A Cochrane Review of Superficial Heat or Cold for Low Back Pain
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
SPINE Volume 31, Number 9, pp 998 –1006
©2006, Lippincott Williams & Wilkins, Inc.
A Cochrane Review of Superficial Heat or Cold
for Low Back Pain
Simon D. French, MPH, BAppSc(Chiro),* Melainie Cameron, PhD, BAppSc(Osteo), MHSc(Research),†
Bruce F. Walker, DC, MPH, DrPH,‡ John W. Reggars, DC, MChiroSc,¶ and
Adrian J. Esterman, PhD, AStat, DLSHTM§
Study Design. Cochrane systematic review. apy provides a small short-term reduction in pain and
Objective. To assess the effects of superficial heat and disability in a population with a mix of acute and subacute
cold therapy for low back pain in adults. low back pain, and that the addition of exercise further
Summary of Background Data. Heat and cold are com- reduces pain and improves function. There is insufficient
monly used in the treatment of low back pain. evidence to evaluate the effects of cold for low back pain
Methods. We searched electronic databases from in- and conflicting evidence for any differences between heat
ception to October 2005. Two authors independently as- and cold for low back pain.
sessed inclusion, methodologic quality, and extracted Key words: systematic review, Cochrane Collabora-
data, using the criteria recommended by the Cochrane tion, heat, cold, low back pain. Spine 2006;31:998 –1006
Back Review Group.
Results. Nine trials involving 1,117 participants were
included. In two trials of 258 participants with a mix of
acute and subacute low back pain, heat wrap therapy Low back pain is a common complaint with the lifetime
significantly reduced pain after 5 days (weighted mean prevalence reported as ranging from 11% to 84%.1 The
difference [WMD], 1.06; 95% confidence interval [CI], cause of pain is nonspecific in about 95% of people pre-
0.68 –1.45, scale range, 0 –5) compared with oral placebo.
senting with acute low back pain, with serious condi-
One trial of 90 participants with acute low back pain found
that a heated blanket significantly decreased pain imme- tions being rare.2 Chronic low back pain is a well-
diately after application (WMD, ⫺32.20; 95% CI, ⫺38.69 to documented disabling condition, costly to both individuals
⫺25.71; scale range, 0 –100). One trial of 100 participants and society.3–5
with a mix of acute and subacute low back pain examined
Different healthcare disciplines commonly use heat
the additional effects of adding exercise to heat wrap and
found that it reduced pain after 7 days. and cold treatments for the treatment of low back
Conclusions. The evidence base to support the com- pain.6 –13 Both therapies are simple to apply and are in-
mon practice of superficial heat and cold for low back expensive. They may be used by people with low back
pain is limited, and there is a need for future higher- pain at home or may be employed by practitioners as
quality randomized controlled trials. There is moderate
evidence in a small number of trials that heat wrap ther-
part of a treatment regimen.
Traditionally, ice has been recommended for acute
injury and heat has been recommended for longer-term
injuries.14,15 Superficial heat methods convey heat by
From the *Australasian Cochrane Centre, Monash Institute of Health
Services Research, Monash University, Clayton, Victoria, Australia; conduction or convection. Superficial heat elevates the
†Centre for Ageing, Rehabilitation, Exercise, and Sport, School of Hu- temperature of tissues and provides the greatest effect at
man Movement, Recreation and Performance, Victoria University;
‡School of Chiropractic, Murdoch University, Murdoch, Western Aus-
0.5 cm or less from the surface of the skin. However,
tralia; §School of Nursing and Midwifery, University of South Austra- deep heating is achieved by converting another form of
lia; and ¶private practice, Ringwood, Victoria, Australia. energy to heat, for example, shortwave diathermy, mi-
Supported by the Australasian Cochrane Centre, Monash Institute of
Health Services Research, Monash University, Australia. crowave diathermy, and ultrasound.16 Superficial heat
This systematic review was originally published in The Cochrane Li- includes such methods as hot water bottles, heated
brary and we acknowledge John Wiley & Sons Ltd., on behalf of the stones, soft heated packs filled with grain, poultices, hot
Cochrane Collaboration, for authorizing this re-publication (French
SD, Cameron M, Walker BF, Reggars JW, Esterman AJ. Superficial towels, hot baths, saunas, steam, heat wraps, heat pads,
heat or cold for low back pain. The Cochrane Database of Systematic electric heat pads, and infra-red heat lamps. Cold ther-
Reviews 2006, Issue 1, Copyright Cochrane Library, reproduced with apy is used to reduce inflammation, pain, and edema.
permission).
The device(s)/drug(s) is/are FDA-approved or approved by correspond- Superficial cold includes cryotherapy, ice, cold towels,
ing national agency for this indication. cold gel packs, ice packs, and ice massage.
No funds were received in support of this work. No benefits in any
form have been or will be received from a commercial party related
Various national guidelines for the management of low
directly or indirectly to the subject of this manuscript. back pain have conflicting recommendation for heat and
There is consensus among the Editorial Board of the Back Review cold therapy. The U.S. Agency for Healthcare Research and
Group that strong evidence can only be provided by multiple higher
quality trials that replicate findings of other researchers in other set-
Quality guidelines found no evidence of benefit from the
tings. application of ice or heat for acute low back pain, however,
Address correspondence and reprint requests to Simon D. French, recommended self-application of heat or cold for patients
MPH, Australasian Cochrane Centre, Monash Institute of Health Ser-
vices Research, Locked Bag 29, Monash Medical Centre, Clayton, to provide temporary relief of symptoms.17 Other guide-
Victoria, Australia; E-mail: simon.french@med.monash.edu.au lines give different recommendations.18 –21
998Superficial Heat or Cold for LBP • French et al 999
Objectives. The objective of this review was to determine the databases were adapted accordingly. We also screened references
efficacy of superficial hot or cold therapies in reducing pain and of identified articles.
disability in low back pain in adults, aged 18 and older.
Methods of the Review
Criteria for Considering Studies for This Review Selection of Studies. One author (S.D.F.) conducted the
Types of Studies. We included randomized controlled trials searches and compiled all of the abstracts retrieved by the
(RCTs) and nonrandomized controlled clinical trials (CCTs) above search strategy. Two authors (S.D.F., B.F.W.) then inde-
comparing superficial hot or cold therapy to placebo, no ther- pendently applied the inclusion criteria to all of these abstracts.
apy, or to other therapies. If the eligibility of the study was not clear from the abstract,
then the full text of the article was obtained and assessed inde-
Types of Participants. Studies were selected that included pendently by the two authors. Any disagreement between the
participants 18 years of age or older, with the complaint of authors was resolved by discussion and consensus.
nonspecific low back pain. Trials that included participants
with pathologic causes of low back pain and low back pain Data Extraction and Management. Two authors (S.D.F. and
with radiculopathy were excluded. For the purpose of this re- M.C.) independently extracted the data onto a standard form.
view, the duration of back pain was defined as acute (⬍6 The data extraction form was pilot tested on one trial to min-
weeks), subacute (6 –12 weeks), or chronic (⬎12 weeks), as imize misinterpretation. Any disagreement between the au-
defined by the Cochrane Back Review Group.22 thors was resolved by discussion and consensus. We requested
additional study details and data from trial authors when the
Types of Interventions. Trials were included in which super- data reported were incomplete. Some data from the Nadler
ficial heat or cold therapy was administered to at least one studies23–25 and from the Mayer study26 were received from
group within the trial. Trials in which cointerventions (e.g., the authors and were incorporated into the table of included
exercise) were given were only included if the cointerventions studies (Table 2) and the results.
were similar across comparison groups. If cointerventions were
given, trials were excluded if we could not isolate the effects of
Assessment of Methodologic Quality of Included Studies.
heat or cold from the effects of the other therapies delivered.
The methodologic quality of the included trials was independently
Trials of spa therapy (balneotherapy) were excluded because
assessed by two authors (S.D.F., M.C.) and checked by a third
that intervention is being assessed by another Cochrane review.
author (J.W.R.). The assessment of methodologic quality was per-
At the time of publication of our review, the protocol for the
formed according to the methodologic criteria list recommended
balneotherapy review had only proceeded to the editorial re-
by the Cochrane Back Review Group22 and scored as a “yes (Y)”,
view stage.
“no (N),” or “don’t know (DK).” There were 11 criteria relevant
Types of Outcome Measures. Trials were included that used to the internal validity of the study, against which each trial was
at least one of the five outcomes considered to be important in assessed, including selection bias, performance bias, attrition bias,
low back pain research: pain (e.g., measured by visual analogue and detection bias. The methodologic quality assessment of the
scale), disability/function (e.g., measured by Oswestry, Roland trials was used to grade the strength of the evidence. Higher-
Disability Scale), overall improvement, patient satisfaction, quality trials were defined as fulfilling 6 or more of the 11 meth-
and adverse effects. The primary outcomes for this review were odologic quality criteria. Lower-quality trials were defined as ful-
pain and physical functional status. Some included trials mea- filling fewer than 6 criteria.
sured other outcomes, e.g., trunk flexibility or skin tempera-
ture; however, these results are not included in the analysis Data Analysis. Only a small proportion of the data in the
because they are out of the scope of this review. included trials were available for pooling. For the majority of
comparisons and outcomes, it was not possible to pool results.
A qualitative method recommended by the Cochrane Back Re-
Search Strategy for Identification of Studies
view Group22 using Levels of Evidence for data synthesis was
Data Sources. The following sources were accessed and performed:
searched:
● Strong evidence1: consistent findings among multiple
1. The Cochrane Controlled Trials Register (CENTRAL) high-quality RCTs
(Cochrane Library Issue 3, 2005) ● Moderate evidence: consistent findings among multiple
2. MEDLINE (1966 to October 2005) low-quality RCTs or CCTs and/or one high-quality RCT
3. EMBASE (1980 to October 2005) ● Limited evidence: one low-quality RCT and/or CCT
4. CINAHL (1982 to October 2005) ● Conflicting evidence: inconsistent findings among multi-
5. PEDro (Physiotherapy Evidence Database: www. ple trials (RCTs and/or CCTs)
pedro.fhs.usyd.edu.au, accessed October 2005) ● No evidence from trials: no RCTs or CCTs
6. Cochrane Back Review Group Specialized register (May
2005) Clinical Relevance. Two authors (S.D.F., J.W.R.) indepen-
7. SPORTDiscus (1830 to October 2005) dently judged the clinical relevance of each trial, using the five
8. OLDMEDLINE (1950 to 1965, searched October 2005) questions recommended by the Cochrane Back Review
Group,22 and scored each one as a yes, no, or don’t know:
Search Strategy. The search strategy was based on that rec-
ommended by the Cochrane Back Review Group.22 1. Are the patients described in detail so that you can decide
The search strategies for CENTRAL, MEDLINE, and EM- whether they are comparable to those that you see in
BASE are seen in Table 1. Search strategies for the remaining your practice?1000 Spine • Volume 31 • Number 9 • 2006
Table 1. Search Strategies for OVID
MEDLINE EMBASE CENTRAL
1. randomized controlled trial.pt. 1. clinical article/ 1. BACK PAIN explode all trees (MeSH)
2. controlled clinical trial.pt 2. clinical study/ 2. back pain or backache or lumbago
3. Randomized Controlled Trials/ 3. clinical trial/ 3. #1 or #2
4. Random Allocation/ 4. controlled study/ 4. CRYOTHERAPY explode all trees (MeSH)
5. Double-Blind Method/ 5. randomized controlled trial/ 5. cryotherapy or ice or cool or cold
6. Single-Blind Method/ 6. major clinical study/ 6. heat* or hot or warm* or infrared or
infra-red or poultice or spa* or sauna*
or shower* or steam*
7. or/1–6 7. double blind procedure/ 7. #4 or #5 or #6
8. Animal/not Human/ 8. multicenter study/ 8. #3 and #7
9. 7 not 8 9. single blind procedure/
10. clinical trial.pt. 10. crossover procedure/
11. exp Clinical Trials/ 11. placebo/
12. (clin$ adj25 trial$).tw. 12. Or/1–11
13. ((singl$ or doubl$ or trebl$ or tripl$) adj25 13. allocat$.ti,ab
(blind$ or mask$)).tw.
14. Placebos/ 14. assign$.ti,ab
15. placebo$.tw. 15. blind$.ti,ab
16. random$.tw. 16. (clinic$ adj25 (study or trial)).ti,ab
17. Research Design/ 17. compare$.ti,ab
18. (latin adj square).tw. 18. control$.ti,ab
19. or/10–18 1 19. cross?over.ti,ab
20. 9 not 8 20. factorial$.ti,ab
21. 20 not 9 21. follow?up.ti,ab
22. Comparative Study/ 22. placebo$.ti,ab
23. exp Evaluation Studies/ 23. prospectiv$.ti,ab
24. Follow-Up Studies/ 24. random$.ti,ab
25. Prospective Studies/ 25. ((singl$ or doubl$ or trebl$ or tripl$) adj25
(blind$ or mask$)).ti,ab
26. (control$ or prospective$ or Volunteer$).tw. 26. trial.ti,ab
27. Cross-Over Studies/ 27. (vs. or vs).ti,ab
28. or/22–27 28. Or/13–27
29. 28 not 8 29. 12 or 28
30. 29 not (9 or 21) 30. human/
31. 9 or 21 or 30 31. nonhuman/
32. low back pain/ 32. animal/
33. low back pain.tw. 33. animal experiment/
34. backache.tw. 34. 31 or 32 or 33
35. lumbago.tw. 35. 30 and 34
36. or/32–35 36. 29 not 34
37. 31 and 36 37. 29 and 35
38. (heat$ or hot or warm$).tw. 38. 36 or 37
39. (infrared or infra-red).tw. 39. low back pain/
40. poultice.tw. 40. back pain.tw
41. (spa$ or sauna$ or shower$ orsteam$).tw. 41. backache.tw
42. Cryotherapy/ 42. lumbago.tw
43. (cryotherapy or ice or cool or cold).tw. 43. Or/39–42
44. or/38–43 44. (heat$ or hot or warm$).tw
45. 37 and 44 45. (infrared or infra-red).tw
46. poultice.tw
47. (spa$ or sauna$ or shower$ or steam$).tw
48. Cryotherapy/
49. (cryotherapy or ice or cool or cold).tw
50. Or/44–49
51. 38 and 43 and 50
2. Are the interventions and treatment settings described able for inclusion. All nine trials were published in En-
well enough so that you can provide the same for your glish. Only 4 of these trials had pain data in a form that
patients? could be extracted and combined in a meta-analysis,23–26
3. Were all clinically relevant outcomes measured and re-
and this was only possible after obtaining further data
ported?
4. Is the size of the effect clinically important? from the authors of the studies. All of these trials exam-
5. Are the likely treatment benefits worth the potential harms? ined a heat wrap as the main intervention. One trial had
pain data that could be extracted27; however, it was ab-
Results
solute pain data as compared with change in pain data
The search strategy identified 1,178 potentially eligible used in the other heat wrap trials. The remaining four
studies. Of these, 123 were retrieved in full text. We trials did not present data in a form that could be ex-
identified nine trials involving 1,117 participants suit- tracted for meta-analysis.28 –31 Despite attempts to con-Superficial Heat or Cold for LBP • French et al 1001
Table 2. Characteristics of Included Studies
Study Participants and settings Interventions Outcomes Authors’ Conclusions
Landen28 (1967) 143 participants with mix of acute and 1. Hot packs: twice daily for 20 min, across 1. Pain: participant reported Ice massage and hot packs
chronic LBP; gender and age, not lumbosacral area (n ⫽ 59) change in symptoms: seem equally effective in
described; 117 participants 2. Ice massage: twice daily with cubes of minimal, moderate, or the symptomatic relief of
completed follow-up; definite ice across lumbosacral area until numb, marked increase or low back pain
diagnosis of disc herniation usually 10–12 min (n ⫽ 58) decrease in pain, or no
excluded; setting: U.S. Army Cointerventions: all participants also change
General Hospital, Germany performed flexion exercises 2. Length of hospitalization
3. Muscular spasm: not
described how measured
Timing of outcome
measures: time of
discharge
Mayer et al 26 100 participants (29 male, 71 female) 1. Heat wrap alone (ThermaCare Heat Wrap; 1. Functional improvement: Combining continuous
(2005) with acute (⬎2 days, ⬍3 mo) Procter and Gamble, Cincinnati, OH), MTAP low-level heat wrap
nonspecific LBP without radiation applied to lumbar area, 40 C for 8 hr per 2. Disability: Roland Morris therapy with directional
below the knee and normal day for 5 consecutive days (n ⫽ 25) 3. Pain relief: 6-point verbal preference-based
neurologic examination, no 2. McKenzie exercise alone (n ⫽ 25) rating scale exercise therapy offers
comorbidities; duration of pain not 3. Heat wrap plus McKenzie exercise Timing of outcome distinct advantages over
reported; mean age, 31.2 yr; setting: (n ⫽ 24) measures: 2 days, 4 days either therapy alone for
outpatient medical facilities in 4. Educational booklet. Participants were and 7 days after the treatment of acute
California advised to closely follow the randomization low back pain
recommendations, except that they were
asked to refrain from performing specific
exercises for the low back, using heat or
cold modalities, and receiving spinal
manipulation (n ⫽ 26)
Cointerventions not allowed, except for
medication as required
Melzack et al 29 44 participants (23 male, 21 female) 1. Ice massage: ice cube gently massaged 1. Pain: McGill pain Ice massage is an effective
(1980) with chronic LBP, unresponsive to on skin for a max 7 min at 3 sites (midline questionnaire; measured therapeutic tool and is
conventional care; mean duration of low back, lateral malleolus, and popliteal immediately after more effective than TES
pain 7.4 yr; age, 18–73 yr space) with 3 min between applications, treatment sessions for some patients
Majority of participants had total treatment time of 30 min; ice 2. Preferred treatment
undergone previous surgery massage was administered by a Timing of outcome
Exclusion criteria: severe emotional “technician” measures: pain measured
problems as determined by 2. TES: 2 treatments of TES at the same 3 before and after each
Minnesota Multiphasic Personality sites and time interval for 30 min; treatment session, then
Inventory; setting: Pain centre, interventions were administered on 2 1–12 mo after completing
Canada occasions at 1- to 2-wk intervals treatment
Treatment for each group was reversed
after the initial 2 treatment sessions with
a further 2 treatments of the alternate
intervention
Cointerventions not reported
Nadler et al 23 371 participants (155 male, 216 female) 1. Heat wrap (ThermaCare Heat Wrap; 1. Pain: 6-point verbal scale Continuous low-level
(2002) with acute (⬍3 mo) nonspecific Procter and Gamble, Cincinnati, OH) for of pain relief topical heat wrap
LBP without radiation below the approximately 8 hr per day (n ⫽ 113) 2. Muscle stiffness: therapy is superior to
knee and normal neurologic 2. Oral ibuprofen: 2 tablets 3 times daily for 101-point scale both acetaminophen and
examination, no comorbidities; a total dose of 1,200 mg, plus oral 3. Disability: Roland-Morris ibuprofen
duration of pain not reported; mean placebo 1 time daily (n ⫽ 106) (0–24 scale)
age, 36.0 yr; setting: clinical 3. Oral acetaminophen: 2 tablets 4 times 4. Lateral trunk flexibility
research sites daily for a total of 4,000-mg dose 5. Adverse effects
(n ⫽ 113) Timing of outcome
4. Oral placebo: 2 tablets 4 times daily measures: pain relief and
(n ⫽ 20) disability measured daily
5. Unheated back wrap: (n ⫽ 19) for 4 days
Cointerventions not allowed post-randomization
Nadler et al 24 219 participants (100 male, 119 female) 1. Heat wrap (ThermaCare Heat Wrap; 1. Pain: 6-point verbal scale Continuous low-level heat
(2003) with acute (⬍3 mo) nonspecific Procter and Gamble, Cincinnati, OH) for 3 of pain relief wrap therapy was shown
LBP without radiation below the consecutive days, approximately 8 hr per 2. Muscle stiffness: to provide significant
knee and normal neurologic day (n ⫽ 95) 101-point scale therapeutic benefits in
examination, no comorbidities; 2. Oral placebo: 2 tablets, 3 times daily, 3. Disability: Roland-Morris patients with acute
duration of pain not reported; mean spaced 6 hr apart (n ⫽ 96) (0–24) nonspecific LBP
age, 36.1 yr; setting: clinical 3. Oral ibuprofen: 200 mg, 2 tablets, 3 times 4. Lateral trunk flexibility No serious or significant
research sites daily, spaced 6 hr apart (n ⫽ 12) 5. Skin quality: 4-point scale adverse effects were
4. Unheated wrap: (n ⫽ 16) Timing of outcome observed during the use
Cointerventions not allowed measures: pain relief and of the heat wrap
disability measured daily
for 4 days post-
randomization
(Continued)1002 Spine • Volume 31 • Number 9 • 2006
Table 2. (Continued)
Study Participants and settings Interventions Outcomes Authors’ Conclusions
Nadler et al 25 76 participants (27 male, 49 female) 1. Heat wrap (ThermaCare Heat Wrap; 1. Pain relief: 6-point verbal Continuous low-level heat
(2003) with acute (⬍3 mo) nonspecific Procter and Gamble, Cincinnati, OH) worn rating scale wrap therapy was shown
LBP without radiation below the during sleep for approximately 8 hr each 2. Muscle stiffness: to provide effective
knee and normal neurologic night for 3 consecutive nights (n ⫽ 33) 101-point numeric rating daytime pain relief after
examination, no comorbidities; 2. Oral placebo: 2 tablets before bed each scale overnight use in subjects
duration of pain not reported; mean night for 3 consecutive nights (n ⫽ 34) 3. Pain effect: 101-point with acute nonspecific
age, 41.4 yr; setting: clinical 3. Oral ibuprofen: 2 tablets; total dose, 400 numeric rating scale LBP
research sites mg, before bed each night for 3 4. Disability: Roland-Morris The heat wrap showed a
consecutive nights (n ⫽ 4) (0–24) good safety profile when
4. Unheated wrap worn during sleep for 5. Lateral trunk flexibility worn during sleep
approximately 8 hr each night for 3 6. Subjective measures of
consecutive nights (n ⫽ 5) sleep quality and
Cointerventions not allowed difficulty in sleep onset:
6-point scale
Timing of outcome
measures: days 1–4 after
commencement of
treatment
Nuhr et al 27 90 participants (57 male, 33 female) 1. Resistive heating of 42 C via a carbon- 1. Pain: 100 point VAS Local active warming is an
(2004) with first episode acute (⬍ 6 hr) fiber electric heating blanket (ThermaMed 2. Tympanic thermocouple effective and easy-to-
LBP without radiation below the GmbH, Bad Oeynhausen, Germany) (core temperature) learn emergency care
knee and normal neurologic covered by a single woolen blanket; the 3. Skin thermosensors and treatment for acute low
examination, no comorbidities; active heating component covered an intracutaneous back pain that could be
mean age, 36.8 yr (⫾8.2 yr); setting: area of 148 ⫻ 40 cm; mean duration of thermosensors used by emergency
emergency site in Austria treatment, 24.8 ⫾ 8.1 min 4. Adverse effects: all physicians as well as by
(n ⫽ 47) measurements were paramedical personnel
2. Passive warming: single woolen blanket; recorded by the same
mean duration of treatment, 26.2 ⫾ 9.3 independent investigator
min (n ⫽ 43) blinded to the treatment
Cointerventions not allowed with the electrical
blanket, which he/she
was forbidden to touch
Timing of outcome
measures: all measured
at arrival to hospital
Roberts et al 30 36 participants (17 male, 19 female), 1. Hot pack (160 F) with 6–8 layers of towels 1. Pain: VAS (0–20) Ice massage was found to
(1992) with chronic low back pain; mean for 20 min immediately after and 1 be significantly more
duration of pain, 4.6 yr; age, 24–72 2. Cold packs (0 F) with 2 layers damp hr after intervention effective than either hot
yr (mean, 40.4 yr); all participants towels for 20 min packs or cold packs for
had failed to respond to traditional 3. Ice massage: lightly rubbing low-back relief of chronic low
medical management; setting: Pain with cake of ice 4.2 ⫻ 2 ⫻ 4.5 inches; back pain
centre, United States over 2-wk period, each participant was
given two applications with each of the
three interventions; cointerventions:
during study all participants were
participating in a comprehensive pain
rehabilitation program, including
medication
St. John 38 participants (12 male, 24 female) 1. Medima angora wool body belt (“giving 1. Preference for type of The findings of this study
Dixon et al 31 with chronic nonspecific low back insulation and warmth without support”) support call into question the
(1972) pain; mean duration of pain, 17 yr 2. Reemploy “instant” lumbosacral corset 2. Pain (but no data common practice of
(range, 0.5–40 yr); age, 30–80 yr (“giving support but little insulation”) reported) prescribing a
(mean, 61 yr) setting: orthopedic 3. Participants instructed to wear each 3. Adverse effects: timing of lumbosacral support for
outpatient department in United intervention for 1 wk; cointerventions not outcome measures: all chronic, nonspecific, low
Kingdom described measured after 1 and back pain; many patients
2 wk fare equally well with a
simple warm body belt
RCT ⫽ randomized controlled trial; CCT ⫽ controlled clinical trial; LBP ⫽ low back pain; NSAIDs ⫽ nonsteroidal anti-inflammatory drugs; VAS ⫽ visual analogue
scale; MTAP ⫽ Multidimensional Task Ability Profile; TES ⫽ transcutaneous electrical stimulation.
tact all authors, only additional data for the three Nadler participants,23–26 three included chronic low back pain
trials and the Mayer trial were obtained. participants,29 –31 and one had a mix of acute, subacute,
Included Studies and chronic participants.28
Table 1 shows characteristics of included studies. The The nature of the interventions differed between the tri-
median number of participants in each trial was 90 als. Two trials compared hot packs to ice massage,28,30 one
(range, 36 –371). According to our definition, one of the trial compared ice massage to transcutaneous electrical
trials included acute low back pain participants,27 four stimulation,29 one trial compared a full body active warm-
included a mix of acute and subacute low back pain ing electric blanket to passive warming by way of a woolSuperficial Heat or Cold for LBP • French et al 1003
Table 3. Methodologic Quality Assessment of Included Trials
Summary Scores and
Study A B C D E F G H I J K Comments
Landen28 (1967) N N DK N N N DK Y Y DK N Score ⫽ 2 (low)
Mayer et al 26 (2005) Y DK Y N N DK Y Y Y Y N Score ⫽ 6 (high)
Melzack et al 29 (1980) N N DK N N DK Y N DK Y N Score ⫽ 2 (low)
Nadler et al 23 (2002) DK DK Y N N Y Y DK Y Y Y Score ⫽ 6 (high)
Nadler et al 24 (2003) DK DK Y N N Y Y DK Y Y Y Score ⫽ 6 (high)
Nadler et al 25 (2003) DK DK Y N N Y Y DK Y Y Y Score ⫽ 6 (high)
Nuhr et al 27 (2004) Y Y Y N N Y Y Y Y Y N Score ⫽ 8 (high)
Roberts et al 30 (1992) N N DK N N N DK DK DK Y N Score ⫽ 1 (low)
St. John Dixon et al 31 (1972) DK N DK N N N DK DK Y Y N Score ⫽ 2 (low)
Y ⫽ yes; N ⫽ no; DK ⫽ don’t know.
A ⫽ Was the method of randomization adequate? B ⫽ Was the treatment allocation concealed? C ⫽ Were the groups similar at baseline regarding the most
important prognostic indicators? D ⫽ Was the patient blinded to the intervention? E ⫽ Was the care provider blinded to the intervention? F ⫽ Was the outcome
assessor blinded to the intervention? G ⫽ Were co-interventions avoided or similar? H ⫽ Was the compliance acceptable in all groups? I ⫽ Was the dropout rate
described and acceptable? J ⫽ Was the timing of the outcome assessment in all groups similar? K ⫽ Did the analysis include an intention-to-treat analysis?
blanket,27 and one trial compared a wool body belt that Methodologic Quality of Included Studies
provided warmth to a lumbar corset.31 The included trials were of varying methodologic quality
Four trials assessed the effect of a disposable heated (Table 3). Applying the criteria of six or more equalling a
lumbar wrap (ThermaCare Heat Wrap; Procter and high-quality study, five of the trials were of high quality
Gamble, Cincinnati, OH) compared with various inter- (range, 6 – 8) and four low quality (range, 1–5). Five of
ventions. Three of these trials compared the heated wrap the trials were reported as randomized.23–27 In the Nadler
with pain relief medication and with a nonheated series of trials, the method of randomization was not
wrap,23–25 and one trial compared the heated wrap alone described. One trial was a nonrandomized CCT28 and
with exercise alone, with heat plus exercise, and with an three were nonrandomized crossover trials.29 –31 Wash-
educational booklet.26 The heat wrap is a disposable out of the interventions was not considered in any of the
crossover trials. Trial population sizes were generally
product made of layers of cloth-like material that con-
small (median sample size, 90; range, 36 –371). Five of
tain heat-generating ingredients (iron, charcoal, table
the trials reported clear inclusion and exclusion crite-
salt, and water). These ingredients heat up when exposed
ria,23–27 two trials reported brief inclusion and exclusion
to oxygen and provide heat (40 C) for at least 8 hours.
criteria,28,29 and two trials did not state exclusion crite-
The heat wrap is applied to the lumbar region of the ria.30,31 Allocation concealment was adequate in only
torso and is secured with a Velcro-like closure, thus al- one of the trials,27 and in the remaining trials was either
lowing it to be worn while remaining mobile. It can be not reported or was inadequate. Blinded outcome assess-
worn during the day or night. The single use heat wrap ment was carried out in four trials23–25,27 and was either
costs approximately U.S. $6.00 to $8.00 for a packet not done or was unclear in the remaining five trials.
of two. Blinding of participants to the interventions of heat or
The outcomes assessed and the timing of outcomes cold was not possible in most cases. Most trials had an
varied. Pain was assessed in all trials, however, for only acceptable loss to follow-up; however, only three of the
five trials were pain data available for meta-analysis. trials reported if an intention-to-treat analysis was un-
Three trials only measured pain immediately after the dertaken.23–25
treatment,27,29,30 four trials measured pain over 4 to 7
Comparison 1: Heated Wrap Versus Oral Placebo
days,23–26 one trial measured pain at the time of hospital
or Nonheated Wrap
discharge,28 and one after 2 weeks.31 A validated disabil-
Four higher-quality trials assessed a heated wrap or
ity measure was used in four trials.23–26 None of the
heated blanket versus either an oral placebo tablet, or a
trials assessed overall improvement or participant satis- nonheated wrap23–25,27 in participants with a mix of
faction. acute and subacute (⬍3 months) low back pain. It was
Four of the trials declared industry funding.23–26 The only possible to combine the data from a maximum of
three Nadler trials were all conducted by the same re- two trials.
search team and were funded by the manufacturer of the Pain relief data were extracted from two of the trials
heat wrap device. The authors of each of these trials were with 258 participants that compared a heated wrap to
either employees or paid consultants of the company that oral placebo.24,25 The short-term pain relief was signifi-
manufactures the heat wrap device. Correspondence cantly greater for the heated back wrap than for the oral
from the authors indicated that these three trials are placebo (weighted mean difference [WMD], 1.06; 95%
completely separate studies with each of them including confidence interval [CI], 0.68 to 1.45 scale; range, 0 –5).
different participants. Pain relief was only measured for up to 5 days after1004 Spine • Volume 31 • Number 9 • 2006
randomization. This result indicates an approximate One high-quality trial of 100 participants compared a
17% reduction in pain after 5 days with a heated back heated back wrap alone with exercise alone and with an
wrap compared with oral placebo. Pain data measuring education booklet.26 Measured at 1 and 4 days after
the degree of “unpleasantness” were only available for randomization, the heat wrap provided significantly
one trial25 and indicated a significant decrease in the more pain relief than an educational booklet (day 2:
short-term (WMD, ⫺13.50; 95% CI, ⫺21.27 to ⫺5.73; WMD, 0.60; 95% CI, 0.05 to 1.15; scale range, 0 –5; day
scale range, 0 –100). 4: WMD, 1.10; 95% CI, 0.55 to 1.65) but not more than
Absolute pain data were only available in one trial of McKenzie exercise (day 2: WMD, 0.40; 95% CI, ⫺1.15
90 participants.27 This trial demonstrated a statistically to 0.95; day 4: WMD, 0.30; 95% CI, ⫺0.41 to 1.01). At
significant benefit of a heated blanket compared with a 7 days after randomization, there were no significant
nonheated blanket immediately after treatment in acute differences in pain relief between the groups. The heat
(⬍6 hours) low back pain (WMD, ⫺32.20; 95% CI, wrap provided significantly improved function com-
⫺38.69 to ⫺25.71; scale range, 0 –100). Pain was only pared with an educational booklet at day 2 (WMD,
measured immediately after the heat was applied for ap- ⫺1.40; 95% CI, ⫺2.79 to ⫺0.01; scale range, 0 –24) and
proximately 25 minutes, and no further follow-up oc- day 4 (WMD, ⫺2.30; 95% CI, ⫺4.24 to ⫺0.36) but not
curred. at day 7 (WMD, ⫺1.70; 95% CI, ⫺3.92 to 0.52). There
Only two trials provided data on disability,24,25 mea- was no significant difference in function between heat
sured with the Roland-Morris Disability Questionnaire. wrap and McKenzie exercise at day 2, day 4, or day 7.
The short-term (4 days) reduction in disability was sig- One lower-quality nonrandomized crossover trial
nificantly greater for the heated back wrap than for oral compared a wool body belt providing warmth to a lum-
placebo (WMD, ⫺2.10; 95% CI, ⫺3.19 to ⫺1.01; scale bar corset in chronic low back pain participants.31 No
range, 0 –24). pain results were provided.
Adverse effects were minor for the heated back wrap.
Two trials provided data on the adverse effect of “skin Comparison 5: Cold Versus Other Interventions
pinkness” after use of the heated wrap.24,25 A total of 6 Only one low-quality nonrandomized crossover trial ex-
of 128 participants experienced this outcome in the amined this comparison.29 This trial compared ice mas-
heated back wrap group, compared with 1 participant sage with transcutaneous electrical stimulation (TES) in
out of 130 in the placebo group. chronic low back pain participants. The trial concluded
No trials were located that examined this comparison that ice massage and TES were equally effective in reduc-
for chronic low back pain or for the medium- or long- ing pain.
term effects of this intervention. Comparison 6: Heat Plus Exercise Versus
Other Interventions
Comparison 2: Cold Versus Placebo or No Cold One higher-quality trial of 100 participants combined a
No studies were located that examined this comparison. heated back wrap with exercise and compared this to
heat alone, exercise alone, and to an educational book-
Comparison 3: Heat Versus Cold let, in participants with a mix of subacute and acute low
Two lower-quality trials evaluated heat versus cold in back pain.26 Heat wrap plus exercise provided signifi-
the form of hot packs versus ice massage.28,30 Unfortu- cantly more pain relief than an educational booklet at
nately, there were very little data available in either of day 4 (WMD, 1.60; 95% CI, 0.89 to 2.31; scale range,
these trials to extract. Both of these trials were nonran- 0 –5) and day 7 (WMD, 2.00; 95% CI, 1.29 to 2.71), and
domized: one a controlled trial28 and the other a cross- also for function (Roland Morris) (day 4: WMD, ⫺2.60;
over trial.30 One trial concluded that hot packs and ice 95% CI, ⫺4.54 to ⫺0.66; day 7: WMD, ⫺4.40; 95% CI,
massage were not significantly different for participants ⫺6.62 to ⫺2.18). Heat wrap plus exercise also provided
with a mix of acute, subacute, and chronic low back significantly more pain relief and improvement in func-
pain. The other concluded that ice massage was superior tion than either heat or exercise alone at day 7. This
to hot packs in participants with chronic low back pain. improvement in pain and function was not evident at the
earlier time periods measured (day 2 or day 4).
Comparison 4: Heat Versus Other Interventions
Three higher-quality trials compared a heated back wrap Clinical Relevance
with oral ibuprofen,23–25 and one of these trials also The median score for the trials was 3 of 5. The higher-
included a comparison group that took oral acetamino- quality trials generally had a higher clinical relevance
phen.23 Unfortunately, none of these trials presented score.
data in a way that could be combined in a meta-analysis.
Discussion
One trial23 found that a heated back wrap provided sig-
nificantly greater pain relief and improved function than Only a few studies have been published evaluating the
oral ibuprofen and oral acetaminophen after both 1 day effects of superficial heat or cold for low back pain. We
and 4 days of treatment. The other two Nadler trials did found nine trials involving 1,117 participants that were
not provide results for this comparison. suitable for inclusion in this review. Of these, six trialsSuperficial Heat or Cold for LBP • French et al 1005
examined heat compared with no heat or other interven- to these interventions. The Nadler series of studies23–25
tions, one compared cold to another intervention, and and the Nuhr study27 attempted to blind participants by
two trials compared heat to cold. The included trials including a nonheated wrap or blanket and an oral pla-
were very heterogeneous in terms of interventions used, cebo group; however, the investigators did not measure
control treatments, outcome measures, timing of follow- whether participants could determine if they were receiv-
up, and presentation of data. Therefore, it was not pos- ing an active therapy or not. Also, outcome assessors
sible to perform any meaningful meta-analyses, and it should be blinded to the allocation of the participants to
was difficult to reach firm conclusions for most types of at least improve the quality of this aspect of the trials. It
treatments. is recommended that these methodologic issues are con-
According to the qualitative criteria for levels of evi- sidered in future trials.
dence, for a mixed population with acute and subacute Heat and cold are methods that are commonly used in
low back pain, there is moderate evidence that a heated practice in conjunction with other interventions, espe-
wrap applied for 8 hours, or an electric blanket applied cially in the physical therapy professions. We only found
for 25 minutes, are both better than no heat for pain in one small study that evaluated the use of heat combined
the short-term (4 days). There is moderate evidence (one with exercise, and we found no study that examined cold
small high-quality RCT) that heat wrap is better for pain in this context. Thus, no conclusions can be drawn re-
and function than an educational booklet during the garding the use of heat or cold in combination with other
early stages of treatment (days 2– 4), but not after 7 days. therapies, other than in combination with exercise.
There is moderate evidence that combining heat wrap
with McKenzie exercises is better for pain relief and
Conclusion
function after 7 days than an educational booklet and
either heat wrap or exercise alone. The effect of these Implications for Practice
treatments was small. If the short-term beneficial effects Heat and cold are commonly recommended by clinicians
of this therapy can be verified in further high-quality for low back pain. The evidence base to support this
trials, then its use would be valuable. common practice is not strong. There is moderate evi-
There is empirical data that indicate that industry- dence that continuous heat wrap therapy reduces pain
funded studies are more likely to be positive than non- and disability in the short-term, in a mixed population
funded studies.32–34 The results of the Nadler series of with acute and subacute low back pain (up to 3 months)
studies and the Mayer study of heat wrap therapy should and that the addition of exercise to heat wrap therapy
be considered with this in mind, and independent studies further reduces pain and improves function. This evi-
would be useful to verify their results. Also, considering dence is limited to a small number of trials using a rela-
the cost of the disposable heat wraps, it would be useful tively small number of participants, and the size of the
to include a cost-effectiveness analysis in future trials. effect is small. The application of cold treatment to low
No RCTs were located that examined the effects of back pain is even more limited, with only three poor-
cold for low back pain. Given that it is a commonly held quality studies located. No conclusions can be drawn
belief that cold is beneficial for recent onset musculoskel- about the use of cold for low back pain. There is conflict-
etal injuries,35 it was surprising that no studies were lo- ing evidence to determine the differences between heat
cated that applied cold treatment to acute low back pain. and cold for low back pain.
Indeed, in the trials conducted with participants with
acute and subacute low back pain, heat was applied. The Implications for Research
trials that were located for cold treatment used cold for Many of the studies were of poor methodologic quality,
chronic low back pain and were of poor methodologic and there certainly is a need for future higher-quality
quality. No conclusions can be drawn for the use of cold RCTs. Also, many trials were poorly reported, and we
treatment in low back pain. recommend that authors use the CONSORT statement
There is conflicting evidence when comparing heat as a model for reporting RCTs (www.consort-
treatment to cold treatment. Two low-quality nonran- statement.org). The results of the majority of the studies
domized trials of chronic low back pain participants could not be pooled with other studies because of the
were located. One concluded that hot packs and cold way the authors reported the results. Therefore, we sug-
packs were equally effective, and the other concluded gest that the publications of future trials report, for con-
that ice massage was better than either hot packs or cold tinuous measures, means with standard deviations or
packs. means with standard error of means, and for dichoto-
There were no major adverse events reported in any of mous measures, number of events and total participants
the trials. Some minor events were reported with the heat analyzed.
wrap therapy, in the form of “skin pinkness” that re- Future research should focus on areas where there are
solved quickly. few or no trials, for example, simple heat applications
There are methodologic challenges when conducting like hot water bottles, ice massage versus no cold and
high-quality trials into these therapies. For example, it is heat versus cold treatment, and trials in chronic low back
questionable whether or not participants can be blinded pain participants. The classification of duration of low1006 Spine • Volume 31 • Number 9 • 2006
back pain was not consistent in the different studies, and 10. Jette DU, Jette AM. Physical therapy and health outcomes in patients with
spinal impairments. Phys Ther 1996;76:930 – 45.
in the future, authors should be clear on the definition of 11. Li LC, Bombardier C. Physical therapy management of low back pain: an
acute, subacute, and chronic low back pain and report exploratory survey of therapist approaches. Phys Ther 2001;81:1018 –28.
the duration of pain in their results. Future studies 12. Rush PJ, Shore A. Physician perceptions of the value of physical modalities in
the treatment of musculoskeletal disease. Br J Rheumatol 1994;33:566 – 8.
should be adequately powered and have both a short-
13. Stanos SP, Muellner PM, Harden RN. The physiatric approach to low back
term follow-up (for acute pain) and a long-term fol- pain. Semin Pain Med 2004;2:186 –96.
low-up (for chronic pain). 14. Grana WA. Physical agents in musculoskeletal problems: heat and cold ther-
apy modalities. Instr Course Lect 1993;42:439 – 42.
Acknowledgments 15. Michlovitz SL. Thermal Agents in Rehabilitation. Philadelphia: F.A. Davis,
1996.
The authors thank Judy Clarke, Laura Coe, Carla Cooper,
16. Vasudevan S. Physical rehabilitation in managing pain. Pain: Clinical Up-
Yngve Falck-Ytter, Anne Keller, Peter Kroeling, Suzanna dates 1997;V. Available at: http//www.iasp-pain.org/PCU97c.html. Ac-
Mitrova, Wu Taixiang, and Simona Vecchi for translation cessed 5th May 2005.
17. Bigos S, Bowyer O, Braen G, et al. Acute Low Back Problems in Adults [Clinical
of non-English language articles; Ashoke Mitra and Kurt
Practice Guideline No. 14; AHCPR Publication No. 95-0642]. Rockville, MD:
Weingand for providing us with information regarding the Agency for Health Care Policy and Research Public Health Service, U.S. Depart-
Nadler studies; John Mayer for the extra data from his trial; ment of Health and Human Services, 1994.
18. New Zealand Guideline Group. Acute Low Back Pain Guide. Wellington,
and the Cochrane Back Review Group, and in particular
New Zealand: New Zealand Guideline Group, 1997.
Vicki Pennick, for their ongoing encouragement and assis- 19. Institute for Clinical Systems Improvement. Adult Low Back Pain. Bloom-
tance while carrying out this review. ington, MN: Institute for Clinical Systems Improvement, 2004.
20. Australian Acute Musculoskeletal Pain Guidelines Group. Evidence-based
management of acute musculoskeletal pain. Available at: http://
www.nhmrc.gov.au. Australian Academic Press: Brisbane 2003.
Key Points 21. van Tulder M, Becker A, Bekkering T, et al. European Guidelines for the
Management of Acute Nonspecific Low Back Pain in Primary Care. Avail-
● Nine trials involving 1,117 participants were in- able at: http://www.backpaineurope.org (accessed 22nd March 2005).
cluded in this systematic review. 22. van Tulder M, Furlan A, Bombardier C, et al. Updated method guidelines for
systematic reviews in the Cochrane Collaboration Back Review Group.
● There is moderate evidence that heat wrap ther-
Spine 2003;28:1290 –9.
apy reduces pain and disability for patients with 23. Nadler SF, Steiner DJ, Erasala GN, et al. Continuous low-level heat wrap
back pain that lasts for less than 3 months. The therapy provides more efficacy than Ibuprofen and acetaminophen for acute
low back pain. Spine 2002;27:1012–7.
relief has only been shown to occur for a short time
24. Nadler SF, Steiner DJ, Erasala GN, et al. Continuous low-level heatwrap
and the effect is relatively small. The addition of therapy for treating acute nonspecific low back pain. Arch Phys Med Rehabil
exercise to heat wrap therapy appears to provide 2003;84:329 –34.
25. Nadler SF, Steiner DJ, Petty SR, et al. Overnight use of continuous low-level
additional benefit.
heatwrap therapy for relief of low back pain. Arch Phys Med Rehabil 2003;
● There is insufficient evidence about the effect of 8:335– 42.
heat for back pain that lasts longer than 3 months. 26. Mayer JM, Ralph L, Look M, et al. Treating acute low back pain with
continuous low-level heat wrap therapy and/or exercise: a randomized con-
● There is insufficient evidence about the effect of
trolled trial. Spine J 2005;5:395– 403.
the application of cold for low back pain of any 27. Nuhr M, Hoerauf K, Bertalanffy A, et al. Active warming during emergency
duration. transport relieves acute low back pain. Spine 2004;29:1499 –503.
28. Landen BR. Heat or cold for the relief of low back pain? Phys Ther 1967;
47:1126 – 8.
References 29. Melzack R, Jeans ME, Stratford JG, et al. Ice massage and transcutaneous
electrical stimulation: comparison of treatment for low-back pain. Pain
1. Walker BF. The prevalence of low back pain: a systematic review of the 1980;9:209 –17.
literature from 1966 to 1998. J Spinal Disord 2000;13:205–17. 30. Roberts D, Walls C, Carlile J, et al. Relief of chronic low back pain: heat
2. Hollingworth W, Todd CJ, King H, et al. Primary care referrals for lumbar versus cold. In: Aronoff GH, ed. Evaluation and Treatment of Chronic Pain.
spine radiography: diagnostic yield and clinical guidelines. Br J Gen Pract Baltimore: Urban & Schwarzenberg, 1992:263– 6.
2002;52:475– 80. 31. St John Dixon A, Owen-Smith BD, Harrison RA. Cold-sensitive, non-
3. Carey TS, Evans A, Hadler N, et al. Care-seeking among individuals with specific, low back pain: a comparative trial of treatment. Clin Trials J 1972;
chronic low back pain. Spine 1995;20:312–7. 4:16 –21.
4. Frymoyer JW, Cats-Baril WL. An overview of the incidences and costs of low 32. Bhandari M, Busse JW, Jackowski D, et al. Association between industry
back pain. Orthop Clin North Am 1991;22:263–71. funding and statistically significant pro-industry findings in medical and
5. Maniadakis N, Gray A. The economic burden of back pain in the UK. Pain surgical randomized trials. CMAJ 2004;170:477– 80.
2000;84:95–103. 33. Djulbegovic B, Lacevic M, Cantor A, et al. The uncertainty principle and
6. Battie MC, Cherkin DC, Dunn R, et al. Managing low back pain: attitudes and industry-sponsored research. Lancet 2000;356:635– 8.
treatment preferences of physical therapists. Phys Ther 1994;74:219 –26. 34. Kjaergard LL, Als-Nielsen B. Association between competing interests and
7. Car J, Sheikh A. Acute low back pain. BMJ 2003;327:541. authors’ conclusions: epidemiological study of randomised clinical trials
8. Geffen SJ. 3: Rehabilitation principles for treating chronic musculoskeletal published in the BMJ. BMJ 2002;325:249 –53.
injuries. Med J Aust 2003;178:238 – 42. 35. Bleakley C, McDonough S, MacAuley D. The use of ice in the treatment of
9. Hamm L, Mikkelsen B, Kuhr J, et al. Danish physiotherapists’ management acute soft-tissue injury: a systematic review of randomized controlled trials.
of low back pain. Adv Physiother 2003;5:109 –13. Am J Sports Med 2004;32:251– 61.You can also read
