Physiotherapy for Parkinson's disease a comparison of techniques

Physiotherapy for Parkinson’s disease: a comparison of

techniques (Review)

Deane K, Jones DE, Ellis-Hill C, Clarke CE, Playford ED, Ben-Shlomo Y

This is a reprint of a Cochrane review, prepared and maintained by The Cochrane Collaboration and published in The Cochrane Library

2009, Issue 1

http://www.thecochranelibrary.com

Physiotherapy for Parkinson’s disease: a comparison of techniques (Review)

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.

http://www.thecochranelibrary.com

T A B L E O F C O N T E N T S

1HEADER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2PLAIN LANGUAGE SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3OBJECTIVES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3METHODS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

4RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

10DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

12AUTHORS’ CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

13ACKNOWLEDGEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

13REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

15CHARACTERISTICS OF STUDIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

23DATA AND ANALYSES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

23WHAT’S NEW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

23HISTORY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

23CONTRIBUTIONS OF AUTHORS . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

23DECLARATIONS OF INTEREST . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

23SOURCES OF SUPPORT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

24INDEX TERMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

iPhysiotherapy for Parkinson’s disease: a comparison of techniques (Review)


[Intervention Review]

Physiotherapy for Parkinson’s disease: a comparison oftechniques

Katherine Deane2, Diana E Jones3, Caroline Ellis-Hill4, Carl E Clarke1, E Diane Playford5, Yoav Ben-Shlomo6

1Department of Neurology, City Hospital, Sandwell and West Birmingham Hospitals NHS Trust, Birmingham, UK. 2Institute of Health

& Society, Newcastle University, Newcastle-upon-Tyne, UK. 3School of Health, Community and Education Studies, Northumbria

University, Newcastle upon Tyne, UK. 4Health and Rehabilitation Research Unit, University of Southampton, Southampton, UK.5Directorate of Musculoskeletal and Rehabilitational Services, The National Hospital for Neurology and Neurosurgery, London, UK.6Dept of Social Medicine, Canynge Hall, Bristol, UK

Contact address: Carl E Clarke, Department of Neurology, City Hospital, Sandwell and West Birmingham Hospitals NHS Trust,

Dudley Road, Birmingham, West Midlands, B18 7QH, UK. [email protected].

Editorial group: Cochrane Movement Disorders Group.

Publication status and date: Edited (no change to conclusions), published in Issue 1, 2009.

Review content assessed as up-to-date: 28 November 2000.

Citation: Deane K, Jones DE, Ellis-Hill C, Clarke CE, Playford ED, Ben-Shlomo Y. Physiotherapy for Parkinson’s disease: a comparison

of techniques. Cochrane Database of Systematic Reviews 2001, Issue 1. Art. No.: CD002815. DOI: 10.1002/14651858.CD002815.


A B S T R A C T

Background

Despite optimal medical and surgical therapies for Parkinson’s disease, patients develop progressive disability. The role of the physio-

therapist is to maximise functional ability and minimise secondary complications through movement rehabilitation within a context

of education and support for the whole person. What form of physiotherapy is most effective in the treatment of Parkinson’s disease

remains unclear.

Objectives

1. To compare the efficacy and effectiveness of novel physiotherapy techniques versus ’standard’ physiotherapy in patients with Parkinson’s

disease. Standard physiotherapy is defined as the type of therapy that the physiotherapist would usually use to treat Parkinson’s disease.

2. To compare the efficacy and effectiveness of one physiotherapy technique versus a second form of physiotherapy.

Search strategy

Relevant trials were identified by electronic searches of MEDLINE, EMBASE, CINAHL, ISI-SCI, AMED, MANTIS, REHABDATA,

REHADAT, GEROLIT, Pascal, LILACS, MedCarib, JICST-EPlus, AIM, IMEMR, SIGLE, ISI-ISTP, DISSABS, Conference Papers

Index, Aslib Index to Theses, the Cochrane Controlled Trials Register, the CentreWatch Clinical Trials listing service, the metaRegister

of Controlled Trials, ClinicalTrials.gov, CRISP, PEDro, NIDRR and NRR; and examination of the reference lists of identified studies

and other reviews.

Selection criteria

Only randomised controlled trials (RCT) were included.

Data collection and analysis

Data was abstracted independently by KD and CEH and differences settled by discussion.

1Physiotherapy for Parkinson’s disease: a comparison of techniques (Review)


mailto:[email protected]

Main results

Seven trials were identified with 142 patients. All used small numbers of patients and the method of randomisation and concealment

of allocation was poor or not stated in all of the trials. These methodological problems could potentially lead to bias from a number of

sources. The methods of physiotherapy varied so widely that the data could not be combined.

Authors’ conclusions

Considering the small number of patients examined, the methodological flaws in many of the studies and the possibility of publication

bias, there is insufficient evidence to support or refute the efficacy of any given form of physiotherapy over another in Parkinson’s

disease. Another Cochrane review, Physiotherapy for patients with Parkinson’s Disease, found that there was insufficient evidence to

support or refute the efficacy of physiotherapy compared to no physiotherapy in Parkinson’s disease.

A wide range of physiotherapy approaches were used in these studies and a survey of UK physiotherapists confirmed that they also use

an eclectic combination of techniques in the treatment of Parkinson’s disease (Plant 1999). Therefore a consensus must be found as to

’best practice’ physiotherapy for Parkinson’s disease.

The efficacy of ’standard’ physiotherapy should be proved first before examining variations in physiotherapy methods. Therefore large

well designed randomised controlled trials are needed to judge the effect of physiotherapy in Parkinson’s disease. After this large RCTs

are needed to demonstrate the most effective form of physiotherapy in Parkinson’s disease. Outcome measures with particular relevance

to patients, carers, physiotherapists and physicians should be chosen and the patients monitored for at least 6 months to determine the

duration of any effect. The trials should be reported according to CONSORT guidelines (CONSORT 1996).

P L A I N L A N G U A G E S U M M A R Y

In spite of the best medical and surgical treatments for Parkinson’s disease, patients develop significant physical problems.

Physiotherapists aim to enable people with Parkinson’s disease to maintain their maximum level of mobility, activity and

independence through the monitoring of their condition and the targeting of the appropriate physical treatment. A range

of approaches to movement rehabilitation, which with education and support are employed to maximise functional ability,

minimise secondary complications and enhance quality of life over the whole course of the disease.

This review will compare the benefits of one form of physiotherapy versus another for people with Parkinson’s disease. Relevant trials

were identified by electronic searches of 21 medical literature databases, various registers of clinical trials and an examination of the

reference lists of the identified studies and other reviews.

Only randomised controlled trials were included in this review. These were studies where two groups of patients were compared, each

group of patients receiving a different form of physiotherapy. The patients were assigned to each of the two groups in a random fashion

to reduce the potential for bias. Data from the selected trials were extracted independently by two reviewers and differences settled by

discussion.

Seven trials were found comparing two forms of physiotherapy in a total of 142 patients. The quality of the trials’ methods was variable

with all the studies failing in at least one critical area. The methods and outcome measures varied so much that the results of the

individual trials could not be combined.

Considering the small number of patients and the methodological flaws in many of the studies, there is insufficient evidence to support

the use of one form of physiotherapy over another for the treatment of Parkinson’s disease.

Another Cochrane review that examined the efficacy of physiotherapy versus placebo (sham) therapy (Physiotherapy for patients with

Parkinson’s Disease) concluded that there was insufficient evidence to support or refute the efficacy of physiotherapy in Parkinson’s

disease.

The benefits of ’standard’ physiotherapy should be proved first before examining variations in physiotherapy methods. Therefore large

well designed randomised controlled trials (RCTs) are needed to judge the effect of physiotherapy in Parkinson’s disease. After this,

large RCTs are needed to demonstrate the most effective form of physiotherapy in Parkinson’s disease. The design of the trials should

minimise bias and be reported fully using CONSORT guidelines. Outcome measures with particular relevance to patients, their carers,

physiotherapists and physicians should be chosen and the patients followed for at least 6 months to determine the duration of any

improvement.



B A C K G R O U N D

The modern management of Parkinson’s disease centres on phar-

macological therapy, principally with levodopa combined with a

dopa decarboxylase inhibitor. However it is now recognised that in

long-term usage, levodopa precipitates motor complications such

as involuntary movements and fluctuations in response. In spite of

optimal pharmacological therapy, the underlying disease contin-

ues to progress. All existing treatments provide only symptomatic

benefit with no conclusive evidence that they are neuroprotective.

The disabilities due to Parkinson’s disease occur at all stages of the

disease and impact upon the patient’s quality of life. Significant

disability can occur in the early stages of the disease with patients

becoming dependent for washing, dressing, eating and other ac-

tivities of daily living. The severity of these disabilities usually in-

creases with disease duration.

It has been suggested that physiotherapy can improve the abilities

of patients with Parkinson’s disease as an adjunct to drug therapy

(Hildick-Smith 1987). The purpose of physiotherapy in Parkin-

son’s disease is to maximise functional ability and minimise sec-

ondary complications through movement rehabilitation within a

context of education and support for the whole person. This work-

ing definition of physiotherapy for Parkinson’s disease has been

generated from the results of an evaluation of best practice phys-

iotherapy in the UK (Plant 1999). Physiotherapy for Parkinson’s

disease covers a number of different treatment techniques, largely

centred on active exercises and re-education of mobility.

A postal questionnaire of 261 Parkinson’s patients in touch with

the Parkinson’s Disease Society in 1982 found that 29% had seen

a physiotherapist (Oxtoby 1982). In Mutch et al’s 1986 commu-

nity-based study of 267 patients, only 7% had seen a physiother-

apist (Mutch 1986). A survey of 72 Parkinson’s patients attending

a movement disorder clinic in 1995 found that 29% had seen a

physiotherapist (Clarke 1995). Finally, a members survey of the

Parkinson’s Disease Society of the United Kingdom with a total

of 1,693 respondents found that 27% of respondents had been

assessed or treated by a physiotherapist (Yarrow 1999). These low

referral rates do not accord with most published guidelines for the

management of Parkinson’s disease. Possible explanations for this

include limited access due to high general demands on physio-

therapy services, perception of lack of benefit and difficulties with

interpretation of trials due to flawed methodology.

This review of randomised clinical trials aims to compare the ef-

ficacy and effectiveness of novel physiotherapy techniques versus

standard physiotherapy in patients with Parkinson’s disease. It will

also compare the efficacy and effectiveness of one novel physio-

therapy technique versus a second novel form of physiotherapy.

Another Cochrane review assesses the effect of active physiother-

apy compared with placebo (Deane).

O B J E C T I V E S

• To compare the efficacy and effectiveness of novel

physiotherapy techniques versus standard physiotherapy in

patients with Parkinson’s disease.

• To compare the efficacy and effectiveness of one

physiotherapy technique versus a second form of physiotherapy.

M E T H O D S

Criteria for considering studies for this review

Types of studies

All randomised controlled trials comparing two types of physio-

therapy were considered for inclusion in the study. Both random

and quasi-random methods of allocation were allowed.

Types of participants

• Patients with a diagnosis of Parkinson’s disease (as defined

by the study authors).

• Any duration of Parkinson’s disease.

• All ages.

• Any drug therapy.

• Any duration of treatment.

Types of interventions

One physiotherapy technique versus a second.

Types of outcome measures

1. Motor impairment, (a) Totaled (e.g. United Parkinson’s Disease

Rating Scale

(UPDRS) motor score, part III).

(b) Individual measures of tremor, rigidity etc.

(c) Timed tests of activities (e.g. walking velocity).

2. Activities of daily living (e.g. UPDRS Activities of Daily Living

(ADL) score, part II).

3. Handicap and quality of life measures, both disease specific, (e.g.

Parkinson’s Disease Questionnaire -39, PDQ-39), and generic,

(e.g. Short Form - 36, SF-36).

4. Depression rating scales (e.g. Hospital Anxiety and Depression

Scale, HADS).

5. Adverse effects.

6. Carer outcomes (e.g. Carer strain index).

7. Economic analysis.

We will examine both short term and long term (e.g. 6-12 months)

effects of the intervention.



Search methods for identification of studies

1. The review is based on the search strategy of the Movement

Disorders Group and also the following more general search strat-

egy:

a. Physiotherapy OR physical therapy OR exercise OR rehabilita-

tion

b. Parkinson OR Parkinson’s disease OR Parkinsonism

c. #a AND #b

Relevant trials were identified by electronic searches of general

biomedical and science databases: MEDLINE (1966-2000), EM-

BASE (1974-2000), CINAHL (1982-2000), ISI-SCI ((1981-

2000); rehabilitation databases: AMED (1985-2000), MAN-

TIS (1880-2000), REHABDATA (1956-2000), REHADAT,

GEROLIT (1979-2000); English language databases of foreign

language research and third world publications: Pascal (1984-

2000), LILACS (1982-2000), MedCarib (17th Century-2000),

JICST-EPlus (1985-2000), AIM (1993-2000), IMEMR (1984-

2000) and hand searching of appropriate journals. Relevant trials

were included on the Group’s specialised register of randomised

controlled trials.

2. The Cochrane Controlled Trials Register, the CentreWatch

Clinical Trials listing service, the metaRegister of Controlled Tri-

als, ClinicalTrials.gov, CRISP, PEDro, NIDRR and NRR, were

also searched for relevant trials.

3. The reference lists of located trials and review articles were

searched.

4. Grey literature (e.g. conference abstracts, theses and inter-

nal reports) were searched. This included The XIII Interna-

tional Congress on Parkinson’s disease (1999), The International

Congress of Parkinson’s Disease and Movement Disorders (1990,

92, 94, 96, 97, 98), The American Academy of Neurology 51st an-

nual meeting (1999). The following grey literature databases were

searched: SIGLE (1980-2000), ISI-ISTP (1982-2000), DISSABS

(1999-2000), Conference Papers Index (1982-2000) and Aslib In-

dex to Theses (1970-2000).

5. National and regional professional associations were asked to

search for relevant trials. Requests for help were placed on bulletin

boards on their web pages.

6. Universities and colleges that carry out degree courses in physio-

therapy were asked to search for any relevant unpublished projects.

7. Patient support groups (the Parkinson’s Disease Society etc.)

were asked if they had funded any relevant trials. Requests for help

were placed on bulletin boards on their web pages.

Further details on this search strategy are available in the Group’s

module within the Cochrane library (www.cochrane.org). This

includes explanations of the acronyms, sources and web sites.

Data collection and analysis

The authors independently assessed the studies identified by the

search strategy. Disagreements about inclusions were resolved by

discussion. The full papers were assessed for methodological qual-

ity by recording the method of randomisation and blinding,

whether an intention-to-treat analysis was used and the number

of patients lost to follow up.

Eligible data was abstracted by two authors (KD and CEH) onto

standardised forms independently, checked for accuracy and amal-

gamated. Disagreements about inclusions were resolved by discus-

sion.

Ordinal data such as UPDRS motor subsection scores were treated

as if they were interval data (i.e. continuous), where we could make

an assumption of equality of intervals e.g. UPDRS part II ADL

and part III motor. Although we recognise that this is controversial,

Popham 1973 reported that ’when parametric procedures have

been employed with ordinal data, they rarely distort a relationship

between variables which may be present in the data’. Other ordinal

data such as UPDRS complications of therapy subsection (and

thus total UPDRS) is based on summation of the scores from

a series of dichotomous questions (which are weighted equally),

there is no ’equality of interval’ and so this data was analysed in a

nonparametric fashion.

R E S U L T S

Description of studies

See: Characteristics of included studies; Characteristics of excluded

studies; Characteristics of ongoing studies.

See Table: Characteristics of Included Studies and Table 1: Key

Characteristics of Included Studies.

Table 1. Key Characteristics of Included Studies

Study Number of

Patients

Mean Age Mean

Hoehn &

Yahr Score

Duration Location Type of

Physio.

Additional

therapy?

Individual or

Group?



Table 1. Key Characteristics of Included Studies (Continued)

Hirsch 96 17 74 2.0 Unspec-

ified time/10

weeks

Outpatients Strength &

balance Vs

balance

None Individual

Homann 98 16 n/a n/a 14 units / 5

weeks

Outpatients Bobath Vs

PNF

None Individual

Marchese

2000

20 66 2.4 urs/6 weeks Outpatients Cued Vs

Standard

None Individual

Mohr 96 41 62 2.1 30 hours/10

weeks

Outpatients Behavioural

(inc. cues)

Vs standard

None Group

Palmer 86 14 65 2.4 36 hours/12

weeks

Outpatients Karate Vs

standard

None Group

Shiba 99 8 65 n/a Unspecified Outpatients Visual cues

Vs auditory

cues

None Individual

Thaut 96 26 72 2.5 10.5 hours/3

weeks

Home Auditory

cues Vs stan-

dard

None Individual

TOTAL 142 6

Outpatients;

1 Home

5 Cued trials 5 Individual;

2 group.

Seven trials were found that compared one form of physiotherapy

with another in a total of 142 Parkinsonian patients.

TRIAL DESIGN

All of the trials were single centre studies. Six of the trials were par-

allel group design, whilst Shiba 99 was a cross-over trial. Physio-

therapy was carried out for 10.5-36 hours over 3-12 weeks. How-

ever it should be noted that three of the trials did not specify the

number of hours of treatment given. One trial (Shiba 99) has re-

cently been published in abstract form. Data from the trial was

therefore limited and the number of weeks over which the phys-

iotherapy was administered was not specified. None of the trials

specified the intensity of the physiotherapy provided.

PARTICIPANTS

The patients were an average of 67 years old. In five out of the seven

trials the treatment groups in each individual trial were internally

comparable. In Shiba 99 the data was unavailable, in Homann

98 only data for the whole patient population of this three arm

trial were available. Only two trials specified individual physical

impairments as part of their inclusion criteria. Hirsch 96 required

patients to have a history of falls or unsteadiness, or to have weak

muscles in the legs. Thaut 96 required their patients to have signif-

icant gait deficits but be able to walk without physical assistance.

INTERVENTIONS

Six out of seven trials were carried out in outpatient settings (this

included research clinics). Thaut 96’s therapy was carried out at

home with the assessments being carried out in a laboratory set-

ting. Four of the trials (Marchese 2000, Homann 98, Shiba 99 &

Thaut 96) carried out physiotherapy interventions with individ-

ual patients. The remaining three interventions (Hirsch 96, Mohr

96 & Palmer 86) involved group work, although Mohr 96 used



unequally size groups (3-4 patients in novel groups versus 5-7 pa-

tients in the standard group).

The therapy was conducted by a physiotherapist in only one trial

(Thaut 96). Mohr 96 used clinical psychologists. Palmer 86 used

a corrective therapist and a rehabilitation nursing student with a

black belt in Karate. Hirsch 96 used volunteers from the Physical

Education and Human Movement Sciences department at Florida

State University and the carers of the patients all of whom were

trained by the principle investigator, but it was unclear what his

qualifications were. It was unclear in the remaining three trials (

Marchese 2000, Homann 98, Shiba 99 ) what the qualifications

were of those conducting the therapy.

Five of the trials used visual or auditory cues to promote move-

ment in the patients, however the amount of emphasis placed on

the cueing and the method of physiotherapy against which it was

compared varied significantly. Overall the methods of physiother-

apy varied so significantly that the studies cannot be combined

and so will be examined individually.

Hirsch 96

Hirsch 96 trained their patients for leg strength and balance and

compared these patients to a group that received balance training

only.

Homann 98

Homann 98 used proprioceptive neuromuscular facilitation

(PNF) which uses cues as part of the technique and directed the

therapy at spinal mobility. This was compared to a Bobath phys-

iotherapy program focusing on proprioceptive skills to improve

posture and gait. The description of the methods used was limited

as this study has only been published in abstract form.

Marchese 2000

Marchese 2000 carried out an individual physical rehabilitation

program including visual and auditory cues and compared these

patients to another group that underwent a similar movement

rehabilitation program but without the cues.

Mohr 96

Mohr 96’s novel group underwent behavioural therapy that in-

cluded relaxation techniques, role playing and motor training us-

ing external cues (visual and auditory). These patients were com-

pared with a ’standard’ group that were given breathing and physi-

cal exercises (without cues), short relaxation and discussion of dis-

ease-related problems. The two methods were described as being

distinct from one another.

Palmer 86

Palmer 86 trained their patients in upper body Karate exercises

and compared these patients to a group that performed stretch

exercises designed for Parkinson’s patients.

Shiba 99

Shiba 99’s study had a cross-over design. Group I had visually stim-

ulated gait training followed by auditory stimulated gait training.

Group II had auditory stimulated gait training first. There was

one week between each training regime.

Thaut 96

Thaut 96 gave auditory stimulated gait training using three differ-

ent tempos of music that increased in pace as the study progressed.

The ’standard’ group were instructed to walk for 30 min/day at

three different paces (normal, quick, fast).

Risk of bias in included studies

See Table 2: Methodological Quality of Included Trials and Table

1: Key Characteristics of Included Trials.

Table 2. Methodological Quality of Included Studies

Study Specified El-

igibility Cri-

teria

Randomisa-

tion Method

Conceal-

ment of Al-

location

Similarity at

Baseline

With-

drawals De-

scribed

Missing Val-

ues

Cointerven-

tions Con-

stant (e.g.

drugs)

Blinded as-

sessors

Hirsch 96 A A B A C C B C

Homann 98 A A A B A A A B

Marchese

2000

A A B A A A A A

Mohr 96 A B B A A A A A



Table 2. Methodological Quality of Included Studies (Continued)

Palmer 86 A B B A A A A A

Shiba 99 A B B B A A B B

Thaut 96 A C B A A A A A

KEY: A:

Adequate B:

Unclear (not

stated) C:

Inadequate

KEY:

A: Good B:

Unclear (not

stated) C:

Weak (eg al-

ternate allo-

cation)

KEY: A:

Adequate B:

Unclear (not

stated) C:

Inadequate

KEY:

A: Good B:

Unclear (not

stated) C:

Poor

KEY: A:

Good,

<10% B:

Unclear (not

stated) C:

Poor, >10%

KEY: A:

Good,

<10% B:

Unclear (not

stated) C:

Poor, >10%

KEY: A:

Constant B:

Unclear (not

stated)

C: Allowed

Variation

KEY: A:

Adequate B:

Unclear (not

stated) C:

Inadequate

METHOD OF RANDOMISATION AND CONCEAL-

MENT OF ALLOCATION

Homann 98 put the patients’ names into alphabetical order before

assigning them a random number. Theoretically if the alphabetical

list was open to the person applying the random number list,

he or she could miss out the next random number if it put the

patient into a group s/he did not want and go on to the next

random number that did. Hirsch 96 and Marchese 2000 used

computer generated random number tables to randomise but did

not state how the allocation was concealed. Also during the course

of the Hirsch 96 study one patient changed therapy group and was

included in the balance group’s data although they were originally

allocated to the balance and strengthening group. This could have

introduced selection and attrition bias. Thaut 96 used a ’random

draw’ method. Whilst this method is random it is susceptible to

tampering. The remaining three studies gave no information as

to their method of randomisation or concealment of allocation so

selection bias cannot be excluded.

ELIGIBILITY CRITERIA

The eligibility criteria for the trials were very broad and varied

considerably between trials. Only one trial (Homann 98) used

strict criteria for the diagnosis of idiopathic Parkinson’s disease.

NUMBERS OF PATIENTS

A total of 142 patients were examined in seven studies. With such

a small number of patients it is unlikely that they were truly rep-

resentative of the Parkinsonian population as a whole. There is a

strong possibility of selection bias with such small numbers. Over-

all only 42 of the 126 patients whose gender was specified were

female (33%), this could introduce difficulty when trying to ap-

ply the results of these trials to the general Parkinson’s population,

where the prevalence is approximately equal (Tanner 1996).

SIMILARITY AT BASELINE

A description of the baseline characteristics of the patients is im-

portant to decide whether the results are generalizable and to com-

pare characteristics of the two groups to ensure that the randomi-

sation was successful. Considering the small number of patients

in all of the studies, the likelihood of an unequal distribution of

patients is high.

The baseline characteristics of the total population studied, rather

than for each treatment group, were given in Homann 98 so no

assessment could be made between the therapy groups. The gender

of the patients in each therapy group was only stated in three trials,

Hirsch 96, Mohr 96 and Thaut 96. The distribution of males and

females was fairly even in these three trials. The age of the patients

was fairly evenly matched in each therapy group in the five trials in

which it was given. Although the Hoehn and Yahr scale is a crude

way of assessing disease severity it does give a rough idea of what

sort of patients were treated in the trials. As the Hoehn and Yahr

scale is not continuous the median should be used, however most

authors quote the mean. The mean Hoehn and Yahr score was

given for each of the therapy groups in four of the trials (Marchese

2000, Mohr 96, Palmer 86 and Thaut 96), and was fairly evenly

matched.

DESCRIPTION OF PHYSIOTHERAPY



The methods of physiotherapy used were generally described only

in a very broad manner. It would have been useful if a fuller de-

scription of the method of therapy had been given so that it would

be easier to compare these trial with others. However further de-

tails of the exact protocol used was available from Thaut 96 on

request. In particular the time spent with the therapist was not

specified in 3 of 6 trials (Marchese 2000, Hirsch 96 and Shiba 99),

although it should be noted that one of these studies (Shiba 99)

was only available to us as an abstract. However lacking this vital

piece of information it would be hard to reproduce these studies.

Mohr 96 used unequally sized groups (3-4 patients in novel ther-

apy vs. 5-7 patients in the standard group). This could have led to

performance bias, in that the patients in the smaller novel therapy

groups would have had more attention from the therapist than

those in the larger standard therapy groups.

OUTCOME MEASURES

A wide variety of outcome measures were used in these trials. A

discussion of their reliability, sensitivity and clinical relevance is

given in the Discussion section. The patients were followed for a

maximum of 8 weeks post-therapy, thus the long-term persistence

of the effect of physiotherapy is hard to determine.

DATA AND ITS ANALYSIS

Four studies analysed their data on an intention-to-treat basis.

However as none of the trials stated the basis of their data analysis,

this is more a reflection of the fact that these four trials had no drop-

outs and therefore by default the data is analysed on an intention-

to-treat basis. The data were analysed on a per protocol basis in

Hirsch 96, and it was unclear how the data in Shiba 99 and Mohr

96 were analysed.

Hirsch 96 violated the random allocation when one of the patients

randomised to the novel (combined therapy) group developed an

inguinal hernia 2 weeks into the trial. In the opinion of the authors

this was not an adverse event due to the physiotherapy. He was

unable to carry out the strengthening portion of the combined

therapy and so was re-allocated to the balance therapy alone group

(standard).

No numerical data were available for the cross-over trial of Shiba

99. If it had been data from the first arm of the trial would be

required so as to remove the possibility of carry-over effects.

Only Thaut 96 specified a primary outcome (walking velocity on

flat and inclined surfaces). Examining multiple outcomes carries

the risk of multiple analyses and a higher risk of finding spurious

but statistically significant results.

Hirsch 96, Marchese 2000, Palmer 86 and Thaut 96 statistically

compared the change in a given outcome measure (i.e. score after

therapy - score at baseline) between the two therapy groups (i.e.

change due to therapy A vs change due to therapy B). The remain-

ing trials statistically compared the change in an outcome for each

therapy group individually over time. This means that these trials

did not examine whether one form of physiotherapy is better than

another, only that improvements occurred after a given therapy.

BLINDING OF PARTICIPANTS

It is practically impossible to blind the patients and the therapists

in trials comparing the efficacy of one form of physiotherapy with

another. This leaves such trials open to performance and attrition

bias. However the assessors were blinded in Marchese 2000, Mohr

96 and Palmer 86 so detection bias is unlikely in these three studies.

Thaut 96 used a computerised gait analysis system and so detection

bias is unlikely in this study. However the assessors were unblinded

in Hirsch 96, and it was not stated in Shiba 99, so detection bias

cannot be excluded in these two studies.

Effects of interventions

Seven trials were identified comparing two forms of physiotherapy

with 142 patients in total. There was significant heterogeneity in

both trial design and the outcomes measures used such that the

results cannot be combined, so the trials results were examined on

an individual quantitative basis.

Hirsch 96, Marchese 2000, Palmer 86 and Thaut 96 provided

data on the mean change in outcome measures and statistically

compared this change across the two arms of the trial. The authors

of the other studies provided the mean and SD at baseline and after

treatment for each therapy group. We are awaiting advice from

the Cochrane Collaboration and other statistical departments on

a valid method of calculating the standard deviation of the change

from this data. Upon receiving this advice we will update this

review.

Hirsch 96

Hirsch 96 measured the number of ’falls’ during testing condi-

tions. Falls were defined as when the patients used their hands for

support or when a stepping strategy was used to avoid a fall. The

tests were carried out for 20 seconds for each of six conditions; a

combination of the patient standing on a fixed or sway-reference

surface, with eyes shut, eyes open, or eyes open and the surround-

ings also sway-referenced. Both methods decreased the number of

falls, but statistical analysis was not given. Those patients receiv-

ing combined therapy had one less fall under test conditions (i.e.

under Equitest conditions) than the balance group immediately

after the course of treatment. This difference persisted at the four

week follow up.

The remainder of the outcomes provided data on the change due

to therapy for each of the two groups. However the author did

not carry out statistical analysis of the differences between the two

forms of physiotherapy. The test conditions described above were

used to assess the degree of sway of the patients using ’Equitest’

Scores. Only one test condition (sway-referenced support and eyes

closed) showed any significant difference (P<0.05) in favour of

the combined balance and strengthening physiotherapy method

at the four week follow-up assessment.



Unsurprisingly, those patients receiving combined strength and

balance training significantly increased the strength of their leg

muscles as compared to the balance training group (P<0.001), a re-

sult that persisted at the four week follow-up assessment (P<0.05).

The activities of daily living ability (ADL) of the patients was as-

sessed using an ad hoc 18 item questionnaire. The score from this

questionnaire was divided into ADL (dressing, toileting, washing

etc., 11 items) and Instrumental ADL (cleaning house, preparing

food, shopping etc., 7 items). The ADL score did not differ be-

tween the two therapy groups. However the Instrumental ADL

was significantly improved by the combined therapy as compared

to the balance therapy alone (P<0.05), and this improvement per-

sisted at the four week follow-up (P<0.01).

Homann 98

This study was only presented as an abstract. No numerical data

were available.

Marchese 2000

Data were available for UPDRS mental, motor and ADL sub-

sections in Marchese 2000 at baseline, immediately after therapy

and six weeks later. None of the UPDRS subsections detected any

significant difference between the cued and uncued physiother-

apy regimes immediately after the course of therapy. However six

weeks later UPDRS motor scores for the cued group differed sig-

nificantly to those for the uncued group (4.9 points, Mann Whit-

ney p<0.02). None of the other subsections’ scores detected sig-

nificant differences between the two treatments.

Mohr 96

This study measured motor impairments using UPDRS motor

subsection. A two-way MANOVA revealed a highly significant dif-

ference with post hoc ANOVAs revealing no performance change

in the control group whilst the group receiving behavioural therapy

showed significant improvements (F(1,19) = 6.5, p = 0.01). The

difference in the change of the mean scores between the groups

was 0.21 points in favour of the behavioural therapy.

They also measured fine motor control impairment specifically in

the right and left hands. This was carried out by using the com-

puter-assisted Motor Performance Test Series (MPS). This test al-

lows for the exact quantification of the speed and accuracy of com-

plex arm, hand and finger movements such as following a winding

line with a pencil-like instrument. The MPS result includes error

scores and reaction times. In order to conduct an overall analysis

of the results in this test series, all data were first converted to z-

scores (with positive values indicating better performance). Subse-

quently these z-scores were subjected to MANOVA. When right

and left hand performance were compared no reliable interaction

was found in the left hand data, but a significant difference was

found for the right hand data. Post hoc ANOVAs showed signifi-

cant improvements after treatment only in the behavioural group

F(1,19) = 8.1, p = 0.01. The difference in the change of the mean

scores between the groups was 0.74 for the right hand and 0.06

for the left hand. It should be noted that all except two patients

(one in each therapy group) were right handed.

A two-way MANOVA of the UPDRS-ADL scores showed no

significant difference. The difference in the change of the mean

scores between the groups was 0.07 in favour of the behavioural

group.

Quality of life was measured using a questionnaire about content-

ment of life, which revealed higher scores after therapy. The differ-

ence in the change of the mean scores between the groups was 0.17

points for general contentment of life and 0.02 points for health-

related contentment of life, both in favour of the behavioural ther-

apy. These differences were not statistically significant.

Patient’s mood was measured using UPDRS-Mental subsection

and the Beck Depression Inventory (BDI). The difference in the

change of the mean scores of the UPDRS mental subsection be-

tween the groups was 0.13 points in favour of the behavioural

therapy. The difference in the change of the mean scores of the

BDI between the groups was 0.63 points in favour of the ’standard’

therapy.

Palmer 86

Palmer 86 administered the Parkinson’s Disease Motor Battery

along with a variety of tests to assess activities of daily living, grip

strength and motor co-ordination. Data was provided on the mean

and standard deviation of the change due to therapy.

Tremor significantly decreased in the left arm in the karate group

compared to the stretch exercises group (p<0.01). However it

should be noted that although not statistically significant the trend

is in the opposite direction for the right arm. The nine-hole peg-

board test improved significantly in both the right (p<0.05) and

left (p<0.001) hands in the karate trained group as compared to

the stretch group. None of the other measures showed any sig-

nificant difference between the two forms of therapy used in this

study.

Shiba 99

This study was presented as an abstract. No numerical data were

available.

Thaut 96

Thaut 96 examined gait parameters after walking to a rhythm

compared to self-paced walking. Data was provided for the per-

centage change score means and standard deviations. ANOVA

comparisons between the groups revealed significant differences

in improvement in walking velocity over flat and inclined surfaces.

Post hoc comparisons showed that the rhythmically cued group

improved significantly over the self-paced timing group (Flat: p

= 0.03; Incline: p = 0.03). Similar analysis of stride length and

cadence showed that changes in stride length did not differ signifi-

cantly between the two therapy groups, whilst changes in cadence

were significantly higher in the cued therapy group (p = 0.03).

OTHER OUTCOMES

None of the studies examined medication dosage, adverse events,

carer outcomes or performed any economic analysis of the thera-

pies.



D I S C U S S I O N

PRINCIPAL FINDINGS

• Seven randomised controlled trials were found comparing

two forms of physiotherapy (142 patients). These studies varied

significantly in their methodology so it was deemed

inappropriate to combine the results of the studies. Individual

studies were too small to demonstrate that any one form of

physiotherapy was significantly better than another.

• Considering the small number of patients examined, the

methodological flaws in the studies, and the possibility of

publication bias, it is unsafe to draw any conclusions regarding

the efficacy of any one form of physiotherapy.

• Considering the wide range of physiotherapy approaches

used in the studies, agreement needs to be found as to ’standard’

or ’optimal’ physiotherapy for Parkinson’s disease.

• Large well designed RCTs are needed initially to

demonstrate ’standard’ physiotherapy’s efficacy and effectiveness

in Parkinson’s disease versus placebo. After that individual

variations in methods can be compared against this standard,

again by using large well designed RCTs.

PHYSIOTHERAPY METHODS

A wide range of physiotherapy methods were used in these tri-

als, e.g. proprioceptive neuromuscular facilitation, Bobath, Karate,

stretches etc. A survey of UK physiotherapists found that there

was a consensus that an eclectic combination of therapies offered

the best treatment approach in Parkinson’s disease (Plant 1999).

However work needs to be done in agreeing a ’standard’ form of

physiotherapy that can then be tested in a trial. This review high-

lights the fact that there is no trial evidence to support any one

form of physiotherapy over another in Parkinson’s disease.

METHODOLOGICAL QUALITY

Overall the methodological quality of the trials and the standard

of the reporting was poor. However it is recognised that four of the

seven studies were published before the CONSORT guidelines

were published (1996), when trial reporting was not formalised.

Two of the studies (Homann 98 and Shiba 99) were only published

as abstracts. Although some more data was obtained from some

of the authors, greater detail may become available when the full

reports are published.

The trials used insufficient numbers of patients to avoid drawing

false negative conclusions and to reduce the possibility of selection

bias. It is now generally accepted that quality of life measures

should be used as the primary outcome of interest in clinical trials.

Only one trial (Mohr 96) used QOL scales, and as only 41 patients

were examined in this study there is insufficient data to inform

sample size calculations for future trials.

The randomisation method used in three trials (Marchese 2000,

Hirsch 96 and Homann 98) was computer generated random

number tables and a random draw in one study (Thaut 96). The re-

maining three trials did not state how randomisation was achieved.

In all seven trials the concealment of the allocation was inadequate,

unclear or not stated.

It is vital that eligibility criteria are well defined so that it is un-

derstood what sort of a population were treated. For example it is

important that the Parkinson’s disease accords with the UK Brain

Bank Parkinson’s Disease criteria (Gibb 1988). This will reduce the

likelihood of including patients with Parkinson’s plus syndrome

which have a significantly different clinical course compared to

idiopathic Parkinson’s disease. The eligibility criteria should also

define the severity of the patients eligible to participate, and state

clearly any exclusion criteria such as severe arthritis. This would

allow an easier assessment of which Parkinson’s disease patients the

trial’s results apply to.

The physiotherapy methods were poorly described, also the time

spent by the therapist with the patient was not specified in four

trials. Also the intensity of the therapy was never defined in any of

the studies examined. This means that it would be hard to repro-

duce the studies’ methodology. It is recognised that physiother-

apy methods can be hard to describe and that the methods vary

according to the patients’ needs and abilities. One way to achieve

this would be to post further details, even videos, on a journal’s

web site.

Only 33% of the patients whose gender was specified were female.

This is in contrast to the general population of Parkinson’s pa-

tients were the prevalence of the disease is equal between the sexes

(Tanner 1996). This is a common finding in Parkinson’s disease

trials but raises the question of whether the results are generalis-

able to the whole Parkinson’s disease population and to women in

particular.

OUTCOME MEASURES

The outcome measures used varied greatly between the trials. The

outcome measures used in five of the seven trials were only assessed

at baseline and immediately after therapy. It would have been valu-

able to know the long-term duration of any improvement follow-

ing therapy. The outcome measures used must be clinimetrically

sound and reliably reflect the impact of physiotherapy on the pa-

tient’s life.

Motor Impairment: Global Scales:



The motor subsection (part III) of the UPDRS is one of the most

popular rating scale used in clinical trials in recent years. The UP-

DRS-Motor subsection combines disability and handicap mea-

sures in the same scale. This scale was used in both Marchese 2000

and Mohr 96. Marchese 2000 found that the cued therapy was

not significantly better than uncued physiotherapy immediately

after the therapy regime. However when the patients were followed

for a further six weeks the cued therapy group differed from the

uncued group by 4.9 points (Mann Whitney p<0.02). A differ-

ence of five points is probably clinically useful. Mohr 96 found a

smaller change in UPDRS-Motor of 0.21 points in favour of the

behavioural therapy. It is unlikely that this difference between the

two therapy methods used in this trial is clinically significant.

Motor Impairment: Gait

The walk index of patients was calculated in Palmer 86 as the

product of the number of seconds required to walk 30 feet mul-

tiplied by the number of steps taken by the right foot. This mea-

sure showed no significant difference between the two therapies

examined. Individual gait parameters were measured Thaut 96.

The change in walking velocity on both flat and inclined surfaces

and in cadence were significantly higher in the group that trained

walking to a rhythmic piece of music. The changes reported may

have clinical significance, they were between 24-26% improve-

ment. Also the authors reported that all the patients receiving the

cued training reported improvements in walking stability, speed

and helped their walking in activities of daily life.

Motor Impairment: Balance

Balance is vital in the prevention of falls which can have a signif-

icant impact upon Parkinson’s disease patients health. Hirsch 96

measured the number of ’falls’ under test conditions. This did not

detect any significant difference between the balance training and

the combined balance and leg strengthening.

Another measure employed by Hirsch 96 was the degree of sway of

the patients under six different conditions designed to test vestibu-

lar and proprioceptive components of balance. Only one measure

showed significant improvement in the combined group at the

four week follow-up assessment; sway referenced support with eyes

closed (P<0.05). This large number of related outcomes may lead

to a problem with examining multiple statistical tests at the 95%

confidence level as 1 in 20 outcomes may be significant by chance.

Re-examining this outcome at the 99% confidence level produces

a non-significant result.

Motor Impairment: Fine motor control

This was assessed in Mohr 96 using a Motor Performance Test Se-

ries (MPS). This is an objective measure of Parkinsonian patients

fine motor control of their hands. Performance in the MPS cor-

relates highly with motor performance relevant in everyday life,

such as handwriting or cutting with a knife and the motor section

of UPDRS (Pinter 1992). The changes due to therapy were not

compared statistically between the groups so we are unable to de-

termine the statistical significance of a difference 0.74 (right hand)

and 0.06 (left hand) in favour of the behavioural therapy group.

Clinical significance seems unlikely with such small changes.

Palmer 86 used a pursuit score calculated as the number of times

in a 50 second period the patient covered a randomly moving

dot on an oscilloscope with a photodetector in his outstretched

hand to assess fine motor control. This measure did not detect any

significant differences between the two therapy modes used in this

study.

Palmer 86 used a nine-hole pegboard test to assess dexterity and

coordination. Over the three months of the therapy the stretch

group got much worse at this test whilst the karate trained group

maintained or slightly improved their scores. This difference was

significant for both the right and left hands. However other tests

used in this study that could also be said to detect changes in

dexterity and coordination such as the Minnesota placing and

turning test or the button board showed no significant differences

between the two therapy techniques.

Motor impairment: Muscle strength

Leg muscle strength was measured in Hirsch 96. The study com-

pared balance training with combined balance and strength train-

ing. As the strengthening exercises were all focused on the legs it

was not surprising that the combined therapy increased the muscle

strength of the legs significantly more than balance training alone.

Grip strength was assessed in Palmer 86 but this did not distinguish

between the two therapies used, although both therapies increased

the grip strength by about 20%.

Activities of Daily Living

The UPDRS ADL subsection was designed for use with Parkin-

son’s disease patients and has been validated with this patient pop-

ulation. This scale was used by Marchese 2000 and Mohr 96.

Only Marchese 2000 statistically examined the change in means

between the two therapy groups. The size of the difference in mean

change was small in both studies, between 0.1 (Marchese 2000)

and 2.2 points (Mohr 96). It is unlikely this has clinical signifi-

cance.

Hirsch 96 used a novel 18 point ADL scale for which we have no

details on its validity, reliability and sensitivity in Parkinson’s dis-

ease. The scale was divided into two parts, ADL and instrumental

ADL. The Instrumental ADL (7 items) was improved significantly

in the combined balance and strength group as compared to the

balance only group. It is hard to determine the clinical significance

of a 3-5 point improvement in this scale.

Palmer 86 timed a number of activities of daily living such as

putting socks and shoes on and off, putting a shirt on and off

and getting up from a chair. None of these measures detected

any significant differences between karate and stretch exercises.



These tests of ADL can be said to be more an assessment of motor

function as they represent an artificial situation, e.g. patients would

rarely put a shirt on and then take it off immediately. Also the

patients did not have access to any aids and adaptations that they

might use in real life e.g. Velcro fasteners.

Quality of Life

Mohr 96 used a ’Contentment of Life’ score to assess general QOL

and health related QOL. We have no knowledge of its validity, re-

liability and sensitivity in Parkinson’s disease patients. The change

in the means was not analysed between the two therapy groups.

Thus it is impossible to determine the statistical significance of a

0.17 point difference for general contentment of life and a 0.02

point difference for health-related contentment. The clinical sig-

nificance of these changes is also impossible to determine.

Depression

The effectiveness of the therapy could potentially be affected by

depression. Depressed patients could be less compliant both dur-

ing the therapy sessions and also in the practice at home. The

therapy itself might affect depression. The patient’s mood may im-

prove due to the attention they are being paid by the therapist, by

getting out of the house and meeting other people. A well designed

placebo intervention would control for the non-therapeutic con-

founders. If the therapy affected the patient’s physical well-being

so that they feel more in control and able to carry out more of

their ADL independently, this could improve the patient’s mood.

Also it is important to measure depression as a number of surveys (

Karlsen 1999, GPDS 2000) have shown that depression accounts

for a significant proportion of the reduction in quality of life due

to Parkinson’s disease.

The Beck Depression Index (BDI) is a highly validated and sen-

sitive score to assess depression that has been used in a number of

studies of Parkinson’s disease patients. The BDI is a 21 item, self-

rated inventory covering a wide variety of cognitive, behavioural

and somatic aspects of depression. It was used in Mohr 96 but

analysis of the mean change between the two groups was not per-

formed. There was a difference of 0.63 points between the groups

in favour of ’standard’ therapy but the clinical significance of such

a small difference is doubtful.

UPDRS mental subsection was used to measure mood in Marchese

2000 and Mohr 96. This subsection consists of only four ques-

tions and is often criticised for being too brief. Only two of the

questions relate to depression the other two assess dementia and

hallucinations. The change of the mean score was compared be-

tween the two therapy groups in only Marchese 2000. The small

difference seen in Marchese 2000 (0.5 points) is not statistically

significant. Mohr 96 detected a difference of 0.13 points. It is

probable that neither difference is clinically significant.

Carer Outcomes

Approximately 75% of patients with Parkinson’s disease live with

a partner, who is usually of a similar age and may have disabilities

of their own (Lloyd 1999). The impact of caring for a person with

Parkinson’s can be severe (O’Reilly 1996), and it would be hoped

that an intervention such as physiotherapy could have a positive

effect on the carer’s life as well as the patient’s.

Health Economics

No health economics analysis of physiotherapy has been per-

formed which precludes an understanding of the economic value

of this therapy.

A U T H O R S ’ C O N C L U S I O N S

Implications for practice

All of the trials reported that they found one form of physiotherapy

superior to another. However the small numbers in all of the trials

and the methodological problems present in most studies prevent

us from drawing any firm conclusion regarding the superiority of

one form of physiotherapy over another in Parkinson’s disease.

Implications for research

Considering the small number of patients and the methodologi-

cal flaws in many of the studies, there is insufficient evidence to

support the use of one form of physiotherapy over another for the

treatment of Parkinson’s disease.

Another Cochrane review that examined the efficacy of physio-

therapy versus placebo or no therapy (Physiotherapy for patients

with Parkinson’s Disease) concluded that there was insufficient ev-

idence to support or refute the efficacy of physiotherapy in Parkin-

son’s disease. Although a recent survey (Plant 1999) has deter-

mined that an eclectic combination of physiotherapeutic methods

offers the best treatment of Parkinson’s disease, more work needs

to be performed to determine the best combination of methods.

These then need to be defined in sufficient detail so as to allow the

examination of the efficacy of physiotherapy in Parkinson’s disease

in a large multi-centre trial.

The efficacy of ’standard’ physiotherapy must be proved first before

examining variations in physiotherapy methods. Therefore large

well designed randomised controlled trials are needed to judge

the effect of physiotherapy in Parkinson’s disease. The placebo

therapy arm should try to compensate for placebo and Hawthorn

effects by this group being ’treated’ for a similar period of time

and in a similar environment as the therapy group. However it

is recognised that double-blinding is impossible in therapy trials

and so some placebo effect will not be controlled for. After this

several large RCTs are needed to demonstrate the most effective

form of physiotherapy in Parkinson’s disease. In these studies, a

rigorous method of randomisation should be used and allocation



adequately concealed. Data should be analysed according to in-

tention-to-treat principles. These trials should be carried out and

reported according to the guidelines set out in the CONSORT

statement (CONSORT 1996).

This review emphasises many methodological shortcomings in

the trials comparing two forms of physiotherapy in Parkinson’s

patients. The issues arising from this review have a significant

bearing on the conduct of future physiotherapy trials in Parkinson’s

disease and other conditions:-

• Firm diagnostic criteria should be used (e.g. UK Parkinson’s

Disease Brain Bank Criteria, Gibb 1988).

• Inclusion and exclusion criteria should be clear and trials

should aim to enrol uniform cohorts of Parkinson disease

patients.

• Investigators should clarify at what stage of the disease

physiotherapy is being evaluated.

• Investigators should define a primary outcome and limit

the number of other outcomes measured so as to reduce the

possibility of finding false positive results.

• Trials must have sufficient numbers of patients to avoid

false negative results.

• Trials must include clear descriptions of the therapeutic

interventions.

• The patients should be followed for at least 6 months after

treatment to assess any long term effects of the physiotherapy

intervention.

• Regardless of the scale used, trials should report whether

scores on impairment and disability refer to the ’on’ or ’off ’

phase.

• The effect of therapy on drug requirements both in the

short and long term should be reported.

• Suitable, clinimetrically sound, outcome measures should

be chosen so that the efficacy and effectiveness of physiotherapy

can be assessed and an economic analysis performed. Outcomes

which have meaning to patients should be used wherever

possible since they need to know the value of physiotherapy in

practical terms.

• The data must be analysed on an intention-to-treat basis

and the change in an outcome measure must be compared

statistically across the two therapy groups.

A C K N O W L E D G E M E N T S

Many thanks to all of the authors of the included studies who

assisted in providing unpublished data and clarification of their

methods. Also thanks to all of the people contacted whilst trying

to locate any other unpublished randomised controlled trials.

R E F E R E N C E S

References to studies included in this review

Hirsch 96 {published and unpublished data}∗ Hirsch, M.A. Activity dependent enhancement of balance

following strength and balance training. PhD Thesis, Florida State

University 1996. [: UMI Number: 9622856]

Homann 98 {published and unpublished data}∗ Homann, C.N, R. Crevenna, H. Kojnig, B. Kurzl, S. Reinprecht,

K. Wenzel, K. Suppan, G. Ivanic, S. Horner, E. Ott. Can

physiotherapy improve axial symptoms in parkinsonian patients? A

pilot study with the computerized movement analysis battery

Zebris. Movement Disorders 1998;13(Supplement 2):234.

Marchese 2000 {published and unpublished data}

Abbruzzese, G, M. Diverio, F. Zucchi, R. Marchese. Comparison

of two physical therapy approaches in the rehabilitation of

parkinsonian patients. Parkinsonism and Related Disorders 1999;5

(Supplement):S49.∗ Marchese R, M. Diverio, F. Zucchi, C. Lentino, G. Abbruzzese.

Comparison of two physical therapy approaches in the

rehabilitation of parkinsonian patients: a comparison of two

physical therapy protocols. Movement Disorders 2000;15(5):

879–883.

Mohr 96 {published data only}∗ Mohr, B, V. Muller, R. Mattes, R. Rosin, B. Federmann, U.

Strehl, F. Pulvermuller, F. Muller, W. Lutzenberger, N. Birbaumer.

Behavioural treatment of Parkinson’s disease leads to improvement

of motor skills and to tremor reduction. Behaviour Therapy 1996;



27:235–255.

Muller, V, B. Mohr, R. Rosin, F. Pulvermuller, F. Muller, N.

Birbaumer. Short-term effects of behavioural treatment on

movement initiation and postural control in Parkinson’s disease: A

controlled clinical study. Movement Disorders 1997;12(3):306–314.

[MEDLINE: 97303399]

Palmer 86 {published data only}

Palmer, S.S, J.A. Mortimer, D.D. Webster, R. Bistevins, G.L.

Dickinson. Comparison of stretch exercises and karate training as

therapy for Parkinson’s disease. Archives of Physical and Medical

Rehabilitaion 1984;65:626.∗ Palmer, S.S, J.A. Mortimer, D.D. Webster, R. Bistevins, G.L.

Dickinson. Exercise therapy for Parkinson’s disease. Archives of

Physical and Medical Rehabilitation 1986;67:741–745.

Shiba 99 {published data only}∗ Shiba, Y, S. Obuchi, H. Toshima, H. Yamakita. Comparison

between visual and auditory stimulation in gait training of patients

with idiopathic Parkinson’s disease. World Congress of Physical

Therapy Conference. 1999.

Thaut 96 {published data only}∗ Thaut, M.H, G.C. McIntosh, R.R. Rice, R.A. Miller, J. Rathbun,

J.M. Brault. Rhythmic auditory stimulation in gait training for

Parkinson’s disease patients. Movement Disorders 1996;11(2):

193–200. [MEDLINE: 96236298]

References to studies excluded from this review

Dam 96 {published data only}∗ Dam M, P. Tonin, S. Casson, F. Bracco, L. Piron, G. Pizzolato, L.

Battistin Dam. M, P. Tonin, et al. (1996). “Effects of conventional

and sensory-enhanced physiotherapy on disability of Parkinson’s

disease patients.” Advances in Neurology 69: 551-555. Dam, M, P.

Tonin, et al. (1996). “Effects of conventional and sensory-enhanced

physiotherapy on disability of Parkinson’s disease patients.”

Advances in Neurology 69: 551-555. Effects of conventional and

sensory-enhanced physiotherapy on disability of Parkinson’s disease

patients. Advances in Neurology 1996;69:551–555.

References to ongoing studies

Stallibrass {published data only}∗ Stallibrass, C. Controlled trial to evaluate the effects of lessons in

the Alexander Technique on the management of disability by

people with Parkinson’s disease. National Research Register End

Date: 01/03/2001.

Wagenaar {published data only}∗ Wagenaar, R. C, R. E. A. van Emmerik, E. E. H. van Wegen, C. J.

Th. de Goede, G. Tissingh, T. W. Koelman. Exercising dynamics

of walking in Parkinson’s disease. 13th International World

Congress of Physical Therapy. 1999:PL–RR-018-25J.

Additional references

Clarke 1995

Clarke CE, Zobkiw RM, Gullaksen E. Quality of life and care in

Parkinson’s disease. British Journal of Clinical Practice 1995;49(6):

288–293. [MEDLINE: 96151530]

CONSORT 1996

Begg, C, M. Cho, S. Eastwood, R. Horton, D. Moher, I. Olkin, R.

Pitkin, D. Rennie, K. F. Schultz, D. Simel, D. F. Stroup. Improving

the quality of reporting of randomized controlled trials. The

CONSORT statement. Journal of the American Medical Association

1996;276(8):637–639.

Deane

Deane K. H. O, D. Jones, C. E. Clarke, E. D. Playford, Y. Ben-

Shlomo. Physiotherapy for patients with Parkinson’s disease. The

Cochrane Library.

Gibb 1988

Gibb W. R. G, A. J. Lees. The relevance of the Lewy body to the

pathogenesis of idiopathic Parkinson’s disease. Journal of Neurology,

Neurosurgery and Psychiatry 1988;51:745–752.

GPDS 2000

The Global Parkinson’s Disease Survey. An insight into quality of

life with Parkinson’s disease. The Parkinson’s Disease Society of the

United Kingdom 2000.

Hildick-Smith 1987

Hildick-Smith M. Has rehabilitation a role in the treatment of

Parkinson’s disease?. In: Clifford-Rose F editor(s). Parkinson’s

Disease. Clinical and Experimental Advances. Vol. 6, London:

Libbey, 1987.

Karlsen 1999

Karlsen K H, Larsen J P, Tandberg E, Maeland J G. Influence of

clinical and demographic variables on quality of life in patients with

Parkinson’s disease. Journal of Neurology, Neurosurgery and

Psychiatry 1999;66(4):431–435.

Lloyd 1999

Lloyd M. The new community care for people for people with

Parkinson’s disease and their carers. In: Percival R, P. Hobson editor

(s). Parkinson’s Disease: Studies in Psychological and Social Care.

London: BPS Books, 1999:13–59.

Mutch 1986

Mutch WJ, Strudwick A, Roy SK, Downie AW. Parkinson’s disease:

disability, review, and management. British Medical Journal 1986;

293:675–677. [MEDLINE: 87001239]

O’Reilly 1996

O’Reilly F, F. Finnan, S. Allwright, G. Davey Smith, Y. Ben-

Shlomo. The effects of caring for a spouse with Parkinson’s disease

on social, psychological and physical well-being. British Journal of

General Practice 1996;46:507–512.

Oxtoby 1982

Oxtoby M. Parkinson’s disease patients and their social needs.

Parkinson’s Disease Society 1982.

Pinter 1992

Pinter M. M, R. J. Helscher, Ch. O. J. Nasel, E. Reidl, G.

Schnaberth. Quantification of motor deficit in Parkinson’s disease

with a motor performance test series. Journal of Neral Transmission

1992;4:131–141.

Plant 1999

Plant RP, Jones D, Ashburn A, Lovgreen B, Kinnear E, Handford F.

Evaluation of physiotherapy in Parkinson’s disease - project update.

The science and practise of multidisciplinary care in Parkinson’s disease

and Parkinsonism. London: British Geriatric Society, 1999.



Popham 1973

Popham, W. J, K. A. Sirotnik. Nonparametric statistics.

EDucational Statistics: Use and Interpretation. 2nd Edition. New

York: Harper & Row Publishers, 1973:267–282. [: ISDN:

06–045252–8]

Tanner 1996

Tanner C. M, J. P. Hubble, P. Chan. Epidemiology and genetics of

Parkinson’s disease. In: Watts R. L, W. C. Koller editor(s).

Movement Disorders. Neurologic Principles and Practise. New York:

McGraw Hill, 1996:137–160. [: ISBN: 0–07–035203–8]

Yarrow 1999

Yarrow S. Members’ 1998 survey of the Parkinson’s Disease Society

of the United Kingdom. In: Percival R, Hobson P editor(s).

Parkinson’s disease: Studies in psychological and social care. Leicester:

BPS Books, 1999:79–92.∗ Indicates the major publication for the study



C H A R A C T E R I S T I C S O F S T U D I E S

Characteristics of included studies [ordered by study ID]

Hirsch 96

Methods Parallel group design.

Randomised according to a computer generated random number list but concealment of allocation not

stated. Analysed on a per protocol basis.

Treated as outpatients for an unspecified period of time (3 days/week) for 10 weeks.

Assessed at baseline, immediately after treatment and 4 weeks later. Assessors were not blinded.

Participants 8 patients in novel combined balance and strength training group and 9 patients in balance training group.

3 drop-outs in combined group.

Patients mean age 75.6 (combined), 72.3 (balance); Male/female 5/3 (combined), 6/3 (balance); Hoehn

and Yahr 2.0 (total).

Inclusion criteria: Parkinson’s disease, Hoehn and Yahr stage I-III, were deemed by neurologist or GP to

be able to follow testing instructions, with 1 or more of following; falls, unsteadiness walking or standing,

difficulty rising from chair/bed, muscle weakness in legs.

Exclusion criteria: Nursing home residents, hospital inpatients.

Interventions Combined: Group training in strengthening and balance exercises. Resistance exercises used Nautilus leg

extension and side-lying leg-flexion machines and therabands. Balance training consisting of gentle sternal

or dorsal perturbation and leaning movements designed to enhance limit of stability whilst standing on

a firm or a compliant surface.

Balance: ’standard’ group balance therapy as described above.

Not stated whether the drug therapy was constant during treatment.

Outcomes ADL

Instrumental ADL

Groningen activity restriction scale (GARS)

Incidence of falls (use of hands or stepping to maintain balance during testing procedure)

Leg muscle strength

Body sway (measured by Equitest).

Notes NB. Randomisation violation; 1 patient who was allocated to the combined therapy group was reassigned

to the balance group after 2 weeks of training due to an inguinal hernia making it impossible for him to

carry out the strength training.

Risk of bias

Item Authors’ judgement Description

Allocation concealment? Unclear B - Unclear



Homann 98


The patient’s names were put into alphabetical number and then randomised using computer generated

random number tables.

Data analysed on an intention to treat basis.

Treated as outpatients for 14 ’units’ over 5 weeks.

Assessed at baseline and immediately after treatment. Not stated whether assessors were blinded.

Participants 8 patients in Bobath physiotherapy group and 8 in PNF physiotherapy group. No drop-outs were noted.

Baseline characteristics only available for all patients in the 3 arms of this trial.

Inclusion criteria: IPD according to UK Brain Bank diagnostic criteria. No exclusion criteria.

Interventions Bobath: Individual Bobath program focusing on proprioceptive skills to improve posture and gait.

PNF: PNF- physiotherapy directed at the mobility of the spine. Therapy carried out by physiotherapist.

Drugs were stable for duration of therapy.

Outcomes UPDRS

Axial symptoms

Stride length

Stride velocity

Stride cadence

Notes Abstract and poster only.

No numerical data available.

Risk of bias


Allocation concealment? No C - Inadequate

Marchese 2000


Randomised according to a pseudo-random number list, concealment of allocation not stated. Analysed

on an intention to treat basis. Treated as outpatients for 18 hours over 6 weeks. Assessed at baseline,

immediately afterwards and 6 weeks later. Assessors were blinded.

Participants 10 patients in novel cued group and 10 in standard physiotherapy group. No drop-outs stated. Patients

mean age 66.9 (cued), 65.0 (standard); total male/female ratio 13/7; Hoehn and Yahr 2.3 (cued), 2.4 (

standard).

Inclusion criteria: Stable IPD with no clinical fluctuations. Non-demented (MMSE>26). No exclusion

criteria.

Interventions Cued: Individual physical rehabilitation program aimed at improving range of movement, trunk rotation,

posture and walking, using external sensory cues. Standard: Individual physiotherapy program described

above but without the cues.

Drug therapy was stable during therapy period.

Outcomes UPDRS, motor, mental and ADL subsections.



Marchese 2000 (Continued)

Notes

Risk of bias



Mohr 96


Randomisation method not stated. Type of analysis of data unclear. Treated as outpatients for 30 hours

over 10 weeks. Assessed at baseline and immediately after treatment. Assessors were blinded.

Participants 20 patients in novel behavioural therapy group and 21 in standard therapy group. 2 patients dropped out

of the standard group. Patients mean age 63.6 (behaviour), 60.4 (standard); male/female 15/5 (behaviour)

, 12/9 (standard); Hoehn and Yahr 2.0 (behaviour), 2.1 (standard).

Inclusion criteria: IPD with no change in medication for 4 weeks prior to study.

Exclusion criteria: History of drug or alcohol abuse, depression, dementia or other psychiatric disorders,

other significant illness.

Interventions Behavioural: Groups of 3-4 patients. Progressive muscle relaxation aimed at reducing tremor, motor

training with use of external cues and internal commands, social interactions training using role playing

and aiming at reducing stress.

Standard: Groups of 5-7 patients. Breathing and physical exercises, short relaxation, given information on

disease and discussion of specific disease-related problems with therapists. Therapy carried out by clinical

psychologists.

Drugs stable during therapy period.

Outcomes UPDRS, motor, mental and ADL subsections.

Hoehn and Yahr

Swab and England

Beck Depression Inventory

Motor performance test series

Assertiveness questionnaire

Contentment with life questionnaire

Notes Muller 1997 reports a subsection of the original population.

Risk of bias





Palmer 86


Randomisation method not stated.

Analysed on an intention to treat basis.

Treated as outpatients for 36 hours over 12 weeks.

Assessed at baseline and immediately after treatment. Assessors were blinded.

Participants 7 patients in novel karate exercise group and 7 patients in standard stretch exercises group. No drop-outs.

Patients mean age 65.9 (karate), 63.9 (standard); 12 males and 2 females total; Hoehn and Yahr stage 2.4

for both groups.

Inclusion criteria: stable drug regime, ability to attend.

Exclusion criteria: any physical problems that might cause them to risk injury during the exercises.

Interventions Karate: Group training in upper body karate exercises with patients seated. Led by a rehabilitation nursing

student with a black belt in karate.

Standard: Group training in stretch exercises from UPF exercise program led by a corrective therapist.

Drugs remained constant during therapy.

Outcomes Parkinson’s disease motor battery

ADL

Grip strength

Motor coordination and speed

Notes

Risk of bias



Shiba 99

Methods Cross-over group design.

Method of randomisation not stated. Method of analysis of the data not stated. Treated as outpatients for

an unknown period of time. 1 week between each training regime. Assessed at baseline and immediately

after treatment. Not stated whether the assessors were blinded.

Participants 4 patients in each group. No drop-outs noted. Total patients mean age 65, 3 males, 5 females. Hoehn and

Yahr score not given.

Inclusion criteria: Stable mild to moderate Parkinson’s disease.

No exclusion criteria.

Interventions Group I: Individual visually stimulated gait training followed by auditory stimulated gait training. Group

II: Individual auditory training followed by visual training. Visual training: patients walked over parallel

lines at 90 degrees to the direction of travel. Distance apart of lines dependant on patients normal stride

length. Auditory stimulation: Patients walked to a rhythm that was 30% higher than their comfortable

walking rhythm. Both type of sessions stopped upon fatigue.

Not stated whether drugs were kept stable.



Shiba 99 (Continued)

Outcomes Stride length

Notes Abstract only. No numerical data available.

Risk of bias



Thaut 96


Randomised by a ’random draw’, but the concealment of allocation was unclear.

Analysed on an intention to treat basis.

Treated at home or in the community, for 10.5 hours over 3 weeks. Assessed in laboratory, at baseline and

immediately after treatment. Not stated whether assessors were blinded.

Participants 15 patients in novel rhythmic auditory stimulation group (RAS) and 11 in standard self paced training

group (SPT). NO drop-outs noted. Patients mean age 69 (RAS), 74 (SPT); male/female 10/5 (RAS), 8/3

(SPT), Hoehn and Yahr 2.4 (RAS), 2.5 (SPT).

Inclusion criteria: IPD with significant gait deficits but able to walk without physical assistance. No

exclusion criteria.

Interventions RAS: Individual. 30 min/day walking to 3 different tempos of music. For 1st week; normal tempo =

pretest cadence, quick = 5-10% faster, fast = an additional 5-10% faster. After each week each tempo was

increased by 5-10% to a maximum pace of 130 steps/min.

SPT: Individual. 30 min/day walking at normal, quick and fast speeds.

Drugs were kept constant throughout trial.

Outcomes Stride velocity

Stride cadence

Stride length

EMG analysis on leg muscles.

Notes 3 arms to trial; RAS, SPT and no treatment. SPT vs. no treatment are compared in ’Physiotherapy for

patients with Parkinson’s disease’ Cochrane review.

Risk of bias





Characteristics of excluded studies [ordered by study ID]

Dam 96 Patients not randomly allocated to the two therapy groups.

Characteristics of ongoing studies [ordered by study ID]

Stallibrass

Trial name or title A controlled trial to evaluate the effect of a course of lessons in the Alexander technique on the management

of disability in Parkinson’s disease.

Methods

Participants 90 patients.

Inclusion criteria: Idiopathic Parkinson’s disease, diagnosed by a consultant neurologist; a minimum score on

a cognitive deficit test; able to get into car unaided; willing to make no changes in their Parkinson’s drugs for

3 months.

Exclusion criteria: Receiving treatment for any other serious disorder affecting the CNS or psychotrophic

medication for depression; previous experience of the Alexander Technique; other individual non-pharmaco-

logical therapies in the last 3 months.

Interventions Group A: 24 lessons in the Alexander Technique.

Group B: Untreated.

Group C: 24 sessions of therapeutic massage.

Outcomes Self report questionnaires:

ADL at best times of day.

Beck Depression Inventory.

Body Concept Scale.

Timed measures.

Medical questionnaire.

Tested at baseline, immediately after and 6 months after intervention.

Starting date Start date: 01/09/1998

End date: 01/01/2001

Contact information Dr C Stallibrass,

Flat 18, Manor Mansions,

Belsize Park,

London,

NW3 4NB,

UK.

[email protected]



Stallibrass (Continued)

Notes www.update-software.com.NRR.htm

Wagenaar

Trial name or title Exercising dynamics of walking in Parkinson’s disease.

Methods

Participants 22 Parkinson’s disease patients.

Eligibility criteria unknown.

Interventions Pilot study examining the effects of gait training in PD patients.

Outcomes Walking velocity.

Other gait parameters.

Starting date Unknown.

Contact information Dr Wagenaar.

[email protected]

Notes



D A T A A N D A N A L Y S E S

This review has no analyses.

W H A T ’ S N E W

Last assessed as up-to-date: 28 November 2000.

13 November 2008 Amended Converted to new review format.

H I S T O R Y

Protocol first published: Issue 4, 2000

Review first published: Issue 1, 2001

29 November 2000 New citation required and conclusions have changed Substantive amendment

C O N T R I B U T I O N S O F A U T H O R S

K H O Deane carried out the majority of the searching for eligible studies. All reviewers were involved in the determination of which

studies were eligible for the review. K H O Deane and C Ellis-Hill extracted the data from the included studies. All reviewers were

involved in the writing of the review. K H O Deane was the primary author.

D E C L A R A T I O N S O F I N T E R E S T

None.

S O U R C E S O F S U P P O R T

Internal sources

• City Hospital NHS Trust, UK.



External sources

• NHS Research and Development Programme for People with Physical and Complex Disabilities; Project PCD2/A1/250, UK.

• Conference grant from The Royal Society, UK.

• Conference grant from Pharmacia Upjohn, UK.

I N D E X T E R M S

Medical Subject Headings (MeSH)

∗Physical Therapy Modalities; Parkinson Disease [∗rehabilitation]; Randomized Controlled Trials as Topic

MeSH check words

Humans



Documents

Physiotherapy for Parkinson's disease a comparison of techniques