Manipulation Contraindicated for Disk Herniation? When the Literature Conflicts

By Ronald L. Rupert

An article was published in the October 1994 issue of Spine¹ which reviews the many interventions for treating lumbar disc herniation. The treatment protocols included were: various surgical procedures, drug regimes and numerous conservative therapies (bed rest, supports, crutches, back school, traction, ice, manipulation, etc.). 
Regarding manipulation, the author simply states, "Manipulation -- abrupt passive movements of a vertebrae beyond its physiologic range, but within its anatomic range -- is contraindicated." It was interesting that no reference was given to support this strong statement and of all the therapies noted, manipulation was the only one considered "contraindicated." The author described the value of all other physiotherapeutic interventions as ranging from "helpful" to "questionable." Let us now turn to a similar review that strictly evaluated the use of manipulation for disc herniation by Cassidy, Thiel and Kirkaldy-Willis published in JMPT in February 1993.²When these authors performed their literature review they concluded, "The treatment of lumbar intervertebral disk herniation by side posture manipulation is both safe and effective."

It is common to find such conflicting viewpoints in biomedicine. Often this is because adequate safeguards are not in place to control bias or there just is not enough information available to the researcher to make an accurate and complete information. Because of the rapid proliferation of research literature, decisions often cannot be made from texts where the primary research data is a couple of years old by the time the work is published. They can't always be made by relying on postgraduate classes where the presenter may not always be current. They can be made by looking directly at the primary journal literature coupled with some critical reading skills. Computerized online searching provides the greatest opportunity for the clinician and researcher to gather information and make intelligent decisions.

So what about manipulation for disc herniations? These two research articles represent diametrically opposing statements from two of the most respected medical and chiropractic peer-reviewed journals. The questions become: first, what is the nature and quality of the articles and second, what point of view does the preponderance of the literature support. The first question involves critical reading and the second involves the use of online searches through databases like Medline and Chirolars.

As we look at the two articles, they are similar in many respects despite their conclusions. They both represent authors expressing views based on their review of the literature and do not provide any new primary research conducted by either author. If one study were a large, well-controlled randomized clinical trial and the other a case report or literature review, we would have to give far greater credence to the findings of the clinical trial. However, how could these authors perform a similar review of the literature and arrive at opposite conclusions about the value of manipulation for treating intervertebral disc herniation? The answer (assuming bias is not present) is that the pool of articles related to the subject available to one or both authors was incomplete. In order to determine if the authors have made a complete literature review, we can perform a simple online search of the appropriate database(s) and compare this to the references used by the authors. This process takes only a few minutes with online computer access to databases. In choosing the appropriate database to search we should know that the National Library of Medicine produces the largest index of medical literature, MEDLINE. However, this database only indexes 18 percent of the biomedical literature. There is very little research included from osteopathic, chiropractic or manual medicine journals dealing with manipulation. One of the strengths of the chiropractic index CHIROLARS is that it does include most of the significant research on adjustments and manipulation from all disciplines. In this instance, CHIROLARS would be the most valuable resource.

When searching CHIROLARS with the appropriate headings (manipulation, orthopedic [and] manipulation, chiropractic [and] intervertebral disk displacement [or] sciatica) we locate approximately 150 articles. As we look at the review articles from JMPT and Spine, we see that the authors have neglected much of the most valuable literature necessary to make an informed decision. Reading the abstracts of these articles from medicine, osteopathy and chiropractic, we find that with few exceptions manipulation is considered effective and safe in the treatment of disk herniations and sciatica. There is however some important work that suggests that the physician should exercise caution with acute disks. There are also many issues including the method of manipulation, patient position and the amount of force used that must be addressed. All forms of manipulation for all disks can hardly be considered "safe" as malpractice claims will verify. The truth then appears to lie somewhere between the two extreme positions published in Spine and JMPT, but heavily favoring the use of manipulation as described by Cassidy et al. Again, assuming no intentional bias, neither author really did a thorough literature review and their conclusions suffered. (A subsequent letter to the editor criticized the weakness in the "selective" kind of review published in JMPT).³

This case exemplifies several points: 1.) The literature can be selectively used (although perhaps not intentionally) to support either side of many clinical issues. 2.) The peer review process does not always insure quality of research, especially when related to reviews of the literature. Despite skills at research design, sampling, statistics, etc., very few researchers exhaust all relevant database sources. Those responsible for peer review often are incapable of evaluating the thoroughness of a literature review because they also lack the skills necessary for proper online searching. 3.) There is no substitute for having a commend of online searching and critical reading in order to resolve these conflicts in the literature. There is also no substitute for each practicing chiropractor to have the power that information access provides. It permits the doctor to help educate patients; to provide expert testimony; to serve as a consultant; to rebut insurance denials and, most importantly, to deliver quality treatment consistent with the literature.

References 

1. Weber H. Spine Update: The Natural History of Disc Herniation and the Influence of Intervention. Spine 1994; 19:2234-8.
   
   
2. Cassidy J, Thiel H and Kirkaldy-Willis W. Side Posture manipulation for Lumbar Intervertebral Disk Herniation. Journal of Manipulative and Physiological Therapeutics 1993; 16(2):96-103.
   
   
3. Slosberg M. Side Posture Manipulation for Lumbar Intervertebral Disk Herniation Reconsidered. Journal of Manipulative and Physiological Therapeutics 1994; 17(4):258-62.

Cubital Tunnel Syndrome - Part I

By Brad McKechnie, DC, DACAN

Cubital tunnel syndrome is the second most common entrapment neuropathy of the upper extremity, with reports of problems related to ulnar compression at the elbow dating back more than 100 years.¹

The cubital tunnel begins at the medial posterior condylar groove of the humerus at the point where the posterior condylar groove is spanned by the cubital tunnel retinaculum.² The cubital tunnel retinaculum runs from the medial epicondyle of the humerus to the olecranon and is transversely oriented to the path of the nerve.

The cubital tunnel retinaculum may be readily palpated by flexing the elbow and applying finger tip pressure along the path of the posterior condylar groove, because the retinaculum is tightened in flexion and relaxed in extension. The cubital tunnel continues distally, as the ulnar nerve passes between the two heads of the flexor carpi ulnaris muscle and enters the forearm.

Most cases of cubital tunnel entrapment of the ulnar nerve occur between 1.5 centimeters and 3.5 centimeters distal to the medial epicondyle. The cubital tunnel's volume decreases with flexion of the elbow, thus increasing the pressure on the ulnar nerve with this position.³ Pressures greater than 100 mm Hg have been documented to occur with elbow flexion when the cubital tunnel retinaculum is artificially tightened. Paresthesia has been induced within 10 minutes in normal subjects when peripheral nerves are artificially subjected to 50 mm Hg of compression.⁴ Most cubital tunnel entrapment syndromes are primarily due to compression of the ulnar nerve from the narrowing of the cubital tunnel with flexion of the forearm.

Work related factors may contribute to cubital tunnel syndrome. Complaints of this condition have been associated with computer keyboard operators working all day with the elbows flexed and the ulnar nerve resting on a hard surface. Truck drivers maintaining the elbows in the flexed posture for long periods with the added insult of elbow pressure versus the arm rest may also develop this syndrome.

Secondly, compression from local pathologies may also cause cubital tunnel syndrome by increasing the size of the nerve, decreasing the available space for the nerve by thickening of the cubital tunnel retinaculum, and by bony overgrowth from osteoarthritis. Additionally, the ulnar nerve may be compromised at this site due to local inflammatory swelling associated with rheumatoid arthritis or from the development of space occupying lesions at this site.

Common complaints associated with this syndrome include complaints of aching pain or discomfort at the medial aspect of the elbow and in the proximal forearm which are sometimes associated with shooting pains into the ulnar side of the hand and little finger. Occasionally there may be proximal radiations of pain. Frequently noted symptoms include numbness, tingling, and coldness of the little finger and the ring finger. Occasionally patients may have no noxious symptoms associated with this entrapment neuropathy. Nocturnal symptoms, while very common in carpal tunnel syndrome, are infrequent complaints in cubital tunnel syndrome.

Hypoesthesia in the ulnar field of the hand (Figure 1) may be intermittent or activity related. Activity related hypoesthesia is usually associated with repetitive movements of the elbow or from the maintenance of a constant flexed elbow posture. Motor complaints associated with cubital tunnel syndrome consist of complaints of clumsiness, weakness, and/or dropping of objects. Muscle wasting is primarily noted in the flexor digitorum profundus muscle to the fourth and fifth fingers, the flexor carpi ulnaris muscle, and the ulnar-innervated intrinsic muscles of the hand.

References 

1. McPherson SA and Meals RA. Cubital tunnel syndrome. Ortho Clin North Am, 23:111-123, 1992.
   
   
2. O'Driscoll SW, Horii E, Carmichael SW and Morrey, BF. The cubital tunnel and ulnar neuropathy. JBJS, 73-B:613-617, 1991.
   
   
3. Sunderland S. Nerves and Nerve Injuries, Second Edition, Churchill Livingstone, NY, 1978.
   
   
4. Buehler MJ and Thayer DT. The elbow flexion test -- a clinical test for cubital tunnel syndrome. Clinical Orthopaedics and Related Research, 233:213-216, 1988.

Management of Vertebral Artery Syndrome: A Conservative Approach

By R. Vincent Davis, DC, PT, DNBPM

Vertebral artery syndrome is considered synonymous with vertebral artery compression syndrome and vertebral-basilar artery insufficiency, and presents with recurrent transient episodes of cerebral symptoms, principal among which are dizziness, nystagmus, with sudden postural collapse without unconsciousness. 
These symptoms may be precipitated by rotation and hyperextension of the cervical spine which may result in temporary occlusion of the vertebral artery following which there is relative ischemia at the base of the brain. This syndrome commonly presents with a combination of cerebrovascular arteriosclerosis and cervical spondylosis as fundamental clinicopathological components.

To be brief, relative to this arterial anatomy, the vertebral artery ultimately supplies components of the brain via the basilar artery, and also provides for twigs to the cervical nerve roots which anastamose with the anterior and posterior spinal arteries. Since the internal carotids and vertebral arteries are the main tributaries to the basilar artery, occlusive arterial disease would gradually reduce arterial flow to a critical point at which further reduction in vascular caliber, prior to the development of an adequate collateral supply, would result in cerebral ischemia with respective clinical symptoms.

Normally, hyperextension with rotation of the cervical spine results in compression and occlusion of the vertebral artery on the contralateral side at the level of the atlas and axis. Occlusion may occur when vessels are subject to atheromatous disease and compression by osteophytes. If collateral blood flow is insufficient, symptoms develop with transient vertebral artery occlusion following rotation and/or hyperextension of the cervical spine. Symptoms are transient and subside as the arterial compression is released and blood flow is re-established. With degenerative disease of the cervical spine, arterial compression is increased due to encroachment by osteophytic projections at the level of the intervertebral foramina on the contralateral side during rotation/hyperextension movements. C5/6 is the site most often subject to osteophytic compression. With advanced degenerative disease, only a limited degree of motion may be necessary to produce complete vertebral artery compression. If the degenerative pathology develops slowly enough, the effects of vertebral artery stenosis may be offset by the formation of sufficient collateral circulation distal to the obstructive site to maintain adequate basilar arterial flow.

Symptoms may include, but not be limited to: dysarthria; headache; paresis of one or more of the extremities; blurred vision; and the "drop attack." Of course, referral for vertebral arteriography is necessary.

In the early stage, when periods of transient episodes are limited in time to only a few seconds and neurological orientation is not involved with "drop attacks" or paresis, conservative care my be initiated.

A cervical collar, or brace, to avoid cervical spine rotation/hyperextension may be applied. Again, in the early stage, cervical spine traction may be applied and should be a prolonged application with the appropriate anterior angles for achieving patency of the intervertebral foraminae. The traction should be directed at 90 degrees to the base line, or straight up, to traction C1 (atlas) and C2 (axis), and a 20 to 30 degree anterior angle for the remainder of the cervical segments. This may be applied b.i.d., or t.i.d. for 30 minutes with as much weight as possible without resulting in counter tractional muscle spasm. Unless manipulation results in a significant clinical change in the patient's condition, it should be avoided, and never performed in the advanced degenerative disease or in the presence of pronounced neurological signs and symptoms. Moist heat therapy may be applied prior to administering traction in the form of hydrocollator packs, or silicone gel packs, and with the usual caution to avoid erythema ab igne. Thermal therapy may be applied for no more than 25 minutes per application.

This author recommends concomitant management of the patient with an internist in order to monitor the administration of an anticoagulant to avoid thromboembolic disease, especially in the presence of paretic extremities, or "drop attack" episodes. Dizziness following rotation of the head to one side, or by looking upward, are initial clinical warnings of vertebral artery syndrome.

References

Davis RV. Therapeutic Modalities for the Clinical Health Sciences, 2nd ed, 1989, Library of Congress Card #TXU 389 661.

Griffin JE and Karselis TC. Physical Agents for Physical Therapists, 2nd ed.. Springfield: Charles C. Thomas, 1982.

Krusen, Kottke, and Ellwood. Handbook of Physical Medicine and Rehabilitation, 2nd ed. Philadelphia: W.B. Saunders Company, 1971.

Schriber WA. A Manual of Electrotherapy, 4th ed. Philadelphia: Lea & Feibiger, 1975.

Turek. Orthopedics -- Principles and their Application, 3rd ed. Lippincott.

Comparison of three active therapies for chronic low back pain: results of a randomized clinical trial with one‐year follow‐up

1. A. F. Mannion, 
 
2. M. Müntener ¹ , 
 
3. S. Taimela ² and 
 
4. J. Dvorak

5. 1. Abstract

      Objectives. To examine the relative efficacy of three active therapies for patients with chronic low back pain.

      Methods. One hundred and forty‐eight subjects with chronic low back pain were randomized to receive, twice weekly for 3 months, (i) active physi otherapy, (ii) muscle reconditioning on training devices, or (ii) low‐impact aerobics. Questionnaires were administered to assess pain intensity, pain frequency and disability before and after therapy and at 6 and 12 months of follow‐up.

      Results. One hundred and thirty‐two of the 148 patients (89%) completed the therapy programmes and 127 of the 148 (86%) returned a questionnaire at all four time‐points. The three treatments were equally efficacious in significantly reducing pain intensity and frequency for up to 1 yr after therapy. However, the groups differed with respect to the temporal changes in self‐rated disability over the study period (P=0.03): all groups showed a similar reduction after therapy, but for the physiotherapy group disability increased again during the first 6 months of follow‐up whilst the other two groups showed a further decline. In all groups the values then remained stable up to the 12‐month follow‐up. The larger group size and minimal infrastructure required for low‐impact aerobics rendered it considerably less expensive to administer than the other two programmes.

      Conclusions. The introduction of low‐impact aerobic exercise programmes for patients with chronic low back pain may reduce the enormous costs associated with its treatment.

      Key words

      • Chronic low back pain
         
      • Exercise
         
      • Aerobics
         
      • Back reconditioning
       
      • Physiotheraphy
         
      • Disability.

      Musculoskeletal disorders, of which back pain accounts for more than half the number of cases, are the most common cause of chronic incapacity in industrialized countries [1]. Chronic low back pain (cLBP), typically defined as low back pain lasting longer than 3 months, represents a particularly costly sociomedical problem because of the expenditure associated with repeated treatment and the long‐term absence from work and need for social support [2]. The development of effective interventions aimed at management of the chronic problem are thus urgently required. Active treatments are increasingly advocated for the treatment of cLBP [3], although few studies have documented the relative efficacies of different types of programme [4]. A number of types of functional restoration programme for patients with cLBP have been established, many of which include exercise routines on special equipment with the aim of reversing the compromised trunk muscle function and mobility of these patients [5–7]. However, it has also been suggested that the specific exercise modality is less important than simply encouraging normal movement and improving general fitness [3, 8]. Nonetheless, this has not been examined within the confines of a randomized clinical trial of different exercise modalities. This is an important issue to address, not least because different programmes are often associated with vastly different implementation costs. In view of the limited resources faced by health‐care providers worldwide, this economic issue cannot be disregarded.

      The aim of the present study was to carry out a randomized clinical trial to examine the relative efficacies of three active therapy programmes for cLBP patients: modern individual physiotherapy, specific trunk‐muscle conditioning using training devices, and group low‐impact aerobics. Outcome was assessed up to 1 yr later in terms of self‐rated pain intensity, pain frequency and disability. The short‐ and medium‐term results of the study have been reported previously [9].

      Previous SectionNext Section

      Methods

      Study population

      Participants were recruited into the study following advertisement in local media. Admission criteria were checked by medical history interview and clinical examination. The main inclusion criteria were: less than 65 yr old; more than 3 months of low back pain with or without referred pain (non‐radicular) serious enough to require medical attention or absence from work; and willingness to comply with the randomly assigned treatment. The exclusion criteria were: constant or persistent severe pain; pregnancy; previous spinal surgery; current nerve root entrapment accompanied by neurological deficit; spinal cord compression; tumours; severe structural deformity; severe instability; severe osteoporosis; inflammatory disease of the spine; spinal infection; severe cardiovascular or metabolic disease; and acute infection.

      The patients gave their signed, informed consent to participation and the study was approved by the University Ethics Committee (University of Zürich).

      Assignment to the treatments

      Before the start of the study, a randomization schedule was drawn up for prestratified groups [stratified by age (less than 40 yr or greater than 40 yr) and sex], using a table of random numbers and a restricted randomization procedure (blocks of 15) [10]. After medical examination and upon agreeing to participate in the study, each patient was assigned a number (in chronological order of acceptance into the study) which would determine their later group membership, according to the randomization table. Once all baseline assessments and questionnaires had been completed, these consecutive numbers were entered into the random numbers table to determine group membership. Patients were assigned to one of the following three treatment groups.

      Physiotherapy group

      The patients had half‐hour individual physiotherapy sessions focused on improving functional capacity using strengthening, co‐ordination and aerobic exercises, and with instruction on ergonomic principles and home exercises.

      Devices group

      Patients had 1‐h sessions for muscle reconditioning using training machines/devices, in groups of two or three. Four exercise devices (DBC International, Finland) provided progressive, isoinertial loading to the trunk in the three cardinal planes. Each session was preceded by 5–10 min of aerobic warm‐up (e.g. cycling), and relaxation/stretching exercises were carried out before and after the use of each device.

      Aerobics group

      Patients took part in low‐impact aerobics classes lasting 1 h, comprising exercises to music, with a maximum of 12 patients per group. A warm‐up of 10–20 min, involving whole‐body stretching and low‐impact aerobic exercises, was followed by 20–30 min of specific trunk and leg muscle exercises. The last 15 min of the class comprised cool‐down and stretching/relaxation exercises.

      The three types of treatment were administered in geographically separate areas of the hospital so as to avoid contact between patients in the different groups. No charge was incurred by the patient or their health insurance provider for receiving the treatment.

      Assessments before and after treatment

      Upon entry to the study (before randomization), after the 3‐month treatment period and at 6 and 12 months of follow‐up, a questionnaire booklet was administered to the patients to complete in their own time, enquiring amongst other things about the following: (i) sociodemographic information; (ii) low back pain intensity [Visual Analogue Scale (VAS) with a score range of 0–10], duration (in months) and frequency (pain‐free, sporadic, often, permanent); (iii) low back disability (Roland and Morris questionnaire [11]); (iv) beliefs about physical activity/work being a cause of back trouble and fears about the dangers of such activities when experiencing low back pain {Fear‐Avoidance Beliefs Questionnaire (FABQ) [12]}; and (v) psychological disturbance [13], using a combined score from the Modified Somatic Perception Questionnaire (MSPQ [14]) and the modified Zung questionnaire [15].

      Immediately after therapy, the questionnaire also enquired about any other treatments for back pain undertaken at the same time as the treatment was received in the study hospital. A list with 11 options was provided: acupuncture, pain medication, injection, physiotherapy, traction, manipulation, chiropractic, massage, corset, strength training, other.

      At the 6‐ and 12‐month follow‐ups, additional questions concerned the duration of the treatment effect and the patient's success in continuing independently with exercises similar to those learnt during the study.

      Statistical analysis

      The required sample size (approximately 54 per group) was determined, assuming a type I error probability of 5%, a type II error probability of 15% (i.e. power of 85%) and 15% dropout, based on the expected change in the clinical measures of pain and disability (determined from other similar exercise programs with similar patients [e.g. 5, 8]) [16]. Calculations were done for a medium effect size (0.55) for group differences after therapy. For the examination of treatment efficacy in low back pain, a sample of 50 volunteers per group after randomization has been considered methodologically adequate [4]. This number is also a manageable quota of additional patients that can be treated simultaneously for the purposes of the study with the resources and space available in the hospital.

      Changes in continuous variables over the four assessment periods were assessed by analysis of variance with repeated measures (group×time of assessment). Contrast analyses were used to identify differences (i) between the various time‐points and (ii) amongst the three groups in their pattern of change over time. Associations between categorical variables were analysed by contingency analysis and group differences in ordinal data were examined with the Wilcoxon rank sum test.

      The data were analysed using the intention‐to‐treat principle, whereby the data from all patients returning a questionnaire at the requested time, including patients who had not completed the full programme, were included. Significance was accepted at the 5% level.

      Previous SectionNext Section

      Results

      Study sample

      A flow diagram summarizing the formation of the final study group is given in Fig.1. From a total of 255 volunteers who responded to the initial recruitment drive, 159 satisfied the admission criteria; 148 of these chose to take part in the study and underwent randomization. One hundred and thirty‐two of the 148 (89.2%) completed the full programme. The majority of dropouts discontinued the programme because of changed work or family commitments or other medical problems and only rarely because they were dissatisfied with the treatment (Fig. 1⇔). There were 127/148 (86%) data sets available for the repeated‐measures analysis of the questionnaire data at all four time‐points (Fig.1). The proportions of participants in each group whose data contributed to this final analysis were as follows: devices 77%, physiotherapy 88%, aerobics 84% (P=0.39).

      Table 1⇔ shows some of the demographic characteristics of the patients; there were no significant differences amongst the three groups for any variable. The dropouts did not differ significantly from those who stayed with the treatment, other than that they were younger (40.1 vs 45.7 yr;P=0.033).

      Sixty‐one per cent of the patients declared they had received no additional treatments for their low back pain during the course of the treatment administered for the study, with no difference between the three groups. Those who declared they had undergone supplementary treatments undertook an average of 1.5 (range 1–3) options from the list of 11 readily available treatments (see Methods). There was no significant difference between the three groups in this respect (P=0.71).

      
      

      View larger version:
      

      • In this page
         
      • In a new window

      • Download as PowerPoint Slide

      FIG. 1.
      

      Flow diagram showing how the study group was formed and the number and group membership of dropouts throughout the course of the study. Shaded boxes represent patients who returned a questionnaire at every time‐point (large boxes represent those completing treatment; small boxes represent treatment dropouts) and whose data were included in the final intention‐to‐treat repeated measures analysis of variance. Open boxes represent patients who returned a questionnaire at any other time‐point.

      View this table:
      

      • In this window
         
      • In a new window

      TABLE 1.
      

      Baseline characteristics of the study participants

      Outcome measurements

      Pain score

      Changes in pain intensity (highest and average VAS score in the last 2 weeks) recorded over the four time‐points for each therapy group are shown in Table 2⇔. For the whole group of patients there was a significant reduction in mean pain intensity immediately after therapy, which was retained 12 months later, with no significant difference amongst the three groups in the extent of the change (Table 2⇔).

      Cut‐off scores were established to categorize clinically significant changes in highest pain intensity (VAS) on the basis of the results of a reliability study in a similar patient group (R. Stärkle et al., unpublished data) as follows: improvement=value reduced by more than 2.8 points; unchanged=values 2.8 higher or 2.8 lower than the pretherapy score; worse=value 2.8 points or more than before therapy. According to this classification, 12 months after therapy 46 patients (36.2%) were improved, 73 (57.5%) unchanged and four (3.15%) worse, with no differences amongst the groups (P=0.91) [four patients (3.15%) had pretherapy scores of less than 2.8 and therefore could not be categorized].

      View this table:
      

      • In this window
         
      • In a new window

      TABLE 2.
      

      Changes in self‐rated pain, disability, fear‐avoidance beliefs and psychological disturbance before therapy (1), after therapy (2) and at 6 and 12 months of follow‐up (3 and 4 respectively) (mean±S.D.) (see footnote for maximum and minimum scores achievable for each questionnaire)

      Pain frequency

      Before therapy, 63 patients (49.6%) suffered from low back pain permanently, 53 (41.7%) often and 11 (8.7%) sporadically. There was a significant reduction in pain frequency in all groups after therapy (Table 2). This was further improved upon during the next 6 months and remained stable up to the 12‐month follow‐up (no significant group differences in the pattern of change over time). At the 12‐month follow‐up, 36 patients (28.35%) suffered from low back pain permanently, 36 (28.35%) often and 46 (36.2%) sporadically; nine patients (7.1%) were pain‐free.

      Disability

      When the whole group of patients was considered, there was a significant reduction in self‐rated disability immediately after therapy, which was retained 12 months later (Table 2⇔). However, a significant interaction suggested that the groups had behaved differently with respect to their patterns of change in disability over time; contrast analyses identified a slight but significant (P=0.03) difference between the physiotherapy group and the other two groups (aerobics and devices; the latter did not differ significantly from one another). This was the result of an increase in disability in the physiotherapy group between the end of therapy and the 6‐month follow‐up; in contrast, during this same period the aerobics and devices groups showed a further reduction. During the final 6 months of the study (months 6–12), the values remained stable at their 6‐month follow‐up levels.

      Working on the same principle as described above for pain intensity, cut‐off scores for the clinically significant change in disability were also calculated: improvement=value reduced by more than 4 points; unchanged=value 4 higher or 4 lower than the pretherapy score; worse=value 4 points or more than before therapy. Twelve months after the therapy, 43 patients (33.9%) were improved, 54 (42.5%) unchanged and nine (7.1%) worse, with no significant differences amongst the groups (P=0.14). Twenty‐one patients (16.5%) had a pretherapy score of less than 4 and therefore could not be categorized.

      Duration of the observed treatment effect

      At the 12‐month follow‐up, the patients were asked to grade the period during which their complaint had been alleviated after the treatment: 1=treatment had no effect in the first place; 2=only a short time; 3=until now. The ratings for each group did not differ significantly (physiotherapy, 16, 49 and 35% respectively; aerobics, 29, 33 and 38%; devices, 34, 24 and 42%; P=0.16). The majority of the patients declared that they had continued independently, at least in part, with exercises similar to those taught in the hospital. There were no group differences [physiotherapy, 81%; aerobics, 86%; devices, 79% (P=0.72)]. There was a low but significant association between continuing with the exercises and the duration of positive effect after 6 months (contingency coefficient=0.25;P=0.016), but this failed to reach significance at 12 months. Those patients who continued with the exercises were more likely, at 12 months, to show a reduction in disability (65% showed a reduction compared with 41% in the group who did not continue exercising; contingency coefficient=0.19, P=0.03) and a reduction in pain intensity (78% showed a reduction compared with 55% in the group who did not continue exercising; contingency coefficient=0.20, P=0.02).

      Psychological parameters

      Fear‐avoidance beliefs about physical activity and about work were significantly reduced in all groups after the treatment (P=0.009), and the values remained significantly lower than those before therapy at both the 6‐ and the 12‐month follow‐up. There was no significant unique group effect regarding the pattern of change.

      There was a slight but significant difference between the pattern of change in psychological disturbance (scores from MSPQ and ZUNG questionnaires combined) for the physiotherapy group compared with that of the other two groups (P=0.015); in the aerobics and device groups these scores declined after therapy, then increased towards pretherapy values over the following 12 months, whilst the physiotherapy group showed no change after therapy, an increase at 6 months and then a reduction to pretherapy values after 12 months.

      Previous SectionNext Section

      Discussion

      The present study is the first clinical trial carried out to examine the relative efficacies of three active therapies for cLBP patients: individual modern physiotherapy, training on machines/devices and low‐impact aerobics. One‐to‐one physiotherapy and device‐training are both considered to be established therapies for cLBP, in that they are recognized and remunerable by health‐care systems, whereas low‐impact aerobics does not at present enjoy this status. The study was carried out as far as was practicable in accordance with previous recommendations [11, 17–19], but certain limitations need to be discussed. The study would naturally have been stronger with the inclusion of a no‐treatment (control) group. However, this was not considered ethical or practicable with the study design chosen. A recent systematic review has provided good evidence of the effectiveness of exercise for cLBP patients [4] and the superiority of an active intervention over a control treatment has also been proved in a randomized controlled trial [5]. The main focus of the present study was therefore to examine whether there were differences in the effectiveness of the different active therapies, and their possible modes of action [20, 21]. Participation was encouraged through media advertisement and was voluntary; as such it would most certainly have been threatened by the inclusion of a no‐treatment group. If the patients had been recruited following general practitioner referral to the hospital and had the waiting lists for treatment been known to be long, the inclusion of a ‘waiting list control group’ would have been facilitated. However, this was not the case. Furthermore, as the majority of the patients had a long history of back pain (over 3 yr in 76% of patients), it was considered unlikely that the observed results simply reflected the natural history of back pain per se rather than the effect of the interventions. The patients displayed pain and disability characteristics comparable to those of the typical cLBP patient described in many previous intervention studies [22–24]. However, as voluntary recruits into the study, they were well‐motivated to undertake one of the active therapies in an attempt to alleviate their prevailing symptoms and were mostly (potential) participants of working life. Whether similar results would have been observed for more severely disabled patients or those with confounding psychosocial problems, who often seek help in the tertiary care setting, remains to be shown.

      The three active treatments proved to be equally efficacious in their ability to reduce pain intensity, pain frequency and disability in tasks of daily living immediately after therapy, even in those patients whose initial values for these three variables were very high. With respect to pain characteristics, these positive effects were well maintained and sometimes even improved upon in all groups over the subsequent 12 months. The mean decrease in highest pain intensity for the whole group after 12 months was 1.8 points on the VAS (effect size 0.9) and, with a quite stringent cut‐off criterion, approximately 36% of the whole group also showed a significant clinical change in pain intensity (reduction of more than 2.8 points on the VAS). Furthermore, when the patients were asked directly after the therapy whether their pain and their ability to perform their everyday functions had worsened, stayed the same or improved, compared with levels before therapy, 57% (pain) and 49% (function) declared some or great improvement [9]. Perhaps, in this sense, the cut‐off criterion was too strict, as individuals appeared capable of detecting a change in their clinical symptoms which was not registered as a clinical improvement by the VAS cut‐off score. The whole issue of statistical and clinical significance is a complicated one, and many methods used previously for the identification of a clinically important difference appear to have been somewhat arbitrary [25].

      With regard to self‐rated disability, the groups showed differences in their course of change in the year after therapy: during the first 6 months the devices and aerobics groups displayed a further decline in disability, whilst the physiotherapy group showed a regression towards pretherapy levels. This divergent behaviour of the groups with regard to disability, but not pain, suggests that the patients’ interpretation of the disabling effects of the pain or adjustment to the pain may have played an important role during this time. It has been shown that fear‐avoidance beliefs [26] and self‐efficacy [27] are significant contributors to the extent that people consider themselves disabled by their chronic pain. In the present study, fear‐avoidance beliefs tended to follow the pattern of change described for disability, especially with respect to the first 6 months after discharge from therapy. In modifying fear‐avoidance beliefs, the manner in which the patients were forced to confront their apprehensions may have played a pivotal role. One‐to‐one physiotherapy perhaps promotes a sense of dependence of the patient on the therapist to guide and govern the most appropriate activity level for them in accordance with their declared level of pain. In contrast, with a group‐exercise approach, this responsibility is rather more centred upon the patient himself or herself: when patients experience themselves behaving differently from their expectations, this can be expected to reduce fear and improve self‐efficacy. Thus, the difference in the behaviour of the groups in the first 6 months after therapy may have, in part, reflected a type of withdrawal effect from the individual guidance given during the one‐to‐one therapy. The corresponding increases in psychological disturbance seen in this group over the first 6 months after therapy tend to support this hypothesis. Nevertheless, when the changes in disability from pretherapy levels to the 12‐month follow‐up were determined using a more stringent criterion based on clinically significant change, the group differences no longer reached significance.

      There was a slight but significant association between continuing with the exercises learnt during the treatment in the hospital and sustained improvement of clinical function, suggesting that the interventions may have provided the necessary impetus to encourage the patients to become more physically active in their daily lives in an attempt to alleviate their pain and disability. This phenomenon has been reported previously [28].

      The costs of administering the different programmes varied widely. The larger group size and minimal investment with regard to infrastructure rendered aerobics considerably less expensive than either one‐to‐one physiotherapy or small‐group device‐training. The charges that would have been made to the patient's health insurance for the different therapy programmes undertaken in the present study were as follows: aerobics, 288 Swiss francs (SFr) (determined from local commercial centres, as they are not at present financed by the insurance companies); physiotherapy 960SFr; devices 1120SFr. This gives a cost ratio of 1:3.3:3.9. In the present study, the respective cost ratio for personnel alone for 1 h of treatment was 1:5:6. Epidemiological studies carried out in the UK have shown that 3–7% of the population report their back problems as being chronic [1], and a significant proportion of these people will seek medical attention continually for their condition. The most common treatments currently employed for cLBP patients are physiotherapy and—with increasing popularity—‘reconditioning’ programmes carried out on training machines. If the results of the present randomized study can be verified by further studies in which the treatment is prescribed rather than undertaken voluntarily, the introduction of low‐impact aerobic exercise programmes for patients with cLBP should allow considerable savings in the direct costs associated with its treatment.

      Previous SectionNext Section

      Acknowledgments

      We are grateful to the Swiss National Science Foundation (grant 32–50979.97) and the Schulthess Klinik Research Fund for funding this work. We thank DBC International for providing staff training and for the loan of the training devices. We would like to thank Melinda Schuetz, Raymond Denzler, Olivia Zimmerli, Tobias Sundberg, Annette van Bolhuis, Julia Reutimann and Geoff Klein for their help and cooperation.

      Previous SectionNext Section

      Footnotes

      • ↵Correspondence to: A. F. Mannion, Department of Neurology, Schulthess Klinik, Lengghalde 2, 8008 Zürich, Switzerland.

      • © British Society for Rheumatology

      Previous Section
      

      References

      1. ↵ 
         Clinical Standards Advisory Group. Epidemiology review: The epidemiology and cost of back pain. Annex to the Clinical Standards Advisory Group's Report on Back Pain. London: HMSO, 1994.
      2. ↵ 
         Hazard RG. Chronic low back pain and disability: The efficacy of functional restoration. Bull Hosp Joint Dis1996;55:213–6.

         Medline
      3. ↵ 
         Abenhaim L, Rossignoi M, Valat JP et al. The role of activity in the therapeutic management of back pain. Report of the International Paris Task Force on Back Pain. Spine2000;25 (4 Suppl.):1S–33S.

         CrossRefMedlineWeb of Science
      4. ↵ 
         van Tulder MW, Koes BW, Bouter LM. Conservative treatment of acute and chronic nonspecific low back pain. A systematic review of randomized controlled trials of the most common interventions. Spine1997;22:2128–56.

         CrossRefMedlineWeb of Science
      5. ↵ 
         Kankaanpää M, Taimela S, Airaksinen O, Hanninen O. The efficacy of active rehabilitation in chronic low back pain. Effect on pain intensity, self‐experienced disability and lumbar fatigability. Spine1999;24:1034–42.

         CrossRefMedlineWeb of Science
      6. Manniche C, Lundberg E, Christensen I, Bentzen L, Hesselsoe G. Intensive dynamic back exercises with or without hyperextension in chronic back pain after surgery for lumbar disc protrusion: A clinical trial. Spine 1993;18:560–7.

         CrossRefMedlineWeb of Science
      7. ↵ 
         Denner A. The trainability of the trunk and neck musculature of deconditioned back pain patients. Manuelle Med1999;37;34–9.
      8. ↵ 
         Frost H, Lamb SE, Klaber Moffett JA, Fairbank JC, Moser JS. A fitness programme for patients with chronic low back pain: 2‐year follow‐up of a randomised controlled trial. Pain1998;75:273–9.

         CrossRefMedlineWeb of Science
      9. ↵ 
         Mannion AF, Müntener M, Taimela S, Dvorak J. A randomized clinical trial of three active therapies for chronic low back pain. Spine1999;24:2435–48.

         CrossRefMedlineWeb of Science
      10. ↵ 
          Kirkwood B. Essentials of medical statistics. Oxford: Blackwell Science, 1988.
      11. ↵ 
          Roland M, Morris R. A study of the natural history of back pain. Part I: development of a reliable and sensitive measure of disability in low‐back pain.Spine1983a; 8:1414.

          MedlineWeb of Science
      12. ↵ 
          Waddell G, Newton M, Henderson I, Somerville D, Main CJ. A Fear‐Avoidance Beliefs Questionnaire (FABQ) and the role of fear‐avoidance beliefs in chronic low back pain and disability. Pain1993;52:157–68.

          CrossRefMedlineWeb of Science
      13. ↵ 
          Greenough CG, Fraser RD. Comparison of eight psychometric instruments in unselected patients with back pain. Spine1991;16:1068–74.

          CrossRefMedlineWeb of Science
      14. ↵ 
          Main CJ. The Modified Somatic Perception Questionnaire (MSPQ). J Psychosom Res1983;27:503–14.

          CrossRefMedlineWeb of Science
      15. ↵ 
          Main CJ, Waddell G. The detection of psychological abnormality in chronic low back pain using four simple scales. Curr Concepts Pain1984;2:10–15.
      16. ↵ 
          Altman DG. Statistics and ethics in medical research: III. How large a sample.Br Med J1980;281:1336–8.
      17. ↵ 
          Deyo RA, Andersson G, Bombardier C et al. Outcome measures for studying patients with low back pain. Spine1994;19(Suppl. 18):2032S–6S.

          Medline