Fostering Friendships - Enhancing Social Bonds in the Classroom
Ā
An Emerging Science Of Improvement In Health Care
1. See discussions, stats, and author profiles for this publication at: http://www.researchgate.net/publication/271856985
An Emerging Science of Improvement in Health
Care
ARTICLE in QUALITY ENGINEERING Ā· JANUARY 2015
Impact Factor: 0.36 Ā· DOI: 10.1080/08982112.2015.968042
CITATION
1
4 AUTHORS:
Bo Bergman
Chalmers University of Technology
94 PUBLICATIONS 899 CITATIONS
SEE PROFILE
Andreas Hellstrƶm
Chalmers University of Technology
16 PUBLICATIONS 47 CITATIONS
SEE PROFILE
Svante Lifvergren
Chalmers University of Technology
19 PUBLICATIONS 33 CITATIONS
SEE PROFILE
Susanne Gustavsson
Chalmers University of Technology
3 PUBLICATIONS 2 CITATIONS
SEE PROFILE
Available from: Svante Lifvergren
Retrieved on: 21 August 2015
2. This article was downloaded by: [ Svante Lifvergren]
On: 25 December 2014, At: 09: 50
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK
Click for updates
Quality Engineering
Publication details, including instructions for authors and subscription information:
http:/ / www.tandfonline.com/ loi/ lqen20
An Emerging Science of Improvement in Health Care
Bo Bergman
ab
, Andreas Hellstrƶm
a
, Svante Lifvergren
ac
& Susanne M. Gustavsson
ac
a
Centre for Healthcare Improvement, Chalmers University of Technology, Gothenburg,
Sweden
b
Meij i University, Tokyo, Japan
c
Skaraborg Hospital Group, VƤstra Gƶtaland Region, Sweden
Published online: 21 Dec 2014.
To cite this article: Bo Bergman, Andreas Hellstrƶm, Svante Lifvergren & Susanne M. Gustavsson (2015) An Emerging Science
of Improvement in Health Care, Quality Engineering, 27:1, 17-34, DOI: 10.1080/ 08982112.2015.968042
To link to this article: http:/ / dx.doi.org/ 10.1080/ 08982112.2015.968042
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the āContentā) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the
Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall not be liable for
any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of
the Content.
This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http: / /
www.tandfonline.com/ page/ terms-and-conditions
3. An Emerging Science of Improvement in
Health Care
Bo Bergman 1, 2
,
Andreas Hellstrā¬
om 1
,
Svante Lifvergren 1, 3
,
Susanne M. Gustavsson 1, 3
1
Centre for Healthcare
Improvement, Chalmers
University of Technology,
Gothenburg, Sweden
2
Meiji University, Tokyo, Japan
3
Skaraborg Hospital Group,
Vā¬
astra Gā¬
otaland Region, Sweden
ABSTRACT The purpose of this article is to describe the emerging science of
improvement in health care and to add a perspective from the industrial quality
improvement movement, the use of data from quality registers, and to give
some personal reļ¬ections and suggestions. Furthermore, we want to introduce
to the broader quality management community what is happening in health
care with respect to quality improvement. We will discuss some of the
challenges of the health care system and the current status of a science of
improvement and give some suggestions for further improvements to the area.
We discuss a possible extension of improvement knowledge and the
theoretical and practical arsenal of a science of improvement, in particular,
understanding variation and implications for the use of, for, example control
charts. In addition, the normative side of a science of improvement is
discussed. The article ends with some brief reļ¬ections of use for future research
agendas.
KEYWORDS health care, quality improvement, improvement science, science of
improvement, profound knowledge
INTRODUCTION
Health care is increasingly facing a lot of challenges. The demands are
increasing and reports about health care shortcomings have become increas-
ingly common. Thus, there is a need to introduce new ways of working. Health
care of today is said to be evidence driven with high requirements for data from
randomized controlled trials and from careful observational studies. However,
even though huge amounts of data are collected in quality registers and by other
means, clinical practice is often not governed utilizing these data for learning,
improvement, and innovation. Thus, even though individual health care work-
ers are providing excellent and dedicated work, the health care system might be
ļ¬awed and patient outcome might be less than optimal, perhaps unsatisfactory,
or even unacceptable without anybody noticing it, more than individual
patients.
During the last decades there has been a movement toward a view that clinics
and health care systems should not only have a systematic way of improving
This article was presented at the
Second Stu Hunter Research
Conference in Tempe, Arizona,
March 2014.
Address correspondence to Bo
Bergman, Chalmers University of
Technology, Gothenburg SE-412 96,
Sweden. E-mail: boberg@chalmers.se
Color versions of one or more of the
figures in the article can be found
online at www.tandfonline.com/lqen.
17
Quality Engineering, 27:17ā34, 2015
Copyright Ć Taylor & Francis Group, LLC
ISSN: 0898-2112 print / 1532-4222 online
DOI: 10.1080/08982112.2015.968042
Downloaded
by
[Svante
Lifvergren]
at
09:50
25
December
2014
4. medical treatment and nursing utilizing evidence from
research but also have a systematic way of improving
how health care is delivered to the patientsāhow the
health care system supports its patients to better health
without risking any harm. An attempt to provide a
basis for such initiatives is called improvement science or
a science1
of improvement. The purpose of this article
is to describe this emerging ļ¬eld and to add a perspec-
tive from the industrial quality improvement move-
ment, the use of data from quality registers, and to give
some reļ¬ections and suggestions. In this introductory
section we will ļ¬rst discuss some challenges of the
health care system and give a short outline of the rest
of the article where the current status of a science of
improvement is discussed as well as some suggestions
for further developments of the area.
The increasing demand on the health care system is
due to, ļ¬rst, the growing population of elderly people
suffering from multiple diseases. At the same time, the
productive part of the population has become rela-
tively smaller and therefore health care funding may be
jeopardized (Weisz et al. 2013). A second reason is the
appearance of new and much improved possibilities to
give life supporting cure and care. Some of these possi-
bilities are very costly, and even if many may indeed
reduce the total costs to society, the cost reductions do
not always beneļ¬t those that have the cost burdens
from new treatments. Systems handling this unbalance
are not common. Furthermore, the rate of innovations
in health care is increasingāit has been argued (Dixon-
Woods, Alamberti et al. 2011; Sjohania et al. 2007)
that the half-life of knowledge is 5.5 years. A third
argument is an increased focus on health and well-
being in the populationāhow does the health care sys-
tem in a productive way take care of an increased health
literacy and requirements on patient autonomy and
patient empowerment? (Anderson and Funnell 2005;
Reach 2014).
Another kind of challenge is the stream of reports on
the shortcomings of health care systems. In the UK, for
example, the Bristol heart scandal (Weick and Sutcliff
2001) at the Bristol Royal Inļ¬rmary a number of years
ago revealed unacceptable shortcomings and later the
Mid Staffordshire scandal attracted a lot of attention,
with many investigations and government reportsā
similar shortcomings were reported in these inquiries:
No clear visions and goals, no culture of patient cen-
teredness, boards and top management not taking the
lead on these questions, insufļ¬cient feedback loops on
what is really going on in the organization with respect
to quality and safety, etc. In a recent study (Dixon-
Woods et al. 2014) of the National Health Service
(NHS) many good things were reported, speciļ¬cally
the ambitions and goodwill of the health care profes-
sionals to provide good care; on the negative side, how-
ever, it reported that the above-mentioned
shortcomings might be more widespread than so far
revealed. We will return to this study later.
It is not only in the UK that these types of problems
are frequent. In Swedish newspapers we read almost
every day about mistreatments, long waiting times, and
scarcity of personnel. Just to mention one recent story
revealed by a Swedish TV program, the deteriorating
screening procedures of cervix cancer samples may
have caused malign tissue samples to remain undiscov-
ered. Even though results were fed into a database,
there was no feedback to the person performing the
screening. In all work correct feedback is impor-
tantāwithout correct feedback any work process may
deteriorate.
The high cost of the U.S. health care system2
is, of
course, also well known. For example, the U.S. health
care expenditure in relation to the gross national prod-
uct is almost double that of Sweden and the medical
results are not better; in many respects, in fact, they are
much worse. Furthermore, in a recent paper (James
2013) it was reported that as many as 200,000 and per-
haps even 400,000 deaths per year in the United States
may be caused by avoidable harm.
Of course, there are no silver bullets to solve these
problems (Shojania and Grimshaw 2004). However, as
one part of a solution, a focus on quality and quality
improvement seems necessary. As illustrated by the
study of the NHS, a mindset, close to what in indus-
trial settings has been known as total quality management
(Bergman and Klefsjā¬
o 2010), should be strived forā
perhaps a new management paradigm in health care. In
this direction we ļ¬nd the emerging science of improve-
ment in health care.
In this article, we will ļ¬rst discuss the ļ¬ndings of
the NHS study mentioned above and relate these to
1
The concept science might be considered problematic; see, for
example, Wilmott (1997).
2
Based on Oganization for Economic Cooperation and
Development data comparisons between the cost and medical
outcomes were compared in a report from Swedish Association of
Local Authorities and Regions (SALAR).
18 B. Bergman et al.
Downloaded
by
[Svante
Lifvergren]
at
09:50
25
December
2014
5. the early discussions on quality in health care by Ernest
Amory Codman3
and Avedis Donabedian (1992,
2002)4
but also to the early discourse on quality
improvement in industry. Based on these discussions
we will inquire into the scope of a science of improve-
ment. During the course of the article we will come
back to the NHS study mentioned above for illustra-
tive purposes. Illustrations from the Swedish health
care will also be used. As part of a brief review, some
earlier discussions on the emerging science of improve-
ment will be given. Then, in the main part of the arti-
cle, the core of a science of improvement will be
discussed. It should, however, be emphasized that a sci-
ence of improvement in health care really is in a prepar-
adigmatic stage, to use the words of Kuhn (1962). Thus,
many interpretations and outlines have been given, per-
haps under different headings, and probably many
more will comeāand should come. Indeed, this article
will give our interpretations and suggestions, which
might go further in scope than many other alternatives.
A science of improvement must, of course, address
how improvement issues are identiļ¬ed and elaborated
upon, eventually leading to successful interventions
with even better outcomes for the patients. It must also
address how these improvements are evaluated and
how lessons are learned from them. But it is equally
important to foster an environment where improve-
ments are seen as normal and inspiring parts of the
daily life in an organization. This environment is a piv-
otal requirement for a transformed health care system.
How is that achieved? These latter questions are much
harder and more pressing than the operating work with
improvements themselves, however important they
are. Of course, these aspectsāthat is, the operational
improvement interventions and the improvement of
the organizationās readiness (capability and capacity) to
improve and innovateāshould go hand in hand, and
no simple recipe can be expected.
THE SCOPE OF A SCIENCE OF
IMPROVEMENT
The NHS study mentioned above was a multime-
thod study of the NHS during the years 2010ā2012
with interviews, surveys, board protocols, and direct
observations together with data from 2007 to 2011 on
staff experiences, patient experiences, and patient mor-
tality rates. A large number of ļ¬ndings in this study
corroborate our view of what is important in a science
of improvement. Some ļ¬ndings were highly encourag-
ing; for example, a consistent focus from front line
people to do their best for their patients to create the
best possible care.
However, in some organizations there were no clear
goals, no vision, and no culture for continually reļ¬ning
patient quality and safety. In such organizations the
front line staff observed severe system deļ¬ciencies but
felt powerless to deal with them, whereas management
perceived problems with quality and safety as caused
by the front line people. That is, the staff and the man-
agement had no common understanding about how to
solve their problems and to improve patient quality
and safety. In addition, there was a lack of intelligen-
ceāthat is, poor follow-up of both their own results
and results from similar unitsāin order to ļ¬nd
improvement opportunities. Interestingly, the research-
ers were able to observe that
. . . hospital standardised mortality ratios were inversely asso-
ciated with positive and supportive organisational climates.
Higher levels of staff engagement and health and wellbeing
were associated with lower levels of mortality, as were staff
reporting support from line managers, well structured apprais-
als (e.g., agreeing objectives, ensuring the individual feels val-
ued, respected and supported), and opportunities to
inļ¬uence and contribute to improvements at work. NSS
[NHS Staff Survey] data also showed that staff perceptions
of the supportiveness of their immediate managers, the extent
of staff positive feeling, staff satisfaction and staff commit-
ment were associated with other important outcomes, including
patient satisfaction. (Dixon-Woods et al. 2014, p. 112)
Let us compare these results with the simpliļ¬ed
model of health care that was suggested by Avedis
Donabedian, who observed that health care (1) out-
comes are produced in (2) processes that are imbedded
in systems and their (3) structures. According to Dona-
bedian (2003), āstructureā is meant āto design the con-
ditions under which care is providedā (p. 46).
Examples include material and human resources, the
presence of teaching and research, performance
reviews, etc. In our interpretation aspects like gover-
nance, leadership, learning mechanisms, organizational
climate, organizational design, intelligence, attractors,
incentives, norms, values, and similar āsoftā issues
should be included. Process signiļ¬es āthe activities that
constitute health care ā including diagnosis, treatment,
3
For a short biography of Codman, see Neuhauser (2002).
4
For a short biography about Donabedian, see Best and Neuhauser
(2004).
Improvement in Health Care 19
Downloaded
by
[Svante
Lifvergren]
at
09:50
25
December
2014
6. prevention, and patient education. . . .ā (Bergman et al.
2011, p. 41). Finally, outcome measures are the resulting
states from care processes, both technical (for instance,
absence of complications) and interpersonal outcomes
(for instance, patient satisfaction).
However, we also want to add how individuals in
the organization would react to different change inter-
ventions; that is, individual predispositions.5
It should
also be noted that we interpret the concept structure
rather freely, not as it is often interpreted as something
that is but rather as something that is in a constant
becoming and ļ¬ux; see, for example, Sztompka (1991).
Indeed, this view on structure is pivotal to our view on
the health care system.6
With a serious ambition to improve outcome for
patientsāthat is, quality and safetyāa science of
improvement has to address improvements on all levels
of the Donabedian model. It is not enough to make
improvements on the process level only. In addition,
soft aspects of the environment of the process (and its
microsystems) have to be addressed: How is the climate
for change generally and how are the individualsā pre-
dispositions for change? How is patient centeredness
supported? How is intrinsic motivation of the staff sup-
ported? What are the norms and values? How is reļ¬ec-
tion and learning supported? It is only at such levels
that we can assure some sort of long-term survivability
of improvement efforts, see e.g. Docherty et al. (2003)
and Docherty and Shani (2008).
From this discussion, the following deļ¬nition of
quality improvement suggested by Batalden and
Davidoff (2007, p. 2) is adequate; that is, quality
improvement is seen
. . . as the combined and unceasing efforts of everyo-
neāhealth care professionals, patients and their families,
researchers, payers, planners and educatorsāto make the
changes that will lead to better patient outcomes (health),
better system performance (care) and better professional
development (learning).
When talking about system performance in the
above deļ¬nition we should also include the predisposi-
tions for change (or readiness for change) of the
individuals within the organization. Depending on
past experiences, these predispositions could look very
different in different organizations. The organizationās
drive for change (i.e., improvements including innova-
tions) should be a very important target of improve-
ment efforts.
A SCIENCE OF IMPROVEMENT: A
BRIEF REVIEW
An important starting point of the quality move-
ment was the seminal books by Shewhart (1931, 1939);
Shewhart was very much inļ¬uenced by pragmatic phi-
losophy (Lewis 1929), eventually leading to the much
used Plan-Do-Study-Act (PDSA) cycle and his thoughts
about predictability and corresponding criteria have
been central. Later, the experiences from the Japanese
wonder and the reawakening (Cole 1999) of the Ameri-
can industry were important. In health care, Ernest
Amory Codman was already a pioneer in advocating
the reporting of failures in order to encourage improve-
ment before WWI. Later, after WWII, Avedis Donabe-
dian developed the model referred to previously.
However, inspiration from the industrial quality move-
ment hit health care in the 1980s, ļ¬rst with quality
circles and later in a much broader scope; see, for exam-
ple, Berwick (1989), Berwick et al. (1990), and Batalden
and Stoltz (1993). In the late 1980s, the Institute of
Healthcare Improvement was founded by Don Berwick
and a number of his colleagues. Many more initiatives
started in the beginning of the 1990s.
Some important milestones were the roundtable
around quality that was arranged by the Institute of Med-
icine; see Chassin and Garvin (1998) and the subsequent
publications To Err Is Human: Building a Safer Health Sys-
tem (Institute of Medicine 1999) and Crossing the Quality
Chasm: A New Health System for the 21st Century (Institute
of Medicine 2001). These publications have had a great
impact; for example, in Sweden they changed the ear-
lier dominant punitive7
view on health care errors and
quality dimensions of health care were identiļ¬ed as
those suggested by the Institute of Medicine (Swedish
National Board of Health and Welfare 2006).
In Sweden, Mā¬
olndals Hospital (Alā¬
ange and Steiber
2011) was a forerunner focusing on quality improve-
ment of its processes; a clinic at Linkā¬
oping University
5
Individual predispositions were suggested by Walker et al. (2007)
as an important factor for the effect of change interventions in an
organization; see also Moore and Buchanan (2013).
6
In the language of researchers on change we have a process view;
see, for example, Pettigrew (1987) and Walker et al. (2007).
However, here we have a somewhat conflicting usage of the word
process; thus, we keep the concept structure and interpret it in a
more open way than it its often done in social sciences.
7
Of course, there were also earlier many proponents for a
nonpunitive view on health care errors, but these views had not
reached the governments, legislation, and media.
20 B. Bergman et al.
Downloaded
by
[Svante
Lifvergren]
at
09:50
25
December
2014
7. Hospital (Bergman and Klefsjā¬
o 2010) was one of the
ļ¬rst organizations, in competition with many otherwise
successful manufacturing companies, to win the Swed-
ish Quality Award; and in Jā¬
onkā¬
oping County they ini-
tiated a journey on quality improvement that is still
going on (Andersson-Gā¬
are and Neuhauser 2007).
Deming (1994) identiļ¬ed that for leaders to work
ļ¬rmly in what he called the āNew Economyā they have
to have some understanding of a number of important
interrelated knowledge areas. He emphasized āa system
of profound knowledgeā and āappreciation for a sys-
tem; knowledge about variation; theory of knowledge;
psychologyā (p. 96). These knowledge areas and their
extensions will be discussed more in the next section.
Here we only mention that these areas were introduced
to the health care community in a paper by Batalden
and Stoltz (1993) under the concept improvement knowl-
edge; see Figure 1. In a seminal book (Langley et al.
1996) on health care improvement the name improve-
ment science seems to have been introduced for the ļ¬rst
time (Perla et al. 2013; see also Berwick 2008).
They also suggested, as illustrated in Figure 2, how
improvement of performance as experienced by
patients (3) is achieved by the application of generaliz-
able scientiļ¬c knowledge (1) to the special context
based on careful planning (4) and then a careful execu-
tion of the plan (5).
In Batalden and Stoltzās approach, eight domains of
interest with associated tools and methods are of
importance. These domains are health care as processes
within systems; variation and measurement; customer/
beneļ¬ciary knowledge; leading, following, and making
changes; collaboration; social context and
accountability; and developing new knowledge. They
are similar to but formulated differently than the prin-
ciples for organizing we present in the section The Nor-
mative Side of a Science of Improvement. In Sweden,
professional societies (nurses, doctors, nutritionists,
physiotherapists, work therapists, etc.) have joined
forces to create a common ground for professional edu-
cation on improvement knowledge.
During April 12ā16, 2010, a colloquium8
on the
epistemology of improving quality convened at Clive-
den (see, e.g., Batalden et al. 2011). One important
outcome of this colloquium was the awareness and
importance of taking many different perspectives on
quality improvement in health care and to base training
programs on multiple epistemologies informing health
care improvement. Not only does traditional evidence-
based medicine, resting on a natural science founda-
tion, have to be called upon but also social science and
deeper understandings of social change processes.
Recently, Perla et al. (2013) suggested some theoreti-
cal and philosophical foundations on which āthe sci-
ence of improvementā rests. For example, learning and
the development of actionable knowledge is stressed
and they refer back to philosophical pragmatism, espe-
cially that of Lewis (1929) and his conceptualistic
pragmatism and subsequent interpretations and transla-
tions of Shewhart and Deming.
Concerns that improvement efforts too often are
lacking in a thorough scientiļ¬c basis has been aired by,
for example, Marshall (2011) and Marshall et al.
(2013). There is also a tendency that too much focus
has been on implementation of solutions and on oper-
ational process improvement rather than on improve-
ment and innovation of the health care system itself;
that is, improvement of the organizational abilities and
transformational changes of the health care system.
Another striking observation is the relative lack of a
patientās perspective. Not that the patient is forgotten,
but quite a lot of the literature is more operations
focused than really patient focused. Works on the
patient perspective (Bergman et al. 2011), patient
empowerment (for a review, see Aujoulat et al. 2007),
patient involvement in health care design and improve-
ment (Bate and Robert 2006; Gustavsson 2013) should,
we think, be much more in focus.
FIGURE 1 This figure is an enhanced version of the illustra-
tion by Batalden and Stoltz (1993). It illustrates how improvement
knowledge should be integrated with professional knowledge
for the improvement of outcomes.
8
The Vin McLoughlin Colloquium on the Epistemology of
Improving Quality, resulting in 23 papers downloadable from
http://qualitysafety.bmj.com/content/20/Suppl_1/i99.full.html#ref-
list-1
Improvement in Health Care 21
Downloaded
by
[Svante
Lifvergren]
at
09:50
25
December
2014
8. It is quite natural for an emerging ļ¬eld that its theo-
ries are discussed quite extensively and also that results
close to the very aim of the ļ¬eld are discussed. Educa-
tional aspects are, of course, also very important. What
would be needed, though, are more empirical papers
contributing to our understanding of how an organiza-
tionās capability and capacity9
for improvement and
innovation are enhanced.10
The NHS study described
earlier is a good illustration. A Ph.D. thesis on a similar
theme but limited to one hospital group including the
surrounding municipalities and the primary care cen-
ters has been presented by Lifvergren (2013).
IMPROVEMENT KNOWLEDGE
As emphasized in Figure 1 based on Demingās āpro-
found knowledge,ā improvement knowledge takes a
central position in most descriptions of a science of
improvement. However, the different knowledge
domains have deepened since the days of Deming and
the need to broaden these areas has become obvious.
In the following we will discuss each one of the basic
knowledge domains separately, emphasizing their deep-
ening and possible ways to extend them. However, we
will leave most of the āunderstanding variationā
domain until the next section.
Appreciation for a System
Some basic features of the understanding that
Deming drew onāfor example the insight that enlarg-
ing the system boundary may give opportunities to bet-
ter solutions to pressing issuesāare still important and,
unfortunately, in these respects health care still has a
lot to do to integrate their different parts as well as to
integrate with other organizations outside health care
for the beneļ¬t of their patients (see the section
Understanding Variation). This is also an important
part of what Deming called wināwin solutions.
However, our understanding of systems has
increased substantially since the days of Deming. The
theories of complexity (see, for example, Waldrop
[1992] and Stacey [2003]) and complex systems had
just started in the beginning of the 1990s. Recent
knowledge from complex adaptive systems research
shows that the complexity increases11
and thus the
manageability and predictability decreases (e.g., Axel-
rod and Cohen 1999; Stacey 2003). In complex systems
the concept of attractorsāthat is, what attracts the sys-
tem to develop toward a certain set of statesāis also
important for our understanding of how change efforts
may work. In order to try to understand social systems
better we also have to draw much more on social scien-
ces in our basic improvement knowledgeāwe will
come back to that later in this section, giving the
knowledge area psychology a much wider deļ¬nition than
in the original one suggested by Deming.
Understanding Variation
Deming emphasized reduction of variation (see also
Bergman 2003); however, there is much more to it. We
will come back to that later but let us ļ¬rst make a note
on the control chart, suitable for ļ¬nding assignable or
(as Deming put it) special cause variation.12
Shewhart
emphasized the need to have predictable processes,
Shewhartās (1939) deļ¬nition of such processes is the
same (see Bergman [2009] and references cited therein)
as that of exchangeability used by de Finetti (1974).
Shewhart (1931) gave ļ¬ve criteria for such processes.13
One of these is based on the control chartāthe one
that is most well known and for which Shewhart is the
FIGURE 2 The āequationā for improvement as suggested by Batalden and Davidoff (2007) (adapted from Batalden and Davidoff 2007).
9
Note that an organization might have the capability to make
improvements but due to resources restrictions their capacity is
limited.
10
This is also emphasized by Marshall et al. (2013); see also Parry
et al. (2013).
11
It could, of course, be argued whether it is the complexity that
increases or our understanding thereof.
12
The use of control charts in health care is discussed in, for example,
Langley et al. (1996), Benneyan et al. (2003), Woodall (2006),
Neuhauser et al. (2011), and Woodall and Montgomery (2014).
13
Interestingly, these have been graphically explained by Ishikawa
(1982, 1985) in the seven quality control tools.
22 B. Bergman et al.
Downloaded
by
[Svante
Lifvergren]
at
09:50
25
December
2014
9. most famous. Violations of these criteria may be used
to identify assignable causes. Speciļ¬cally, if we have
made an intervention hoping for an improvement,
then that should result in an assignable cause indicating
a positive result. That used to be one of the corner-
stones in the philosophy of Joseph Juran.14
We will dis-
cuss variation more in the next section.
Knowledge Theory
This theory originated in the conceptualistic prag-
matic philosophy of Clarence I. Lewis, professor of
philosophy at Harvard University. It provided a back-
ground for the PDSA cycle (see, for example, Mauleon
and Bergman 2009; Peterson 1999; Wilcox 2004). In
addition, it is emphasized that theory is the basis for
predictionāprediction is the basis for action and if the
action does not lead to the predicted result there is a
need for reļ¬ection and revision of the theory (i.e.,
learning) in order to perform better actions in the
future. Lewis emphasized that knowledge is for action
(Lewis 1929). Indeed, this is what Dixon-Woods, Bosk
et al. (2011) wrote in their reļ¬ections on the Michigan
case. Furthermore, when different humans observe the
same phenomenon, they may experience it very differ-
ently due to their a priori understanding. This under-
standing sometimes leads to conļ¬icts in a group if it is
not understood that the differences depend on differ-
ent mental models, as Peter Senge (1990) would state it.
Closely related is how we as individuals create mean-
ingāand even more important how individuals trans-
form their ways of creating meaningāsomething that is
important in many quality improvement efforts, espe-
cially on the structural and system levels in the Dona-
bedian model. When there is a need for second-order
type of learning, how is that achieved? We have to call
in much more from current research around meaning
creation, transformative learning (Mezirow & Associ-
ates 2000), and learning to learn (Argyris 2003; Visser
2003). Note that ļ¬rst-order learning means that we
learn something within our earlier understanding of
reality, and second-order learning means to change our
way of understanding reality, to get a new way of creat-
ing meaning. Third-order learning means achieving a
way of reļ¬ecting on self and question the current way
of creating meaning. In this learning, reļ¬ection is an
important process: to use a statistical metaphor, the
Bayesian learn more and more about the current model
when the given likelihood function changes due to an
increasing amount of data (single loop learning). How-
ever, to change the shape of the likelihood function or
recognizing that no likelihood function exists due to
assignable causes of variation (not in themselves under
statistical control; see Bergman and Chakhunashvili
2007) is an illustration of a kind of second-order learn-
ing. Within this metaphor, third-order learning should
mean to have a systematic approach to ļ¬nd out whether
the current assumed likelihood function really is relevant
for the problem at hand or not. Usually, the above forms
of learning are taken as individual learning; however, in
an organization it is important to attach these types
of learning to the organization. Learning is the achieve-
ment of knowledgeāknowledge is for actionāand com-
mon knowledge is for common action.15
Psychology
What Deming emphasized here was the importance
of intrinsic motivation. Currently, there is much more
evidence about the importance of intrinsic motivation
and how that has been put into systematic use in some
organizations. A popular book by Dan Pink (2010)
illustrates this nicely. Closely related to research results
on intrinsic motivation and self-determination theory
(see the introductory chapter of Deci and Ryan 2002)
is the new research ļ¬eld called positive psychology (see,
for example, Csikszentmihalyi and Nakamura 2011)
with an interesting forerunner, Mihaly Csikszentmiha-
lyi (1990), and his research on creativity and the con-
cept of optimal experiences, also known as ļ¬ow.
Under the heading psychology, Deming emphasized
not only individual psychology but also interactions
between people. This should lead us into a much
broader research ļ¬eld to integrate into a science of
improvement. Indeed, a lot of social science should be
incorporated here; there might be a need for a new
heading for this broad knowledge ļ¬eld, which has been
proven to be of great interest to a science of improve-
ment (see, e.g., Dixon-Woods, Bosk et al. 2011). For
example, an important research tradition with much in
common with the quality movement is action research;
14
This is called the Juran trilogy: designācontrolāimprovement; see
Juran (1986).
15
This is a reformulation of a comment in Lewis (1929); an
interesting framework connecting the individual with the
common is given by Crossan et al. (1999).
Improvement in Health Care 23
Downloaded
by
[Svante
Lifvergren]
at
09:50
25
December
2014
10. see, for example, Lewin (1948) and Coghlan and Bran-
nick (2010). For some recent applications within health
care see, for example, Lifvergren, Gremyr et al. (2010),
Lifvergren, Docherty et al. (2010), Lifvergren et al.
(2011), Lifvergren et al. (2012), Gustavsson (2014), and
Lifvergren (2013).
A System of Knowledge
Deming emphasized that profound knowledge is a
system of knowledgeāthe different parts are interre-
lated. Even though Dixon-Woods, Bosk et al. (2011) do
not explicitly use this conceptualization in their study
of the Michigan project, it is an excellent illustration.
Even though Demingās profound knowledge serves
well as a starting point, we now need a broader set of
basic scientiļ¬c knowledgeāperhaps under different
headings. Indeed, this is underway even if the headings
have remained the same. However, it is important to
understand that the theories we use in our interpreta-
tions are important for our continual learning how to
better achieve our organizational goals; as Deming
(1994) said, āExperience by itself teaches nothing. . . .
Without theory, experience has no meaning. Without
theory, one has no questions to ask. Hence, without
theory, there is no learningā (p. 106).
UNDERSTANDING VARIATION
In the Beginning There Was Variation
Already with the Big Bang variation was thereāif it
had not, we would probably still have a homogenous
universeāand no āweā! And so it goes onāheavier
and heavier atoms, the creation of stars and galaxies,
and eventually some things that replicate, vary, inter-
act, and compete for energyāthe organic life has
begun.16
And eventually intelligent creatures who cre-
ate ideas and thereby a new evolution starts. And here
we areāthe results of a long chain of results from varia-
tion, selection, and interaction.17
Indeed, variation is a
central part of our lives.
In discussions of variation it is important to see dif-
ferences and similarities between two different kinds of
variation: variation between units when time is not in
focus, or synchronic variation, and variation over time,
or diachronic variation. Too often, only the synchronic
variation is discussed, though in many cases diachronic
variation is the most interestingāhow is it possible to
predict future outcomes given the data? Did an inter-
vention give the intended result? Much medical
research in terms of randomized controlled trials
emphasizes synchronic variation. A stronger emphasis
on diachronic variation would be an important contri-
bution to medical research. And, of course, diachronic
variation is of great importance for the improvement
of health care. An illustration where a quality improve-
ment intervention seems to have had a very important
effect is illustrated in Figure 3. For some further reļ¬ec-
tions see the following subsection.
Variation in Health Care
We are all differentāpatients are all different. Most
often health care professionals have a good sense for
thatāmeeting each patient here and now on the condi-
tions of this patient. Thus, on an individual level there
is often a deep understanding of variation. However,
when it comes to clinical research, variation has often
been kept under the carpetāwhat have been counted
are averagesāmean values. At least, that has very often
been the case in randomized controlled clinical studies,
known as the golden standard of clinical research. Vari-
ation is used mostly through the standard error of the
mean to judge whether mean values are statistically sig-
niļ¬cant or not. If it is not mean values it is often per-
centages of individuals reaching a certain target; HbA1c
is a typical case18
where changes have been made a
FIGURE 3 Number of sudden infant death syndrome cases; a
reasonable assumption is that these deaths occur as in a Poisson
process.
16
Richard Dawkins has written many popular books on this theme,
starting with The Selfish Gene in 1976.
17
Some interesting references are Dawkins (1976); Jablonka and
Lamb (2005) on evolution in four dimensions, and Dennett (1995).
18
HbA1c measures the long-range blood sugar level and is an
important measure in diabetes care.
24 B. Bergman et al.
Downloaded
by
[Svante
Lifvergren]
at
09:50
25
December
2014
11. number of timesānot only target values but also mea-
surement methods. This is, of course, very problemat-
icātarget values change and thus operating deļ¬nitions
change and possibilities to compare with history are
lost. This may partly be due to a lack of insight con-
cerning diachronic variation.
Nowadays, with the insights and the possibilities for
genetic testing, an increasing understanding for varia-
tion and that randomized testing is not always the best
answer is emerging. But genetic variation is just one
source of variationāalso important are epigenetic vari-
ation, behavioral variation, and perhaps even cultural
variation (Jablonka and Lamb 2005). These types of
variation must be taken into account. That is often not
done in the current clinical research paradigm. As
many interventions and improvement activities have to
be adopted in a social system, the variation in contexts
is very important and can explain sometimes confusing
results and large variations; see, for example, Pettigrew
and Whipp (1991), Bate (2014), and the Health Foun-
dation (2014). Today, there is quite a lot of focus on
quality registers and thereby a potential to evaluate
new interventions. From the registers random selec-
tions taking different factors into accountāthe use of
design of experiments based on registers and follow-up
through the registers and then subsequent analysis of
resultsāseems to be a very interesting possibility not
yet utilized.
Quality Registers
The worldās ļ¬rst national quality register was created
by Ove Guldberg HĆøegh to register patients suffering
from leprosy in Norway 1856. In the beginning of the
20th century, Ernest Amory Codman created a register
on shoulder surgery. However, it was not until the end
of the 20th century that registers for a large number of
diseases became common. In Sweden, we have around
100 such registers todayāeach recording a number of
different diagnoses and corresponding process and out-
come measures. They were ļ¬rst created by enthusiastic
physicians to get better hold of their specialities as well
as for research purposes. An illustration is the ļ¬rst regis-
ter in Sweden, a knee replacement register, initiated in
1975 and shortly after followed by a hip replacement
register. Due to the registers it has been possible to
increase reliability of these types of replacements with
great beneļ¬ts both for patients and for replacement
costs. The situation in Sweden, where all citizens have
a unique personal number, is especially beneļ¬cial for
research. It is, for example, possible to combine results
from different types of registers. Of course, efforts are
needed to assure that integrity aspects are taken into
account.
Registers are increasingly thought of as a huge possi-
bility for improvement activities. However, it has still
very much stayed at a rhetoric level and much has to
be done to educate health care professionals to utilize
these registers in their daily improvement work (see,
e.g., Bergman 2013). Among the best registers to pro-
mote improvement work is the diabetes register, where
today it is possible to get information online concern-
ing different hospitals and their results with respect to
smoking cessation, blood pressure, HbA1c, patient
recorded outcome measures, etc.
Reflections on Variation, Statistics,
and the Theory of Probability
As we observed in an earlier section, variation is a
basic concept in our understanding of the universe. In
this section, we want to explore further some of the
basic aspects of knowledge theory and variation. When
observing the world we ļ¬nd variation, and that gives
rise to uncertainty when we try to make predictions
about the futureāand, in a way, such predictions lie
behind our decisions and subsequent actions, even if
not always as rational as we would like to think; see, for
example, Kahneman (2011). In this way of looking at
things, which is very much in line with the ideas of C.
I. Lewis that inļ¬uenced Deming and Shewhart, proba-
bilities describing our uncertainty exist in our theories
about the world, not in the world itself. In fact, that
leads us to a subjective (or Bayesian) interpretation of
probabilities as promoted early by Ramsay (1926), de
Finetti (1974), and Savage (1954). Bayesian statistics
has gained a lot of attention in medical statistics lately
(see, e.g., Ashby 2006).
The Subjectivist Interpretation of
Probabilities and Conceptualistic
Pragmatism
Lewis (1929), in his conceptualistic pragmatism,
emphasized that theories are a priori, which however,
we are free to choose. He also emphasized that it is
only in the light of this a priori that we experience what
Improvement in Health Care 25
Downloaded
by
[Svante
Lifvergren]
at
09:50
25
December
2014
12. happens in the world. According to Lewis, experiences
are an interpretation of the sensuously given based on
the a priori. It is indeed also a basic ingredient in our
possibility to reļ¬ect upon and learn from our experien-
ces. We ļ¬nd a striking similarity between the thinking
of Lewis and that of Bayesians.
There is another interesting connection between
Bayesian statistics and traditional quality improvement
a la Shewhart and Deming. The control chart makes an
interesting bridge investigated by Bergman (2009).
When Shewhart (1939) made a formal deļ¬nition of a
process under statistical controlāthat is, about pre-
dictabilityāhe used a criterion that in the Bayesian/
subjectivist literature is called exchangeability.19
It was
introduced by de Finetti (1974) as a prerequisite for a
predictive distribution to exist. However, in the Bayes-
ian/subjectivist literature the judgment of a process
being exchangeable is left to the decision maker and no
tools are provided. However, here the Shewhart control
chart comes in handy. It is designed exactly as a tool for
judging predictability of a process. We will come back
to this in the next section. First, however, we will make
an important note that is sometimes forgotten when
discussing the contributions of Shewhart. He not only
suggested the control chart to be used for judging if a
process is under statistical control, but he also pointed
to four more criteria to be fulļ¬lled for a process to be
free from assignable causes of variation.20
Four of these
ļ¬ve criteria of Shewhartās were later transformed into a
graphical form by Kaoru Ishikawa (1982) to form some
of the well-known seven QC tools: histogram, stratiļ¬-
cation, scatter diagram, and control chart.
Control Charts
When Shewhart introduced the control chart in his
seminal book from 1931 he recognized that not all pro-
cesses were predictable in a statistical senseāsome-
times assignable causes of variation interfered and
predictability was lost. It is important to identify such
assignable causes of variation and to eliminate them in
order to get a predictable process. But not all predict-
able processes give an acceptable result. We have to
make more drastic improvements in the process, not
only to remove assignable causes of variation but to
more radical changes in the process/system. This was
illustrated in what Juran called the Juran quality tril-
ogy (Juran 1986): planning a process, controlling it
(and take away assignable causes of variation), and
then improving the process. After a successful
improvement we go into a control phase again; see
Figure 4.
As seen in Figure 4, the control chart works in three
different ways. First, to judge whether a process is pre-
dictable, take away assignable causes, estimate parame-
ters, etc. (phase I); second, to survey the process (phase
II) to guard against new assignable causes that hope-
fully will be eliminated to give a predictable process;
and thirdly, to work as a conļ¬rmation that an inter-
vention, a change of some kind, really was an
improvement (phase III). It should be mentioned
that phase III is usually not mentioned in a system-
atic way in the literature on control charts even
though it should be regarded as a very important
aspect of the use of control charts. In Figure 3 on
sudden infant death syndrome we saw that a sub-
stantial improvement in the process had occurred.
The occurrence of an assignable cause of variation
of a positive kind could be taken as a strong indica-
tion that the intervention (in this case a campaign)
really was successful. Of course, it is important to
make sure that no other changes have been active
that could have affected the process.
In health care today, both process measures and out-
come measures are documented in quality registers.
This provides us with an excellent possibility to evalu-
ate improvement interventionsāif the relevant meas-
ures come from a process under statistical control it is
possible to conclude whether an intervention really is
FIGURE 4 The quality trilogy (Planning, Control, Improve)
suggested by Juran (1986) with a more refined division in phases
I-IV.
19
It could also be called permutabilityāeach permutation of the
stochastic variables has the same distribution.
20
Note that sometimes the process may be predictable even
though assignable causes exist.
26 B. Bergman et al.
Downloaded
by
[Svante
Lifvergren]
at
09:50
25
December
2014
13. an improvement or not because it should show up as a
positive assignable cause. In many respects, such a strat-
egy for evaluating an intervention has many advantages
over the usual gold standard of clinical medical
research; however, we will not go further into this
debate here.
When reļ¬ecting on the ideas behind Figure 4 it
becomes obvious that the usual phase I and phase
II descriptions of the work with a control chart are
not quite enough. A phase III should be added
when improvements are in focus. The special assign-
able cause that might have occurred due to an inter-
vention usually takes some time to mature (if
positive as intended). How long do we need to run
the process until we can judge predictability (and
the new level)? A possible phase IV should be very
similar to phase II. We will come back to this later
in this section.
A Bayesian Control Chart
In the literature on Bayesian control charts we do
not see the type of control charts that Shewhart advo-
cated; that is, a chart for judging whether a process is
under statistical control or not. Rather, the state under
statistical control is determined from the beginning
and an out-of-control state is deļ¬ned. The argument
for the name Bayesian is because these states are given
subjective probabilities and if the out-of-control state
gets a high enough probability an out-of-control situa-
tion is declared. In fact, the out-of-control state is in
itself assumed to be a process under statistical control.
Depending on the purpose, this might be perfectly all
right. However, we just want to determine whether the
process is under statistical control or not; that is,
whether it is reasonable to judge it predictable or not.
Based on the nature of the process, some sort of model
might be judged if the process is under statistical
control.21
When we observe the process, we also increase our
knowledge about the process and increase our possibil-
ity to predict what will happen in the future of the
processāat least if it is under statistical control; that is,
still predictable. But how do we judge that? A simple
way would be that the observation at time t C 1 is
within what could be called the predictive interval as
judged from observations up to time t. And that should
also be the case for all of the past observations due to
the fact that the process under statistical control is
exchangeable.22
The only thing to ļ¬nd out is how the
predictive interval should be calculated. In the spirit of
Shewhart it may be taken as the 3s limits of the predic-
tive distribution. However, just as Shewhart indicated,
another choice could be relevant depending on the sit-
uation. It should be noted that it is necessary to go
back also to old observations because in the beginning
of the observation period the knowledge about the pro-
cess is not as strong as later on. That means that an
observation at time s might be within the predictive
interval based on observations up to time s Ā” 1 but
when time has passed to time t s we know much
more and therefore the observation at time s may be
outside the new predictive interval and judged to be
due to an assignable cause of variation. This is an inter-
pretation of what Lewis (1934) meant by ā[. . .] know-
ing begins and ends in experience; but it does not end
in the experience in which it beginsā (p. 134). Based on
more knowledge, we may reinterpret the history that
gave us the knowledge we have.
We still have a task to think about how the process
should be handled after an improvement has been
made and found effective as described by the Juran tril-
ogy. In fact, we could proceed just as beforeāwe have
an improvement and we have to reformulate a priori in
order to predict the future. When starting to observe
the process it was natural to assume a noninformative a
priori distribution. However, after observing a positive
effect of an intervention it is not quite clear how a
renewed a priori would look like. A pessimistic choice
would probably be to use a noninformative a priori
before the observation that has just been seen to be
outside the predictive limitsāmost probably due to
the intervention madeāand then start anew from
there. As is often the case, the process has not yet
reached a steady state and thus it would be natural to
ļ¬nd the next point outside its predictive interval. The
procedure goes on until a new steady state may be
concluded.
21
Such an assumption could be based on a judgment concerning
statistics that are judged to be sufficient.
22
A small problem is that the prediction interval for an observation
at time s t is based on all observations including the one we want
to have a predictive interval for. If a recalculation should be made
for the predictive interval for the observation at time s, utilizing all
observations but that at time s is a matter of being purist or not.
Improvement in Health Care 27
Downloaded
by
[Svante
Lifvergren]
at
09:50
25
December
2014
14. Evidence-Based Medicine
The gold standard of clinical research has been the
randomized controlled trial for a long time. However,
some problems are surfacing. Too much focus has
been on differences in averages (or median), though of
course variation is also very important, as well as factors
underlying that variation. The Plavix case is an interest-
ing illustrationādue to genomics, 30% of patients do
not have the possibility to transform the drug into its
active component. Emphasis is also almost always on
one-factor-at-a-time experimentation. Furthermore, the
population used for the test was different from the one
in which the results will be utilized, and the controls
may be different and not adequate for the application
area. For a long time treatments that have been shown
to have efļ¬cacy for one type of population have been
used for a very different population (for example, the
elderly or small children). Thus, much more emphasis
should be put on methods utilized in industrial set-
tings: design of experiments, control charts, data taken
from the treatment processes, etc.
Control charts are an excellent illustration: Assume
that an intervention to a process under statistical con-
trol has been performed (after Phase II in Figure 4),
and that after the intervention the process again is
under statistical control but on a better level (better
mean, lower standard deviation etc.). Then there are
good reasons to assume that the intervention had a pos-
itive effect on the process, at least if no other changes
to the process are found.
We will not go further into these interesting areas
here, but there is a great potential in letting improve-
ment methodologies stimulate the development of
complementary medical research methodology.
THE NORMATIVE SIDE OF A SCIENCE
OF IMPROVEMENT
A science of improvement is not a science that is
purely objective and without a purpose; on the con-
trary, its purpose is quite clear. The knowledge needs
that a science of improvement has to fulļ¬ll are how
improvements are achieved and sustained. In a health
care context, it should be understood that improve-
ments are directed toward health in society: For
patients here and now and in the future and, more gen-
erally, for the health of the citizens here and now and
in the future. In the quality movement this has been
formulated as a value or a principle for organizing. As a
support to this main principle some other principles
have been promoted (see, e.g., Anderson et al. 1994;
Bergman and Klefsjā¬
o 2010; Cole and Scott 2010; Dean
and Bowen 1994; Hackman and Wageman 1995; ISO-
9000, International Organization for Standardisation
(ISO) 2000). As we emphasized earlier, it has to address
operational improvements not only at the operational
process level but on the organization as well as on the
overall health system level. An illustration is given in
Figure 5. The different principles are related to those
presented in Bergman and Klefsjā¬
o (2010) but with a
slightly different wording. These and similar principles
have been suggested in many other publications, in
research papers on the quality movement, and in more
popular books as well as in ISO-9000. Sometimes these
principles are formulated as values. Here we prefer to
talk about principles for organizingāprinciples that
should be possible to test for when decisions and other
kinds of actions are taken in the organization. Do they
follow or are they breaking against these principles? In
the next section we will discuss these principles. Then
we will discuss the relations between principles, practi-
ces, and tools.
Principles for Organizing
In Figure 5 the principles advocated in Bergman and
Klefsjā¬
o (2010) are displayed but with slightly different
formulations. Below, we only give a very brief descrip-
tionāmore detailed descriptions are found in text-
books like Bergman and Klefsjā¬
o (2010).
FIGURE 5 Principles for organizing; see Bergman and Klefsjā¬
o
(2010).
28 B. Bergman et al.
Downloaded
by
[Svante
Lifvergren]
at
09:50
25
December
2014
15. Take a Customer Perspective
This means that we always should try to understand
things from the patientās point of view and also to sup-
port the empowerment23
of the patient. We judge this
to be a huge revolution in health care because better
informed patients are able to be equal partners in the
curing and caring processes, resulting in higher quality
and lower costs. Of course, not all patients have the
possibility, but the lower the resource needs for capable
patients, the more there is for those who are not as
capable and therefore in need of more support. It is
also worth considering that health care in a strict meaning
seldom create direct value to the patientsāthe value is
created in the life world of the person who is, has been,
or is close to the current or former patient. Health careās
contribution is of cause very important, but it is only one
of the components in the creation of value.
As observed by Normann (2001), it seems to be a
general trend in different industries to change the way
customers are looked upon. First, customers are looked
upon as only consumers on a market, mass production
is in focus and the decisive competence is production
competence. Later on, customers are seen as a source of
information (i.e., what is quality from a customer point
of view), a customer basis becomes important and the
decisive competence is the ability to build relations
with the customers. In the ļ¬nal stage, the customer is
seen as a co-producer and even as a co-creator of the
value creation process. Value creation networks are in
focus and the creation of such networks becomes a
decisive competence.
Patient empowerment, co-production of care, and
even in some cases codesign of care are currently dis-
cussed.24
A case of the last kind has been performed as
an action research project on the rebuilding of neonatal
care at the Skaraborg Hospital Group (SkaS). The
method applied has been that of experienced based
co-design developed by Bate and Robert (2006) based
on concepts from the design industry. The method is
based on four key steps: capturing experiences from
patients and staff, understanding these experiences,
improving based on these experiences, and ļ¬nally mea-
suring the results. In all of these steps patients are
actively working ļ¬rst in their own group (the ļ¬rst and
part of the second step) and then together with the
staff. The experiences from the project at SkaS led to a
number of changesāexperiences from the patients
indicated improvement possibilities of not only trivial
kinds but regarding quite complex issues such as the
cooperation between different wards, which showed
new ways to coordinate work. Interestingly, the issues
noted by the patients differed markedly from those of
the staff (for a discussion of this and similar projects
see Gustavsson 2013).
Use Intelligence for Decisions about Future
Actions
How do we know about our customers preferences
and needs and wantsāhow are our decisions in line
with the life situation of the patient? Understanding
the situation of the patient is an important starting
point even though in decisions regarding treatment
many other factors have to be taken into account. It is
also important to understand the value creation pro-
cesses and their functioning. How do we measure and
with which results? But also intelligence concerning
the processes and results of providers that are similar
to us. From whom is it possible to learn? How do we
keep track of new research results of importance to
our activities? Earlier (Bergman and Klefsjā¬
o 2010) we
called this ābase decisions on factsā; however, replac-
ing this with āintelligenceā is a quest for an active
process for gathering necessary information of impor-
tance for decisions. In addition, a focus on a high
degree of intelligence is pivotal for learning in the
organization.
Focus on the Value Creation Processes
The value creation processes in a system is all of those
activities that repetitively create value. These activities
may be value streams as emphasized in Lean manufactur-
ing, more complex networks of activities, or shop solu-
tions.25
Value is more and more created in networks of
individualsāit has to be emphasized that these networks
not only consist of the patient and other patients and
health care personnel but also of others in society. For
23
Patient empowerment has grown to a very intense area of both
practical applications and theoretical research; see, for example,
Nordgren (2008), Ekman et al. (2012), and Gustavsson (2013).
24
For illustrations of how patients have been involved in the
creation of new health care solutions, see, for example, Bate and
Robert (2007), Iedema et al. (2010), and Maher and Baxter (2009).
25
Unfortunately, the concepts here have become blurredāin many
treatises processes are thought of as linear streams of activities;
that is, value streams. Here it is interpreted as all of those
arrangements that repetitively create value.
Improvement in Health Care 29
Downloaded
by
[Svante
Lifvergren]
at
09:50
25
December
2014
16. further discussions on processes see e.g. Hellstrā¬
om
(2007), and Hellstrā¬
om et al. (2010).
Improve Continuously
Improvement is, of course, at the heart of a science
of improvement but is also very demanding for an
organization that has to be able to, at the same time,
exploit the knowledge, capabilities, and capacities to
deliver value to its customers. Simultaneously, the
organization must explore new ways of working both
in small incremental improvement steps but also in
more innovative ways. Improvement in many cases
means innovationāit is important for health care to
rethink its capability and capacity for innovation; see
also Dixon-Woods, Alamberti et al. (2011). Continual
improvement is also very close to organisational learn-
ing.26
From a management point of view it is important
to understand and help the organization to build
learning mechanisms of different kinds: cognitive,
structural, and procedural learning mechanisms (see
Shani and Docherty 2003).
Support Employee Intrinsic Motivation
Earlier we discussed the importance of intrinsic
motivation. As described by Dixon-Woods et al.
(2014), most people in health care really want to help
their patientsāthey have a strong intrinsic motivation.
However, this is not always a principle that has been
used for organizing, which is a serious waste.
Take a Systems Perspective
Everything goes on in systemsātodayās health care
is a very complex system that might sometimes make it
unpredictable. System understanding should be uti-
lized as a basis for decision making.
Lead According to All of the Other
Principles
Good management and visionary leadership are
important assets in all organizations. The support of
employeesā inner motivation is an art. It is also a radical
shift from many contemporary leadership and manage-
ment ideals.
It should be noted that these principles are close to
those advocated in ISO-9000 (2000) and not very far
from what Anderson et al. (1994) described as concepts
in a theory underlying the Deming management
method.
Improvement Domains
It is easy to see that the principles described above
are closely related to the domains described in the brief
review above. The reason we have chosen to describe
similar things as principles rather than domains is to
clearly show the normative side of a science of
improvement. Of course, the selection of domains is
also governed by a normative will. However, it is not
very clearly expressed.
Principles, Practices, and Tools
A large number of practices and tools are related to
the principles and are necessary for the principles to be
applied in everyday work. In many treatises on quality
management, practices are not as clearly stated as
opposed to principles and tools. One reason for this
may be that practices have to be tailored to the context
of the organization and its operation. On the high level
of principles it is easy to give these principles a near-
universal applicability. The same goes for the precise
tools, whether quantitative or qualitative. However,
when it comes to practicesāthe actual work processes
applied in an organizationāthey have to be tailored to
the speciļ¬c situation and the tasks to be performed in
the organization. To give universally valid descriptions
of work practices would be hard even though recom-
mendations could be given.
REFLECTIONS
When reļ¬ecting upon an emerging science of
improvement in health care, a number of issues arise.
Some of these reļ¬ections are very practical with respect
to the development of health care in the future and
some more philosophical. Only a few points will be
given an initial reļ¬ection and some important ques-
tions and agendas for future research will be raised.
26
Learning and organizational learning are a controversial area
from a theoretical point of view; what should be meant by double
loop, triple loop, or deuteron learning are discussed by Argyris
(2003), Visser (2003), and Tosey et al. (2012). A related concept is
that of transformative learning as discussed by Mezirow and Asso-
ciates (2000).
30 B. Bergman et al.
Downloaded
by
[Svante
Lifvergren]
at
09:50
25
December
2014
17. Practical Reflections
Reļ¬ecting on the challenges that health care faces,
it could be asked whether it is really enough with a
science of improvement. Will the use of improve-
ment activities alone solve the problem or perhaps
only lock health care even stronger in the current
dead-end trajectory? Probably not. But technologies
exist that could provide us with a completely differ-
ent health care, but there seems to be no strong pull
for these new technologiesāhealth care is the last
industry taking information technology solutions to
heart. With a stronger focus on improvement
issuesāareas for reļ¬ection and an awareness of the
possibility to ļ¬nd better waysāwe believe that the
ability to adopt new ways of working utilizing new
technologies will be improved. Transformative learn-
ing is needed at all levels of health care and some-
times even on the patient side. But how is
transformative learning supported in a health care
system that is fully occupied with todayās problem
(and sometimes yesterdayās omissions)?
As emphasized by Walker et al. (2007), working with
improvements makes it easier to adopt more improve-
ments and even disruptive innovations (see also Moore
and Buchanan 2013). Today, skilful people work so
hard that they have no time to ļ¬nd solutions just
around the corner. āBest efforts are not enoughā
(Deming 1986, p. 19).
Philosophical Questions on an
Improvement Science
Is a science of improvement really a science? What is
a science? What is its ontology? What about its episte-
mology? How is it related to other kinds of sciences?
What about its universality? As a combination of dif-
ferent knowledge areas, it could possibly be character-
ized by what Gibbons and collaborators (1994) called
Mode 2 research. Many of its methods have been
inspired by statistical improvement from other areas of
application, but it is also about social change. How are
these different ontologies and epistemologies com-
bined in noncontradictory ways?
If we look upon health care as a service industry, we
should perhaps learn more from the knowledge area of
service management. It is often emphasized that value
is not created by the deliverer only but together with
the customer and in the customerās own processes.
This seems to be very much related to issues in the
patient empowerment movement. How far can we go
in that direction?
What about fashions and fads? Many researchers of
the critical management school are talking about the
problems with what they call ānew public manage-
ment.ā Is there something to take care of and be wor-
ried about from their critique? How is that done in a
constructive way? And whatever theories and sciences
we come up with, they are in a sense socially construc-
tedāthey cannot be otherwise (Hacking 1999). But the
real test is their usefulnessāstill a science of improve-
ment is only in an initial phase, but it is important that
its usefulness is proven not only on local scales but on
more universal ones!
. . . and the world moves onāwho wants to be left
behind?
ABOUT THE AUTHORS
Bo Bergman is partly retiring from a chair in Quality
Sciences at Chalmers University of Technology and
currently also a guest professor at Meiji University,
Tokyo. His career started with 15 years in aerospace
industry during which period he also became a PhD in
Mathematical statistics from Lund University 1978 and
was a part time professor in Reliability at Royal Insti-
tute of Technology, Stockholm. In 1983 he became a
professor in Quality Technology at Linkā¬
oping Univer-
sity and 1999 the SKF professor in Quality Manage-
ment at Chalmers University of Technology. He was a
co-founder of Centre for Healthcare Improvement
(CHI) at Chalmers and its ļ¬rst director (2004ā2009).
As a professor he has supervised a large number of PhD
students many of which now are themselves professors.
He is a member of the International Statistical Institute
(ISI) and an Academician of the International Academy
for Quality (IAQ).
Andreas Hellstrā¬
om is a senior lecturer at the Depart-
ment of Technology Management and Economics at
Chalmers University of Technology, Sweden and co-
director of Centre for Healthcare Improvement (CHI)
at Chalmers ā a research and education center focusing
on quality improvement, innovation, and transforma-
tion in health care. In 2007, he successfully defended
his Ph.D. thesis on the diffusion and adoption of man-
agement ideas. In 2013, he was one of four selected in
the ļ¬rst cohort of Vinnva
rdās Improvement Science
Improvement in Health Care 31
Downloaded
by
[Svante
Lifvergren]
at
09:50
25
December
2014
18. Fellowship. He has led research projects in both the pri-
vate sector and the public sector.
Svante Lifvergren, MD, is a specialist in internal and
respiratory medicine. He is quality director at the Skar-
aborg Hospital Group (SkaS), Sweden, having also
worked as a senior physician at SkaS since 1998. Fur-
ther, he holds a Ph.D. in Quality Sciences and he is
co-director for the Centre for Healthcare Improvement
(CHI) at Chalmers University of Technology. His
research mainly focuses on organizational development
in complex healthcare organizations.
Susanne M. Gustavsson, RN and Midwife, is Nurs-
ing Director at the Skaraborg Hospital Group (SkaS),
Sweden. She is also an external PhD student at the
Centre for Healthcare Improvement (CHI) at Chalm-
ers University of Technology, Gothenburg Sweden.
Her research is focused on experience-based co-design
with patients and relatives in health care processes.
ACKNOWLEDGMENTS
First, we want to thank all those devoted health care
personnel who have participated in our research and in
our courses and have given us tremendous support. In
particular, the Skaraborg Hospital Group has been very
important as well as the Vā¬
astra Gā¬
otaland Region.
FUNDING
We want to acknowledge the important support
from the research foundation VINNVA
RD, providing
resources for our health care improvement research
journey.
REFERENCES
Alā¬
ange, S., Steiber, A. (2011). Diffusion of organisational innovations: an
empirical test of an analytical framework. Technology Analysis
Strategic Management, 23(8):881ā897.
Anderson, J. C., Rungtusanatham, M., Schroeder, R. G. (1994). A theory
of quality management underlying the Deming management
method. The Academy of Management Review, 19(3):472ā509.
Anderson, R. M., Funnell, M. M. (2005). Patient Empowerment: reflec-
tions on the challenge of fostering the adoption of a new para-
digm, Patient Education and Counceling, 57: 153ā157.
Andersson-Gā¬
are, B., Neuhauser, D. (2007). The health care quality jour-
ney of Jonkoping County Council, Sweden. Quality Management
in Health Care, 16(1):2ā9.
Argyris, C. (2003). Retrospective organizational learning: A theory of
action perspective. Academy of Management Executive, 17(2):
37ā55.
Ashby, D. (2006). Bayesian statistics in medicine: A 25 year review. Statis-
tics in Medicine, 25:3589ā3631.
Aujoulat, I., dāHoore, W., Deccache, A. (2007). Patient empowerment in
theory and practice: Polysemy or cacophony? Patient Education
and Counseling, 66:13ā20.
Axelrod, R., Cohen, M. D. (1999). Harnessing Complexity Organizational
Implications of a Scientific Frontier. New York: Free Press.
Batalden, P. B., Davidoff, F. (2007). What is āquality improvementā and
how can it transform healthcare? Quality and Safety in Health
Care, 16:2ā3.
Batalden, P., Davidoff, F., Marshall, M., Bibby, J., Pink, C. (2011). So
what? Now what? Exploring, understanding and using the episte-
mologies that inform the improvement of healthcare. BMJ Quality
and Safety, 20:i99āi105.
Batalden, P. B., Stoltz, P. (1993). A framework for the continual
improvement of health care; building and applying professional
and improvement knowledge to test changes in daily work.
The Joint Commission Journal on Quality Improvement, 19
(10):432ā452.
Bate, P. (2014). Context is everything, 1ā30. London: The Health
Foundation.
Bate, P., Robert, G. (2006). Experience-based design: From redesigning
the system around the patient to co-designing services with the
patient. Quality and Safety in Health Care, 15:307ā310.
Bate, P., Robert, G. (2007). Experience-based design and a case study
toward more user-centric OD: Lessons from the field of experience-
based design and a case study. Journal of Applied Behavioral Sci-
ence, 43(41):41ā66.
Benneyan, J. C., Lloyd, R.C., Plsek, P. E. (2003). Statistical process control
as a tool for research and healthcare improvement. Quality and
Safety in Healthcare, 12(6): 458ā464.
Bergman, B. (2003). On the profound knowledge of variation. In The
Best on Quality Book Series of the International Academy for
Quality, 2ā13. Milwaukee: American Society for Quality, Quality
Press.
Bergman, B. (2009). Conceptualistic pragmatism: A framework for
Bayesian analysis? IIE Transactions, 41:86ā93.
Bergman, B. (2013). ā¬
Annu Bā¬
attre Va
rdāVad Kan vi Lā¬
ara fra
n Variationen i
ā¬
Oppna Jā¬
amfā¬
orelser? [Even better careāWhat can we learn from
variation in the Open Comparisons?]. Stockholm, Sweden: SKL.
Bergman, B., Chakhunashvili, A. (2007). In weak statistical control? Inter-
national Journal of Six Sigma and Competitive Advantage, 3(1):
91ā101.
Bergman, B., Klefsjā¬
o, B. (2010). Quality from Customer Needs to Cus-
tomer Satisfaction. 3rd ed. Lund, Sweden: Studentlitteratur.
Bergman, B., Neuhauser, D., Provost, L. (2011). Five main processes in
healthcareāA citizen perspective. BMJ Quality and Safety, 20:i41ā
i42.
Berwick, D. M. (1989). Continuous improvement as an ideal in health
care. New England Journal of Medicine, 320(1):53ā56.
Berwick, D. M. (2008). The science of improvement. JAMA, 299
(10):1182ā1184.
Berwick, D. M., Blanton Godfrey, A., Roessner, J. (1990). Curing Health
Care. New Strategies for Quality Improvement. 1st ed. San Fran-
cisco: Jossey-Bass.
Best, M., Neuhauser, D. (2004). Avedis Donabedian: Father of quality
assurance and poet. Quality and Safety in Health Care, 13:472ā473.
Chassin, M. R., Garvin, R. W. (1998). The urgent need to improve health
care quality. Institute of Medicine National Roundtable on Health
Care Quality. JAMA, 280: 1000ā1005.
Coghlan, D., Brannick, T. (2010). Doing Action Research in Your Own
Organization. 3rd ed. Thousand Oaks, CA: Sage.
Cole, R. E. (1999). Managing Quality Fads. How American Business
Learned to Play the Quality Game. Milwaukee, WI: ASQ Quality
Press.
Cole, R. E., Scott, W. R. (2000). The Quality Movement Organization
Theory. London: Sage.
Crossan, M. M., Lane, H. W., White, R. E. (1999). An organizational
learning framework: From intuition to institution. The Academy of
Management Review, 24(3):522ā553.
32 B. Bergman et al.
Downloaded
by
[Svante
Lifvergren]
at
09:50
25
December
2014
19. Csikszentmihalyi, M. (1990). Flow: The Psychology of Optimal Experi-
ence. New York: Harper and Row.
Csikszentmihalyi, M., Nakamura, J. (2011). Positive psychology: Where
did it come from, where is it going? In Sheldon, K. M., Kashdan, T.
B., Steger, M. F., Eds. Designing Positive Psychology. New York:
Oxford University Press, 2ā9.
Dawkins, R. (1976). The Selfish Gene. Oxford: Oxford University Press.
Dean, J. W., Bowen, D. E. (1994). Management theory and total quality:
Improving research and practice through theory development.
Academy of Management Review, 19(3):392ā418.
Deci, E. L., Ryan, R. M. (2002). Handbook of Self-Determination Theory.
Rochester, NY: University of Rochester Press.
de Finetti, B. (1974). Theory of Probability. Vols. 1ā2. New York: Wiley.
Deming, W. E. (1986). Out of the Crisis. Cambridge, MA: MIT Press.
Deming, W. E. (1994). The New Economics. 2nd ed. Cambridge, MA: MIT
Press.
Dennett, D. (1995). Darwinās Dangerous Idea: Evolution and the Mean-
ings of Life. New York: Simon Schuster.
Dixon-Woods, M., Amalberti, R., Goodman, S., Bergman, B., Glasziou, P.
(2011). Problems and promises of innovation in healthcare: Why
healthcare needs to rethink its love/hate relationship with the new.
BMJ Quality and Safety, 20(Suppl. 1):i47āi51.
Dixon-Woods, M., Baker, R., Charles, K., et al. (2014). Culture and
behaviour in the English National Health Service: Overview of les-
sons from a large multimethod study. BMJ Quality and Safety,
23:106ā115.
Dixon-Woods, M., Bosk, C. L., Aveling, E. L., Goeschel, C. A., Pronovost,
P. J. (2011). Explaining Michigan: Developing an ex post theory of a
quality improvement program. Milbank Quarterly, 89:167ā205.
Docherty, P., Huzzard, T., de Leede, P., Totterdill, P. (2003). Home and
Away. Learning in and Learning from Organizational Networks in
Europe. London: Innoflex.
Docherty, P., Shani, A. B. (2008). Learning mechanisms as means and
ends in collaborative management research. In Shani, A. B., Mohr-
man, S. A., Pasmore, W. A., Stymne, B., Adler, N., Eds. Handbook
of Collaborative Management Research, 163ā181. Thousand Oaks,
CA: Sage.
Donabedian, A. (1992). The Role of Outcomes in Quality Assessment and
Assurance. Quality Review Bulletin (QRB), 18: 356ā360.
Donabedian, A. (2000). Evaluating Physician Competence. Bulletin of the
World Health Organization, 78(6): 857ā860.
Donabedian, A. (2003). An Introduction to Quality Assurance in Health
Care. Oxford: Oxford University Press.
Ekman, I., et al. (2012). Effects of person-centred care in patients with
chronic heart failure: The PCC-HF study. European Heart Journal,
33(9):1112ā1119.
Gibbons, M., Limoges, C., Nowotny, H., Schwartzman, S., Scott, P.,
Trow, M. (1994). The new production of knowledge: The dynamics
of science and research in contemporary societies. London: Sage.
Gustavsson, S. (2013). The customer in focus: Patient roles in designing
and improving care processes. Licentiate dissertation. Chalmers
University of Technology, Gothenburg.
Gustavsson, S. (2014). Improvements in neonatal care using experience-
based co-design. International Journal of Health Care Quality Assur-
ance, 27(5):427ā438.
Hacking, I. (1999). The Social Construction of What? New York: Harvard
University Press.
Hackman, J. R., Wageman, R. (1995). Total quality management. Empiri-
cal, conceptual, and practical issues. Administrative Science Quar-
terly, 40:309ā342.
Health Foundation. (2014). Perspectives on Context. London: The Health
Foundation.
Hellstrā¬
om, A. (2007). On the diffusion and adoption of management
ideas. Ph.D. dissertation. Chalmers University of Technology, Goth-
enburg, Sweden.
Hellstrā¬
om, A., Lifvergren, S., Quist, J. (2010). Process management in
healthcareāInvestigating why it is easier said than done. Journal of
Manufacturing Technology, 21:499ā511.
Iedema, R., Merrick, E., Britton, K., Gray, J., Verma, R., Manning, N.
(2010). Codesigning as a discursive practice in emergency health
services: The architecture of deliberation. Journal of Applied Behav-
ioral Science, 46:73ā91.
Institute of Medicine. (1999). To Err Is Human: Building a Safer Health
System. Washington: National Academy Press.
Institute of Medicine. (2001). Crossing the Quality Chasm: A New Health
System for the 21st Century. Washington: National Academy Press.
International Organization for Standardisation (ISO). (2000). ISO
9000:2000 Quality management systems. Fundamentals and
vocabulary. Geneva: International Organization for Standardisation.
Ishikawa, K. (1982). Guide to Quality Control. Tokyo: Asian Productivity
Press.
Ishikawa, K. (1985). What Is Total Quality Control? The Japanese Way.
1st ed. Englewood Cliffs, NJ: Prentice-Hall.
Jablonka, E., Lamb, M. J. (2005). Evolution in Four Dimensions: Genetic,
Epigenetic, Behavioural, and Symbolic Variation in the History of
Life. Cambridge, MA: MIT Press.
James, J. T. (2013). A new, evidence-based estimate of patient harms
associated with hospital care. Journal of Patient Safety, 9(3):127ā
142.
Juran, J. M. (1986). The quality trilogy, a universal approach to managing
for quality. Quality Progress, 19(8):18ā24.
Kahneman, D. (2011). Thinking, Fast and Slow. New York: Farrar, Strauss
Giroux.
Kuhn, T. S. (1962). The Structure of Scientific Revolutions. Chicago: Uni-
versity of Chicago Press.
Langley, G. J., Nolan, K. M., Norman, T. W., Provost, L. P. (1996). The
Improvement Guide. San Francisco: Jossey Bass.
Lewin, K. (1948). Action research and minority problems. In Lewin,
G. E., Ed. Resolving Social Conflicts Field Theory in Social Sci-
ence, Washington, DC: American Psychological Association,
201ā216.
Lewis, C. I. (1929). Mind and the World Order. New York: Dover
Publications.
Lewis, C. I. (1934). Experience and meaning. The Philosophical Review,
43(2):125ā146.
Lifvergren, S. (2013). Quality improvement in healthcare experien-
ces from two longitudinal case studies using an action
research approach. Ph.D. dissertation. Centre for Healthcare
Improvement, Chalmers University of Technology, Gothen-
burg, Sweden.
Lifvergren, S., Andin, U., Huzzard, T., Hellstrā¬
om, A. (2012). The path to
sustainability in healthcareāExploring the role of learning micro-
systems. In Mohrman, S., Shani, A. B., Eds. Organizing for Sustain-
able Healthcare. Vol. 2. Bingley, UK: Emerald, 169ā197.
Lifvergren, S., Chakhunashvili, A., Bergman, B., Docherty, P. (2008).
Online statistical monitoring of critical patient data increases
patient safety. Proceedings of the PICMET-Conference, Capetown,
South Africa, July 27ā31.
Lifvergren, S., Docherty, P., Hellstrā¬
om, A. (2010). Management by dia-
logue: Joint reflection, sense making and development. Proceed-
ings of A Time for Change: Restructuring Americaās Health Care
Delivery SystemāCornell Universityās International Health Care
Conference, May 11ā12, New York.
Lifvergren, S., Docherty, P., Shani, A. B. (2011). Toward a sustainable
healthcare system: Transformation through participation. In Mohr-
man, S., Shani, A. B., Eds. Organizing for Sustainable Effectiveness,
Vol. 1. Bingley, UK: Emerald, 99ā125.
Lifvergren, S., Gremyr, I., Hellstrā¬
om, A., Chakunashvili, A., Bergman, B.
(2010). Lessons from Swedenās first large-scale implementation of
Six Sigma in healthcare. Operations Management Research, 3:117ā
128.
Maher, L., Baxter, H. (2009). Working in partnership with service users.
British Journal of Healthcare Management, 15(4):172ā175.
Marshall, M. (2011). Seeing with new eyes: What can general practice
learn from the science of improvement. British Journal of General
Practice, 61(586): 352ā355.
Improvement in Health Care 33
Downloaded
by
[Svante
Lifvergren]
at
09:50
25
December
2014
20. Marshall, M., Pronovost, P., Dixon Woods, M. (2013). Promotion of
improvement as a science. Lancet, 318:419ā421.
Mauleon, C., Bergman, B. (2009). Exploring the epistemological origins
of Shewhartās and Demingās theory of quality. Influences from
Lewis conceptualistic pragmatism. International Journal of Quality
and Service Sciences, 1(2):160ā171.
Mezirow Associates (2000). Learning as Transformation, Critical Per-
spectives on a Theory in Progress. San Francisco, CA: Jossey-Bass.
Moore, C., Buchanan, D. A. (2013). Sweat the small stuff: A case study of
small-scale change processes and consequences in acute care.
Health Service Management, 26(1):9ā17.
Neuhauser, D. (2002). Heroes and martyrs of quality and safety: Ernest
Amory Codman MD. Quality and Safety in Health Care, 11:104ā
105.
Neuhauser, D., Provost, L., Bergman, B. (2011). The meaning of variation
to healthcare managers, clinical and health-services researchers,
and individual patients. BMJ Quality and Safety, 20:i36āi40.
Nordgren, L. (2008). The performativity of the service management dis-
course: āValue creating customersā in health care. Journal of
Health Organization and Management, 22(5):510ā528.
Normann, R. (2001). Reframing Business: When the Map Changes the
Landscape. Chichester, UK: Wiley.
Parry, G. J., Mate, K., Perla, R. J., Provost, L. (2013). Promotion of
improvement as a science. Lancet, 381:1901ā1902.
Perla, R. J., Provost, L. P., Parry, G. J. (2013). Seven propositions of the sci-
ence of improvement: Exploring foundations. Quality Management
Health Care, 22(3):170ā186.
Petersen, P. B. (1999). Total quality management and the Deming
approach to quality management. Journal of Management History,
5(8):468ā488.
Pettigrew, A. (1987). Context and action in the transformation of the
firm. Journal of Management Studies, 24(6): 649ā670.
Pettigrew, A. M., Whipp, R. (1991). Managing Change for Competitive
Success. Oxford: Blackwell.
Pink, D. H. (2010). Drive: The Surprising Truth about What Motivates Us.
New York: Riverhead Books.
Ramsay, F. (1926). Truth and probability. In The Foundations of Mathe-
matics and other Logical Essays, Ch. VII, p.156ā198, edited by R. B.
Braithwaite, London: Kegan, Paul, Trench, Trubner Co., New
York: Harcourt, Brace and Company.
Reach, G. (2014). Patient autonomy in chronic caring: Solving a paradox.
Patient Preference and Adherence, 2014(8): 15ā24.
Savage, L. J. (1954). The Foundations of Statistics. New York: John Wiley
and Sons.
Senge, P. M. (1990). The Fifth Discipline; the Art and Practice of the
Learning Organization. New York: Doubleday Currency.
Shani, A. B., Docherty, P. (2003). Learning by Design: Building Sustain-
able Organizations. Oxford: Blackwell.
Shewhart, W. A. (1931). Economic Control of Quality of Manufactured
Product. New York: Van Nostrand.
Shewhart, W. A. (1939). Statistical Method from the Viewpoint of Quality
Control. Washington, DC: Graduate School of the Department of
Agriculture.
Shojania, K. G., Grimshaw, J. M. (2004). Still no magic bullets: Pursuing
more rigorous research in quality improvement. American Journal
of Medicine, 116:778ā780.
Sjohania, K., et al. (2007). How quickly do systematic reviews go out of
date? A survival analysis. Annals of Internal Medicine, 147:224ā
233.
Stacey, R. D. (2003). Strategic Management and Organizational Dynam-
ics. The Challenge of Complexity. 4th ed. Harlow, UK: Prentice Hall.
Swedish Association of Local Authorities and Regions (SALAR). (2008).
Swedish healthcare in international comparison (in Swedish.
http://webbutik.skl.se/bilder/artiklar/pdf/7164-363-6.pdf (accessed
November 11, 2014).
Swedish National Board of Health and Welfare. (2006). God Va
rdāOm
ledningssystem fā¬
or kvalitet och patientsā¬
akerhet I hā¬
also- och sjukva
r-
den [Good careāOn management system for quality and safety in
healthcare]. Stockholm: Swedish National Board of Health and
Welfare.
Sztompka, P. (1991). Society in Action: The Theory of Social Becoming.
Chicago: University of Chicago Press.
Tosey, P., Visser, M., Saunders, M. N. K. (2012). The origins and concep-
tualizations of ātriple-loopā learning: A critical review. Manage-
ment Learning, 43(3):291ā307.
Visser, M. (2003). Gregory Bateson on deutero-learning and double bind:
A brief conceptual history. Journal of History of the Behavioral Sci-
ences, 39(3):269ā278.
Waldrop, M. M. (1992). Complexity, The Emerging Science at the Edge of
Order and Chaos. New York: Simon and Schuster.
Walker, J. H., Armenakis, A. A., Bernerth, J. B. (2007). Factors influencing
organizational change efforts: An integrative investigation of
change content, context, process and individual differences. Jour-
nal of Organizational Change Management, 20(6):761ā773.
Weick, K. E., Sutcliffe, K. M. (2001). Managing the Unexpected: Assuring
High Performance in an Age of Complexity. University of Michigan
Pressing Problem Series. San Francisco: Jossey-Bass.
Weisz, J. A., Mohrman, S. A., McCracken, A. (2013). Itās A Great Time to
Be a Physician: Building a Healthcare System That Works. Bozeman,
MT: Second River Healthcare.
Wilcox, M. (2004). Prediction and pragmatism in Shewhartās theory of
statistical control. Journal of Management History, 42(1):152ā165.
Wilmott, H. (1997). Management and organization studies as science?
Organization, 4:3309ā3344.
Woodall, W. H. (2006). The use of control charts in health-care and pub-
lic-health surveillance, with discussion. Journal of Quality Technol-
ogy, 38(2):89ā134.
Woodall, W. H., Mongomery, D. C. (2014). Some current directions in
the theory and application of statistical process monitoring. Journal
of Quality Technology, 46(1):78ā94.
34 B. Bergman et al.
Downloaded
by
[Svante
Lifvergren]
at
09:50
25
December
2014