SA JOURNAL OF PHYSIOTHERAPY 2002 VOL 58 NO 3          3

(Hoffmann et al 1994). Outcomes research
has emerged from a growing concern
about which medical treatments work
best and for whom (Andersson &
Weinstein 1994), and it is a focus of
attention which the physiotherapy pro-
fession needs to acknowledge and
address. Outcomes research is an analysis
of clinical practice as it actually occurs
for the purpose of determining effec-
tiveness of clinical methods (Domholdt
2000).  This research is affordable, highly
feasible in the clinical situation and has
the potential to improve the quality of
research into the clinical management of
LBP. It would appear to be a valuable
tool in determining the best use of limited
resources for the management of a 
condition that places enormous demands
on health-care budgets. Applied in the
clinical situation, outcomes research has
the potential to play a role in evaluating
whether beneficial outcomes are due to
physiotherapeutic interventions (Bardin
1998), or to time, chance or coincidental
occurrences (Twomey 1990).  Further-
more, the results and conclusions from
outcomes research are likely to contribute
to further research questions, the for-
mulation of hypotheses and the planning
of randomised, controlled trials (RCT’s)

ABSTRACT: Low back pain (LBP) is one of the most common and costly
conditions treated by physiotherapists and is acknowledged as a major
health problem. Much published research on LBP is of poor design and 
optimal outcome measures are not selected for LBP patients. Effective and
cost-effective interventions for LBP, particularly chronic LBP, need to be
identified using appropriate, valid, reliable and responsive measures of 
outcome.  These outcome measures should reflect the biopsychosocial model
necessary for evaluating the broad impact of LBP, in particular chronic LBP, on a patient’s life. Outcomes research is
a feasible and affordable analysis of clinical practice as it occurs, and provides an opportunity to evaluate the effec-
tiveness of interventions for LBP. This is in contrast to a randomised, controlled trial (RCT) that evaluates efficacy
under controlled conditions that often do not reflect clinical practice. Using a battery of outcome measures appropriate
for measuring change in the LBP population, outcomes research has the potential to identify effective and cost-effec-
tive interventions, promote and influence further research, and contribute to the demand for evidence-based practice. 

KEY WORDS: LOW BACK PAIN, OUTCOMES RESEARCH, EVIDENCE-BASED PRACTICE,
OUTCOME MEASURES, BIOPSYCHOSOCIAL MODEL, EFFECTIVENESS, EFFICACY

PHYSIOTHERAPY AND LOW BACK PAIN - PART 1:
OUTCOMES RESEARCH IN THE

QUEST FOR EVIDENCE

BARDIN LD,
Cert. OMT (SA), M.Sc. (Physio.)1

1
Private practitioner, Park Orchards Physiotherapy
and Ringwood Physiotherapy and Spinal clinic,
Melbourne, Australia.

CORRESPONDENCE TO:
L Bardin
Park Orchards Physiotherapy,
616 Park Road, Park Orchards,
Melbourne, Victoria, 3114, Australia.
Tel: 9-6-13-9876196
Fax: 9-6-13-98790711                            
E-mail: lynnbardin@yahoo.co.uk

L I T E R A T U R E
R E V I E W

INTRODUCTION
Low back pain (LBP) is the most 
common diagnosis for which patients
are treated in outpatient physiotherapy
settings (Jette and Davis 1991), and
70%-80% of people in Western society
have at least one episode of LBP in a
lifetime (Biering-Sorensen 1983). It is 
a well-known major health problem that
causes the greatest disability in patients
under the age of 45 years (Deyo 1983),
as well as the greatest disability in role
function (Jette and Jette 1996). With
high prevalence and recurrence rates
(Croft et al 1997; Taimela et al 1997), it
has enormous implications for health-
care. Furthermore, there appears to be an
associated increase in chronic incapacity
and up to 35% of those with LBP develop
a chronic problem (Taimela et al 1997),
which further adds to the burden of
providers and purchasers of health-care.
Thus LBP, in particular chronic/recurrent
low back pain (CLBP), is a growing
problem that places increasing demands
on health budgets and has raised ques-
tions about the validity of current treat-
ments (Waddell 1998).   

Much of the published research on
therapies for LBP is flawed by poor
design and suboptimal outcome measures

currently regarded as the gold standard
of evidence and the only ‘scientific’
way of providing evidence of effective
intervention and management (Mawson
2002).

OUTCOMES RESEARCH AND
EVIDENCE-BASED PRACTICE
Although RCTs are regarded as the most
robust form of evidence (Sackett et al.
1996), it has been suggested that they
are not a universally attainable or even
desirable gold standard of evidence in
physiotherapy (Bithell 2000). They may
also not be suitable for rehabilitation
(Mawson 2002), which is highly relevant
in the approach to management of
CLBP (Bardin 1998 and 2000b).  The
Research Committee (Victorian Branch)
of the Australian Physiotherapy Asso-
ciation (1999) recently advocated that



4          SA JOURNAL OF PHYSIOTHERAPY 2002 VOL 58 NO 3

the use of evidence concerning efficacy
derived under the restricted, experi-
mental conditions of a RCT requires
clinical skill, judgement and experience
for application to individual patients.
Because of the large number of variables
that can impact on a clinical intervention,
the RCT may not always be the best
methodology for evaluating the efficacy
of many of the modalities and multi-
modalities used in clinical practice
(Moore and Petty 2001). It has been 
suggested that within the chronological
growth of the health professions, evi-
dence-based practice is still in it’s 
infancy (Moore and Petty 2001).  The
RCT is at the top of the hierarchy of 
evidence, which ranges from clinical
experience up to systematic review of
RCTs (Fig.1). It has further been sug-
gested that there is a paradox regarding
the RCT. It is recognised as the best way
to assess whether an intervention works,
but is arguably the worst way to assess
who will benefit from it (Mant 1999).  

Few RCTs exist in the evaluation of
orthopaedic procedures as ethical and
practical issues limit the possibility of
randomisation.  Outcomes research is an
alternative clinical investigation where
the balance of rigour and relevance is
more easily attained.  Liang et al (1994)
defined outcomes research as research
on the management of patients that asks
what treatment is effective, and for
whom, in more realistic settings than

ones used in RCTs.  Outcomes research
provides a feasible analysis of clinical
practice as it occurs (Bardin 2000a).
Domholdt (2000) suggests that research
questions should arise from problems
experienced in the clinical practice of
physiotherapy. Outcomes research can
be conducted in everyday clinical phy-
siotherapy settings and, prospectively
planned, is well suited to raise and
explore various aspects of clinical
research that can contribute to the body
of evidence from grade V to III levels
(Fig. 1). 

In the quest for credibility of the
value of treatment for LBP, the role of
outcomes research should be considered,
possibly as a forerunner to develop
research questions, to test hypotheses
and to refine aspects of the research
process.   Preliminary research using the
outcomes research model should have
the potential to improve the quality of
research implemented in RCTs, which
are often difficult and expensive to 
conduct. In spite of difficulties in the
application of the evidence of efficacy
(Australian Physiotherapy Association,
Victorian Branch 1999) and of to whom
to apply treatment, the RCT is likely 
to continue to be acknowledged as the
highest level of evidence.  For this reason
it is of cardinal importance that future
protocols for this rigorous research 
are based on appropriate preliminary
research, of which outcomes research 

is one example.  It would appear that
outcomes research has a significant role
to play in our endeavours to refine the
quality of scientific inquiry into the
effective management of LBP. 

MEASURING THE OUTCOME AND
ASSESSING EFFECTIVENESS
Physiotherapists need to adopt outcome
measures that will document effective-
ness of interventions and guide clinical
decisions and management. In outcomes
research effectiveness is the usefulness
of a particular treatment to the indivi-
dual(s) receiving it, under typical cli-
nical conditions usually determined by
nonexperimental methods (Domholdt
2000). In contrast, clinical efficacy is the
biological effect of treatment under
carefully controlled conditions, usually
determined by RCT.  Measuring the
effectiveness of our treatments is vital to
the survival of the physiotherapy pro-
fession and to satisfy the demands for 
evidence-based practice.  It is only pos-
sible if we measure change by careful
documentation of the outcome of care
provided in clinical practice. Although it
has been suggested that evidence-based
physiotherapy is possibly seeking the
unattainable (Clemence 1998), the pur-
suit of best evidence is of paramount
importance to our future credibility. It 
is not optional if the profession is to
compete in the health-care arena for the
provision of evidence-based treatments

In
cr

ea
si

ng
 s

tr
en

gt
h 

of
ev

id
en

ce

Strong evidence from systematic review of multiple RCTs

Strong evidence from RCT of appropriate size

Evidence from well-designed trials without randomisation

Well-designed non-experimental studies from more
than 1 centre

Opinions of respected authorities, based on clinical evidence,
descriptive studies or reports of expert committees

I

II

III

IV

V

Figure 1: Hierarchy of evidence ranging from clinical experience up to systematic review of RCTs. A broad contribution from levels 
V - III is possible through outcomes research. (Modified from Moore and Petty 2001)



SA JOURNAL OF PHYSIOTHERAPY 2002 VOL 58 NO 3          5

for LBP.  Outcomes research using appro-
priate outcome measures in a typical
clinical setting is attainable, affordable
and should provide a measure of evi-
dence. It will undoubtedly emphasise
accountability for standards of care and
selection of treatment intervention, and
is likely to promote and influence fur-
ther research.

What is an outcome measure?
Broadly speaking, an outcome measure
quantifies a change in the patient’s status
over time (Kane 1994), and makes it
possible to examine the end results of
health-care and the effects of the health
care process on the well-being of either
individual patients or whole populations
(Andersson and Weinstein 1994).  It has
also been described as a measure of
change, the difference from one point in
time (usually before an intervention) to
another point in time (usually following
an intervention) (Kendall 1997).  Mayo
et al (1994) further specify a physiothe-
rapy outcome measure as a test or 
scale, administered and interpreted by
physiotherapists, that has been shown to 
measure accurately a particular attribute
of interest to patients and therapists 
and is expected to be influenced by the
intervention.

Purpose of outcome measures
The physiotherapy profession is being
challenged to record the outcome of 
care provided to patients.  The purpose of
using outcome measures is an attempt 
to define closely the subjective and
objective information we usually record,
but in a more robust manner (Hammond
2000).  This can be achieved by the use
of appropriate and valid outcome mea-
sures which, by providing reliable docu-
mentation, will guide clinical decisions,
treatment planning and prove effective-
ness of interventions in a clinical setting
(Bardin 1998).  

Measuring outcomes enables the
clinician to determine:
• The impact of an overall treatment on

an individual
• The impact of a specific treatment

approach on an individual
• The overall impact of care on all

clients within a programme
• Productivity - i.e. the relationship

between resources used and treatment
outcomes is assessed in order to eva-
luate the effectiveness and efficiency
of the service provided
It has been suggested that, in order to

assess progress in patient management,
important outcomes to measure are those
of pain relief and disability (Liebenson
and Yeomans 1997). Reduction in LBP,
improvement in function and prevention
of disability are core aspects of physio-
therapy management of LBP.  It is impor-
tant for the credibility of our management
approach to LBP that we measure the
effect of our treatments on these outcomes.

Selection of outcome measures
Outcome measures should be clinically
appropriate, functionally relevant, valid,
reliable and responsive to clinical change
(Schoppink et al 1996; Bardin 1997;
Bardin 1998). Pain scales useful in the
clinical situation, for example, are the
Numerical Rating Scale, Visual Analogue
Scale and McGill Pain Questionnaires
(MPQ) (long- and short-forms) (Jensen
et al 1986; Melzack 1975). The Roland
Morris Disability Questionnaire (Roland
and Morris 1983), Aberdeen Back Pain
Scale (Ruta et al 1994) and Oswestry
Disability Questionnaire (Fairbank et al
1980) for pain and functional impair-
ment/disability, and the condition-
specific functional status questionnaire
for lumbar spinal stenosis (Stucki et al
1996) are examples of questionnaires
measuring the broad biopsychosocial
impact of interventions for LBP.  Bouter
et al (1998) suggest that researchers
identify the factors to be measured and
measure them with high-quality, stan-
dardised parameters with explicit instruc-
tions for administration and scoring.
This will enable practitioners to make
comparisons about the health status of
patients at the end of their intervention.
Several authors suggest that outcome
measures need to be inexpensive and
time-efficient to minimize the burden 
to health professionals and patients
(Liebenson and Yeomans 1997; Deyo et
al 1998; Wright et al 1998), while
Hammond (2000) is of the opinion that
in order to be acceptable to patients 
they should be comfortable and painless.
A spectrum of outcomes measures
reflecting the biopsychosocial model

will be reviewed in Physiotherapy and
low back pain: Parts II and III.

OUTCOMES RESEARCH IN THE ASSESSMENT
OF THE QUALITY OF HEALTH CARE
Outcomes research is a focus of atten-
tion on the assessment of health-care
quality that the physiotherapy profes-
sion needs to acknowledge and address.
It has emerged from a growing concern
about which medical treatments work
best and for whom (Andersson and
Weinstein 1994), and will assist analysis
of the relative effectiveness and cost-
effectiveness of different interventions
for LBP.  Furthermore it should facilitate
physiotherapists to identify subgroups 
to be treated (Bardin 1998 and 2000c;
O’Sullivan 2000), and to provide pur-
chasers of health-care with evidence 
of effectiveness (Chesson et al 1996;
Watson 1999). 

Health-care consumers want evidence
of effectiveness when choosing what to
purchase.  In the context of cost contain-
ment, there is a need to analyse the rela-
tive effectiveness of different interventions
especially for recurrent/chronic LBP
(Bardin 1998 and 2000b).

Outcomes research will facilitate the
practice of outcomes management which
relies on an analysis of outcomes to
examine the cost and quality of health-
care.  These two inter-related issues are
of great interest to clinicians as well 
as to government health departments,
insurance departments, insurance compa-
nies and consumers.  Outcomes manage-
ment offers great potential for identifying
client needs, improving outcomes, utilis-
ing evidence-based treatments and allo-
cating scarce resources (Hazard 1995;
Mayer et al 1995; Bardin 1998), and is
considered a crucial component in pro-
moting quality without sacrificing cost.

Outcomes research, as defined by
Liang et al (1994), focuses on investi-
gating what treatments are effective in
realistic clinical settings (in contrast to
RCTs) and for which groups of patients.
These researchers suggest that the
emphasis should be on the patient’s
assessment of pain, function, quality of
life and satisfaction with the results of
the intervention.  Outcomes research can
be described as research that evaluates
the broad biopsychosocial impact of



6          SA JOURNAL OF PHYSIOTHERAPY 2002 VOL 58 NO 3

treatments implemented in typical clinical
settings (Keller et al 1994; Bradham
1994; Bardin 2000a).  This is in sharp
contrast to a clinical efficacy study
where the biological effect of treatment
is monitored under the carefully con-
trolled conditions possible with labora-
tory research.  A recent study by Bardin
(1998 and 2000b) is an example of out-
comes research.  This research included
inter alia patients’ assessments, as sug-
gested by Liang et al (1994), to evaluate
the effectiveness of a spinal rehabilitation
programme utilising patient education
and group exercises for the treatment of
selected patients with chronic/recurrent
LBP.  Results from this study showed
significant reduction in pain and handi-
cap, as well as significant improvement
in function and flexibility following
patient participation over a one-month
period. Importantly, these improvements
in outcomes were significantly main-
tained at the three and six month follow-
ups and there was an exceptionally
high (91%) level of patient satisfaction
with this cost-effective approach to the
management of CLBP. Domholdt (2000)
suggests that in outcomes research prac-
titioners and researchers supplement
measures of pathology and impairment
with person-level measures of functional
limitation, disability or handicap. This
type of research is feasible in most clini-
cal situations and is necessary to deter-
mine a valid approach to the manage-
ment of LBP, particularly CLBP. In a
study on the outcome of 438 patients
treated surgically for lumbar spinal
stenosis, the authors acknowledged that
the most important shortcoming of their
study was that the initial levels of the
patients’ pain, depression and disability
were not known (Olavi et al 1997). This
acknowledged shortcoming illustrates
that prospectively planned documen-
tation of outcomes is essential if we are
to gather even lower levels of evidence,
without which the higher levels may not
achieve an appropriate focus of inquiry.

Outcomes research is a feasible,
affordable analysis of the outcome of
health-care that utilises measures reflect-
ing the biopsychosocial model necessary
for evaluating the impact of LBP, in 
particular CLBP, on a patient’s life.
Outcome measures that reflect the bio-

psychosocial model are important to
evaluate change in LBP patients (Bardin
1998; Cherkin 1998; Deyo et al 1998) as
objective outcome measures related to
the biological component have been
found to be only weakly correlated with
patient behaviour or symptoms. Health
related quality of life measures, which
assess multiple dimensions of life, are
thought to be an inexpensive, valuable
source of quantitative data for quality
assurance and research of the CLBP
patient.

The physiotherapy profession should
respond positively to the challenge of
analysing the quality of our care for acute
and chronic LBP patients.  Outcomes
research is feasible and affordable.  It
should be enthusiastically embraced as
an opportunity to demonstrate our scien-
tific foundations, to promote quality of
care and to provide a necessary and
valuable contribution to our quest for
evidence-based practice. 

CONCLUSION
The high prevalence, recurrence rates
and enormous cost to society of LBP,
particularly CLBP, places a responsibi-
lity on the physiotherapy profession 
to provide evidence for management
approaches to LBP patients and for
effectiveness of interventions.  In the
quest for evidence, physiotherapists can
utilise outcomes research - an analysis
of clinical practice that focuses on
patients’ assessments of pain, function,
quality of life and satisfaction with the
results of the intervention. Prospectively
planned documentation of outcomes
using valid, reliable and responsive out-
come measures is necessary to provide
evidence of effectiveness, and to con-
tribute to further research in the field 
of LBP.  These outcome measures should
be clinically appropriate, functionally
relevant and reflect the broad biopsy-
chosocial impact occurring in the LBP
patient as a result of this disabling con-
dition. The RCT, conducted under con-
trolled conditions, is the highest form 
of evidence, however treatment under
these conditions often does not reflect
clinical practice and the application of
evidence may be difficult. The role of
outcomes research to assess health-care,
to evaluate effectiveness and  cost-effec-

tiveness, and to promote further research
and scientific inquiry should be consid-
ered. It is a feasible and affordable option
to complement RCTs in the demand 
for evidence-based practice. Outcomes
research of LBP is an important aspect
of clinical practice that will assist the
physiotherapy profession to meet the
demand for evidence-based practice, to
demonstrate our scientific foundations
and to validate that this profession has
now come of age.

REFERENCES

Andersson GBJ, Weinstein JN 1994 Outcomes
related to low back pain. Spine 19: 2026S -
2027S

Australian Physiotherapy Association, Victorian
Branch Research Committee and invited 
contributors 1999. Evidence-based practice.
Australian Journal of Physiotherapy 45:
167- 171 

Bardin LD 1997 Outcome measurements in
spinal rehabilitation - a case report.  
Proceedings of International congress of
South African Society of Physiotherapy, Cape
Town, South Africa, April 1997

Bardin LD 1998 A study to evaluate the effec-
tiveness of a total back care programme for
group rehabilitation of selected patients with
chronic low back pain. MSc thesis, University
of Stellenbosch, South Africa

Bardin LD 2000a Outcomes research of a
spinal rehabilitation programme for chronic/
recurrent low back pain.  Proceedings of the
South African Spine Society Congress, Cape
Town, June 2000

Bardin LD 2000b A Spinal Rehabilitation
Programme for selected patients with recur-
rent/chronic low back pain:  a one year prospec-
tive study.  Proceedings of the 7th Scientific
Conference of the International Federation of
Orthopaedic Manipulative Therapists, Perth,
Australia, November 2000

Bardin LD 2000c Spinal rehabilitation of
chronic low back pain using patient education
and group exercise therapy. Course manual,
pre-conference course, 7th Scientific Conference
of the International Federation of Orthopaedic
Manipulative Therapists, Perth, Australia,
November 2000

Biering-Sorensen F 1983 A prospective study
of low back pain patients in a general popu-
lation: I. Occurrence, recurrence and aetiology.



SA JOURNAL OF PHYSIOTHERAPY 2002 VOL 58 NO 3          7

Scandinavian Journal of Rehabilitation Medi-
cine 15: 71-79

Bithell C 2000 Evidence-based physiotherapy.
Physiotherapy 86: 58 - 60

Bouter LM, van Tulder M, Koes B 1998.
Methodologic Issues in Low Back Pain
Research in Primary Care. Spine 23: 2014 -
2019

Bradham DD 1994 Outcomes research in
orthopaedics: history, perspectives, concepts
and future.  Arthroscopy 10: 493 - 501

Cherkin DC 1998 Primary Care Research on
Low Back Pain. Spine 23: 1997-2002

Chesson R, Macleod M, Massie S 1996
Outcome measures used in therapy departments
in Scotland.  Physiotherapy 82: 673-679

Clemence ML 1998 Evidence-based physio-
therapy: seeking the unattainable? British
Journal of Therapy and Rehabilitation 5: 
257-260

Croft PR, Papageorgiou AC, McNally R 1997
Low back pain: Healthcare needs assessment.
Radcliffe Medical Press, Oxford

Deyo RA 1983 Conservative therapy for 
low back pain: distinguishing useful from use-
less therapy.  Journal of the American Medical
Association 250: 1057-1062

Deyo RA, Battie M, Beurskens AJHM,
Bombardier C, Croft P, Koes B, Malmivaara
A, Roland M, Von Korff M, Waddell G 
1998 Outcome Measures for Low Back Pain
Research: a proposal for standardised use.
Spine 23: 2003-2013

Domholdt E 2000 Physical Therapy Research.
WB Saunders, Philadelphia

Fairbank J, Couper J, Davies JB, O’Brien JP
1980 The Oswestry low back pain question-
naire.  Physiotherapy 66: 271-273 

Hammond R 2000 Evaluation of physiothe-
rapy by measuring outcome. Physiotherapy 86:
170-172

Hazard RG 1995 Spine update: functional
restoration.  Spine 20: 2345-2348

Hoffmann RM, Turner JA, Cherkin DC, Deyo
RA, Herron LD 1994 Therapeutic trials for
low back pain.  Spine 19: 2068S - 2075S

Jensen MP, Karoly P, Braver S 1986 The 
measurement of clinical pain intensity: a 
comparison of six methods. Pain 27: 117-126

Jette AM, Davis KD 1991 A comparison of
hospital-based and private outpatient physical
therapy practices.  Physical Therapy 74: 366-375
Jette DU, Jette AM 1996 Physical therapy 

and health outcomes in patients with spinal
impairments. Physical Therapy 76: 930-945 

Kane R 1994 Looking for physical therapy
outcomes. Physical Therapy 74: 425-429

Keller RB, Rudicel SA, Liang MH 1994
Outcomes research in orthopaedics.   Instruc-
tional course lectures 43: 599 - 611

Kendall N 1997 Developing outcome assess-
ments: A step-by-step approach.  New Zealand
Journal of Physiotherapy, December 11-17

Liang MH, Andersson GBJ, Bombardier MD,
Cherkin DC, Deyo MD 1994 Strategies for
outcomes research in spinal disorders. Spine
19: 2037S-2040S 

Liebenson C, Yeomans S 1997 Outcomes
assessment in musculoskeletal medicine.
Manual Therapy 2: 67 - 74

Mant D 1999 Can randomised trials inform
clinical decisions about patients? Lancet 353:
743-746

Mawson S 2002 Teler versus MAS: validating
TELER indicator definitions for use in the
measurement of physiotherapy outcomes in
acute stroke rehabilitation. Physiotherapy 88:
67-76

Mayer TG, Polatin P, Smith B 1995 Contem-
porary concepts in spine care: spine rehabili-
tation- secondary and tertiary non-operative
care.  Spine 20: 2060-2066  

Mayo N, Cole B, Dowler J, Gowland C, Finch
E 1994 Uses of outcome measures in physio-
therapy:  Survey of current practice.  Canadian
Journal of Rehabilitation 81:82

Melzack R 1975 The McGill Pain Question-
naire: major properties and scoring methods.
Pain 277-299

Moore A and Petty N 2001 Evidence-based
practice - getting a grip and finding a balance.
Manual Therapy 6: 195-196

Olavi A, Arto H, Veli T, Tapani S, Olavi S
1997 Surgical outcome of 438 patients treated
surgically for lumbar spinal stenosis. Spine
22: 2278-2282

O’Sullivan PB 2000 Lumbar segmental 
‘instability’: clinical presentation and specific
stabilising exercise management. Manual
Therapy 5: 2-12

Roland M, Morris R 1983 A study of the 
natural history of low back pain: I Development
of a reliable and sensitive measure of disabi-
lity in low back pain. Spine 1983: 141-144
Ruta DA, Garratt AM, Wardlaw D, Russell IT
1994 Developing a valid and reliable measure

of health outcome for patients with low back
pain. Spine 19: 1887-1896

Sackett DL, Rosenberg WMC, Gray JAM,
Haynes BR, Richardson WS 1996 Evidence
based medicine: what it is and what it isn’t.
British Medical Journal 312: 71-72

Schoppink  LEM, Van Tulder HW, Koes BW,
Behrskens SAJHM, De Bie RA 1996
Reliability and validity of the Dutch adap-
tation of the Quebec Back Disability Scale.
Physical Therapy 76: 268 - 275

Stucki G, Daltroy L, Liang M, Lipson S, Fossel
A, Katz JN 1996 Measurement properties of a
self-administered outcome measure in lumbar
spinal stenosis. Spine 21: 796-803

Taimela S, Kujala UM, Salminem JJ, Viljanen T
1997 The prevalence of low back pain among
children and adolescents: A nationwide, cohort-
based questionnaire survey in Finland. Spine
22: 1132-1136

Twomey L 1990 A growing commitment to
research and evaluation.  Australian Journal of
Physiotherapy 36:83.

Waddell G 1998 The Back Pain Revolution,
Churchill Livingstone, London.

Watson PJ 1999 Psychosocial Assessment.
Physiotherapy 85: 530-535

Wright J, Cross J, Lamb S 1998 Physiotherapy
outcome measures for rehabilitation of elderly
people. Physiotherapy 84: 216-221