A Review of the Literature on Shortwave Diathermy as Applied to Osteo-arthritis of

A Review of the Literature on Shortwave Diathermy as Applied to Osteo-arthritis of the Knee Summary This review specifically examined the efficacy of shortwave diathermy (SWD) for alleviating the main symptoms of osteo-arthritis (OA) of the knee. To this end, the available databases were searched, and then the outcomes of the 11 relevant non-randomised comparative and randomised controlled clinical studies detailing the application of SWD for treating knee OA were critically evaluated using specified criteria, and their outcomes categorised in terms of their favourable, non-favourable and questionable effects on pain and mobility.

Given the equivocal findings and poor methodological quality of most studies reviewed, we conclude further controlled studies are essential to establish whether either continuous or pulsed SWD is efficacious for treating patients with knee OA.

Marks, R, Ghassemi, M, Duarte, R and Van Nguyen, J P (1999). 'A 1.c.vic.w of thr lit.eratiirc o i l shortwave diathermy as iq)pl i cd 10 o s teo-arthrit i s

85, (i, 304-3 16. ( ) f ' t l 1 c It11 rr ' , z%y.sioth/r/gly,

Introduction Osteo-arthritis (OA), the most prevalent of the rhcurnatic diseases, affects more than 60% ofwestern World adults over the age of 65 years (Lawrence et al, 1986), with the knee being one of the most corrinionly afflicted .joints (Davis, 1988). Knee OA, marked by pain, deformity, inflammation, stiffness, muscle atrophy and damage, and progressive loss of independence (Threlkeld and Currier, 1988) is also considered a leading cause of functional disability in the elderly (Hochberg, 1984).

No curative treatment has yet been found for knee OR (Puet.t and Grif'ien, 1994). Treatment is therefore directed towards symptom relicf arid the prevention of further functiorial deterioration (Dekker et al, 1992; Munice, 1986) and often includes a number of' e lec tro th erapeu tic modalities (Marks and Cantin, 1997). As yet, however, it is relatively unclear whether any o f these modalities is efficacious, over and above the placebo effect.

In view of this, we initially set out to examine and document all the research findings concerning the unique value of those e 1 e c t r o t h e r ap e u t i c inter ve n t.i o n s commonly described in the English literature for treating OA of the knee. Particular attention was paid riot only to the study results, but to analysing the quality of

the research designs used by the various investigators in support of these. However, given that many of the investigations reviewed at that time were found to have used a combination of electrotherapeutic modalities, and that these were poorly controlled, with very low methodological scores (Ghassemi and Marks, 1995), we were not able to determine their separate effects with any clarity.

In the present paper, we thus attempted to remedy this situation by specifically describing the results of those studies which employed shortwave diathermy (SWD), one form of electrotherapy widely applied to alleviate the symptoms associated with OA joint disease (Vanharanta, 1982), separately from studies describing other forms of electrotherapy. We have also attempted to classify the published trials according to their implied findings and to examine each published trial with respect to method- ological issues, using a standardised format and three separate reviewers who were essentially blind to the authors and ins ti tu ti o n s . Notwithstanding the poor quality of most of these studies, the discussion then attempts to summarise the present findings s o that future work may be directed towards clarifying and overcoming the present limitations in this important area.

The data were retrieved from a com- puterised Medicine, Cumulative Index to Nursing and Allied Health, and Excerpta Medica database search plus a manual search o i bibliographies of original and review articles and appropriate Internet resources. Studies on both continuous SWD which has been used in the treatment of many conditions for a considerable time by physiotherapists and those on pulsed SWD, often referred to as pulsed electroniagnetic field (PEMF) treatments, first introduced in the early 1950s (Kitchen and Partridge,

1992) were evaluated. Overall, the time period 1955-97 was considered and the key words used were osteo-arthritis, arthritis, knee joint, shortwave diathermy, electromagnetic fields, physical therapy, physiotherapy, diathermy, diapulse, shortwave therapy, pain, articular cartilage, .joints, and electrotherapy. The articles had to be published in English or have an English abstract.

‘This literature search revealed 11 relevant treatment studies, and in addition several basic studies, relevant reviews concerning the thermal and non-thermal effects of SWD, and the application of SWD and PEMF applications in several animal models of arthritis and related soft tissue abnormalities (see tables 1 and 2 ) .

T h e retrieved clinical studies were reviewed qualitatively and in narrative form t o highlight their distinctive protocols and varied outcomes (see table l ) , and quantitatively according to the criteria of Beckerman et al (1992) and Gam and Johannsen (1995), to assess their method- ological deficiencies (see tables 3 and 4).

Osteo-arthritis Osteo-arthritis, a progressive degenerative disease affecting synovial joints such as the knee, is characterised by the focal loss ofthe articular cartilage lining of the joint, by sclerosis of the underlying subchondral bone, and by osteophyte formation at the joint margins (Doherty and Jones, 1994; Souhami and Moxham, 1990). A number of processes that may be involved in the development of OA include a failure of c arti 1 age re m o de 11 in g , in flarnm a tio n , ligamentous damage, altered neurological and muscle function, muscle damage, and pathological changes in the surrounding soft tissues which can increase articular compression and promote further joint damage (Marks, 1993).

As a consequence of these pathological changes, people diagnosed as having O A may present clinically with considerable pain, stiffness, .joint swelling, and secondary muscle wasting in and around the affected joirits (Doherty and Jones, 1994). Function may also be significantly impaired over time and considerable disability may result (Munice, 1986).

Owing to the limits of medical treatments for reducing these disease symptoms, and claims that SWD may have beneficial effects in this respect, this form of physiotherapy is often recommended for the treatment of OA.

Shortwave Diathermv Shortwave diathermy, a form of electromagnetic therapy, produces a n oscillating electromagnetic field in the frequency range of 27.12 MHz. These oscillations, applied in either the continuous o r pulsed modes, are thought to cause movement of ions, distortion of molecules and creation of eddy currents within the field (Goats, 1989a, b ) . The therapeutic effects of these oscillations lies in their ability to decrease tissue viscosity and with this muscular (Thom, 1966) and tendinous contractures (Lehmann et a l , 1970). Additionally, the deep heating effect of continuous SWD may induce an anti- inflammatory response (Kitchen and Partridge, 1992; Nadasdi, 1960); reduce joint stiffness (Wright, 1973); stimulate connective tissue repair (Aaron and Ciomber, 1993); reduce muscle spasm and pain, restore the action potential of traumatised muscle and aid healing of muscle tissue (Bansal et al, 1990) and of bone (Aaron et al, 1989). It may also alleviate tendinitis (Andrew and Bassett, 1993) and/or render a joint more amenable to physical exercise (Vanharanta et al, 1982). Additionally, at the cellular/biochemical level, SWD has been shown to increase the uptake of 35s-sulphate by capsular tissues of rabbit knees, the glycosaminoglycan concentration of treated cartilage (Liu et al, 1996; Threlkeld, 1984), and the galacto- samine and glucosamine concentrations of ligamentous tissue (Vanharanta et al, 1982) which could be highly beneficial in mediating repair in damaged OAjoints.

In the pulsed mode, which is usually of the same frequency as continuous shortwave but in pulse trains varying from 25-400 microseconds delivered at a rate of between 15 and 800 per second (Kitchen and Partridge, 1992), the thermal effect of SWD is said to be reduced, although not entirely lost. However, pulsed mode SWD therapeutic effects which may include the induction of osteogenesis ( Binderman et al, 1985), and cytodifferentiation of cartilage (Grigorieva et al , 1980; Liu et al , 1996; Sanseverino, 1980), are generally attributed to a non-thermal effect (Chapman, 1991). In addition, the use of low frequency PEMF may produce a potentiated anti-inflammat- ory effect (Fabbri and Lucchese, 1980; Nadasdi, 1960) and reduce pain (Wdrnke, 1983) without the unfavourable effects of excessive heat production (Nadasdi, 1960).

Due to the potential benefits of either

Authors Ray Marks PT is director of clinical rcsearch in the Osteo-arthritis Research Centre in Toronto, and a rcsearch fellow and lecturer at Columbia Univcrsity’s Tkachers (hllegc Department of Mcalth and Beliaviour. Masoumeh Ghassemi BSc PT and Richard Duarte BSc PT are physical therapists in Toronto, Ontario. John P Van Nyugen BSc PT is a physical therapist in Peel, Ontario.

This article was rcceived onJuly 29, 1997, and accepted on.July 27, 1998.

Address for Correspondence li Marks, PO Box 1153 Adelaide Postal Smtiori, Toronto M5C 2K5, Ontxio, Canada.

Physiotherapy,June 1999/vol85/no 6

306

Table 1: Published studies of shortwave diathermy for treatment of knee OA

Authors Sample

Basil and Joshi (1975)

Chambet . i n eta/ (1982)

Clarke eta/ (1 974)

Hamilton eta/ (1959)

Jan and Lai (1991)

Klaber Moffett eta/ (1996)

Lankhorst eta/ (1982)

60 OA knee patients (40 F, 20 M; 20 uniMO bilateral, ages 40-85 years)

30 SWD 30 US

42 OA knee patients (10 M, 32 F)

24 SWD 18 Ex

48 OA knee patients (15 M, 33 F, mean age 61 yr)

13 placebo SWD 17 SWD 15 Ice

100 patients

18 RA knee 26 OA knee 33 RA hand

61 female OA knee patients (28 uni133 bilateral, ages 40-74 years)

21 knees US 28 SWD 20 US + Ex 25 SWD + Ex

92 patients (34 M, 58 F, 35-80 years)

46 OA hip 46 OA knee

30 active pulsed SWD

30 placebo SWD

30 no treatment

24 OA knee patients

12 SWD +Ex 12 SWD

Design Controls Outcome measures Results

Prospective randomised mixed factor experiment comparing coplanar SWD for 20 min and US for 3-7 min

Prospective equivalent groups randomised mixed factor experiment comparing SWD and ex vs ex alone

Prospective equivalent groups randomised mixed factor design of SWD versus ice

Repeated treatment design with cross-over of SWD versus IR versus Faradism vs wax

Prospective equivalent groups randomised mixed factor design of SWD versus US versus SWD and Ex versus US and Ex

Prospective placebo controlled double blind trial of the comparative efficacy of active pulsed SWD, placebo SWD, and no treatment

Prospective randomised controlled trial using independent observer

relief, ROM, ability 6 squat, cross-leg sitting, walking, stair climbing, presence of swelling, crepitus and deformity

No Pain, discomfort, deqree of 1, Subjectively SWD was slightly more effective than US in cases of acute pain

2. Objectively, US seemed marginally superior

No Pain, function (walking, stairs, kneeling, use of cane), ROM, endurance

Yes Pain (0-4 scale), stiffness (0-3 scale), ten%erness and swelling (0-3 scale), knee girth, walk time, doctor's assessment

Yes Girth, ROM, strength, walk time, stairs

No Functional incapacity, knee torque

Yes Pain, general health, activities of daily living

No Maximal knee extensor torque, walking speed, number of steps, stair climbing

1. SWD + Ex versus Ex alone were equally effective in decreasing pain and increasing function a t 4 weeks

2. Effect was maintained at 12 weeks only in those who continued Ex

1. Group receiving ice showed best improvement at 3 weeks

2. Equal improvement in all groups at 3 months

1. Improvements in outcome with SWD, IR, Faradism, and cold SWD

2. No significant difference between the outcomes measured for any electrotherapeutic modality and cold SWD

1. All groups had similar decreases in pain and functional improvements

2. Ex in addition to SWD promoted treatment effect

1, No difference between active or placebo treatment in effect on pain

2. Patients in placebo group reported more benefit from treatment than those receiving active treatment

1. Marked improvement in knee torque and function for both groups (p < 0.001)

2. No group difference in any outcome measure

Physiotherapy June IYl)l)/vol85/no 6

Prof essiona I a r t ides

Table 1 continued

307

Authors Sample Design

Quirk eta / (1985)

Svarcova eta / (1988)

Valtonen and Alaranta (1971)

Wright (1964)

38 OA knee patients (9 M, 29 F; mean age approx 60 years)

21 IFC + Ex 12 SWD + Ex 14 Ex only

180 OA hip and knee patients (mean age 63 years)

60 US 50 galvanic current 60 SWD

132 OA knee patients (15 M, 117 F; mean age 62 years)

70 SWD 62 LWD

38 OA knee patients (3 M, 35 F; mean age 62 years)

13 placebo tablets

12 placebo intra-articular injections

13 SWD

Pilot study of SWD versus IFC versus Ex using a randomised mixed factor design

Mixed factor design comparing pulsed SWD + medication, US, galvanic current

Randomised trial comparing the effects of SWD versus LWD 3 x week with average course of 13-14 treatments

Randomised trial comparing the effects of placebo tablets 2 x day, fortnightly placebo injections, and SWD 3 x week for 6 weeks

Controls Outcome measures

No ROM, Ex tolerance, knee girth, rest pain (verbal and VAS)

Yes Pain relief using VAS, therapeutic effect as evaluated by patient and physician

N o Subjective distress level

Yes Walk time, tenderness, pain, number of analgesic tablets taken daily

Results

1. Post-treatment improvement in mean pain score for all 3 groups.

2. No significant outcome difference between regimes was found

1. No difference between therapies or combined effect of physiotherapy and drugs and physiotherapy alone, after 5 treatments

2. After 10 treatments pain intensity (VAS) decreased and the combined effect of physio and drug therapy was significant (p < 0.05). No outcome difference between SWD, US or galvanism was found

1.One-fifth patients markedly improved

2. Three-fifths slightly improved

3. One-fifth no benefit at all

4 SWD provided similar results to that of LWD

1. Significantly greater benefits derived from SWD than from placebo

2. More patients improved after SWD than after injections, and this was more noticeable in long- term improvement, but differences did not achieve significance

Abbreviations: Ex = exercise; F = females; IR = infra-red; IFC = interferential current; LWD = longwave diathermy; M = males; ROM = range of motion; US = ultrasound; SWD = shortwave diathermy; VAS = visual analogue scale

Table 2: Animal studies of shortwave diathermy applied to knees in experimental arthritis

Authors Sample Methods Controls Outcome measures Results

Nadasdi 30 rats Athermal pulsed SWD (1960) applied at 400 pulses

per sec and penetration rate o f 4 for 10 min at a time

Vanharanta 9 rabbits SWD 55 times for (1 982) 5 min at 50 watts for

11 weeks t o one knee

Yes Rate o f inflammation SWD significantly inhibited inflammation during experimental joint damage Volume o f paw

Yes Range o f motion 1. SWD increased development of extension deficiency at knee o f the treated group

2. Small decreases in flexion mobility

(goniometer)

X-rays

Physiotherapy June IY99/vol85/no 6

Table 3: Results of methods assessment of clinical trials of shortwave diathermy for knee OA and their criterion scores (adapted from Beckerman et a/ (1992) and from criteria described by authors of reported studies)

Criteria Points awarded Possible Banshi Chamber- Clarke Hamilton Jan and Klaber Lakhorst Quirk Svarcova Valtonen Wright points and Joshi lain e ta / e ta / e ta / l a ; (1991) Moffett e ta / e ta / e ta / and Alaranta (1964)

(1975) (1982) (1974) (1959) (1996) (1985) (1985) (1988) (1971)

Randomisation correct 1 1 1 1 1 0 1 1 1 0 1 1

No of patients in smallest group after randomisation 3* 1 0 0 1 1 1 0 0 2 2 0

Percentage lost t o follow-up 2+ 2 1 2 0 2 2 2 2 2 2 2

No selective loss t o follow-up 1 1 0 1 1 1 1 1 1 1 1 1

I I

Restriction t o homogeneous group 1 1 1 1 1 1 0 1 1 0 0 1

Relevant baseline characteristics described 1 1 0 1 1 1 1 1 1 1 0 0

Co-interventions similar in all groups 1 0 0 0 1 0 0 1 0 0 0 0

Com para bi I i ty of prognoses groups 1 1 1 1 1 1 0 1 0 0 1 1

Correction for imbalance at baseline 1 0 1 1 0 0 0 0 0 0 0 0

Patients blinded 1 0 0 1 0 0 1 0 0 0 0 0

Therapists blinded 1 0 0 0 0 1 1 0 0 0 0 0

Evaluator blinded 1 0 1 1 0 0 0 1 1 0 0 0

Outcome measured relevant and well described 2* 1 1 1 1 2 1 1 0 1 1 1

relevant points in trial 1 1 1 1 1 0 1 0 0 0 1 1

Outcome measured after treatment period (follow-up) 1 0 1 1 1 0 1 1 1 0 1 1

Analysis blinded 1 0 0 0 0 0 0 0 1 0 0 0

Adverse effects investigated 1 0 1 1 1 0 0 0 1 0 0 0

Intention-to-treat analysis 1 0 0 0 1 0 0 0 1 0 0 0

Outcome measures at

Frequencies o f most important outcomes presented for each group 1 1 1 1 0 1 1 0 1 1 1 1

explicitly described 1 0 0 0 0 0 1 1 1 1 0 0

Other interventions standardised and described 1 0 1 1 1 1 0 1 0 1 1 1

Statistical analysis not correct -1 0 0 -1 0 -1 0 0 0 0 0

Power calculations not performed -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1

Total 25 10 11 14 12 10 12 12 11 9 8 1 0

Therapy standardised and

-

*2 25 = 1 point, * Valid and reliable according t o reviewer = 1 point, valid and reliable according t o relevant publications = 2 points

Note: One point was awarded i f criterion was attained, with a maximum score o f 25 points, unless otherwise stated. Where reviewers were uncertain about whether a criterion was implemented or criteria were clearly not met, no points were awarded.

t 50 = 2 points, t 75 = 3 points 5 20% = 1 point, 5 10% = 2 points L - ~~ .. ... -

PhysiotherapyJunc l'.N9/vol 85/no 6

Prof essiona I a rt ides

Table 4: Information abstracted from 11 clinical trials examining shortwave diathermy applications in knee OA with available data expressed as a percentage of the 11 reviewed studies (adapted f rom Gam and Johannsen, 1995)

Percentage

Sample size Sex Age Selection criteria inclusion criteria Exclusion criteria Description o f drop-outs

Description of randomisation method Comparison group Placebo SWD Control group and placebo SWD Co-interventions

Description of apparatus Name of apparatus Description of placebo-SWD apparatus Frequency (MHz) Intensity (W) Mode of delivery

Treatments Number Treated area Dosehime per treatment Total treatment period

Assessments Subjective measures - pain*, patientlphysician

assessments Objective measures -function, knee range,

knee strength Subjective and objective measures

100 73 82 90 45 64 45

55 90 27 9

63

55 45 9

27 18 55

100 100

18 81

100

31

81 54

*Pain instruments included unsegmented visual analogue scale, verbal scoring of rest, post-exercise, nocturnal pain; pain at rest, level walking, ascending and descending stairs and squatt ing on a three- po int scale; pain relief on a four-point scale; pain and subjective reports on pain intensity and pain related distress using a 0-100 numerical scale; crude scoring of diurnalhocturnal pain with weighting score according t o severity and duration.

continuous or pulsed electromagnetic fields for reducing the symptoms and signs of OA i n c 1 u d i n g pa i n , in fl am m a ti o n , caps u 1 ar thickening and contracture, muscular and tcndinous contractures, and cartilage proteoglycan and ligamentous collagen losses, plus the high incidence of functional disability associated with this disease at the knee jo in t , we specifically elected to establish whether SWD, in any form (continuous or pulsed), is efficacious for treating knee OA. To this e n d , the 11 published studies we retrieved and analysed in depth, with respect to their conclusions, experimental methodologies and limitations, are shown in tables 1, 3 and 4. These trials and their outcomes are also described in detail, in the following section.

Clinical Trials with SWD . . . . . . Reasonably’ Favourable Results T h e earliest evidence of a reasonably favourable outcome for the application of SWD for treating knee OA was that reported by Wright (1964). In that seminal study, Wright compared the outcome of six weeks of placebo tablet treatment, fortnightly in,jections of normal saline, and SWD treatments applied for 20 minutes three times per week to 38 cohorts diagnosed as having unilateral or bilateral knee OA. The treatments were randomly allocated and the four efficacy measures documented were pain, tenderness, analgesic intake levels, and walking time.

The assessments were conducted im- mediately before the start of the trial and at fortnightly intervals for 26 weeks. At each assessment a patient was considered improved if two of the four efficacy para- meters showed improvement. The results showed that more knees improved after a course of SWD than after a course of placebo tablets. Although there was n o significant difference between the improvements observed after SWD and those after placebo injections, more patients showed long-term improvement after SWD than after placebo injections, even though this difference did not reach the level of significance (p < 0.1). However, the sample size was small and a power analysis was not forthcoming. Unfortunately, too, it was difficult to establish whether optimal SWD parameters were used in the trial, as the type, frequency and intensity of SWD used were not recorded. The SWD group also seemed to be more disabled than the other groups.

Another early study suggesting a positive effect of SWD applications for treating knee OA was that of Valtonen and Alaranta (1971). A strength of that study was the large sample of 160 patients, of whom 132 had radiologically verified knee OA. After being treated with a self-tuning SWD machine with a frequency of 27.33 megacycles for 15-20 minutes three times weekly for an average of 13-14 treatments with a n intensity that did not exceed a comfortable sensation of warmth, approximately one-fifth of these patients improved markedly (ie a complete or almost complete disappearance of the patient’s main symptom and the disturbances associated with it) ; three-fifths were slightly improved ( ie a clear decrease of the patient’s subjective distress) ; and only one-

-

PhysiotherapyJune 1999/vol %/no 6

310

fifth showed no beneficial treatment effects. The results were comparable to those obtained with long-wave diathermy, said to be much more cumbersome to apply. The main study limitations were the concurrent application of exercise therapy; and the lack of'a control group.

In a study by Lankhorst et a1 (1982), 24 patients with OA of one or both knees who were able to walk were randomly divided into two equally sized and comparable groups with no baseline clinical, radiological or age differences.

Patients in the first group received a combination of SWD treatments applied with condensator electrodes, 11 m wavesbat 2'7.12 MHz and 150 watts, two to three times a week for 15 minutes for six weeks, and individually tailored mobilising, strengthening and stabilising exercises, co- ordination training and functional walking training for 30 minutes immediately after diathermy during the last four weeks.

The second group was treated according to the same time schedule, but received diathermy only.

Knee extension torque, endurance and walking speed measures on level ground and stairs were made by an independent observer, under standardised conditions, one week before treatment and two weeks after cessation of treatment.

Although maximal knee extensor torque and the functional measures improved significantly after therapy for both groups, analysis of variance showed no significant differences between the time courses of the variables of either group.

While it is possible the marked functional improvements in both groups were due simply to a learning and/or a Hawthorne effect, since SWD treatment was given to both groups, it is possible the application of SWD alone contributed to these very favourable results.

Questionable Results O n e widely cited controlled trial of physiotherapy which examined the effects of several forms of electrotherapy, including SWD, for treating degenerative knee joints has been that ol' Hamilton et al (1959). In that repeated trial 26 patients were treated with o n e of four randomly assigned treatment modalities as follows: SWD for 20 minutes three times per week; infra-red radiation for 20 minutes three times per week; faradism to the knee for the same duration; and 'cold SWD' (control treatment

which appeared to apply SWD bu t the current was not switched on).

Although knee strength and walking time up and down four stairs and on level ground generally showed post-treatment improvements for the SWD applications, all treatments as well as 'cold SWD' yielded similar results. The overall improvement irrespective of' t reatment or dummy exposure may have been confounded by the fact that patients were on a basic exer- cise regimen, analgesics, and splintage throughout the study.

B a n d and Joshi (1975) compared the effectiveness of three to seven minutes of ultrasound (US) and 20 minutes contra- planar SWD for the treatment of' knee OA using a randomised mixed factor research design, That is, 60 patients aged 40 to 85 years with knee OA of one to ten years' duration were randomly allocated to one of two comparable treatment groups. After a single treatment session, analysis suggested that SWD using the contraplanar method for 20 minutes was subjectively more effective in decreasing acute OA pain than US delivered for three to seven minutes with an intensity ranging from 2-3 watts/cmL and a frequency of' 1 MHz. However, although no statistical evidence was forthcoming in that study, the converse result was said to occur when the investigators used objective outcome measures such as swelling.

A single-blinded study by Chamberlain et a1 (1982) determined whether treatment for knee OA involving SWD by inductotherm coil and exercise three times a week administered by trained personnel, plus instructions regarding addition a1 home exercises, was better than home exercise instruction alone. The results indicated that after four weeks there were significant improvements in function, maximum weight lifted and endurance (p < 0.01) irrespective of treatment group. The dropout rate was higher for the exercise group, however, and these data were not included in the analysis. The SWD group was also significantly weaker than the exercise-only group at baseline which suggests some benefit was achieved from the SWD applications. It is also possible that pain relief attributable to the SWD applications was masked by permitting the patients to use unlimited analgesic medication, o r by a suboptimal mode, frequency, durat ion, and intensity of application as these characteristics were not documented.

In a study of 180 hip or knee OA patients,

PhysiotherapyJunc IW9/vol 85/no t i

Professional articles 311

Svarcova et aZ(1988) cornpared the analgesic effects of three different types of physiotherapy: US; galvanic current; or pulsed SWD, frequency 46 M H z and iriaxirnum peak intensity 700 W. The groups were further sub-divided: with half of the patio n t s receiving an ti -inflarnm ato ry rri c dic a ti o n , and half p lac eh o tab1 e ts . Patients from all groups received ten treatments at two-day intervals for three weeks. Results on a visual analogue pain scale showed n o main effect for any treatment. However, physiotherapy and drug therapy were found significantly better than physiotherapy and placebo treatments alone. Further, it was concluded that plvanic current alone was as effective as any individual treatment when combined with d rug therapy, although the unclear doc 11 men ta ti o n m e tho d , grea te r disc as e duration of the SWD group compared to the galvanic g roup , and the higher than standard SWD frequency and two-minute treatment duration used made it difficult to justify this conclusion.

In 1991, Jan and Lai examined the effects of‘ US and SWD with and without exercise with respect to OA knees. Sixty-one women (with 94 OA knees) were allocated to one of four groups: SWD for 20 minutes in the ‘traditional’ way; ten minutes continuous US alone; SWD plus exercise and US plus a t lcast 200 straight leg raise exercises per day. Treatment was terminated when a patient’s predominant symptom was significantly relieved for one to two weeks. Following an average of 41.2 (range 24-69) treatments, a11 patients improved their fiinctioiial scores significantly, but the SWD treatment alone was not as cffective as the excrciscs and SWD or the exercises and US. Patients in thc latter groups may, however, h avc: 1 ) e e r i th os e w h o r c c e ive d bi 1 ate r a1 LJS o r SWD treatments and therefore expcrienced an additive treatment dosage and greater attention as a whole.

Negative Results I 11 a p r o s p e c ti ve rand o m i s e d c o n t r o 11 e d (rial contliicted by Clarke et aZ (1974), the investigators cornpared t h e outcome of applying ice, continuous SWD administered according to standard practice, and placebo SWD to 48 osteo-arthritic knees. All treat- ments were administered three times a week lbr three weeks and patients were re- assessed then, and after a further three months. Dependent variables assessed included pain, stiffness, knee circumference,

range of motion, walking time, radiographic grading, physician assessment and self- assessment. After three weeks, the ice group showed significant within-group improve- ments in pain and stiffness (p < 0.05). It was also shown at three months that ice, SWD and untuned SWD all had a similar effect on improving the subjects’ pain.

Comparable findings were those of’ Quirk et aZ(l985) who compared the effects on the OA knees of continuous SWD using the condenser field method and exercise, interf‘erential stimulation and exercise, and exercise alone. Patients in the first two groups were treated for 20 minutes three times per week and patients in the exercise- only group were treated twice a week for two weeks, and then once a week for two weeks. Results showed all three groups had similar decreases in pain intensity and an improved clinical conditions suggesting no benefit of either the SWD, the interferential treatments, or the extra attention given to patients in these groups. Yet, as in the study by Clarke et al (1974), history was not controlled for.

Similarly, in a more recent randomised controlled trial evaluating the efficacy of pulsed SWD for alleviating OA hip and knee pain, Klaber Moffett et a1 (1996) found no significant differences between groups receiving active, placebo and no treatment protocols after nine SWD treatment sessions of 15 minutes duration using a pulse frequency of 82 pulses/sec x 7. Some patients given placebo SWD treatments were said to have had more benefit from these than those receiving active treatment, as did pre-surgical patients. Reported outcome measures were average sensory and affective pain diary report scores, self-reported benefit scores, and a general health questionnaire, indicative of minor psy- c h i a tr i c disturb an c e s . The t reatment dosage, based o n the non-significant outcome oi’ the pain responses of a sample of 45 hip O A patients t.o pulsed SWD may not have been optimal for treating patients with knee OA, especially those requiring surgery. Further, all groups received instruction in exercise, and walking aids as required, and the placebo group may, due to their lengthier disease duration (ie 103.07 versus 62.00 months), along with those awaiting surgery, have had a higher exercise and/or medication compliance rate, or a preferential response to the advice and attention afforded by the ‘treatment’ situation.

Physiotherapy,Jirne 1999/vol 85/rio 6

312

Overview of Study Methods The 11 papers reporting on the application of SWD- for treating the symptoms of knee OA were found lacking in most cases when assessed according to criteria used for similar purposes in the recent literature. Table 3 and data in appendix 1 highlight some of these deficiencies. They include inadequate information concerning group assignment, the number of subjective outcome variables used, the number of different variables used to assess pain and function, possible bias due to the lack of adequate blinding procedures, the study of non-homogeneous groups, failure to include results of failures or ‘drop-outs’ into the analyses, use of co-interventions which were not necessarily standardised, failure to conduct power analyses to estimate the impact of small sample size and measurement variability on insignificant results, failure to report the rationale for the SWD parameters used, and failure to specify in detail the SWD parameters used. Also, although we found nine studies which stated that randomisation had been used, only Iaaber Moffett el al (1996) described their precise method of randomisation, so we cannot evaluate the randomisation quality of the remaining studies.

Due to their failure in general to use a control g roup and their insufficient reporting of the outcome parameters, we were unable to apply techniques of meta- analysis to the present data. As a result, the outcomes reported above must be seen only as suggestive, irrespective of the reported effect. The papers we studied also give no information about dose-responsiveness or mechanisms of action.

Discussion In the present review, 11 English-language articles were retrieved and analysed to answer the question: Is SWD efficacious for treating knee OA? However, given their generally low methodological quality, their incomparable co-interventions, protocols, outcome measures, and outcomes, n o definite conclusions could bc reached. Indeed, even if we overlooked some SWD trials in our literature search, the scrutinised results varied from extremely positive (eg Lankhorst P t al, 1982), to extremely negative (eg Klaber Moffett P t al, 1996). The same variability was observed with respect to the outcomes of the few animal models used to test the effects of SWD on joint inflamm- ation and motion limitation (eg Nadasdi,

1960; Vanharanta, 1982) and muscle injury (Bansal et al, 1990; Brown and Baker, 1987).

Factors that could influence the extent of the benefit of SWD treatments applied to OA knees, even if a consistent treatment mode such as continuous SWD is used, might be the size of the treatment area; degree of inflammation; treatment method; frequency, duration and intensity; unreliable outcome measures; other co-interventions such as exercise which were not controlled for; and the stage of OA studied. Neither a reasonably positive nor a negative study result could be attributed to any single study parameter.

The widespread use of subjective outcome measures may suggest, however, that some of the results we reviewed were subject to reporting, recall or rater error. As indicated in table 3, in most cases the reliability and validity of the outcome measures employed must remain in question. Table 3 also shows that neither the evaluators nor the therapists and patients were necessarily blinded, and the quality of blinding when it did take place was unclear. Along with inadequate reporting of the data and selection criteria, it was therefore impossible to pool these results and hence ei ther to justify the common usage of SWD by physiotherapists for treating knee OA, o r to refute the commonly held view that SWD is efficacious for reducing OA knee stiffness and pain. Indeed, potential clinically beneficial effects which might accrue from either thermal or non-thermal SWD applications to an OA knee might have been overlooked in the currently reported studies d u e to sub- optimal treatments, inadequate power, switch-overs, sample heterogeneity, carry- over effects, inappropriate effect measures and /o r lack of adequate follow-up pro- cedures.

Clearly, to improve practice, and to rationalise and optimise physiotherapeutic strategies for knee OA, research results based on sound methodologies, using comparable samples and protocols which provide support for postulated mechanisms, must be emphasised.

Outline of Future Efficacy Trial Design To determine more precisely whether SWD is efficacious for treating the OA knees, and under what conditions, we propose that the following experimental approach be adopted to examine this issue selectively with respect to patients with varying degrees of disease severity and chronicity using both

Physiotherapy,Jime l9W/vol 85/no ti

Prof essiona I a r t ides 313

pulsed arid continuous SWD without any extraneous interventions.

IjeJign: The design advocated is double- an d / o r p 1 ace b o- controlled, with an adequately blinded and valid randomisation procedure.

t r ip 1 e- b 1 in d e d an d

Sample: The prospective study sample should be homogeneous with respect to disease duralion, activity and stage, past treatments, age, and knee involvement, arid constituted by persons conforming to the American R h c 11 matol ogy Asso cia ti on cli ni cal an d radiographic criteria for a single diagnosis 01’ knee OA (Altman et al, 1986) with pain of at least one year’s duration.

Size: The sample should be large enough to account for potential drop-outs and loss to follow-up, stratification of patients according to sex, age arid disease severity, and normal day-today variations in disease presentation, as these impact o n measurement error magnitudes.

Main exclusion criteria: Evidence of current or recent participation in SWD or other forms of physiotherapy, recent knee jo in t in.jections o r surgery, and patients not stabilised for at least three months on nnchanged medication.

15 . . f f i ca cy criteria: O u tc o m e s i n c 1 u d i n g measnres of daily activities and work capacity; stair climbing and walking ability; weighlbearing pain; knee strength, stability, joint stiffness, range of‘ motion, and swelling; kntx: joint synovial fluid keratin sulphate composition, and f’ernorotibial bone corn- position must be assessed with reliable and validated indices or instruments.

E:valutntor~s: Assessors should be well-trained, consistent and blinded to the study hypo- theses and group assignment.

lhv-ation: Treatments should be applied three times per week Ibr a maximum of six weeks or on the basis of‘ a pilot study of a Gniilar sample.

l hwgp: Treatment dosages should be those suggested by previous research, for example 11 ni waves at 27.12 MH7 and 150 watts for 15 minutes with condcnsater electrodes (Lankhorst rt al, 1982); or indicated by pilot work on a similar sample; or specified by the dcvice manufacturer.

StatiJtical procedures: Appropriate analyses, inclnding power analyses where results are non-conclusive, should be carried out by blinded analysts using a sufficiently discriminating significance level. Group comparability, prior to and during the experimental schedule, should also be validated and calculations should be made according to the ‘intention-to-treat’ principle. That is ‘drop-outs’ and ‘losses’ are assumed to represent treatment failures.

Measurementx Pretreatment measurements should be implemented at least twice over a one-week period and comparable mid-point and post-experiment follow-up measures should be implemented by the same blinded evaluators at bi-monthly intervals for at least one year.

firording: Finally, a complete description of the apparatus and placebo-apparatus, mode of delivery, dosage parameters, randomi- sation method, specification of the joint selected for symptom evaluation in bilateral cases, and other methodological variables used by any future investigators of this question must be clearly documented to permit study replication and facilitate interpretation of findings.

lur ther studies: If shown t.o be efficacious, dose-effects, the effects of SWD on the pathology of knee OA and the numerous explanations suggested as to why SWD is effective in treating knee OA should be investigated, preferably through collabor- ation with researchers in biomechanical, radiological and biochemical laboratory settings.

Summary and Conclusions Although strong theoretical arguments can be made for the potential benefits of SWD on the underlying pathological processes found in OA (see Andrew and Bassett, 1993; Harris, 1963; Weinberger, 1988), the prevailing clinical studies concerning the application of SWD for treating painful knee OA are essentially non-conclusive, given their poor overall methodological quality as highlighted in table 3. In particular, several had inadequately \ized Famples and no non-treatment control group, and used c om p e tin g i n t e rve n ti o n 9. Fur t h e r, few documented their p re c i 5 e therapeutic paradigms, and the reliability and validity of their treatments and outcome measures remain in question. Table 4 also highlights

Physiotherapy,lurle lSOS/vol 85/no 6

314

their lack of stringent inclusion criteria, u n k n o w ri treat m e n t par am e t e rs , and possible failure to exclude patients with acute ixiflarnmatory OA from some studies. Consequently, we conclude additional study is essential to determine whether SWI) is indeed efficacious for knee OA in either its continuous or pulsed mode or both, and if so , for which osteo-arthritic symptoms Favourable treatment effects are likely to occur. Additionally, if a specific form of SWD is found beneficial for treating knee OA patients, the mechanism (s) underlying this should be explored with a view to exploiting appropriate cell-specific field intensities. Also, if short-term treatment efficacy can be demoristrated, long-term follow-up stud SWD for treatment of knee OA woii indicated.

Kising health care costs as our population ages and the substantial contribution to this cost due to the growing number of individuals with disabling knee OA, plus the need to satisfji clients' treatment object- ives, have placed a strong dcniand on physiotherapy providers to ensure optimal treatments for their patients. In this regard, i t is the present authors ' view that to increase productivity and con t air1 disease costs, while alleviating the symptoms of knee Oh, the value of' using SWD as cit.hcr a primary or an adjunctive physiotherapeulic measure ii)r the treatment of knee OR, if any, should not he overlooked due to the lack of' rigorously controlled trials with large homogenous samples arid clinically reliable and valid endpoints.

Shortwave diathermy, unlike several other elcctrical modalities, requires very little t1ic:rapist time beyond its initial set-up. Evidence from several basic scicncc exper- iments also indicates that improving our ability t o apply selectively modes of SWL) and parameters of' application would have a beneficial elftct on the ccllirlar pa OA knee joints including thcir muscle pathology (eg I3;rrisal el al, 1990). lridcetl, the very positive hidings of a post- (reatment. gain in knee ext,ensor torque and

walking ability as described by Lankhorst et al (1982) in a small group of patients who received n o t reatment o the r than continuous SWD warrant further study to rule out placebo attention, learning or other effects.

O n the o the r hand , although Klaber Moffett et a1 (1996) have shown that pulsed SWD may have no more than a placebo ef'fect on OA knee pain when applied for three weeks, positive effects with respect to treatments of longer duration and different dosages cannot be ruled out.

Further, while not necessarily considered as a form of SWD, very positive benefits might yet be attained for these patients by stepwise application of extremely low li-equency PEMF as indicated by Trock el nl (1993, 1994) in two recent multi-centre double blind trials and by Nadasdi (1960).

In view of this possibility, plus the urgent need fbr physiotherapists to validate their practice and provide optimal treatments to their clients, we would like to stress our belief in the importance of future efforts to establish unequivocally the true efficacy of continuous and pulsed electromagnetic field treatment applications used by physiothera- pists for treating knee OA. In particular, i n ad di t i o r i to do 11 b 1 e b 1 in d , rand o rn is e d cont.rolled trials o f large homogencous samples using reliable arid validaled indices, we would encourage preliminary ex p c ri me n tal s tu di e s to tie te rni i ne pos- sible positive and negative dose-effect relationships and optimal set t ings I'or SWD machint: parameters. Sex arid age effects that could influence dose-resporis- iveriess to SWD of patients with knee OA also need investigation.

Implications for Practice (kneralisable conclusions which lead to effective control by physiotherapists of OA symptoms will not only help to reduce the personal disability associated with this disease, but should help t o control health care costs arid could ensure our future role in the market place.

Prof essiona I a r t i cles 315

Andrew, C and Bassett, L (1993). ‘Beneficial effects of electromagnetic fields’, Journal of Cellular Biochemistry, 51, 410-416.

Bansal, P S, Sobti, V K and Roy, K S (1990). ‘Histomorphochemical effects of shortwave diathermy on healing of experimental injury in dogs’, Indian Journal of Experimental Biology, 28,766-770.

Bansil, C K and Joshi, J B (1975). ‘Effectiveness of shortwave diathermy and ultrasound in the treatment of osteo-arthritis of the knee joint’, Medical Journal of 7 ~ m b i a , 9, 138-139.

Beckerman, H, Bie, R A de, Bouter, L M, De Cuyper, H J and Oostendorp, R A B (1992). ‘The efficacy of laser therapy for musculoskeletal and skin disorders: A criteria-based meta-analysis of randomised clinical trials’, Physical Therapy,

Binderman, I, Somjen, D, Shimshoni, Z, Levy, J, Fischler, H and Korenstein, R ( 1985). ‘Stimulation of skeletalderived cell cultures by different electric field intensities is cell-specific’, Biochemica et Biophysics Acta, 844, 273-279.

Brown, M and Baker, R D (1987). ‘Effects of pulsed short wave diathermy on skeletal muscle injury in rabbits’, Physical Therapy, 67, 208-214.

Chamberlain, M A, Care, G and Harfield, B (1982). ‘Physiotherapy in osteo-arthrosis of the knees: A controlled trial of hospital versus home exercises’, International Journal of Rehabilitation Medicine, 4, 101-106.

Chapman, C E (1991). ‘Can the use of physical modalities for pain control be rationalised by the research evidence?’ Canadian Journal of Physiological Pharmacology, 69, 70471 2.

Clarke, G R, Willis, L A, Stenner, L and Nichols, P J R (1974). ‘Evaluation of physiotherapy in the treatment of osteo-arthrosis of the knee’, Rheumatology and Rehabilitation, 13, 190-197.

Davis, M A (1988). ‘Epidemiology of osteo- arthritis’, Clinics in Geriatric Medicine, 4. 241-255.

Dekker, J, Boot, B, Woude Luc, H V and Bijlsma, J W J (1992). ‘Pain and disability in osteo-arthritis: A review of biobehavioral mechanisms’, Journal of Behavioral Medicine, 15, 189-214.

Doherty, M and Jones, A C (1994). ‘Osteo-arthritis’, Medicine International,

72, 483-491.

52, 129-135.

Fabbri, F and Lucchese, V (1980). “on-thermal effects of pulsed magnetic fields: An experimental study of 22 laboratory parameters’, ~blinn-ua Ortopedica, 31, 251-256.

Gam, A N and Johannsen, F (1995). ‘Ultrasound therapy in musculosLc.leta1 disorders: A meta- analysis’, Pain, 63, 85-91. Ghassemi, M and Marks, R (1995). ‘The efficacy of electrotherapy for osteo-arthritis of the knee’, Proceedings of the 12th. International Congtess of the I\‘m-ld Confehation for Physical Therapy,

Washington DC, American Physical Therapy Association, Alexandria, page 85.

Goats, G C (1989a). ‘Continuous short-wave (radio-frequency) diathermy’, British Journal of Sports Medicine, 23, 213-216.

Goats, G C (198913). ‘Pulsed electromagnetic (short-wave) energy therapy’, British Journal of Sports Medicine, 23, 123-127.

Grigorieva, V D, Tsarfis, P G, Gerasimenko, A A, Fivieskaya, A V, Shubina, G J, Proskurova, G J et aZ(1980). ‘Use of a low frequency permanent magnetic field in patients with deforming osteo- arthrosis and rheumatic arthritis’, Voposy Kurortologii Fizioterapi, 4, 29-35.

Hamilton, D E, Bywaters, E G L and Please, N W (1959). ‘A controlled trial of various forms of physiotherapy in arthritis’, British Medical Journal, 2, 54‘2-545.

Harris, R (1963). ‘The effect of various forms of physiotherapy on radiosodium clearance from the normal and arthritic knee-joint’, Annals of Physical Medicine, 7, 1-10.

Hochberg, C (1984). ‘Osteo-arthritis: Pathophysiology, clinical features, management’, Hospital Practice, 19, 41-53.

Jan, M and Lai, J (1991). ‘The effects of physiotherapy on osteo-arthritic knees of females’, Journal of Formosan Medical Association, 90, 1008-13.

Kitchen, S and Partridge, C (1992). ‘Review of shortwave diathermy continuous and pulsed patterns’, Physiotherapy, 78, 243-252.

Klaber Moffett, J A, Richardson P H, Frost, H and Osborn, A (1996). ‘A placebo-controlled double-blind trial to evaluate the effectiveness of pulsed short wave therapy for osteo-arthritic hip and knee pain’, Pain, 67, 121-127.

Lankhorst, G J, van de Stadt, R J, van der Korst, J K, Hinlopen-Bonrath, E, Griffioen, F M M and de Boer, W (1982). ‘Relationship of isometric knee extension torque and functional variables in osteo-arthrosis of the knee’, Scandinavian~~ournal of Rehabilitation Medicine, 14, 7-10.

Lawrence, J S, Bremner, J M and Bier, F (1986). ‘Osteo-arthritis - Prevalence in the population and relationship between symptoms and X-ray changes’, Annals of The Rheumatic Diseases, 25, 1-24.

Lehmann, J F, Masock, A J, Warren, C G and Koblanski, J N (1970). ‘Effects of therapeutic temperatures on tendon extensibility’, Archives of Physical Medicine and Rehabititation, 51, 481-487.

Liu, H, Abbott, J and Bee, J A (1996). ‘Pulsed electromagnetic fields influence hyaline articular cartilage extracellular matrix composition without affecting molecular structure’, Osteo- arthritis and Cartilage, 4, 63-76.

Marks, R (1993). ‘Muscles as a pathogenic factor in osteo-arthritis’, Physiotherapy Canada, 45, 251-260.

Physiotherapy June 1999/vol85/no 6

Marks, Rand Cantin, D (1997). ‘Symptomatic osteo-arthritis of the knee: The efficacy of physiotherapy’, Physiotherapy, 83, 306-312.

Munice, H I (1986). ‘Summary: Applications of the multidisciplinary management of osteo-arthritis’, Clinical Therapeuticc, 9, 67-68.

Nadasdi, M (1960). ‘Inhibition of experimental arthritis by athermic pulsating shortwaves in rats’, Orthopedics, 2, 105-107.

Puett, D W and Griffen, M R (1994). ‘Published trials of non-medical and non-invasive therapies for hip and knee osteo-arthritis’, Annals of Internal Medicine, 121, 133-140.

Quirk, A S, Newman, R J and Newman, K J (1985). ‘An evaluation of interferential therapy, shortwave diathermy and exercise in the treatment of osteo-arthrosis of the knee’, Physiotherafl, 71, 55-57.

Souhami, R L and Moxham, J (1990). Textbook of Medicine, Churchill Livingstone, Edinburgh.

Sanseverino, E R (1980, November). ‘Membrane phenomena and cellular processes under the action of pulsating magnetic fields’, Lecture, 2nd International Congress for Magneto-Medicine, Rome, Italy.

Svarcova, J, Trnavsky, K and Zvarova, J (1988). ‘The influence of ultrasound, galvanic currents, and shortwave diathermy on pain intensity in patients with osteo-arthritis’, Scandinavian Journal of Rheumatology Supplement, 67, 83-85.

Thom, H (1966). Introduction to Shortwave and Microwave Therapy, Thomas, Springfield, Illinois.

Threlkeld, A J (1984). The effects of altered weight bearing and pulsed ekctromapeticjelds on the articular cartilage of the rat femoral head, PhD thesis, University of Kentucky.

Threlkeld, A J and Currier, D P (1988). ‘Osteo-arthritis: Effect on synovial joint tissues’, Physical Therupy, 68, 364370.

Trock, D H, Boktt, A J, Dyer, R H J, Fielding, L P, Miner, W K and Markoll, R (1993). ‘A double- blind ti-ial of the clinical effects of pulsed electromagnetic fields in osteo-arthritis’,,/ournal of Rheumatology, 20, 456-460.

Trock, D H, Bollett, A J and Markoll, R (1994). ‘The effect of pulsed electromagnetic fields in the treatment of osteo-arthritis of the knee and cervical spine: Report of randomised, double- blind placebo-controlled trials’, Journal of Rheumatology, 21, 1903-11.

Valtonen, E J and Alaranta, H (1971). ‘Comparative clinical study of the effect of shortwave and longwave diathermy on osteo-arthritis of the knee and hip’, Scandinavian Journal of Rehabilitation Medicine, 3, 109-1 12.

Vanharanta, H (1982). ‘Effect of shortwave diathermy on mobility and radiological stage of the knee in the development of experimental osteo-arthritis’, American,/ournal of Physical

Vanharanta, H, Eronen, I and Videman, T (1982). ‘Shortwave diatherniy effects 011 “S-sulphate uptake and glycosaminoglyc-an concentration in rabbit knee tissue’, Archivrs of Physical Mrdicinr and Rehabilitation, 63, 25-28.

Warnke, U (1983). ‘The possible role of pulsating magnetic fields in the reduction of pain’ in: Rizzi, R and Voseentin, M (eds) Pain 7’herapy, Elsevier Biomedical Press, New York.

Weinberger, A, Faddah, R, Lev, A, Levi, A and Pinkhas, J (1988). ‘Deep heat in the treatment of inflammatory joint disease’, Medical HyjIothesPs, 25,

Wright, V (1964). ‘Treatment of osteo-arthritis of the knees’, Annals of the Rheumatic Diseases, 23, 389-391.

Wright, V (1973). ‘Stiffness: A review of its measureinent and physiological importance’, Physi5thera~, 59, 107-1 1 1.

T~PRz~IJ , 61, 59-65.

231-233.

wiotherapy June 1999/vol85/no 6