open university press handbook of research methods in health jun 2005

Handbook of
Health research methods
Investigation, measurement and analysis
Edited by
Ann Bowling and Shah Ebrahim
HHAANNDDBBOOOOKK OOFF HHEEAALLTTHH RREESSEEAARRCCHH MMEETTHHOODDSS
Investigation, Measurement and Analysis
• Which research method should I use to evaluate services?
• How do I design a questionnaire?
• How do I conduct a systematic review of research?
This handbook helps researchers to plan, carry out, and analyse health
research, and evaluate the quality of research studies. The book takes
a multidisciplinary approach to enable researchers from different
disciplines to work side-by-side in the investigation of population
health, the evaluation of health care, and in health care delivery.
Handbook of Health Research Methods
is an essential tool for
researchers and postgraduate students taking masters courses, or
undertaking doctoral programmes, in health services evaluation,
health sciences, health management, public health, nursing, sociolo-
gy, socio-biology, medicine and epidemiology. However, the book
also appeals to health professionals who wish to broaden their knowl-
edge of research methods in order to make effective policy and prac-
tice decisions.
CCoonnttrriibbuuttoorrss::
Joy Adamson, Geraldine Barrett, Jane P. Biddulph, Ann Bowling, Sara
Brookes, Jackie Brown, Simon Carter, Michel P. Coleman, Paul
Cullinan, George Davey Smith, Paul Dieppe, Jenny Donovan, Craig
Duncan, Shah Ebrahim, Vikki Entwistle, Clare Harries, Lesley
Henderson, Kelvyn Jones, Olga Kostopoulou, Sarah J. Lewis, Richard
Martin, Martin McKee, Graham Moon, Ellen Nolte, Alan O’Rourke, Ann

Oakley, Tim Peters, Tina Ramkalawan, Caroline Sanders, Mary Shaw,
Andrew Steptoe, Jonathan Sterne, Anne Stiggelbout, S.V.
Subramanian, Kate Tilling, Liz Twigg, Suzanne Wait.
AAnnnn BBoowwlliinngg
is Professor of Health Services Research in the
Department of Primary Care and Population Sciences at University
College London, and has a part secondment to the MRC Health
Services Research Collaboration, University of Bristol. Her other
publications with Open University Press include:
Measuring Disease
(2001),
Research Methods in Health
(2002) and
Measuring Health
(2004).
SShhaahh EEbbrraahhiimm
is Professor of Epidemiology of Ageing and Head of the
Department of Social Medicine, University of Bristol, and also Honorary
Professor of Epidemiology at University College London. He is coordi-
nating editor of the Cochrane Heart Group and co-editor of the
International Journal of Epidemiology
. He is widely published in the
fields of geriatric medicine, stroke and cardiovascular epidemiology.
Cover design: Kate Prentice
9 780335 214600
ISBN 0-335-21460-6
Handbook of Health research methods
Bowling and Ebrahim
BowEbrah606pb.qxd 1/30/2007 2:23 PM Page 1
HANDBOOK OF HEALTH

RESEARCH METHODS

HANDBOOK OF HEALTH
RESEARCH METHODS
Investigation, measurement and
analysis
Edited by Ann Bowling and
Shah Ebrahim
Open University Press
Open University Press
McGraw-Hill Education
McGraw-Hill House
Shoppenhangers Road
Maidenhead
Berkshire
England
SL6 2QL
email:
world wide web: www.openup.co.uk
and Two Penn Plaza, New York, NY 10121-2289, USA
First published 2005
Copyright © Ann Bowling and Shah Ebrahim 2005
Individual chapters © The Contributors 2005
All rights reserved. Except for the quotation of short passages for the purpose of
criticism and review, no part of this publication may be reproduced, stored in a
retrieval system, or transmitted, in any form or by any means, electronic, mechanical,
photocopying, recording or otherwise, without the prior written permission of the
publisher or a licence from the Copyright Licensing Agency Limited. Details of
such licences (for reprographic reproduction) may be obtained from the Copyright
Licensing Agency Ltd of 90 Tottenham Court Road, London, W1T 4LP.

A catalogue record of this book is available from the British Library
ISBN–10: 0 335 21460 6 (pb) 0 335 21461 4 (hb)
ISBN–13: 978 0335 214600 (pb) 978 0335 2146 1 7 (hb)
Library of Congress Cataloging-in-Publication Data
CIP data applied for
Typeset by RefineCatch Limited, Bungay, Suffolk
Printed in the UK by Bell and Bain Ltd, Glasgow
Contents
List of contributors ix
Preface xi
Part 1 Introduction 1
1 Research on health and health care 3
Paul Dieppe
2 Describing and evaluating health systems 12
Ellen Nolte, Martin McKee and Suzanne Wait
Part 2 Multidisciplinary methods of investigation 45
3 Evidence-based health care: systematic reviews 47
Paul Cullinan
4 Critical appraisal 62
Alan O’Rourke
5 Features and designs of randomized controlled trials and
non-randomized experimental designs 85
Kate Tilling, Jonathan Sterne, Sara Brookes and Tim Peters
6 Epidemiological study designs for health care research and
evaluation 98
Richard M. Martin
7 Finding and using secondary data on the health and health
care of populations 164
Mary Shaw
8 Quantitative social science: the survey 190

Ann Bowling
9 Approaches to qualitative data collection in social science 215
Simon Carter and Lesley Henderson
10 Combined qualitative and quantitative designs 230
Joy Adamson
11 Design and analysis of social intervention studies in health
research 246
Ann Oakley
12 Area-based studies and the evaluation of multilevel influences
on health outcomes 266
Graham Moon, S.V. Subramanian, Kelvyn Jones, Craig Duncan and
Liz Twigg
13 Mathematical models in health care 293
Jane P. Biddulph
14 Economic evaluation of health care 314
Jackie Brown
Part 3 Multidisciplinary research measurement 329
15 Psychological approaches to measuring and modelling
clinical decision-making 331
Clare Harries and Olga Kostopoulou
16 Approaches to measuring patients’ decision-making 362
Clare Harries and Anne Stiggelbout
17 Techniques of questionnaire design 394
Ann Bowling
18 Measuring health outcomes from the patient’s perspective 428
Ann Bowling
19 Genetics, health and population genetics research 445
Sarah J. Lewis, George Davey Smith and Shah Ebrahim
20 Tools of psychosocial biology in health care research 471
Andrew Steptoe

Part 4 Data analysis 495
21 Key issues in the statistical analysis of quantitative data in
research on health and health services 497
Kate Tilling, Tim Peters and Jonathan Sterne
vi Contents
22 Key issues in the analysis of qualitative data in health services
research 515
Jenny Donovan and Caroline Sanders
Part 5 Essential issues to consider when conducting research 533
23 Involving service users in health services research 535
Vikki Entwistle
24 Ethical and political issues in the conduct of research 555
Geraldine Barrett and Michel P. Coleman
25 Training for research 584
Tina Ramkalawan
General glossary 601
Index 611
Contents vii

List of contributors
Dr Joy Adamson, Department of Health Sciences, University of York, UK.
Dr Geraldine Barrett, Department of Health and Social Care, Brunel University, UK.
Dr Jane P. Biddulph, Department of Primary Care and Population Sciences,
University College London, University of London, UK.
Professor Ann Bowling, Department of Primary Care and Population Sciences,
University College London, University of London, UK.
Ms Sara Brookes, Department of Social Medicine, University of Bristol, UK.
Dr Jackie Brown, MRC Health Services Research Collaboration, Department of
Social Medicine, University of Bristol, UK.
Dr Simon Carter, Sociology Group, Department of Public Health and Policy,

London School of Hygiene and Tropical Medicine, University of London,
UK.
Professor Michel P. Coleman, Non-Communicable Disease Epidemiology Unit,
London School of Hygiene and Tropical Medicine, University of London, UK.
Dr Paul Cullinan, Department of Occupational and Environmental Medicine,
National Heart and Lung Institute, Imperial College London, University of
London, UK.
Professor George Davey Smith, Department of Social Medicine, University of Bristol,
UK.
Professor Paul Dieppe, MRC Health Services Research Collaboration, Department
of Social Medicine, University of Bristol, UK.
Professor Jenny Donovan, Department of Social Medicine, University of Bristol, UK.
Dr Craig Duncan, Institute for the Geography of Health, University of Portsmouth,
UK.
Professor Shah Ebrahim, Department of Social Medicine, University of Bristol, UK.
Dr Vikki Entwistle, Health Services Research Unit, Department of Public Health,
University of Aberdeen, UK.
Dr Clare Harries, Department of Psychology, University College London,
University of London, UK.
Dr Lesley Henderson, Department of Human Sciences, Brunel University, UK.
Professor Kelvyn Jones, School of Geographical Sciences, University of Bristol, UK.
Dr Olga Kostopoulou, Department of Primary Care and General Practice, University
of Birmingham, UK.
Dr Sarah J. Lewis, Department of Social Medicine, University of Bristol, UK.
Dr Richard Martin, Department of Social Medicine, University of Bristol, UK.
Professor Martin McKee, European Centre on Health of Societies in Transition,
London School of Hygiene and Tropical Medicine, University of London, UK.
Professor Graham Moon, Institute for the Geography of Health, University of
Portsmouth, UK.
Dr Ellen Nolte, European Centre on Health of Societies in Transition, London

School of Hygiene and Tropical Medicine, University of London, UK.
Dr Alan O’Rourke, Institute of General Practice, School of Health and Related
Research (ScHARR), University of Sheffield, UK.
Professor Ann Oakley, Social Science Research Unit, Institute of Education,
University of London, UK.
Professor Tim Peters, Academic Unit of Primary Health Care, Department of
Community Based Medicine, University of Bristol, UK.
Dr Tina Ramkalawan, MRC Health Services Research Collaboration, Department
of Social Medicine, University of Bristol, UK.
Ms Caroline Sanders, Department of Social Medicine, University of Bristol, UK.
Dr Mary Shaw, Department of Social Medicine, University of Bristol, UK.
Dr Andrew Steptoe, Psychobiology Group, Department of Epidemiology and Public
Health, University College London, University of London, UK.
Dr Jonathan Sterne, Department of Social Medicine, University of Bristol, UK.
Dr Anne Stiggelbout, Department of Medical Decision Making, Leiden University,
The Netherlands.
Dr S.V. Subramanian, Harvard School of Public Health, Harvard University, USA.
Dr Kate Tilling, Department of Social Medicine, University of Bristol, UK.
Dr Liz Twigg, Institute for the Geography of Health, University of Portsmouth, UK.
Dr Suzanne Wait, Judge Institute of Management Studies, University of Cambridge,
UK.
x List of contributors
Preface
This book aims to assist researchers from clinical and non-clinical disciplines to
plan, carry out, analyse and evaluate research on population health, health outcomes
and health care delivery. A sound knowledge of research methods is important to all
professionals involved in health policy and the delivery of health care. It is increas-
ingly common for researchers from different disciplines to work together, and this
book also aims to provide insight into their different research perspectives and
methods. The focus of the book therefore reflects a multidisciplinary approach to

research that is relevant to a wide range of students and researchers.
The book includes an impressive number of authors, all of whom are active and
experienced investigators, with international reputations in their area of expertise.
While the length of each chapter varies, depending on its aims and subject matter,
the authors have each provided a comprehensive guide to their specialist topics,
pitched at a level suitable for a multidisciplinary readership. Where appropriate,
authors have included a list of further reading and resources to point the reader
towards more detailed material. It is hoped that this book will introduce readers to
research methodology across disciplines, and increase awareness of some of the
critical issues involved in investigating health and health services.
Ann Bowling and Shah Ebrahim

Part 1
Introduction

Research on health and
health care
Paul Dieppe
He who has choice has trouble.
(Dutch proverb)
Doubt is not a pleasant condition, but certainty is absurd.
(Voltaire)
All democratic governments are rightly concerned with the education and health of
the populations they serve. But modern democracies are finding it increasingly
difficult to maintain the health of their people in equitable and cost-effective ways.
They need research on the effectiveness, outcomes, access to and costs of the vary-
ing options available to us to improve the health of individuals and populations. This
book is about the methods needed to do that research. In this introductory chapter,
one of the principles behind all health problems, as well as health care research –
uncertainty – is explored.

Health, disease and illness
Health is difficult to define. Literally it means ‘wholeness’ and it can be thought of
as the ability of an individual to fulfil their potential. In practice, health has often
been used to denote the absence of disease and, as outlined below, modern health
services and health research have concentrated on the prevention or treatment of
disease rather than on health. But this may not be sufficient for our increasingly
demanding and wealthy western populations. They may prefer the World Health
Organization’s definition of health as ‘A state of complete physical, mental and
social wellbeing’ (WHO 1948). But this utopian state is difficult to achieve, except
perhaps fleetingly (Skrabanek and McCormick 1998).
Disease can be defined as an abnormality of the structure or function of the body
(a definition that raises the difficult question of what we think of as normal, and
how much diversity we are willing to accept). An illness is a symptom experience,
which can include features such as pain or distress, restriction of normal activities
(disability), or reduced ability to participate in life in the ways in which an indi-
vidual would like. It is now customary to consider disease and illness within a
‘biopsychosocial model’, which stresses the importance of environmental and per-
sonal factors, and their interactions with a disease on health. Disease and illness can
lead to sickness – which is the role played by people with illness in our society. But
the illness experience and sickness are not always caused by disease; they depend on
psychosocial factors as well. Health care professionals spend a lot of their time with
people with illness but not disease.
1
Therefore, health problems are about sensations that we consider to be less than
ideal, and/or about what our society regards as abnormalities of body or mind. They
also embody things that we see as an internal threat to our identity and ourselves.
An external threat can come from something like the perceived risk of a terrorist
attack, and can engender symptoms associated with disease, such a pain and anxiety.
An internal threat occurs when our body or mind appears to let us down or
malfunction, resulting in uncertainty about who or what we are and about our

future.
The WHO (1999) model (see Figure 1.1) suggests that a health condition (any
disease or illness) is one of the factors that dictates what one is able to do (activ-
ities) and how much one is able to take part in society (participation). A key health-
related variable is impairment (loss of a certain bodily function). Other contextual
factors also affect activities and participation; they include external (environ-
mental) factors, such as culture, housing and income, and internal (personal) factors
such as psychological status, motivation and educational attainments. All factors
interact.
Health care
As Horacio Fabrega (1997) and others have said, illness can be regarded as embody-
ing an increase in uncertainty among individuals and societies. Members of the
society try to reduce or eliminate that uncertainty through knowledge structures
related to disease, illness and healing. Two main systems emerge in most societies –
the informal, lay or folk systems of healing, and the formal or professional structures.
Most health care, in every country in the world, takes place within the informal
sector.
Societies are diverse, and often seem to need more than one type of professional
health care system to accommodate the varying views, priorities and approaches of
the different individuals within them. This has been termed ‘medical pluralism’ and
leads to a situation in which a sick individual can choose help from a system that fits
best with their own perceptions of the problem. However, over the last two centur-
ies, the biomedical model has become hugely dominant in western society, eclips-
ing most other concepts of health and health care. This dominance has been aided
by regulation and research. Alternatives are available through what we pejoratively
call ‘complementary and alternative medicine’ – but we now appear to be doing
Figure 1.1
Source: Adapted, with permission from The International Classification of Impairment,
Disability and Handicap (WHO 1999).
4 Introduction

our best to incorporate them into the biomedical model by carrying out scientific
research on them and rejecting those that do not seem to work in ways that
biomedicine understands. The result is three main systems of health care:
1 lay advice and folk remedies;
2 biomedicine, working within state-funded, regulated systems;
3 complementary and alternative traditions and healers, who tend to be marginal-
ized by the biomedical movement.
The biomedical revolution has been enormously successful. Many conditions
that would have been fatal in the past can now be treated successfully with modern
drugs and surgery. However, within this success story there appear to be two
significant problems:
1 our society remains uncertain about its health and health care;
2 the systems that we have built up around the biomedical model of health care are
not well equipped to deal with chronic illness.
The persisting uncertainty about health in our population is apparent from the
extent of the utilization of complementary and alternative practices; this often takes
place in conjunction with the utilization of modern biomedicine. We go to alterna-
tive practitioners, who work outside the biomedical model even though we do not
always think that they can ‘cure’ us in the conventional sense. We are looking for
something else in relation to our uncertainty about our health.
During a time of enormous advances in biomedicine and biomedical research,
and in the number and complexity of the interventions available to help people
with disease and illness, two major new trends have put increasing pressure on
western health systems:
1 the development of expensive technologies for health care delivery in place of
simple, cheaper options;
2 the rise in the prevalence of chronic disease.
The two issues are interrelated.
Take, for example, the case of kidney failure. Until the 1960s all you could offer
someone whose kidneys were failing was supportive care until they died. Then

along came dialysis, followed by kidney transplantation. These are great success
stories, allowing huge numbers of people to be kept alive for much longer. But they
are expensive, and they do not always ‘cure’ the individual with the kidney problem,
who may remain ill for years, and in need of human caring and expensive drugs, in
spite of the ‘successful’ renal transplantation.
Scientific research often results in more expensive options in health care, for
diagnosis as well as treatment. For example, we have replaced the humble, cheap
X-ray machine, housed in a van or shed and operated by a single radiographer, with
the expensive and complex magnetic resonance imaging suite in which you can
find an army of physicists, radiographers and radiologists.
Disease and illness varies in different cultures and continents, and changes over
time. In many parts of the world malnutrition, infection and injury are still the
dominant health problems. However, in the rich western countries chronic disease
is becoming the major issue. There are several reasons for this:
1 diminishing importance of malnutrition and infection;
2 increasing age of the population;
3 increasing prevalence of chronic disease risk factors such as obesity;
4 partial treatment of acute conditions (the ability to keep more people with
diseases alive for longer, but without completely ‘curing’ them);
Research on health and health care 5
5 the creation of chronic diseases without illness (hypertension for example – the
diagnosis of which can lead to great anxiety);
6 the invention of chronic illness without disease (e.g. awareness of environmental
hazards and the cultural belief that they are making you ‘ill’).
We are adapting quickly to the influx of modern new technologies in health care,
in part because of the love affair that health care professionals have developed for
expensive bits of equipment that they can use to try and help their patients. But we
are not adapting so well to the rise in chronic health problems, and health care
professionals are still taught more about the management of acute crises than they
are about chronic health care.

Our systems of hospitals and clinics was designed and set up to deal with acute
crises. They remain excellent at dealing with people who have a myocardial infarc-
tion, or get knocked down and fracture a leg. Triage takes place in a well-equipped
accident and emergency department, followed by referral to the specialist cardiolo-
gist or orthopaedic surgeon respectively, who then administers the necessary tech-
nology, according to current, evidence-based protocols. But these systems are not so
good at dealing with the person with chronic disease, or the individual who is ill, but
not diseased. Indeed, super-specialization and increasing fragmentation of health
care into ‘cells’ built around systems or diseases may be acting to the direct disadvan-
tage of such people. Although primary care remains strong in the UK and many
other countries, it is increasingly difficult to find a ‘general physician’ who can look
at all aspects of the health and care of someone with multiple, difficult problems.
Health care research
In the previous sections the concept of illness as ‘uncertainty’ (about your body,
your mind, your identity and your future) was introduced. Uncertainty is also the
underlying principle behind all research.
Although physicists may be able to predict the next eclipse of the sun with
astounding accuracy and very little uncertainty, in the biological and social sciences
uncertainty always remains high, however much evidence is brought to bear on the
problem. We do research to reduce the degree of uncertainty, but we cannot abolish
it. We do not ‘know’ anything for certain (sadly this concept seems lost on the
general public, who are encouraged to believe that medical scientists can come up
with definitive answers to their problems).
As Kerlinger (1986) pointed out, we get to know things in one of three main
ways:
1 by authority (we are told the ‘truth’ by someone we believe in);
2 by intuition (it is our judgement that, or it seems to stand to reason that,
something is true);
3 by scientific methods (which involve the key principle of self-correction).
The scientific method finds things out by research that is characterized by control

and replication:
• control means that the central observations or experiments take place in a known
framework, so that the causes of the results can be identified;
• replication means that if the work was repeated the answer should be the same.
In the physical sciences quantitative methods are pre-eminent: hypotheses are
constructed and tested in experiments that take place in tightly controlled condi-
tions, outcomes are measured with high precision, and the findings are always
6 Introduction
carefully replicated. The same positivist methodology is valuable in the biological
and social sciences, although a wide range of different quantitative methods and
approaches are needed to help acquire reliable, replicable data on complex biological
or health care systems. But the great complexity of these systems, the huge variations
in individual human behaviour and outcomes, and the positivist belief in a single
truth for all, can limit the value and applicability of classical quantitative methods if
used in isolation. For these reasons qualitative as well as quantitative research
methods have to be used in biological and social sciences, including health research.
Qualitative methods, such as the collection of narratives, interviews, focus groups
and ethnographic work, can provide rich insights into the experience of individuals,
the meaning and interpretation of those experiences, and the likely relationships
between different factors. Although often working in a hypothesis-free environ-
ment they can also be very helpful in the formation of new hypotheses to be tested
by quantitative methods. Qualitative research methods uncover a different type of
truth, but one that is no less important. Its limitations stem largely from problems of
replication and in the time taken to acquire and interpret the data.
The methods used to undertake health and health care research are diverse. But
all lead to the acquisition of new data, and share the need for storage, analysis and
interpretation of those data. Computers have revolutionized our ability to deal with
the vast quantities and diverse types of data obtained within health research, and to
increase output. The massive rise in research outputs over recent years has led to a
further division of health research into ‘primary’ and ‘secondary’ types. Primary

research involves observation and experiment to gain new data. Secondary research
involves finding and analysing research done by others.
Health and health care research are needed in order to reduce the uncertainty
associated with the diagnosis, treatment and delivery of health care to all of those
people in our society who are in need of it. There is a massive spectrum of
such research, ranging from laboratory investigations on single molecules in
tightly-controlled conditions, to observations on the complex behaviours within
populations or systems of health care delivery.
This book is concerned with applied research on health and health services,
rather than basic biological investigations. This spans a continuum from studies on
individuals to those on groups of subjects and finally on populations and systems. It
should be seen as a broad range of research techniques rather than a speciality or
discipline. It covers all aspects of health care, including prevention, diagnosis and
treatment and includes research on patients as well as healthy volunteers.
Health services research
The term ‘health services research’ (HSR) has been used to cover most of the
methods and approaches to research described in this book. In general terms, HSR
seeks knowledge and evidence that will lead to improvements in the delivery of
health care; it is not a distinct discipline or profession, rather it is a set of techniques
used in applied health research with the aim of improving health, health care and its
delivery – it covers a huge range of activities.
As part of an attempt to provide a framework for HSR, and help delineate its
borders with clinical research, audit and quality assurance (see below), the Medical
Research Council’s Health Services Research Collaboration and others have sug-
gested that the main purpose of HSR should be to attempt to integrate the four
main requirements of a good health service – i.e. that it should be effective, efficient,
equitable and acceptable – and to research methods of implementing such services.
Research on health and health care 7
Effectiveness
It is obvious that anything we try to do to improve health and health care should

work. And yet, many of the things we do probably have no effect. In medical
literature two different aspects of whether something ‘works’ or not are dis-
tinguished – efficacy, which is whether an intervention helps a clearly defined
group of people on whom it is tested, and effectiveness, which denotes the ability of
that intervention to work for everyone who might need it. This raises the issue of
whether the results of a trial (carried out using the scientific method, in carefully
controlled conditions) is generalizible to the whole population, as discussed later in
this book (see Chapter 5).
Efficiency
Efficiency means value for money. As interventions become increasingly expensive,
efficiency becomes increasingly important. Health economists have developed a
variety of ways of comparing the cost-effectiveness of different interventions and
strategies. One of HSR’s major challenges is to find out what strategies are likely to
have the greatest pay-off to society as a whole, as well as to sick individuals. For
example, it may be that we would be better off spending most of a country’s health
budget on the prevention of smoking and obesity, rather than on the provision of
services for people who have developed diseases that result from these problems.
But it is clear that we have not yet learnt how to control smoking and obesity, and
another aspect of HSR is research into how behavioural change might be achieved.
Equity
Although not a central issue in health care for all advanced nations, the concept that
health care should be available to everyone who needs it was one of the founding
principles behind the establishment of the UK National Health Service. But equity
of health services could mean a variety of different things: it could mean equity of
access, equity of opportunity or equity of outcome. In the UK, equity of access,
rather than equity of opportunity, has come to dominate the political debate.
Concerns about equity are growing, with increasing awareness of the problems of
diversity in society and the strong link between disadvantage and health problems.
Acceptability
Over the last few decades we have become increasingly aware of the need for any

intervention to be acceptable to the public, as well as it being effective and efficient.
This first came to prominence with the realization that many effective drugs pre-
scribed by doctors are not taken by the patients – opening up research on what is
now called adherence. This realization has turned into a modern movement in
health care with increasing emphasis given to patient choice, patient-centred care
and the empowerment of the public and patients to take control of their health and
health care. The lesson for HSR is that it needs to make sure that any recommended
new approach to health care is acceptable: qualitative research methods are
particularly valuable in establishing this.
Implementation
Once we have an intervention, strategy or policy that is clearly effective, efficient
and acceptable, our problems are still not over. We then have to implement it. In
8 Introduction
other words, we have to find ways of ensuring that the health care professionals,
managers or policy-makers take up the options that research findings find most
appropriate. A lot of work has been undertaken on the production of evidence-
based guidelines and protocols, in the naïve belief that the production of a guideline
would ensure good practice. HSR has shown that this is not the case – doctors, in
particular, do not necessarily follow the guidelines of best practice. There are many
good reasons for this. For example, guidelines are based on evidence from the mean
of groups of people with well-defined problems, and may not be appropriate for an
individual with a multiplicity of problems. An emerging challenge for HSR is to
find ways of helping to individualize the care of people with chronic disease.
Another way of defining HSR is to say that it is about appropriateness and quality –
the delivery of the most appropriate and highest quality interventions and health
care services for individuals and populations respectively. Appropriateness means
doing the things that are most likely to help and avoiding those that do not. Quality
is more difficult to define – but has been said to encompass the domains of
effectiveness, efficiency, equity and accessibility, plus acceptability as outlined above.
Doing research on these topics is difficult. That is what this book is about. An

even greater challenge is to find ways of integrating the different aspects of HSR
and achieving genuine interdisciplinarity. As shown in Figure 1.2, each of the key
domains overlap. HSR is about trying to make sure that health care technology and
services are effective, acceptable, efficient, equitable and that such services can be
and are implemented – i.e. that interventions and services are appropriate and of
Figure 1.2 The main domains of HSR (as defined by the Medical Research
Council and Health Services Research Collaboration)
Research on health and health care 9
high quality. These domains overlap and need integrating through multidisciplinary
research.
HSR, clinical research, audit and quality assurance
HSR overlaps with clinical research, clinical audit and quality assurance, but there
are distinctions between them.
Clinical research can be defined as any research that involves sick people. Such a
definition would include things like clinical research, work on adherence and many
other types of research also called HSR. But in addition, one of its most important
functions is applied physiology: as we learn more about biology and disease pro-
cesses from laboratory-based investigations, we have an increasing need to under-
stand whether the mechanisms that have been uncovered are responsible for disease
and illness in our patients, thus defining new targets for therapy. This involves
careful observation of disease and the measurement of physiological and patho-
logical variables in patients during the course of disease and before and after
interventions – research activities not covered by the term HSR.
Audit is about maintaining high quality of care in health services. It involves the
systematic analysis of procedures used and outcomes of all forms of health care
delivered on a day-to-day basis by professionals and in medical institutions such as
hospitals. It also has a major educational function, as it helps engender a critical and
enquiring approach to health care delivery. The health care professionals generally
carry it out.
Quality assurance can be defined as the definition of standards, the measurement of

their achievement and the mechanisms employed to improve performance. It is
more a managerial function than audit, and implies the presence of a planned
service and agreed standards or targets.
Why do we need HSR?
During the nineteenth century and the first half of the twentieth, health research
was dominated by the description of disease and clinical investigations. However,
over the last 50 years there has been a huge shift to laboratory-based science. The
vast majority of health research now involves genetic, molecular or cellular investi-
gations in laboratories, predicated on the belief that a better understanding of the
ways in which the body works in health and disease will lead to improved health.
But the translation of knowledge gained in laboratories to the improved health of
individuals and populations will not occur without HSR. HSR is a relatively new
development within health research, and one of the main reasons for its recent and
continuing growth is the huge expansion of the technological advances available to
us. As more and more options become available it is increasingly important to assess
them properly.
But there are many other reasons for expanding HSR, including:
• The use of inappropriate care: it is clear that interventions and systems of health care
delivery that do not work, are inefficient or downright dangerous continue to be
practised. We need to root out such practices and replace them with more
appropriate ones.
• Variations and inequities in health care: although the UK National Health Service
(NHS) and other health systems aspire to equitable care delivery this has not yet
been achieved. Large variations in the delivery of health care occur, and the
reasons for them need to be explored.
• Limitation of resources: the potential for health care to consume ever increasing
10 Introduction
proportions of a country’s gross national product became apparent in the USA
towards the end of the last century, before the introduction of managed care and
other mechanisms designed to reduce spending. No matter how much a com-

munity might wish to pay for good health care and health research for all its
citizens, decisions have to be made, and HSR can inform these.
Multidisciplinarity and this book
This book is about the methods used in HSR and related applied health research. It
focuses on the need for a multidisciplinary approach to the evaluation of health and
health care. Research multidisciplinarity, or the involvement of people from more
than one discipline or approach in a project, is an essential part of good HSR for the
reasons outlined in the second section of this chapter. It has also been one of the
principles behind the development of the Medical Research Council’s (MRC)
Health Services Research Collaboration (HSRC) (see www.hsrc.ac.uk), and the
majority of the authors of the chapters that follow have been closely involved in the
development of that organization. Most of the research that we currently undertake
is led by one discipline, but involves others; in the future we need to do better than
this and make sure that the design, management and execution of the research is
interdisciplinary, i.e. that each discipline is of equal importance to the project.
Further reading
Bowling, A. (1997) Research Methods in Health: Investigating Health and Health Services. Buck-
ingham: Open University Press.
Crombie, I.K. (1997) Research in Health Care. Chichester: Wiley.
Dieppe, P. (2000) To cure or not to cure, that is not the question, Journal of the Royal Society of
Medicine, 93: 611–13.
Good, G.J. (1994) Medicine, Rationality and Experience. Cambridge: Cambridge University
Press.
Medical Research Council (1998) Health Services Research Collaboration, www.hsrc.ac.uk.
Wade, D.T. and Halligan, P.W. (2004) Do biomedical models of illness make for good health-
care systems?, British Medical Journal, 329: 1398–401.
References
Fabrega, H. (1997) Evolution of Sickness and Healing. Berkley, CA: University of California
Press.
Kerlinger, F. (1986) Foundations of Behavioural Research. New York: Holt.

Skrabanek, P. and McCormick, J. (1998) Follies and Fallacies in Medicine, 3rd edn. Whitehorn:
Tarragon Press.
WHO (World Health Organization) (1948) Preamble to the Constitution of the World Health
Organization as adopted by The International Health Conference, New York 19–22 June
1946. Geneva: World Health Organization.
WHO (World Health Organization) (1999) International Classification of Impaired,
Disability and Handicap. Geneva: World Health Organization.
Research on health and health care 11
Describing and evaluating
health systems
Ellen Nolte, Martin McKee and Suzanne Wait
Introduction: what is a health system?
Before one can even begin to discuss evaluation of health systems, it is first necessary
to decide what a health system is. There is, unfortunately, no simple answer. A
pragmatic view interprets a health system as being ‘made up of users, payers, pro-
viders and regulators [that] can be defined by the relations between them’ (McPake
et al. 2002), with ‘relations’ referring to four key functions of health systems: regula-
tion, financing, resource allocation and provision of services (Mills and Ranson
2001). In practice, however, health care systems are often defined by national bor-
ders, exemplified by the remark made frequently by journalists since the publication
of the World Health Report (WHO 2000) that ‘the French health care system [is]
judged by the World Health Organization . . . to be the best in the world’ (British
Broadcasting Corporation 2000). Yet within each country there is almost always a
complex mixture of different systems, in which some people use different ways to
pay for health care and in turn receive different benefits (McKee and Figueras
1997). For example, while many people would identify the British health system
with its National Health Service (NHS), a system established in 1948 to provide
universal coverage paid from general taxation, that interpretation would miss the
growing differences in the way in which health care is organized in the four con-
stituent parts of the UK, with Scotland, in particular, moving increasingly away from

the model evolving in England. Similarly, it would miss the substantial volume of
health care provided in the private health care sector, both to those that have private
health insurance and, increasingly, for those who choose to pay directly. And the
UK is, in comparison with some countries, remarkably homogeneous. What, for
example, is meant by the term ‘American health care system’, with its myriad of
payment plans for those in employment, superimposed upon Medicare, for the
elderly, and Medicaid (with its many variations from state to state) for the poor, to
say nothing of a range of other federally funded programmes such as those for
the armed forces, for veterans and for native Americans? Even the Soviet health
care system, which might be thought to have been more homogenous than most,
contained a large number of parallel systems for those employed in the armed
forces, the railways, Aeroflot (the Soviet airline), as well as the nomenklatura (the
Communist Party elite).
Then there is the problem of defining the boundaries of a health system. There
are many activities that contribute, directly or indirectly, to the provision of health
2