people Leg pain, measured on a 100- Reported as not significant
mm VAS from 0 = no pain to
RCT
100 = worst possible pain , 6
months
9 with microdiscectomy plus
Not significant
physiotherapeutic instructions
18 with conservative treatment
(physiotherapeutic instructions
plus continued isometric exercis-
es)

[50]
56 people Leg pain, measured on a 100- Reported as not significant
mm VAS from 0 = no pain to
RCT
100 = worst possible pain , 1
year
6 with microdiscectomy plus
Not significant
physiotherapeutic instructions
9 with conservative treatment
(physiotherapeutic instructions
plus continued isometric exercis-
es)

[50]
56 people Leg pain, measured on a 100- Reported as not significant
mm VAS from 0 = no pain to
RCT
100 = worst possible pain , 2
years
6 with microdiscectomy plus
Not significant
physiotherapeutic instructions
15 with conservative treatment
(physiotherapeutic instructions
plus continued isometric exercis-
es)

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 38

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
[51]
283 people with 6 Leg pain, measured on a 100- Reported as not significant
to 12 weeks of per- mm VAS from 0 = no pain to
RCT
sistent sciatica and 100 = worst possible pain ,
radiologically con- baseline
firmed disc hernia-
67.2 with early microdiscectomy
tion
(scheduled within 2 weeks of Not significant
randomisation)
64.4 with conservative care
For full details about interventions
used, see further information on
studies

[51]
283 people with 6 Leg pain, measured on a 100- Difference 17.7
to 12 weeks of per- mm VAS from 0 = no pain to
RCT 95% CI 12.3 to 23.1
sistent sciatica and 100 = worst possible pain , 8
radiologically con- weeks
firmed disc hernia-
10.2 with early microdiscectomy
tion early microdiscecto-
(scheduled within 2 weeks of
randomisation) my

27.9 with conservative care

For full details about interventions
used, see further information on
studies

[51]
283 people with 6 Leg pain, measured on a 100- Difference 6.1
to 12 weeks of per- mm VAS from 0 = no pain to
RCT 95% CI 2.2 to 10.0
sistent sciatica and 100 = worst possible pain , 6
radiologically con- months
firmed disc hernia-
8.4 with early microdiscectomy
tion early microdiscecto-
(scheduled within 2 weeks of
randomisation) my

14.5 with conservative care

For full details about interventions
used, see further information on
studies

[51]
283 people with 6 Leg pain, measured on a 100- Difference 0
to 12 weeks of per- mm VAS from 0 = no pain to
RCT 95% CI –4.0 to +4.0
sistent sciatica and 100 = worst possible pain , 1
radiologically con- year
firmed disc hernia-
11.0 with early microdiscectomy
tion
(scheduled within 2 weeks of Not significant
randomisation)
11.0 with conservative care
For full details about interventions
used, see further information on
studies

[51]
283 people with 6 Leg pain, measured on a 100- Difference –2
to 12 weeks of per- mm VAS from 0 = no pain to
RCT 95% CI –6.0 to +2.0
sistent sciatica and 100 = worst possible pain , 2
radiologically con- years
firmed disc hernia-
11.0 with early microdiscectomy
tion
(scheduled within 2 weeks of Not significant
randomisation)
9.0 with conservative care
For full details about interventions
used, see further information on
studies

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 39

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
Back pain
[50]
56 people Back pain, measured on a 100- Reported as not significant
mm VAS from 0 = no pain to
RCT
100 = worst possible pain ,
baseline
53 with microdiscectomy plus
Not significant
physiotherapeutic instructions
47 with conservative treatment
(physiotherapeutic instructions
plus continued isometric exercis-
es)

[50]
56 people Back pain, measured on a 100- Reported as not significant
mm VAS from 0 = no pain to
RCT
100 = worst possible pain , 6
weeks
21 with microdiscectomy plus
Not significant
physiotherapeutic instructions
28 with conservative treatment
(physiotherapeutic instructions
plus continued isometric exercis-
es)

[50]
56 people Back pain, measured on a 100- Reported as not significant
mm VAS from 0 = no pain to
RCT
100 = worst possible pain , 3
months
15 with microdiscectomy plus
Not significant
physiotherapeutic instructions
22 with conservative treatment
(physiotherapeutic instructions
plus continued isometric exercis-
es)

[50]
56 people Back pain, measured on a 100- Reported as not significant
mm VAS from 0 = no pain to
RCT
100 = worst possible pain , 6
months
13 with microdiscectomy plus
Not significant
physiotherapeutic instructions
20 with conservative treatment
(physiotherapeutic instructions
plus continued isometric exercis-
es)

[50]
56 people Back pain, measured on a 100- Reported as not significant
mm VAS from 0 = no pain to
RCT
100 = worst possible pain , 1
year
19 with microdiscectomy plus
Not significant
physiotherapeutic instructions
17 with conservative treatment
(physiotherapeutic instructions
plus continued isometric exercis-
es)

[50]
56 people Back pain, measured on a 100- Reported as not significant
mm VAS from 0 = no pain to
RCT
100 = worst possible pain , 2
years
11 with microdiscectomy plus
Not significant
physiotherapeutic instructions
21 with conservative treatment
(physiotherapeutic instructions
plus continued isometric exercis-
es)

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 40

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
[51]
283 people with 6 Back pain, measured on a 100- Reported as not significant
to 12 weeks of per- mm VAS from 0 = no pain to
RCT
sistent sciatica and 100 = worst possible pain ,
radiologically con- baseline
firmed disc hernia-
33.8 with early microdiscectomy
tion
(scheduled within 2 weeks of Not significant
randomisation)
30.8 with conservative care
For full details about interventions
used, see further information on
studies

[51]
283 people with 6 Back pain, measured on a 100- Difference 11.3
to 12 weeks of per- mm VAS from 0 = no pain to
RCT 95% CI 5.6 to 17.4
sistent sciatica and 100 = worst possible pain , 8
radiologically con- weeks
firmed disc hernia-
14.4 with early microdiscectomy
tion
(scheduled within 2 weeks of microdiscectomy
randomisation)
25.7 with conservative care
For full details about interventions
used, see further information on
studies

[51]
283 people with 6 Back pain, measured on a 100- Difference +2.3
to 12 weeks of per- mm VAS from 0 = no pain to
RCT 95% CI –3.6 to +8.2
sistent sciatica and 100 = worst possible pain , 6
radiologically con- months
firmed disc hernia-
15.5 with early microdiscectomy
tion
(scheduled within 2 weeks of microdiscectomy
randomisation)
17.8 with conservative care
For full details about interventions
used, see further information on
studies

[51]
283 people with 6 Back pain, measured on a 100- Difference +2.3
to 12 weeks of per- mm VAS from 0 = no pain to
RCT 95% CI –3.6 to +8.2
sistent sciatica and 100 = worst possible pain , 1
radiologically con- year
firmed disc hernia-
14.2 with early microdiscectomy
tion
(scheduled within 2 weeks of microdiscectomy
randomisation)
16.5 with conservative care
For full details about interventions
used, see further information on
studies

[51]
283 people with 6 Back pain, measured on a 100- Difference +1.4
to 12 weeks of per- mm VAS from 0 = no pain to
RCT 95% CI –4.5 to +6.3
sistent sciatica and 100 = worst possible pain , 2
radiologically con- years
firmed disc hernia-
15.9 with early microdiscectomy
tion
(scheduled within 2 weeks of Not significant
randomisation)
17.3 with conservative care
For full details about interventions
used, see further information on
studies

Short Form-36 bodily pain questionnaire

[51]
283 people with 6 Short Form (SF)-36 bodily pain Reported as not significant
to 12 weeks of per- questionnaire, measured on a Not significant
RCT
sistent sciatica and scale from 0 to 100; increasing

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 41

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
radiologically con- score indicates less-severe
firmed disc hernia- symptoms , baseline
tion
21.9 with early microdiscectomy
(scheduled within 2 weeks of
randomisation)
23.9 with conservative care
For full details about interventions
used, see further information on
studies

[51]
283 people with 6 SF-36 bodily pain question- Difference –8.4
to 12 weeks of per- naire, measured on a scale
RCT 95% CI –13.5 to –3.2
sistent sciatica and from 0 to 100; increasing score
radiologically con- indicates less-severe symp-
firmed disc hernia- toms , 8 weeks
tion
62.8 with early microdiscectomy
(scheduled within 2 weeks of microdiscectomy
randomisation)
54.4 with conservative care
For full details about interventions
used, see further information on
studies

[51]
283 people with 6 SF-36 bodily pain question- Difference –3.3
to 12 weeks of per- naire, measured on a scale
RCT 95% CI –8.4 to +1.8
sistent sciatica and from 0 to 100; increasing score
radiologically con- indicates less-severe symp-
firmed disc hernia- toms , 6 months
tion
76.1 with early microdiscectomy
(scheduled within 2 weeks of Not significant
randomisation)
72.8 with conservative care
For full details about interventions
used, see further information on
studies

[51]
283 people with 6 SF-36 bodily pain question- Difference –2.7
to 12 weeks of per- naire, measured on a scale
RCT 95% CI –7.9 to +2.6
sistent sciatica and from 0 to 100; increasing score
radiologically con- indicates less-severe symp-
firmed disc hernia- toms , 1 year
tion
81.2 with early microdiscectomy
(scheduled within 2 weeks of Not significant
randomisation)
78.5 with conservative care
For full details about interventions
used, see further information on
studies

[51]
283 people with 6 SF-36 bodily pain question- Difference +2.3
to 12 weeks of per- naire, measured on a scale
RCT 95% CI –2.7 to +7.3
sistent sciatica and from 0 to 100; increasing score
radiologically con- indicates less-severe symp-
firmed disc hernia- toms , 2 years
tion
78.4 with early microdiscectomy
(scheduled within 2 weeks of Not significant
randomisation)
80.7 with conservative care
For full details about interventions
used, see further information on
studies

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 42

 Musculoskeletal disorders
Herniated lumbar disc
Functional improvement
Compared with conservative treatment We don't know whether microdiscectomy is more effective at improving Os-
westry Disability index at 6 weeks to 2 years (very low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Oswestry Low Back Pain Disability Score
[50]
56 people Oswestry Low Back Pain Dis- Reported as not significant
ability Score, measured on a
RCT
scale of 0 to 100; increasing
score indicates greater lower
back pain-related disability ,
baseline
Not significant
39 with microdiscectomy plus
physiotherapeutic instructions
39 with conservative treatment
(physiotherapeutic instructions
plus continued isometric exercis-
es)

[50]
56 people Oswestry Low Back Pain Dis- Reported as not significant
ability Score, measured on a
RCT
scale of 0 to 100; increasing
score indicates greater lower
back pain-related disability , 6
weeks
Not significant
16 with microdiscectomy plus
physiotherapeutic instructions
22 with conservative treatment
(physiotherapeutic instructions
plus continued isometric exercis-
es)

[50]
56 people Oswestry Low Back Pain Dis- Reported as not significant
ability Score, measured on a
RCT
scale of 0 to 100; increasing
score indicates greater lower
back pain-related disability , 3
months
Not significant
16 with microdiscectomy plus
physiotherapeutic instructions
22 with conservative treatment
(physiotherapeutic instructions
plus continued isometric exercis-
es)

[50]
56 people Oswestry Low Back Pain Dis- Reported as not significant
ability Score, measured on a
RCT
scale of 0 to 100; increasing
score indicates greater lower
back pain-related disability , 6
months
Not significant
8 with microdiscectomy plus
physiotherapeutic instructions
12 with conservative treatment
(physiotherapeutic instructions
plus continued isometric exercis-
es)

[50]
56 people Oswestry Low Back Pain Dis- Reported as not significant
ability Score, measured on a
RCT
scale of 0 to 100; increasing
score indicates greater lower
back pain-related disability , 1
year Not significant
10 with microdiscectomy plus
physiotherapeutic instructions
11 with conservative treatment
(physiotherapeutic instructions

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 43

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
plus continued isometric exercis-
es)

[50]
56 people Oswestry Low Back Pain Dis- Reported as not significant
ability Score, measured on a
RCT
scale of 0 to 100; increasing
score indicates greater lower
back pain-related disability , 2
years
Not significant
6 with microdiscectomy plus
physiotherapeutic instructions
11 with conservative treatment
(physiotherapeutic instructions
plus continued isometric exercis-
es)

Modified Roland disability questionnaire

[51]
283 people with 6 Modified Roland disability Reported as not significant
to 12 weeks of per- questionnaire, measured on a
RCT
sistent sciatica and scale of 0 to 23; increasing
radiologically con- score indicates worse function-
firmed disc hernia- al status , baseline
tion
16.5 with early microdiscectomy
(scheduled within 2 weeks of Not significant
randomisation)
16.3 with conservative care
For full details about interventions
used, see further information on
studies

[51]
283 people with 6 Modified Roland disability Difference 3.1
to 12 weeks of per- questionnaire, measured on a
RCT 95% CI 1.7 to 4.3
sistent sciatica and scale of 0 to 23; increasing
radiologically con- score indicates worse function-
firmed disc hernia- al status , 8 weeks
tion
6.1 with early microdiscectomy
(scheduled within 2 weeks of microdiscectomy
randomisation)
9.2 with conservative care
For full details about interventions
used, see further information on
studies

[51]
283 people with 6 Modified Roland disability Difference +0.8
to 12 weeks of per- questionnaire, measured on a
RCT 95% CI –0.5 to +2.1
sistent sciatica and scale of 0 to 23; increasing
radiologically con- score indicates worse function-
firmed disc hernia- al status , 6 months
tion
4.0 with early microdiscectomy
(scheduled within 2 weeks of Not significant
randomisation)
4.8 with conservative care
For full details about interventions
used, see further information on
studies

[51]
283 people with 6 Modified Roland disability Difference +0.4
to 12 weeks of per- questionnaire, measured on a
RCT 95% CI –0.9 to +1.7
sistent sciatica and scale of 0 to 23; increasing
radiologically con- score indicates worse function-
firmed disc hernia- al status , 1 year
Not significant
tion
3.3 with early microdiscectomy
(scheduled within 2 weeks of
randomisation)
3.7 with conservative care

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 44

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
For full details about interventions
used, see further information on
studies

[51]
283 people with 6 Modified Roland disability Difference +0.5
to 12 weeks of per- questionnaire, measured on a
RCT 95% CI –0.8 to +1.8
sistent sciatica and scale of 0 to 23; increasing
radiologically con- score indicates worse function-
firmed disc hernia- al status , 2 years
tion
3.1 with early microdiscectomy
(scheduled within 2 weeks of Not significant
randomisation)
2.6 with conservative care
For full details about interventions
used, see further information on
studies

Short Form-36 physical functioning questionnaire

[51]
283 people with 6 Short Form (SF)-36 physical Reported as not significant
to 12 weeks of per- functioning questionnaire,
RCT
sistent sciatica and measured on a scale from 0 to
radiologically con- 100; increasing score indicates
firmed disc hernia- less-severe symptoms , base-
tion line
33.9 with early microdiscectomy Not significant
(scheduled within 2 weeks of
randomisation)
34.6 with conservative care
For full details about interventions
used, see further information on
studies

[51]
283 people with 6 SF-36 physical functioning Difference –9.3
to 12 weeks of per- questionnaire, measured on a
RCT 95% CI –14.2 to –4.4
sistent sciatica and scale from 0 to 100; increasing
radiologically con- score indicates less-severe
firmed disc hernia- symptoms , 6 weeks
tion
71.2 with early microdiscectomy
(scheduled within 2 weeks of microdiscectomy
randomisation)
61.9 with conservative care
For full details about interventions
used, see further information on
studies

[51]
283 people with 6 SF-36 physical functioning Difference –1.5
to 12 weeks of per- questionnaire, measured on a
RCT 95% CI –6.4 to +3.4
sistent sciatica and scale from 0 to 100; increasing
radiologically con- score indicates less-severe
firmed disc hernia- symptoms , 6 months
tion
79.1 with early microdiscectomy
(scheduled within 2 weeks of Not significant
randomisation)
77.6 with conservative care
For full details about interventions
used, see further information on
studies

[51]
283 people with 6 SF-36 physical functioning Difference –2.2
to 12 weeks of per- questionnaire, measured on a
RCT 95% CI –7.2 to +2.8
sistent sciatica and scale from 0 to 100; increasing
radiologically con- score indicates less-severe
firmed disc hernia- symptoms , 1 year Not significant
tion
84.2 with early microdiscectomy
(scheduled within 2 weeks of
randomisation)

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 45

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
82.0 with conservative care
For full details about interventions
used, see further information on
studies

[51]
283 people with 6 SF-36 physical functioning Difference +1.3
to 12 weeks of per- questionnaire, measured on a
RCT 95% CI –3.7 to +6.3
sistent sciatica and scale from 0 to 100; increasing
radiologically con- score indicates less-severe
firmed disc hernia- symptoms , 2 year
tion
82.3 with early microdiscectomy
(scheduled within 2 weeks of Not significant
randomisation)
83.6 with conservative care
For full details about interventions
used, see further information on
studies

-
Quality of life
Compared with conservative treatment We don't know whether microdiscectomy is more effective at 6 weeks to 2
years at improving quality-of-life scores or the subjective ability to work (low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Quality of life
[50]
56 people Oswestry Low Back Pain Dis- Reported as not significant
ability Score, measured on a
RCT
scale of 0 to 100; increasing
score indicates greater lower
back pain-related disability ,
baseline
Not significant
0.83 with microdiscectomy plus
physiotherapeutic instructions
0.84 with conservative treatment
(physiotherapeutic instructions
plus continued isometric exercis-
es)

[50]
56 people Oswestry Low Back Pain Dis- Reported as not significant
ability Score, measured on a
RCT
scale of 0 to 100; increasing
score indicates greater lower
back pain-related disability , 6
weeks
Not significant
0.92 with microdiscectomy plus
physiotherapeutic instructions
0.89 with conservative treatment
(physiotherapeutic instructions
plus continued isometric exercis-
es)

[50]
56 people Oswestry Low Back Pain Dis- Reported as not significant
ability Score, measured on a
RCT
scale of 0 to 100; increasing
score indicates greater lower
back pain-related disability , 3
months
Not significant
0.94 with microdiscectomy plus
physiotherapeutic instructions
0.91 with conservative treatment
(physiotherapeutic instructions
plus continued isometric exercis-
es)

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 46

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
[50]
56 people Oswestry Low Back Pain Dis- Reported as not significant
ability Score, measured on a
RCT
scale of 0 to 100; increasing
score indicates greater lower
back pain-related disability , 6
months
Not significant
0.95 with microdiscectomy plus
physiotherapeutic instructions
0.90 with conservative treatment
(physiotherapeutic instructions
plus continued isometric exercis-
es)

[50]
56 people Oswestry Low Back Pain Dis- Reported as not significant
ability Score, measured on a
RCT
scale of 0 to 100; increasing
score indicates greater lower
back pain-related disability , 1
year
Not significant
0.95 with microdiscectomy plus
physiotherapeutic instructions
0.94 with conservative treatment
(physiotherapeutic instructions
plus continued isometric exercis-
es)

[50]
56 people Oswestry Low Back Pain Dis- Reported as not significant
ability Score, measured on a
RCT
scale of 0 to 100; increasing
score indicates greater lower
back pain-related disability , 2
years
Not significant
0.95 with microdiscectomy plus
physiotherapeutic instructions
0.93 with conservative treatment
(physiotherapeutic instructions
plus continued isometric exercis-
es)

-
[51]
No data from the following reference on this outcome.

-
Patient perception of improvement
Compared with conservative treatment Microdiscectomy may be more effective at improving patients' perceived re-
covery at 8 weeks but may be no more effective at 6 months to 2 years (very low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Patient perception of improvement
[51]
283 people with 6 Short Form (SF)-36 physical Difference 44.7
to 12 weeks of per- functioning questionnaire,
RCT 95% CI 34.2 to 55.0
sistent sciatica and measured on a scale from 0 to
radiologically con- 100; increasing score indicates
firmed disc hernia- less-severe symptoms , 8
tion weeks
81.2 with early microdiscectomy microdiscectomy
(scheduled within 2 weeks of
randomisation)
36.5 with conservative care
For full details about interventions
used, see further information on
studies

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 47

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
[51]
283 people with 6 SF-36 physical functioning Difference +6.6
to 12 weeks of per- questionnaire, measured on a
RCT 95% CI –3.7 to +17.0
sistent sciatica and scale from 0 to 100; increasing
radiologically con- score indicates less-severe
firmed disc hernia- symptoms , 6 months
tion
77.4 with early microdiscectomy
(scheduled within 2 weeks of Not significant
randomisation)
70.8 with conservative care
For full details about interventions
used, see further information on
studies

[51]
283 people with 6 SF-36 physical functioning Difference +3.2
to 12 weeks of per- questionnaire, measured on a
RCT 95% CI –5.4 to +11.9
sistent sciatica and scale from 0 to 100; increasing
radiologically con- score indicates less-severe
firmed disc hernia- symptoms , 1 year
tion
85.7 with early microdiscectomy
(scheduled within 2 weeks of Not significant
randomisation)
82.5 with conservative care
For full details about interventions
used, see further information on
studies

[51]
283 people with 6 SF-36 physical functioning Difference +2.4
to 12 weeks of per- questionnaire, measured on a
RCT 95% CI –7.2 to +12.0
sistent sciatica and scale from 0 to 100; increasing
radiologically con- score indicates less-severe
firmed disc hernia- symptoms , 2 years
tion
81.3 with early microdiscectomy
(scheduled within 2 weeks of Not significant
randomisation)
78.9 with conservative care
For full details about interventions
used, see further information on
studies

-
[50]
No data from the following reference on this outcome.

-
Adverse effects

-
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
Adverse effects
[50]
56 people Urosepsis
RCT with microdiscectomy plus phys-
iotherapeutic instructions
with conservative treatment
(physiotherapeutic instructions
plus continued isometric exercis-
es)
Absolute results not reported
The RCT reported that 1 person
(1/28 [4%]) in the microdiscecto-
my group contracted urosepsis,
requiring intravenous antibiotics
and a prolonged hospital stay

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 48

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
[51]
283 people with 6 Short Form (SF)-36 physical
to 12 weeks of per- functioning questionnaire,
RCT
sistent sciatica and measured on a scale from 0 to
radiologically con- 100; increasing score indicates
firmed disc hernia- less-severe symptoms , 8
tion weeks
with early microdiscectomy
(scheduled within 2 weeks of
randomisation)
with conservative care
Absolute results not reported
For full details about interventions
used, see further information on
studies
The RCT did not report any data
on harms of microdiscectomy
versus conservative treatment. It
reported complications in 3/187
(2%) of all surgically treated peo-
ple between the two groups (in-
cluding 2 dural tears and 1 wound
haematoma), none of which re-
quired further intervention

-
-
Microdiscectomy versus standard discectomy:
See option on standard discectomy, p 51 .

-
-
Video-assisted arthroscopic microdiscectomy versus standard discectomy:
[52]
We found one RCT.

-
Pain
Compared with standard discectomy We don't know how video-assisted arthroscopic microdiscectomy and standard
discectomy compare for reducing pain (low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Pain
[52]
60 people with Mean pain score (visual ana- Reported as not significant
confirmed lumbar logue scale: 0 = no pain,
disc herniation and 10 = severe and incapacitating
associated radicu- pain) , about 31 months
Not significant
lopathy after failed
1.2 with video-assisted arthro-
conservative treat-
scopic microdiscectomy
ment
1.9 with standard discectomy

-
Patient perception of improvement
Compared with standard discectomy We don't know whether video-assisted arthroscopic microdiscectomy is more
effective at increasing the number of people "very satisfied" as measured on a 4-point scale in people with confirmed
lumbar disc herniation and associated radiculopathy after failed conservative treatment (low-quality evidence).

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 49

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
Patient perception of improvement
[52]
60 people with Proportion of people “very RR 1.10
confirmed lumbar satisfied” on a 4-point satisfac-
95% CI 0.71 to 1.34
disc herniation and tion scale , about 31 months
associated radicu-
22/30 (73%) with video-assisted Not significant
lopathy after failed
arthroscopic microdiscectomy
conservative treat-
ment 20/30 (67%) with standard dis-
cectomy

-
Adverse effects

-
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
Adverse effects
[52]
60 people with Adverse effects
confirmed lumbar
with video-assisted arthroscopic
disc herniation and
microdiscectomy
associated radicu-
lopathy after failed with standard discectomy
conservative treat-
ment The RCT reported that 1 person
having open discectomy had
leakage of spinal fluid from the
dural sac 2 weeks after the oper-
ation. No other postoperative
complications or neurovascular
injuries were observed in either
the standard discectomy or mi-
crodiscectomy groups

-
-
Microdiscectomy versus automated percutaneous discectomy:
See automated percutaneous discectomy, p 56 .

-
-
-
Further information on studies
[51]
Conservative care included prescription of painkillers (details not given), advice to resume daily activities, rec-
ommendation of a mobilisation scheme based on time rather than pain (compliance not checked), and referral
to a physiotherapist if fearful of movement. Subsequent microdiscectomy was considered for the conservative-
care group if sciatica persisted 6 months after randomisation, or earlier (within 6 months) in case of increasing
leg pain that was not responsive to drugs and progressive neurological deficit. A total of 125/141 (89%) people
in the early microdiscectomy group had microdiscectomy as intended. The remaining 16 people spontaneously
recovered. A total of 55/142 (39%) people in the conservative-care group went on to have microdiscectomy in
the first year, and one further 7 (5%) had microdiscectomy in the second year after randomisation. The results
presented above are based on an intention-to-treat analysis. The interventions in the two groups may have
been too similar to detect a significant difference in the outcomes measured at 6 months' to 2 years' follow-up.
[52]
The mean duration of postoperative recovery was almost twice as long with open surgery as with microdiscec-
tomy (27 days with microdiscectomy v 49 days with standard discectomy; P value not reported).

-
-
[53]
Comment: We found one further trial on microsurgery lumbar discectomy that was written in Chinese. We
are currently awaiting full text translation and we will assess this for inclusion in our next update.

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 50

 Musculoskeletal disorders
Herniated lumbar disc
OPTION STANDARD DISCECTOMY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

• For GRADE evaluation of interventions for Herniated lumbar disc, see table, p 62 .
• Both standard discectomy and microdiscectomy seem to increase self-reported improvement to a similar extent.

Benefits and harms

Standard discectomy versus conservative treatment:
[23]
We found one systematic review (search date 2007, 2 RCTs).

-
Pain
Compared with conservative treatment We don't know whether standard discectomy is more effective at improving
pain at 1 to 2 years in people with lumbar disc herniation (low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Pain
[54]
126 people with Proportion of people reporting RR 1.79
symptomatic L5/S1 their improvement as "good" ,
RCT 95% CI 1.30 to 2.18
disc herniation 1 year
[23] NNT 3, 95% CI 2 to 9
In review 39/60 (65%) with standard dis-
cectomy Contributors' own calculations
The randomisation
procedure used in 24/66 (36%) with conservative
this study was not treatment (physiotherapy for 6 standard discecto-
clear weeks) my

Improvement graded in terms of

pain and function into 4 cate-
gories: "good" (completely satis-
fied), "fair", "poor", and "bad"
(completely incapacitated for
work because of pain)

[54]
126 people with Proportion of people reporting RR 1.29
symptomatic L5/S1 their improvement as "good" ,
RCT 95% CI 0.96 to 1.56
disc herniation 4 years
[23] Contributors' own calculations
In review 40/60 (67%) with standard discec-
tomy
The randomisation
procedure used in 34/66 (51%) with conservative
this study was not treatment (physiotherapy for 6 Not significant
clear weeks)
Improvement graded in terms of
pain and function into 4 cate-
gories: "good" (completely satis-
fied), "fair", "poor", and "bad"
(completely incapacitated for
work because of pain)

[54]
126 people with Proportion of people reporting RR 1.04
symptomatic L5/S1 their improvement as "good" ,
RCT 95% CI 0.73 to 1.32
disc herniation 10 years
[23] Contributors' own calculations
In review 35/60 (58%) with standard discec-
tomy
The randomisation
procedure used in 37/66 (56%) with conservative
this study was not treatment (physiotherapy for 6 Not significant
clear weeks)
Improvement graded in terms of
pain and function into 4 cate-
gories: "good" (completely satis-
fied), "fair", "poor", and "bad"
(completely incapacitated for
work because of pain)

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 51

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
[55]
501 surgical candi- Short Form (SF)-36 Bodily Pain Difference +2.9
dates; mean age mean improvement in pain on
RCT 95% CI –2.2 to +8.0
42 years; 42% fe- a scale from 0 to 100 from
male, with imaging- baseline , 3 months
confirmed lumbar
30.5 with standard open discecto-
intervertebral disc Not significant
my
herniation and at
least 6 weeks of 27.6 with non-operative treatment
radicular symp-
toms
[23]
In review

[55]
501 surgical candi- SF-36 Bodily Pain mean im- Difference +2.8
dates; mean age provement in pain on a scale
RCT 95% CI –2.3 to +7.8
42 years; 42% fe- from 0 to 100 from baseline , 1
male, with imaging- year
confirmed lumbar
39.7 with standard open discecto-
intervertebral disc Not significant
my
herniation and at
least 6 weeks of 36.9 with non-operative treatment
radicular symp-
toms
[23]
In review

[55]
501 surgical candi- SF-36 Bodily Pain mean im- Difference +3.2
dates; mean age provement in pain on a scale
RCT 95% CI –2.0 to +8.4
42 years; 42% fe- from 0 to 100 from baseline , 2
male, with imaging- years
confirmed lumbar
40.3 with standard open discecto-
intervertebral disc Not significant
my
herniation and at
least 6 weeks of 37.1 with non-operative treatment
radicular symp-
toms
[23]
In review

-
Functional improvement
Compared with conservative treatment We don't know whether standard discectomy is more effective at improving
function or Oswestry Disability Index at 1 to 2 years in people with lumbar disc herniation (low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Short Form-36 Physical Function scores
[55]
501 surgical candi- Short Form (SF)-36 Physical Difference +2.8
dates; mean age Function mean improvement
RCT 95% CI –2.5 to +8.1
42 years; 42% fe- on a scale from 0 to 100 from
male, with imaging- baseline score , 3 months
confirmed lumbar
27.7 with standard open discecto-
intervertebral disc Not significant
my
herniation and at
least 6 weeks of 24.9 with non-operative treatment
radicular symp-
toms
[23]
In review

[55]
501 surgical candi- SF-36 Physical Function mean Difference +2.8
dates; mean age improvement on a scale from
RCT 95% CI –2.5 to +8.1
42 years; 42% fe- 0 to 100 from baseline score ,
male, with imaging- 1 year
confirmed lumbar
27.7 with standard open discecto-
intervertebral disc Not significant
my
herniation and at
least 6 weeks of 24.9 with non-operative treatment
radicular symp-
toms
[23]
In review

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 52

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
[55]
501 surgical candi- SF-36 Physical Function mean Difference 0
dates; mean age improvement on a scale from
RCT 95% CI –5.4 to +5.5
42 years; 42% fe- 0 to 100 from baseline score ,
male, with imaging- 2 years
confirmed lumbar
35.9 with standard open discecto-
intervertebral disc Not significant
my
herniation and at
least 6 weeks of 35.9 with non-operative treatment
radicular symp-
toms
[23]
In review

Oswestry Disability Index

[55]
501 surgical candi- Oswestry Disability Index mean Difference –4.7
dates; mean age reduction in disability score
RCT 95% CI –9.3 to –0.2
42 years; 42% fe- from baseline on a scale from
male, with imaging- 0 to 100 , 3 months
confirmed lumbar
–26.0 with standard open discec-
intervertebral disc open discectomy
tomy
herniation and at
least 6 weeks of –21.3 with non-operative treat-
radicular symp- ment
toms
[23]
In review

[55]
501 surgical candi- Oswestry Disability Index mean Difference –3.2
dates; mean age reduction in disability score
RCT 95% CI –7.8 to +1.3
42 years; 42% fe- from baseline on a scale from
male, with imaging- 0 to 100 , 1 year
confirmed lumbar
–30.6 with standard open discec-
intervertebral disc Not significant
tomy
herniation and at
least 6 weeks of –27.4 with non-operative treat-
radicular symp- ment
toms
[23]
In review

[55]
501 surgical candi- Oswestry Disability Index mean Difference –2.4
dates; mean age reduction in disability score
RCT 95% CI –7.4 to +1.9
42 years; 42% fe- from baseline on a scale from
male, with imaging- 0 to 100 , 2 years
confirmed lumbar
–31.4 with standard open discec-
intervertebral disc Not significant
tomy
herniation and at
least 6 weeks of –28.7 with non-operative treat-
radicular symp- ment
toms
[23]
In review

-
Adverse effects

-
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
Adverse effects
[55]
501 surgical candi- Perioperative complications
dates; mean age
RCT with standard open discectomy
42 years; 42% fe-
male, with imaging- with non-operative treatment
confirmed lumbar
intervertebral disc The most common intraoperative
complication was dural tear in Not significant
herniation and at
least 6 weeks of 10/243 (4%) people; 230/243
radicular symp- (95%) people reported no intraop-
toms erative complications. Superficial
wound infection was the most
[23]
In review

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 53

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
common postoperative complica-
tion in 4/243 (2%) people;
226/243 (93%) people reported
no postoperative complications.
The reoperation rate for recurrent
herniation was 5/243 (2%) at 1
year and 8/243 (3%) at 2 years

-
[54]
No data from the following reference on this outcome.

-
-
Standard discectomy versus epidural corticosteroid injection:
See option on epidural corticosteroid injections, p 4 .

-
-
Standard discectomy versus microdiscectomy:
[23] [56] [57]
We found one systematic review (search date 2007, 3 RCTs, 219 people) and two subsequent RCTs
comparing standard discectomy versus microdiscectomy. The review did not perform a meta-analysis of the three
RCTs because outcomes were not comparable.

-
Pain
Compared with microdiscectomy We don't know how standard discectomy and microdiscectomy compare at reducing
pain in people with herniated disc (very-low quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Pain
[58]
60 people with Pre- and postoperative pain Reported as "similar"
lumbar disc hernia- scores measured on visual
RCT P value not reported
tion analogue scale (VAS)
[23]
In review with standard discectomy
with microdiscectomy
Absolute results not reported

[59]
79 people with Pain in the legs or back mea-
lumbar disc hernia- sured on VAS , 6 weeks
RCT
tion
with standard discectomy
[23]
In review
with microdiscectomy
Absolute results not reported

[23]
80 people "Clinical outcomes" (not fur- Reported as "similar"
ther specified) , 15 months
RCT Data from 1 RCT Significance not assessed
with standard discectomy
with microdiscectomy
Absolute results not reported

[56]
119 people Mean intensity of sciatic pain P = 0.27
scores
RCT
Not significant
1.3 with macrodiscectomy
1.2 with microdiscectomy

[56]
119 people Mean change in Japanese Or- P = 0.08
thopaedic Association (JOA) Not significant
RCT
score from baseline: scale

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 54

 Musculoskeletal disorders
Herniated lumbar disc
Ref Results and statistical Effect
(type) Population Outcome, Interventions analysis size Favours
from −6 to +29; higher scores
indicating better outcomes
27 with macrodiscectomy
27 with microdiscectomy

[57]
40 people with sci- Pain measured on VAS: 0 = no P = 0.15
atica that did not pain, 10 = worst pain ever expe-
RCT
respond to conser- rienced , 24 months
vative treatment,
mean 0, range (0−6) with open Not significant
and posterolateral
discectomy
herniated lumbar
disc observed on mean 1, range (0−3) with mi-
MRI scans crodiscectomy

-
Functional improvement
Compared with microdiscectomy Standard discectomy and microdiscectomy may be equally effective at reducing
disability and enabling return to work at 1 month (low-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Oswestry Disability Index
[57]
40 people with sci- Oswestry Disability Index P = 0.87
atica that did not (score range 0−100) , 24
RCT
respond to conser- months
vative treatment,
Median (range) score: 10 (0−30) Not significant
and posterolateral
with open discectomy
herniated lumbar
disc observed on Median (range) score: 10 (0−22)
MRI scan with microdiscectomy

Return to work and normal activities

[57]
40 people with sci- Mean time to return to work P = 0.79
atica that did not and normal activities between
RCT
respond to conser- groups
vative treatment,
21 days with open discectomy Not significant
and posterolateral
herniated lumbar 21 days with microdiscectomy
disc observed on
MRI scans

-
[23] [56]
No data from the following reference on this outcome.

-
Patient perception of improvement
Compared with microdiscectomy Standard discectomy and microdiscectomy seem equally effective at increasing
the number of people with lumbar disc herniation who rate their surgeries as "good", "almost recovered", or "totally
recovered" at 1 year (moderate-quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Patient perception of improvement
[58]
60 people with Proportion of people who rated RR 1.08
lumbar disc hernia- their operative outcome as
RCT 95% CI 0.78 to 1.20
tion "good", "almost recovered", or
[23] "totally recovered" , 1 year The RCT also found similar
In review
changes in both groups in preop-
26/30 (87%) with standard dis-
erative and postoperative pain Not significant
cectomy
scores, and in time taken to re-
24/30 (80%) with microdiscecto- turn to work (pain scores: visual
my analogue scale [VAS]; P value
not reported; time taken to return
to work: 10 weeks in both groups)

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 55

 Musculoskeletal disorders
Herniated lumbar disc
-
[23] [56] [57]
No data from the following reference on this outcome.

-
Adverse effects

-
-
[23] [56] [57] [58]
No data from the following reference on this outcome.

-
-
-
Further information on studies
[54]
The RCT comparing standard discectomy versus conservative treatment had considerable crossover between
the two treatment groups. Of 66 people randomised to receive conservative treatment, 17 received surgery; of
60 people randomised to receive surgery, one refused the operation. The results presented above are based
on an intention-to-treat analysis.
[55]
This RCT had nearly 50% crossover in both directions. Of 232 people randomised to surgery and included in
the analysis, only 140/232 (60%) had surgery. Of the 240 people randomised to non-operative care and included
in the intention-to-treat analysis, 107/204 (52%) had surgery. The 3-year and 4-year follow-up results from this
[60]
study were published separately. The follow-up at these end points was <80% of randomised participants,
so data are not reported above. Similar results for Short Form (SF)-36 scores measuring improvement in pain
and Oswestry Disability Index measuring reduction in disability were observed between the group of people
who had surgery and the group of people who had non-surgical treatment at both 3 and 4 years.
[56]
The RCT analysed the difference in scores between groups after surgery, without comparing the change in
score from baseline to end point between groups. The baseline scores for sciatic pain intensity and Japanese
Orthopaedic Association scores did not differ significantly at baseline or after surgery. There was, however, a
significant difference in leg pain scores at baseline as well as after surgery. Therefore, analysis of the data
found neither surgery better than the other.
[57]
The RCT stated that only those participants with a final postoperative follow-up period of at least 2 years were
included in this study. The RCT reported no information on the number of people who withdrew. It is unclear
whether 40 people were originally recruited for the study, or whether this was adjusted based on the follow-up
rate.
[58]
The RCT also found similar changes in both groups in time taken to return to work (10 weeks in both groups).

-
-
Comment: Standard discectomy versus epidural corticosteroid injection:
See comment in epidural corticosteroid injections, p 4 .

OPTION AUTOMATED PERCUTANEOUS DISCECTOMY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

• For GRADE evaluation of interventions for Herniated lumbar disc, see table, p 62 .
• We found no clinically important results from RCTs about automated percutaneous discectomy compared with
either conservative treatment, standard discectomy, or microdiscectomy.

Benefits and harms

Automated percutaneous discectomy versus conservative treatment:
We found no systematic review or RCTs.

-
-
Automated percutaneous discectomy versus standard discectomy:
One systematic review (search date not reported) identified no RCTs comparing automated percutaneous discectomy
[61]
versus standard discectomy.

-
© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 56
 Musculoskeletal disorders
Herniated lumbar disc
-
Automated percutaneous discectomy versus microdiscectomy:
[23] [62]
We found one systematic review (search date 2007), which identified one RCT that met our inclusion criteria.
The review did not perform a meta-analysis. One identified RCT did not meet our inclusion criteria due to a high follow-
up loss (>20%) and is not discussed further.

-
Pain
Compared with microdiscectomy Automated percutaneous discectomy may be less effective at increasing treatment
success rates (very-low quality evidence).

Ref Results and statistical Effect

(type) Population Outcome, Interventions analysis size Favours
Treatment success
[62]
71 people with ra- Proportion of people with out- P <0.001
diographical confir- come classified as "success"
RCT Trial stopped prematurely, after
mation of disc her- by clinician and masked ob-
an interim analysis at 6 months
niation server (details not reported)
microdiscectomy
9/31 (29%) with automated per-
cutaneous discectomy
32/40 (80%) with microdiscecto-
my

-
Adverse effects

-
-
[62]
No data from the following reference on this outcome.

-
-
-
Further information on studies

-
-
Comment: None.

OPTION LASER DISCECTOMY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

• For GRADE evaluation of interventions for Herniated lumbar disc, see table, p 62 .
• We found no direct information from RCTs about laser discectomy for the treatment of people with symptomatic
herniated lumbar disc.

Benefits and harms

Laser discectomy:
[61] [23] [63] [64]
Four systematic reviews (search dates not reported, 2007, 2000, and 2009 ) found no RCTs on the
[64]
effectiveness of laser discectomy that met Clinical Evidence reporting criteria. One of the reviews identified ob-
servational studies, ranging from case reports to large non-randomised studies (see further information on studies).

-
-
-

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 57

 Musculoskeletal disorders
Herniated lumbar disc
Further information on studies
[64]
This systematic review found many observational studies on percutaneous disc decompression with laser as-
sisted disc removal. Most studies did not meet the quality reporting criteria of the systematic review (lumbar
disc pain of at least 3 months' duration; treatment with percutaneous laser disc compression; minimum follow-
up of 12 months; at least 50 participants included), but of the 10 that did, all showed a positive effect on pain
relief. Several studies reported adverse effects or complications. Overall the most frequently reported complication
was spondylodiscitis, which ranged from 0% (4 studies) to 1.2% (1 study). In one study of 164 people, there
was 1 case of an instrument tip being faulty, 12 cases of postoperative dermatomal dysaesthesia, and 2 cases
of reflex sympathetic dystrophy. In one retrospective study of 658 people, 1.1% reported intraoperative compli-
cations and 1.5% reported postoperative complications, including 4 radicular deficits, 3 incidences of L5 nerve
root injury, 2 incidences of vascular injuries, 1 incidence of sigmoid artery injury, 1 incidence of anomalous ili-
olumbar artery injury, and 1 incidence of transverse process injury. There was a case report of subacute cauda
equine syndrome.

-
-
Comment: None.

OPTION PERCUTANEOUS DISC DECOMPRESSION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

• For GRADE evaluation of interventions for Herniated lumbar disc, see table, p 62 .
• We found no direct information from RCTs about percutaneous disc decompression for the treatment of people
with symptomatic herniated lumbar disc.

Benefits and harms

Percutaneous disc decompression:
We found one systematic review (search date 2006), which found no RCTs of percutaneous disc decompression
[65]
for lumbar disc herniation.

-
-
-
Further information on studies
[65]
The systematic review also searched for non-experimental descriptive studies, expert opinion, and clinical ex-
perience of respected authorities. These data are not included in this review.

-
-
Comment: We also found a systematic review on percutaneous disc decompression that was not written in
[66]
English. We are currently awaiting full text translation and we will assess this for inclusion in
our next update.

GLOSSARY
Autotraction The person provides the traction force on the traction table by pulling on the bar on the head of the
table while his or her pelvis is held by a girdle and chain to the lower end of the table.
Laser discectomy The surgeon places a laser through a delivery device that has been directed under radiographic
control to the disc, and removes the disc material using the laser. It uses many of the same techniques used in au-
tomated percutaneous discectomy.
Microdiscectomy Removal of protruding disc material, using an operating microscope to guide surgery.
Automated percutaneous discectomy Percutaneous disc decompression using a combined irrigation, suction,
and cutting device inserted through a cannula.
Cauda equina syndrome Compression of the cauda equina, causing symptoms that include changes in perineal
sensation (saddle anaesthesia) and loss of sphincter control. The cauda equina is a collection of spinal roots de-
scending from the lower part of the spinal cord, which occupy the vertebral canal below the spinal cord.
Japanese Orthopaedic Association (JOA) score This score is for clinical symptoms in people with herniated
lumbar disc. Functionality and pain are measured across 4 parameters, on a scale from −6 to +29, with higher scores
indicating better outcomes: first, subjective symptoms (0–9 points; low back pain leg pain, tingling gait, or both);

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 58

 Musculoskeletal disorders
Herniated lumbar disc
second, clinical signs (0–6 points; straight leg raising test sensory disturbance motor disturbance); third, restriction
in activities (0–14 points; turn over while lying, standing, washing, leaning forward, sitting for about 1 hour, lifting or
holding a heavy object, walking); and fourth, urinary bladder function (–6 points maximum).
Lasègue's sign The limitation of straight leg raising in a supine position usually associated with lumbar nerve root
compression. Also, in sciatica, added foot dorsiflexion to a straight leg raise results in more pain.
Low-quality evidence Further research is very likely to have an important impact on our confidence in the estimate
of effect and is likely to change the estimate.
Manual traction A form of passive traction. The person lies supine on a plinth with varying degrees of flexion in the
hip and knee joints. The traction force is exerted by the therapist using a belt placed around the therapist's back or
hips and attached behind and below the person's knees. The traction force is adjusted by the therapist according to
the patient's symptoms, with a maximum force of about 30 kg as measured by a force transducer in the belt.
Moderate-quality evidence Further research is likely to have an important impact on our confidence in the estimate
of effect and may change the estimate.
Oswestry Disability Index Back-specific, self-reported questionnaire measuring pain and function in completing
physical and social activities. The scale score ranges from 0 (no disability) to 100 (maximum disability).
Passive traction The person lies supine on a traction table with thighs flexed and supported by pillow over knees.
The traction force is adjusted manually by the therapist to about 35% of person's body weight, measured by a dy-
namometer, and then maintained by a chain connection to the foot of the bed. The traction force is adjusted regularly
during the treatment session.
Percutaneous disc decompression Any technique for discectomy performed through percutaneous portals inserted
with x-ray control, generally removing intradiscal fragments rather than sequestrated extradiscal fragments.
Roland Morris Disability Questionnaire A 24-item, self-reported, disability scale specific to back pain recommended
for use in primary care and community studies. Measures daily function in completing activities affected by back
pain. The scale score ranges from 0 (no disability) to 24 (severe disability).
Short Form (SF)-36 A health-related quality-of-life scale across 8 domains: limitations in physical activities (physical
component), limitations in social activities, limitations in usual role activities owing to physical problems, pain, psy-
chological distress and wellbeing (mental health component), limitations in usual role activities because of emotional
problems, energy and fatigue, and general health perceptions.
Standard discectomy Surgical removal, in part or whole, of an intervertebral disc, generally with loop magnification
(i.e., eyepieces).
Very low-quality evidence Any estimate of effect is very uncertain.

SUBSTANTIVE CHANGES
[17] [18] [19] [21]
Corticosteroids (epidural injections) New evidence added. Categorisation unchanged (Unknown
effectiveness), as there remains insufficient evidence to judge the effects of this intervention because the evidence
is inconsistent.
[64]
Laser discectomy New evidence added. Categorisation unchanged (Unknown effectiveness), as there remains
insufficient evidence to judge the effects of this intervention.

REFERENCES
1. Fardon DF, Milette PC. Nomenclature and classification of lumbar disc pathology: 10. Kelsey JL, Githens P, O'Connor T, et al. Acute prolapsed lumbar intervertebral
recommendations of the Combined Task Forces of the North American Spine disc: an epidemiologic study with special reference to driving automobiles and
Society, American Society of Spine Radiology, and American Society of Neuro- cigarette smoking. Spine 1984;9:608–613.[PubMed]
radiology. Spine 2001;26:E93–E113.[PubMed] 11. Pedrini-Mille A, Weinstein JN, Found EM, et al. Stimulation of dorsal root ganglia
2. Boden SD. The use of radiographic imaging studies in the evaluation of patients and degradation of rabbit annulus fibrosus. Spine 1990;15:1252–1256.[PubMed]
who have degenerative disorders of the lumbar spine. J Bone Joint Surg Am 12. Deyo RA, Weinstein JN. Low back pain. N Engl J Med
1996;78:114–125.[PubMed] 2001;344:363–370.[PubMed]
3. Borenstein DG, O'Mara JW Jr, Boden SD, et al. The value of magnetic resonance 13. Altman DG, Machin D, Bryant TN, et al, eds. Statistics with confidence. 2nd ed.
imaging of the lumbar spine to predict low-back pain in asymptomatic subjects. London: BMJ Books, 2000.
J Bone Joint Surg Am 2001;83-A:1306–1311.[PubMed]
14. Abramson JH, Gahlinger PM. Computer programs for epidemiologists. Salt Lake
4. Andersson G. Epidemiology of spinal disorders. In: Frymoyer JW, Ducker TB, City: Sagebrush Press, 2001.
Hadler NM, et al, eds. The adult spine: principles and practice. New York, NY:
15. Vroomen PC, de Krom MC, Slofstra PD, et al. Conservative treatment of sciatica:
Raven Press, 1997:93–141.
a systematic review. J Spinal Disord 2000;13:463–469.[PubMed]
5. Heliovaara M. Epidemiology of sciatica and herniated lumbar intervertebral disc.
16. DePalma MJ, Bhargava A, Slipman CW. A critical appraisal of the evidence for
Helsinki, Finland: The Social Insurance Institution, 1988.
selective nerve root injection in the treatment of lumbosacral radiculopathy. Arch
6. Postacchini F, Cinotti G. Etiopathogenesis. In: Postacchini F, ed. Lumbar disc Phys Med Rehabil 2005;86:1477–1483.[PubMed]
herniation. New York: Spring-Verlag, 1999:151–164.
17. Conn A, Buenaventura RM, Datta S, et al. Systematic review of caudal epidural
7. Friberg S, Hirsch C. Anatomical and clinical studies on lumbar disc degeneration. injections in the management of chronic low back pain. Pain Physician
Acta Orthop Scand 1949;19:222–242. 2009;12:109–135.[PubMed]
8. Schultz A, Andersson G, Ortengren R, et al. Loads on the lumbar spine. Validation 18. Buenaventura RM, Datta S, Abdi S, et al. Systematic review of therapeutic lumbar
of a biomechanical analysis by measurements of intradiscal pressures and myo- transforaminal epidural steroid injections. Pain Physician
electric signals.J Bone Joint Surg Am 1982;64:713–720.[PubMed] 2009;12:233–251.[PubMed]
9. Jensen MC, Brant-Zawadzki MN, Obuchowski N, et al. Magnetic resonance 19. Parr AT, Diwan S, Abdi S, et al. Lumbar interlaminar epidural injections in man-
imaging of the lumbar spine in people without back pain. N Engl J Med aging chronic low back and lower extremity pain: a systematic review. Pain
1994;331:69–73.[PubMed] Physician 2009;12:163–188.[PubMed]

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 59

 Musculoskeletal disorders
Herniated lumbar disc
20. Valat JP, Giraudeau B, Rozenberg S, et al. Epidural corticosteroid injections for 44. Waddell G, Feder G, Lewis M. Systematic reviews of bed rest and advice to stay
sciatica: a randomised, double blind, controlled clinical trial. Ann Rheum Dis active for acute low back pain. Br J Gen Prac 1997;47:647–652. Search date
2003;62:639–643.[PubMed] 1996.[PubMed]
21. Tafazal S, Ng L, Chaudhary N, et al. Corticosteroids in peri-radicular infiltration 45. Rattanatharn R, Sanjaroensuttikul N, Anadirekkul P, et al. Effectiveness of lumbar
for radicular pain: a randomised double blind controlled trial. One year results traction with routine conservative treatment in acute herniated disc syndrome. J
and subgroup analysis. Eur Spine J 2009;18:1220–1225.[PubMed] Med Assoc Thai 2004;87:S272–S277.[PubMed]
22. Buchner M, Zeifang F, Brocai DR, et al. Epidural corticosteroid injection in the 46. Ljunggren AE, Weber H, Larsen S. Autotraction versus manual traction in patients
conservative management of sciatica. Clin Orthop Relat Res with prolapsed lumbar intervertebral discs. Scan J Rehabil Med
2000;375:149–156.[PubMed] 1984;16:117–124.[PubMed]
23. Gibson JN, Waddell G. Surgical interventions for lumbar disc prolapse. In: The 47. Tesio L, Merlo A. Autotraction versus passive traction: an open controlled study
Cochrane Library, Issue 5, 2014. Chichester, UK: John Wiley & Sons Ltd. Search in lumbar disc herniation. Arch Phys Med Rehabil 1993;74:871–876.[PubMed]
date 2007.[PubMed] 48. van der Heijden GJ, Beurskens AJ, Koes BW, et al. The efficacy of traction for
24. Buttermann GR. Treatment of lumbar disc herniation: epidural steroid injection back and neck pain: a systematic, blinded review of randomized clinical trial
compared with discectomy. A prospective, randomized study. J Bone Joint Surg methods. Phys Ther 1995;75:93–104. Search date 1992.[PubMed]
Am 2004;86-A:670–679.[PubMed] 49. Chen W, Yang AT, Dai MT, et al. Observation on therapeutic effect of elec-
25. Korhonen T, Karppinen J, Paimela L, et al. The treatment of disc herniation-in- troacupuncture under continuous traction for treatment of lumbar disc herniation.
duced sciatica with infliximab: results of a randomized, controlled, 3-month follow- Zhongguo Zhen Jiu 2009;29:967–969. [in Chinese][PubMed]
up study. Spine 2005;30:2724–2728.[PubMed] 50. Osterman H, Seitsalo S, Karppinen J, et al. Effectiveness of microdiscectomy
26. Genevay S, Viatte S, Finckh A, et al. Adalimumab in severe and acute sciatica: for lumbar disc herniation: a randomized controlled trial with 2 years of follow-up.
a multicenter, randomized, double-blind, placebo-controlled trial. Arthritis Rheum Spine 2006;31:2409–2414.[PubMed]
2010;62:2339–2346.[PubMed] 51. Peul WC, van den Hout WB, Brand R, et al. Prolonged conservative care versus
27. Wang BX, La JL. Therapeutic effects of electroacupuncture and diclofenac on early surgery in patients with sciatica caused by lumbar disc herniation: two year
herniation of lumbar intervertebral disc. Chin J Clin Rehabil 2004;8:3413–3415. results of a randomised controlled trial. BMJ 2008;336:1355–1358.[PubMed]
28. Solomon SD, McMurray JJ, Pfeffer MA, et al. Cardiovascular risk associated with 52. Hermantin FU, Peters T, Quartararo L, et al. A prospective, randomized study
celecoxib in a clinical trial for colorectal adenoma prevention. New Engl J Med comparing the results of open discectomy with those of video-assisted arthro-
2005;352:1071–1080.[PubMed] scopic microdiscectomy. J Bone Joint Surg Am 1999;81:958–965.[PubMed]
29. Bresalier RS, Sandler RS, Quan H, et al. Cardiovascular events associated with 53. Lu X, Peng H, Ling S, et al. Therapeutic effect of microsurgery lumbar discectomy
rofecoxib in a colorectal adenoma chemoprevention trial. New Engl J Med on single-level lumbar disc protrusion. Zhongguo Xiu Fu Chong Jian Wai Ke Za
2005;352:1092–1102.[PubMed] Zhi 2009/8;23:909–912. [in Chinese][PubMed]
30. Snelling NJ. Spinal manipulation in patients with disc herniation: a critical review 54. Weber H. Lumbar disc herniation: a controlled, prospective study with ten years
of risk and benefit. Int J Osteopath Med 2006;9:77–84. of observation. Spine 1983;8:131–140.[PubMed]
31. Santilli V, Beghi E, Finucci S. Chiropractic manipulation in the treatment of acute 55. Weinstein JN, Tosteson TD, Lurie JD, et al. Surgical vs nonoperative treatment
back pain and sciatica with disc protrusion: a randomized double-blind clinical for lumbar disk herniation: the Spine Patient Outcomes Research Trial (SPORT):
trial of active and simulated spinal manipulations. Spine J a randomized trial. JAMA 2006;296:2441–2450.[PubMed]
2006;6:131–137.[PubMed] 56. Katayama Y, Matsuyama Y, Yoshihara H, et al. Comparison of surgical outcomes
32. Shekelle PG, Adams AH, Chassin MR, et al. Spinal manipulation for low-back between macro discectomy and micro discectomy for lumbar disc herniation: a
pain. Ann Intern Med 1992;117:590–598. Search date not reported.[PubMed] prospective randomized study with surgery performed by the same spine surgeon.
33. Stevinson C, Ernst E. Risks associated with spinal manipulation. Am J Med J Spinal Disord Tech 2006;19:344–347.[PubMed]
2002;112:566–570. Search date 2001.[PubMed] 57. Righesso O, Falavigna A, Avanzi O. Comparison of open discectomy with mi-
34. Oliphant D. Safety of spinal manipulation in the treatment of lumbar disk hernia- croendoscopic discectomy in lumbar disc herniations: results of a randomized
tions: a systematic review and risk assessment. J Manipulative Physiol Ther controlled trial. Neurosurgery 2007;61:545–549.[PubMed]
2004;27:197–210.[PubMed] 58. Tullberg T, Isacson J, Weidenhielm L. Does microscopic removal of lumber disc
35. Liu J, Zhang S. Treatment of protrusion of lumbar intervertebral disc by pulling herniation lead to better results than the standard procedure? Results of a one-
and turning manipulations. J Tradit Chin Med 2000;20:195–197.[PubMed] year randomized study. Spine 1993;18:24–27.[PubMed]
36. Oppenheim JS, Spitzer DE, Segal DH. Nonvascular complications following 59. Henriksen L, Schmidt V, Eskesen V, et al. A controlled study of microsurgery
spinal manipulation. Spine J 2005;5:660–666.[PubMed] versus standard lumbar discectomy. Br J Neurosurg 1996;10:289–293.[PubMed]
37. Zhang WB, Cao Y, Sun YA, et al. Manipulative reduction for lumbar intervertebral 60. Weinstein JN, Lurie JD, Tosteson TD, et al. Surgical versus nonoperative treat-
disc herniation: a controlled clinical trial. Zhongguo Gu Shang 2008;21:273–275. ment for lumbar disc herniation: four-year results for the Spine Patient Outcomes
[in Chinese][PubMed] Research Trial (SPORT). Spine 2008;33:2789–2800.[PubMed]
38. Smith LA, Oldman AD, McQuay HJ, et al. Teasing apart quality and validity in 61. Hoffman RM, Wheeler KJ, Deyo RA. Surgery for herniated lumbar discs: a liter-
systematic reviews: an example from acupuncture trials in chronic neck and back ature synthesis. J Gen Intern Med 1993;8:487–496. Search date not report-
pain. Pain 2000;86:119–132. Search date 1998.[PubMed] ed.[PubMed]
39. Wang Y-Q, Tang L-X, Pan C-Q. Amelioration of ventral acupuncture therapy on 62. Chatterjee S, Foy PM, Findlay GF. Report of a controlled clinical trial comparing
the pain symptom in patients with lumbar disc herniation. Zhongguo Linchuang automated percutaneous lumbar discectomy and microdiscectomy in the treatment
Kangfu 2005;9:122–123. of contained lumbar disc herniation. Spine 1995;20:734–738.[PubMed]
40. Ljunggren AE, Walker L, Weber H, et al. Manual traction versus isometric exer- 63. Boult M, Fraser RD, Jones N, et al. Percutaneous endoscopic laser discectomy.
cises in patients with herniated intervertebral lumbar discs. Physiother Theory Aust N Z J Surg 2000;70:475–479. Search date 2000.[PubMed]
Pract 1992;8:207–213. 64. Singh V, Manchikanti L, Benyamin RM, et al. Percutaneous lumbar laser disc
41. Zhao Y. Assessment of the short-term and long-term effects of exercise prescrip- decompression: a systematic review of current evidence. Pain Physician
tion and rehabilitation education on lumbar intervertebral disc herniation patients. 2009;12:573–588.[PubMed]
Chin J Clin Rehabil 2004;8:8314–8315. 65. Goupille P, Mulleman D, Mammou S, et al. Percutaneous laser disc decompres-
42. Zhang JF, Chen WH. Curative effect of nonoperative therapy for the lumbar disc sion for the treatment of lumbar disc herniation: a review. Semin Arthrit Rheum
herniation. Chin J Clin Rehabil 2004;8:2006–2007. 2007;37:20–30.[PubMed]
43. Vroomen PC, de Krom MC, Wilmink JT, et al. Lack of effectiveness of bed rest 66. Felder Puig R, Falkner E, Geiger Gritsch S, et al. Percutaneous nucleotomy and
for sciatica. N Engl J Med 1999;340:418–423.[PubMed] percutaneous laser disk decompression. Systematic Review. Vienna: Ludwig
Boltzmann Institut fuer Health Technology Assessment, 2009.

Joanne L Jordan
Research Information Manager
Arthritis Research UK Primary Care Centre
Primary Care Sciences, Keele University
Keele
UK

Kika Konstantinou
Senior Clinical Lecturer and and Spinal Physiotherapy Specialist
Arthritis Research UK Primary Care Centre
Primary Care Sciences, Keele University
Keele
UK

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 60

 Musculoskeletal disorders
Herniated lumbar disc
John O'Dowd
Consultant Spinal Surgeon
RealHealth Institute
London
UK

Competing interests: JJ, KK, and JOD declare that they have no competing interests.

Disclaimer

The information contained in this publication is intended for medical professionals. Categories presented in Clinical Evidence indicate a
judgement about the strength of the evidence available to our contributors prior to publication and the relevant importance of benefit and
harms. We rely on our contributors to confirm the accuracy of the information presented and to adhere to describe accepted practices.
Readers should be aware that professionals in the field may have different opinions. Because of this and regular advances in medical research
we strongly recommend that readers' independently verify specified treatments and drugs including manufacturers' guidance. Also, the
categories do not indicate whether a particular treatment is generally appropriate or whether it is suitable for a particular individual. Ultimately
it is the readers' responsibility to make their own professional judgements, so to appropriately advise and treat their patients. To the fullest
extent permitted by law, BMJ Publishing Group Limited and its editors are not responsible for any losses, injury or damage caused to any
person or property (including under contract, by negligence, products liability or otherwise) whether they be direct or indirect, special, inci-
dental or consequential, resulting from the application of the information in this publication.

© BMJ Publishing Group Ltd 2011. All rights reserved. .......................................................... 61

 Musculoskeletal disorders
Herniated lumbar disc
GRADE Evaluation of interventions for Herniated lumbar disc.
-
Important out-
comes Functional improvement, Need for surgery, Pain, Patient perception of improvement, Quality of life
Type
Studies (Partici- of evi- Consis- Direct- Effect
pants) Outcome Comparison dence Quality tency ness size GRADE Comment
What are the effects of drug treatments for herniated lumbar disc?
[16] [17]
8 (705) Pain Epidural corticosteroid injections 4 –1 –1 0 0 Low Quality point deducted for incomplete reporting of results.
[19] [20] [21]
versus no epidural corticosteroid in- Consistency point deducted for different results at different
jection end points
[20] [19]
4 (386) Functional improve- Epidural corticosteroid injections 4 –1 0 0 0 Moderate Quality point deducted for incomplete reporting of results
[21]
ment versus no epidural corticosteroid in-
jection
[15] [20]
2 (417) Patient perception Epidural corticosteroid injections 4 0 –1 –1 0 Low Consistency point deducted for different results at different
of improvement versus no epidural corticosteroid in- end points. Directness point deducted for not defining outcome
jection measured
[16] [19]
2 (213) Need for surgery Epidural corticosteroid injections 4 –1 –1 –1 0 Very low Quality point deducted for sparse data. Consistency point
versus no epidural corticosteroid in- deducted conflicting results among trials. Directness point
jection deducted for narrow included population
[22]
1 (36) Pain Epidural corticosteroid plus conser- 4 –1 0 –1 0 Low Quality point deducted for sparse data. Directness point de-
vative non-operative treatment ver- ducted for wide range of interventions used in comparison,
sus conservative treatment alone making the results difficult to apply in clinical practice
[22]
1 (36) Functional improve- Epidural corticosteroid plus conser- 4 –1 0 –1 0 Low Quality point deducted for sparse data. Directness point de-
ment vative non-operative treatment ver- ducted for wide range of interventions used in comparison,
sus conservative treatment alone making the results difficult to apply in clinical practice
[22]
1 (36) Need for surgery Epidural corticosteroid plus conser- 4 –1 0 –1 0 Low Quality point deducted for sparse data. Directness point de-
vative non-operative treatment ver- ducted for wide range of interventions used in comparison,
sus conservative treatment alone making the results difficult to apply in clinical practice
[24]
1 (100) Pain Epidural corticosteroid injection ver- 4 –2 –1 0 0 Very low Quality points deducted for sparse data and incomplete report-
sus discectomy ing of results. Consistency point deducted for different results
at different end points
[24]
1 (100) Functional improve- Epidural corticosteroid injection ver- 4 –2 –1 0 0 Very low Quality points deducted for sparse data and incomplete report-
ment sus discectomy ing of results. Consistency point deducted for different results
at different end points
[25]
1 (41) Pain Infliximab versus placebo 4 –1 0 0 0 Moderate Quality point deducted for sparse data
[25]
1 (41) Functional improve- Infliximab versus placebo 4 –2 0 0 0 Low Quality points deducted for sparse data and incomplete report-
ment ing of results at 12 weeks
[25]
1 (41) Need for surgery Infliximab versus placebo 4 –1 0 0 0 Moderate Quality point deducted for sparse data
[15]
3 (321) Pain NSAIDs versus placebo 4 0 0 –2 0 Low Directness points deducted for limited range of NSAIDs as-
sessed and for use of unclear outcome measure in meta-
analysis

© BMJ Publishing Group Ltd 2011. All rights reserved. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

 Musculoskeletal disorders
Herniated lumbar disc
Important out-
comes Functional improvement, Need for surgery, Pain, Patient perception of improvement, Quality of life
Type
Studies (Partici- of evi- Consis- Direct- Effect
pants) Outcome Comparison dence Quality tency ness size GRADE Comment
[27]
1 (40) Pain NSAIDs versus electroacupuncture 4 –1 0 –2 0 Very low Quality point deducted for sparse data. Directness points de-
ducted for possible inclusion of people without disc herniation
and uncertainty about generalisability of outcomes measured
[27]
1 (40) Functional improve- NSAIDs versus electroacupuncture 4 –1 0 –1 0 Low Quality point deducted for sparse data. Directness point de-
ment ducted for possible inclusion of people without disc herniation
What are the effects of non-drug treatments for herniated lumbar disc?
[31]
1 (102) Pain Spinal manipulation versus placebo 4 –1 0 0 0 Moderate Quality point deducted for sparse data
or sham treatment
[31]
1 (102) Functional improve- Spinal manipulation versus placebo 4 –2 0 0 0 Low Quality points deducted for sparse data and for incomplete
ment or sham treatment reporting of results
[30]
1 (233) Patient perception Spinal manipulation versus heat 4 –4 0 0 0 Very low Quality points deducted for incomplete reporting of results
of improvement treatment and for methodological flaws (not reporting group baseline
characteristics, uncertainty about intention-to-treat analysis,
poor follow-up, and uncertainty about groups receiving equal
number of treatments)
[30]
1 (322) Pain Spinal manipulation versus exercise 4 –3 0 –1 0 Very low Quality points deducted for incomplete reporting of results
therapy and methodological flaws (not reporting group baseline char-
acteristics and uncertainty about blinding). Directness point
deducted for inclusion of people without herniated disc
[30]
1 (322) Patient perception Spinal manipulation versus exercise 4 –3 0 –1 0 Very low Quality points deducted for incomplete reporting of results
of improvement therapy and methodological flaws (not reporting group baseline char-
acteristics, uncertainty about blinding). Directness point de-
ducted for inclusion of people without herniated disc
[30]
1 (322) Patient perception Spinal manipulation versus traction 4 –3 0 –1 0 Very low Quality points deducted for incomplete reporting of results
of improvement and methodological flaws (not reporting group baseline char-
acteristics and uncertainty about blinding). Directness point
deducted for inclusion of people without herniated disc
[35]
1 (112) Functional improve- Spinal manipulation versus traction 4 –2 0 0 0 Low Quality points deducted for sparse data and uncertainty about
ment end point
[38]
1 (30) Pain Acupuncture versus sham acupunc- 4 –2 0 –2 0 Very low Quality points deducted for sparse data and incomplete report-
ture ing of results. Directness points deducted for inclusion of
people without disc herniation
[38]
1 (42) Pain Laser acupuncture versus sham 4 –1 0 –2 0 Very low Quality point deducted for sparse data. Directness points de-
laser acupuncture ducted for no long-term results and for inclusion of a wide
population making it unclear whether the data are generalis-
able to herniated disc
[39]
1 (58) Pain Adding acupuncture to manipulation 4 –2 0 –1 0 Very low Quality points deducted for sparse data and for unspecified
compared with manipulation alone follow-up time. Directness point deducted for no long-term
results

© BMJ Publishing Group Ltd 2011. All rights reserved. ............................................................................................................ 63

 Musculoskeletal disorders
Herniated lumbar disc
Important out-
comes Functional improvement, Need for surgery, Pain, Patient perception of improvement, Quality of life
Type
Studies (Partici- of evi- Consis- Direct- Effect
pants) Outcome Comparison dence Quality tency ness size GRADE Comment
[40] [30]
2 (372) Pain Exercise therapy versus traction 4 –2 0 –2 0 Very low Quality points deducted for incomplete reporting of results
and lack of blinding in 1 RCT. Directness points deducted for
poorly defined outcome measure in 1 RCT and for inclusion
of people without herniated disc
[41]
1 (40) Functional improve- Adding exercise plus education to 4 –1 –1 0 0 Low Quality point deducted for sparse data. Consistency point
ment conventional non-surgical treatment deducted as result sensitive to different methods of calculation
versus conventional non-surgical
treatment alone
[42]
1 (110) Pain Massage/manipulation versus mas- 4 –1 0 –2 0 Very low Quality point deducted for sparse data. Directness points de-
sage/manipulation plus functional ducted for unclear measurement of outcomes and for including
training exercises versus traction spinal massage techniques (uncertainty about whether results
using spinal techniques are comparable with results using
other massage techniques)
[42]
1 (110) Pain Massage/manipulation versus trac- 4 –1 0 –2 0 Very low Quality point deducted for sparse data. Directness points de-
tion ducted for unclear measurement of outcomes and for including
spinal massage techniques (uncertainty about whether results
using spinal techniques are comparable with results using
other massage techniques)
[43]
1 (183) Pain Bed rest versus no treatment 4 –1 0 –1 0 Low Quality point deducted for sparse data. Directness point de-
(watchful waiting) ducted as results were only in people with sciatica, so there
is uncertainty about generalisability of results to people with
herniated lumbar disc
[43]
1 (183) Functional improve- Bed rest versus no treatment 4 –1 0 –1 0 Low Quality point deducted for sparse data. Directness point de-
ment (watchful waiting) ducted as results were only in people with sciatica, so there
is uncertainty about generalisability of results to people with
herniated lumbar disc
[43]
1 (183) Patient perception Bed rest versus no treatment 4 –1 0 –1 0 Low Quality point deducted for sparse data. Directness point de-
of improvement (watchful waiting) ducted for uncertainty about generalisability of results for
people with herniated lumbar disc
[15]
1 (329) Pain Traction versus no traction or sham 4 –1 0 –2 0 Very low Quality point deducted for incomplete reporting of results. Di-
traction rectness points deducted for inclusion of people without disc
herniation and for inclusion of wide range of traction tech-
niques and comparators
[45]
1 (102) Functional im- Traction versus no traction or sham 4 –1 0 –1 0 Low Quality point deducted for sparse data. Directness point de-
provement traction ducted for use of co-intervention
[45]
1 (102) Patient perception Traction versus no traction or sham 4 –1 0 –1 0 Low Quality point deducted for sparse data. Directness point de-
of improvement traction ducted for use of co-intervention
[46] [47]
2 (93) Functional improve- Autotraction versus passive traction 4 –3 –1 0 0 Very low Quality points deducted for sparse data, incomplete reporting
ment of results and no intention-to-treat analysis. Consistency point
deducted for conflicting results, perhaps owing to different
measures of outcome used

© BMJ Publishing Group Ltd 2011. All rights reserved. ............................................................................................................ 64

 Musculoskeletal disorders
Herniated lumbar disc
Important out-
comes Functional improvement, Need for surgery, Pain, Patient perception of improvement, Quality of life
Type
Studies (Partici- of evi- Consis- Direct- Effect
pants) Outcome Comparison dence Quality tency ness size GRADE Comment
What are the effects of surgery for herniated lumbar disc?
[50] [51]
2 (339) Pain Microdiscectomy versus conserva- 4 –1 –1 –1 0 Very low Quality point deducted for methodological flaw (high crossover
tive treatment between interventions). Consistency point deducted for differ-
ent results at different end points. Directness point deducted
for multiple interventions in comparison
[50] [51]
2 (339) Functional improve- Microdiscectomy versus conserva- 4 –1 –1 –1 0 Very low Quality point deducted for methodological flaw (high crossover
ment tive treatment between interventions). Consistency point deducted for differ-
ent results at different end points. Directness point deducted
for multiple interventions in comparison
[50]
1 (56) Quality of life Microdiscectomy versus conserva- 4 –1 0 –1 0 Low Quality point deducted for sparse data. Directness point de-
tive treatment ducted for multiple interventions in comparison
[51]
1 (283) Patient perception Microdiscectomy versus conserva- 4 –1 –1 –1 0 Very low Quality point deducted for methodological flaw (high crossover
of improvement tive treatment between interventions). Consistency point deducted for differ-
ent results at different end points. Directness point deducted
for multiple interventions in comparison
[52]
1 (60) Pain Video-assisted arthroscopic microdis- 4 –2 0 0 0 Low Quality points deducted for sparse data and incomplete report-
cectomy versus standard discectomy ing of results
[52]
1 (60) Patient perception Video-assisted arthroscopic microdis- 4 –1 0 –1 0 Low Quality point deducted for sparse data. Directness point de-
of improvement cectomy versus standard discectomy ducted for unclear outcome measure
[54] [55]
2 (627) Pain Standard discectomy versus conser- 4 –1 –1 0 0 Low Quality point deducted for high crossover between treatments.
vative treatment Consistency point deducted for different results at different
end points
[54] [55]
2 (627) Functional improve- Standard discectomy versus conser- 4 –1 –1 0 0 Low Quality point deducted for high crossover between treatments.
ment vative treatment Consistency point deducted for different results at different
end points
[23] [56]
5 (378) Pain Standard discectomy versus mi- 4 –1 0 –2 0 Very low Quality point deducted for sparse data. Directness points de-
[57]
crodiscectomy ducted for uncertainty about outcomes in 1 study and for un-
certainty about baseline differences in another study
[57]
1 (40) Functional improve- Standard discectomy versus mi- 4 –2 0 0 0 Low Quality points deducted for sparse data and unclear follow-up
ment crodiscectomy rate
[58]
1 (60) Patient perception Standard discectomy versus mi- 4 –1 0 0 0 Moderate Quality point deducted for sparse data
of improvement crodiscectomy
[62]
1 (71) Pain Automated percutaneous discectomy 4 –2 0 –1 0 Very low Quality points deducted for sparse data and premature termi-
versus microdiscectomy nation of the trial. Directness point deducted for unclear out-
come measure
We initially allocate 4 points to evidence from RCTs, and 2 points to evidence from observational studies. To attain the final GRADE score for a given comparison, points are deducted or added from this initial
score based on preset criteria relating to the categories of quality, directness, consistency, and effect size. Quality: based on issues affecting methodological rigour (e.g., incomplete reporting of results, quasi-
randomisation, sparse data [<200 people in the analysis]). Consistency: based on similarity of results across studies. Directness: based on generalisability of population or outcomes. Effect size: based on magnitude
of effect as measured by statistics such as relative risk, odds ratio, or hazard ratio.

-
© BMJ Publishing Group Ltd 2011. All rights reserved. ............................................................................................................ 65