40 Quality in Health Care 2001;10:40–48

Management matters: the link between hospital

organisation and quality of patient care
Elizabeth West

Abstract
Some hospital trusts and health authori- Key messages
ties consistently outperform others on dif- + Studies linking the organisation and
ferent dimensions of performance. Why? management of health care to patient and
There is some evidence that “manage- staV outcomes, mainly conducted in the
ment matters”, as well as the combined USA, can be criticised both theoretically
eVorts of individual clinicians and teams. and methodologically.
However, studies that have been con- + There are currently no high quality stud-
ducted on the link between the organis- ies of these relationships in the UK.
ation and management of services and + This paper identifies key review articles
quality of patient care can be criticised of studies, both in health care and in
both theoretically and methodologically. A business, that might throw new light on
larger, and arguably more rigorous, body the determinants of hospital perform-
of work exists on the performance of firms ance.
in the private sector, often conducted + Research on the performance of business
within the disciplines of organisational firms suggests the importance of decen-
behaviour or human resource manage- tralised decision making, staV participa-
ment. Studies in these traditions have tion and involvement, innovative work
focused on the eVects of decentralisation, practices, and the “fit” between struc-
participation, innovative work practices, ture, strategy, and environment.
and “complementarities” on outcome + Future research could be improved by
variables such as job satisfaction and per- greater attention to the mechanisms that
formance. The aim of this paper is to might plausibly link, for example, staV
identify a number of reviews and research variables to patient outcomes by adopt-
traditions that might bring new ideas into ing longitudinal rather than cross sec-
future work on the determinants of hospi- tional research designs and by using
tal performance. Ideally, future research appropriate statistical methods such as
should be more theoretically informed multilevel modelling.
and should use longitudinal rather than
cross sectional research designs. The use
of statistical methods such as multilevel empirical evidence, ways of organising that
modelling, which allow for the inclusion of could improve patient care.
variables at diVerent levels of analysis, One of the criticisms of research on hospital
would enable estimation of the separate performance is that it has been rather insular,
contribution that structure and process paying little attention to developments in
make to hospital outcomes. related fields such as organisational sociology,
(Quality in Health Care 2001;10:40–48) organisational behaviour, management studies,
or human resource management. Most of these
Keywords: hospital organisation; hospital performance; disciplines study organisational performance in
management; quality of care the context of a market and their dependent
variables are usually profitability, productivity,
or market share which are very diVerent from
Introduction many of the proxies for quality of care—such as
Organisational researchers have long sought to mortality or morbidity—used in studies of hos-
establish the impact of organisational struc- pital performance. However, these reports are
tures and managerial processes on outcomes similarly concerned with issues of motivating,
such as profitability,1 eVectiveness,2 perform- engaging, and rewarding staV which may be
ance,3 and organisational growth and survival.4 linked to patient outcomes as well as to
Organisational researchers have also focused business success. Greater attention to the work
Royal College of on the public sector, particularly hospitals, in that has been done on organisational perform-
Nursing, RadcliVe an eVort to link organisational characteristics ance, broadly defined, could illuminate our
Infirmary, Woodstock to a number of important outcomes for attempts to link the characteristics of hospitals
Road, Oxford patients and staV.5 6 Although few would now and units to the kind of care they are able to
OX1 6HE, UK
E West, senior research
question that “management matters” in deliv- provide to patients.
fellow ering quality health care, knowledge about the Of course, the disciplines of organisational
nature of the relationship is incomplete. The sociology and human resource management
Correspondence to: fact that we know so little about the relation- are vast and the aims of this paper are modest.
Dr E West ship between structures, processes, and out- It is impossible to treat the literature on these
elizabeth.west@rcn.org.uk
comes within hospitals makes it diYcult to re- subjects in great depth here. The main aim of
Accepted 21 December 2000 commend, on the basis of sound theory and this paper is to identify a number of “landmark

www.qualityhealthcare.com
Management matters 41

studies”, defined as frequently cited review and fatigued doctors, are not a recipe for excel-
articles, that try to make sense of the burgeon- lence in patient care.10 So, how can we improve
ing literature on organisational performance. the quality of working life in ways that will
These studies could contribute to the develop- enhance the ability of the NHS to recruit and
ment of theory in this area. A second aim is to retain staV? Pay, flexible hours, and job
identify variables at diVerent levels of prospects are obviously central, but improving
analysis—individual, organisational, and the quality of working life also means helping
environmental—that could be used in future individuals to develop their potential, to
models of hospital organisation and quality of increase their sense of autonomy, and the abil-
patient care. ity to achieve their goals. At the same time,
attention needs to focus on organisational
Health policies motivating organisational development. Creating an environment that is
research perceived as “a good place to work” requires
The message from the current UK government multiple interventions at diVerent levels.
that quality of care must be given greater prior- Clinical governance and better human re-
ity than in the past has been widely welcomed source management practices are important
by the professions. Some of the main policy planks in the current health policies emphasis-
documents relating to quality of care in the UK ing quality of patient care. Both planks demand
National Health Service are described in table attention, not just to the individual level of
1. Within the quality initiative there is a clear analysis, but to the ways that clinical directo-
recognition that only so much can be achieved rates, divisions, trust boards, and professions
by appealing to individual practitioners, and work together to achieve quality. These goals
that more eVort needs to be expended on move organisational research onto the centre
understanding how the organisation and man- stage.
agement of care aVects outcomes. Many of the
goals of the new NHS—including clearer lines Organisational research focusing on
of accountability and responsibility, better hospitals
communication, and improved conditions for Studies of the organisation and management of
staV—require interventions at the level of the hospitals have examined the impact of a dizzy-
organisation. ing array of factors on the quality of patient
One of the most important planks in the care. Flood,11 in a wide ranging review of
quality platform is the policy of clinical organisational research conducted mainly in
governance. Clinical governance has been the USA in the 1980s, identified the basic
defined as “a framework through which NHS sources of variation that were found to be asso-
organisations are accountable for continuously ciated with quality of patient care.
improving the quality of their services and A number of studies have found a weak rela-
safeguarding high standards of care by creating tionship between doctors’ training and experi-
an environment in which excellence in clinical ence and quality of care. Flood11 has inter-
care can flourish”.7 Buetow and Roland8 noted preted this to mean “. . . not that physicians are
that the “duty of quality” relates to the organ- unimportant for quality but that organisational
isation, not just to individuals within the context is far more important in setting limits
organisation. Although a named individual, (upper and lower) for physicians than formerly
most often the Chief Executive, will assume recognised . . .”. Medical staV organisation—
statutory responsibility for quality, many trusts including peer review, selection and continued
have already implemented structural changes, review of new staV members, and participation
creating new layers of management and estab- in policy making committees—have also been
lishing new committees to enable them to meet shown to be positively related to quality of
the challenge of clinical governance. Clinical patient care.
governance also demands cultural change Few studies have examined whether a similar
towards openness, participation, staV empow- set of relationships hold for other staV, but stud-
erment, partnership, and collaboration; an ies of coordination and communication have
important goal is to move away from a culture focused on nurses and ancillary staV. Coordina-
based on blame to one that emphasises tion appears to be particularly significant, and a
learning from mistakes.9 The emphasis on the series of studies conducted in intensive care
need for structural and cultural change in both found that “conflict management skills, includ-
organisations and professions recognises that ing communication, problem solving and
not all the quality goals of the NHS can be leadership, combined with a patient orientation”
achieved by inducing or exhorting individual were positively related to quality of patient
clinicians and managers to change their own care.11 Flood suggests that one promising area
practice. for future research will be the extent to which the
The quality of patient care may be related in boundary between the two traditional authority
an important way to the quality of life structures—professional and administrative—
experienced by staV at work. Partly as a result are breached in hospital organisations.
of the quality initiative, concern about the way There is a well established relationship
the NHS treats its employees has increased. between the volume of patients passing
Issues of human resource management have through a health care unit and the quality of
also been highlighted by the projected crisis in care delivered,12 although there is disagreement
the number of nurses and by the dissatisfaction as to the mechanism generating this relation-
of junior doctors with their working hours. Too ship. The literature proposes at least five plau-
few trained nurses, combined with overworked sible hypotheses,11 two of which rely on the idea

www.qualityhealthcare.com
42 West

Table 1 Quality initiatives for the NHS in England introduced since 1997 (excluding documents relating to the development and regulation of the professions)

Year Source Type of document Title Quality initiatives

1997 DoH White paper The new NHS: modern, dependable Describes a 10 year plan to improve the NHS, including the
replacement of the internal market with a system of
integrated care. NHS to be primary care led. Introduced
national service frameworks, new organisational structures
to promote evidence-based care and monitor standards
(NICE and CHI), the policy of clinical governance and
NHS Direct
1998 NHSE Consultation document A first class service: quality in the new NHS Announced a three part approach:
+ National standards to be set by National Service
Frameworks and the National Institute for Clinical
Excellence
+ Dependable local delivery systems to be achieved through
clinical governance, lifelong learning for NHS staV, and a
system of self-regulation
+ Monitoring by the Commission for Health Improvement,
a national framework for performance assessment and an
annual survey of patients’ and users’ experiences
Changes to be implemented through partnerships among
government, NHS and related organisations and patients
1998 DoH Green paper Our healthier nation Key aims are to improve the health of the population as a
whole by increasing the length of people’s lives and number
of years people live free from illness. Describes the “third
way” between blaming individuals and nanny state social
engineering. Third way is a contract whereby the
government, local communities, and individuals join in
partnership to improve health
1998 DoH Working together: securing a quality workforce A strategic approach to managing human resources in the
for the NHS NHS. Three aims:
+ to ensure that we have a quality workforce, in the right
numbers, with the right skills and diversity, organised in the
right way, to deliver the government’s service objectives for
heath and social care;
+ to demonstrate that we are improving the quality of
working life for staV;
+ to address the management capacity and capability
required to deliver this agenda and the associated
programme of change.
1999 NHSE Implementation paper The NHS performance assessment framework Introduced a broader based approach to assessing
performance in six areas:
+ health improvement;
+ fair access to services, irrespective of geography,
socioeconomic group, ethnicity, age or sex;
+ eVective delivery of appropriate health care—care must
be eVective, appropriate and timely, and must comply with
agreed standards;
+ eYciency—to ensure that eVective care is delivered and
that the NHS achieves value for money;
+ patient/carer experience—to assess the way people view
their care to ensure the NHS is sensitive to individual
needs;
+ health outcomes—to assess the contribution of the NHS
to the health of the population.
1999 NHSE Improving quality and performance in the new NHS: First set of data on high level performance indicators and
NHS performance indicators clinical indicators to enable health authorities, primary care
groups, and NHS trusts to monitor and compare their
performance
1999 DoH Implementation paper Clinical governance: quality in the new NHS Sets out the arrangements for implementing clinical
governance. Regulates NHS organisations to put in place:
+ clear lines of accountability and responsibility for the
quality of clinical care;
+ a comprehensive programme of quality improvement
activities;
+ clear policies aimed at managing risk;
+ procedures for all professional groups to identify and
remedy poor performance
1999 DoH White paper Saving lives: our healthier nation An action plan to tackle poor health. Focused attention on
cancer, coronary heart disease and stroke, accidents, and
mental illness, and set targets for reduction of incidence.
Established the Health Development Agency and the
Public Health Development Fund
2000 DoH White paper The NHS plan: a plan for investment, a plan for reform + A modernisation board, 10 task forces and a performance
working group to oversee, advise, and drive forward the
implementation of the plan
+ New incentives for better performance and reward
schemes
+ New modern and better equipped buildings
+ More staV working in better conditions with
opportunities for career development and advanced training
+ New educational systems and expansion of medical
students, nurses and other health professionals
+ New structures for regulating health professions
+ New ways of involving and representing patients views
+ New strategy for public health and prevention of disease
+ New technologies, such as Care Direct
2000 DoH Consultation document A health service for all talents: developing the NHS A review of the workforce and its educational requirements.
workforce Established a National Workforce Development Board
2000 NHSE Improving quality and performance in the new NHS: Second set of data on performance
NHS performance indicators

www.qualityhealthcare.com
Management matters 43

that “practice makes perfect”—that is, the skills the network of relationships in which hospitals
of individual practitioners are enhanced by operate. The omission of environmental and
specialisation and by repeated performance of relational variables would be particularly egre-
the same or similar tasks. Highly skilled and gious in models of quality in the NHS where
specialised practitioners also provide better the links between the organisation and the
peer review. A third mechanism involves units healthcare system are particularly important.
with good reputations attracting more referrals Flood also criticises the lack of attention to
and consequently having a high volume. It has culture as an important influence on manage-
also been suggested that high volumes are rial decision making. Future research should
associated with a more preventative orientation try to make some theoretical progress in this
among a group of doctors, with patients being area which will help to explain how organisa-
treated at an earlier stage of their illness. tional structures and processes, as well as the
Finally, some studies have suggested that “. . . internal and external environments, are related
volume of similar cases leads to benefits to quality of care. The problem of causal
because the organisation and its staV become ordering, which is ubiquitous in organisational
more practised in managing and caring for research, can only really be addressed by longi-
these patients or because certain eYciencies tudinal rather than by cross sectional research
can be introduced with suYcient volume much designs.
akin to economies of scale”.11 Whereas Flood11 focused her review on the
Flood surmises that many diVerent mecha- independent variables, a recent review has
nisms may be operating at once to produce the examined the variety of dependent variables
relationship between volume and quality, but it that are frequently used in studies of hospital
is clear that extent of specialisation of staV, vol- performance. Mitchell and Shortell13 set out to
ume of patients, and case mix are important examine “the state of the science with respect
variables in relation to quality of patient care. to morbidity, mortality and adverse events
Complexity can take many forms—for exam- indicative of organisational variables in care
ple, the severity of each individual patient’s ill- delivery systems”. They found a total of 81
ness, the frequency of multiple diagnoses, and studies in this area, most of which were
the number of patients who have combined conducted in acute care settings in the USA
health and social problems—which require the during the 1990s. The independent variables in
coordination of a large number of clinicians these studies were more frequently features of
and services. It also includes characteristics of the organisational structure (high technology,
the work, such as whether or not admission nurse staYng, professional expertise, size,
patterns are predictable. Complexity could ownership, urban/rural location, teaching hos-
plausibly be related to quality of care and con- pital status) than organisational or clinical
tingency theory would suggest that some processes (collaboration and care coordina-
managerial approaches will work for some tion, volume of patients).
groups of patients and types of services and not In general, they found that this body of work
for others. provides some support for the conclusion that
A number of stable characteristics of hospi- nursing surveillance, quality of working envi-
tals have also been related to outcomes. One ronment, and quality of interaction among
consistent finding is that quality of care is bet- professionals distinguish hospitals with lower
ter in hospitals aYliated to a major medical mortality and complications from those with
school. Findings over the last 30 years have higher rates of adverse events. However,
shown, at least in the USA, that teaching empirical results are ambiguous. As researchers
hospitals aYliated to a major medical school become increasingly adept at controlling for
tend to be associated with higher costs, better patient factors such as severity of illness, the
quality outcomes, and more sophisticated organisation and managerial variables tend to
techniques, after taking into account patient decrease in significance.
mix. The main contribution of this review is the
Flood11 concluded from her review of studies fact that they distinguish between the diVerent
of health care organisations that, although kinds of dependent variables—mortality, com-
much of this research can be criticised both plications (surgical complications, infections),
theoretically and methodologically, there is at and other adverse events (falls, pressure sores,
least some support for the relationship between and medication errors). They show that these
quality of care and a number of variables. The three are not interchangeable and suggest that
most serious deficiencies in this body of there are problems in using each as a proxy for
research lie in the failure to specify the quality of care. The authors argue that
mechanism linking organisational characteris- mortality and adverse events are important
tics to outcomes, and in failing to show that outcome measures because they can alert us
organisational and managerial factors come when things are going badly wrong in a health-
logically before quality. It is still possible to care setting. However, if we want to understand
infer from many studies of this type that qual- how the organisation and management of hos-
ity of care might have caused changes in the pitals aVects patient outcomes, mortality in
structure of the organisation, managerial pro- particular may not be the best dependent vari-
cesses, or in the kind of staV who chose to work able because death rates are so heavily depend-
there, rather than the other way round. Many ent on patient characteristics.
studies also focus exclusively on the internal Mitchell and Shortell13 suggest that “. . . given
structure and processes of the hospital and fail that adverse events appear more closely related
to consider the wider environment, particularly to organisational factors than to mortality,

www.qualityhealthcare.com
44 West

researchers need to refine and better define such characteristics of the organisation and out-
events that are logically related to the co- comes for patients and staV. This research pro-
ordinative organisational processes among gramme has now expanded to include an inter-
caregivers.” The “failure to rescue” measure national sample including hospitals and nurses
developed by Silber et al14 is a significant in Scotland and England. The results of this
development in this area. This conditional study, which is currently underway, will have
probability—death rates following much potential to inform policies for changing
complications—has been found to be more the organisation of nursing work to promote
closely related to hospital factors than raw mor- positive patient outcomes.19
tality figures. The idea is that, while the
likelihood that a patient will develop a complica- Research on non-hospital organisations
tion is largely due to factors such as their age and Although there are many diVerences between
severity of illness, the likelihood that they survive hospitals and other kinds of organisations such
following development of a complication is at as business firms and industries, research on
least partly a function of the care they receive. organisational outcomes provides support for
Finally, Mitchell and Shortell13 recommend that some of the independent variables identified by
future research should focus on smaller care giv- Flood and Aiken and suggests some additional
ing units rather than on the hospital because variables that might be considered. Clues from
units within a hospital vary greatly. Using the the literature on industry, firms, and other
hospital as the unit of analysis may be masking businesses suggest that decentralisation and
the eVect of organisational and managerial vari- participation in management, which are or-
ables as the amount of variation within a hospi- ganisational level variables related to autonomy
tal may be greater than that which exists and control at the individual level, should be
between hospitals. considered as contenders for a place in a causal
model. Some of these variables refer to organi-
sational structures and others to processes, and
these will be discussed in turn.
Mintzberg20 explains the importance of
Organisation of nursing work
structure in the following way: “Every organ-
In the early 1980s the American Nurses’
ised human activity—from the making of pots
Association identified a group of hospitals that
to placing a man on the moon—gives rise to
were known by reputation as “good places to
two fundamental and opposing requirements:
work”.15 Designated as “magnet” hospitals
the division of labour into various tasks to be
because they had little diYculty in recruiting performed, and the coordination of these tasks
and retaining staV, they were found to share a to accomplish the activity. The structure of an
number of organisational features, including: organisation can be defined simply as the sum
+ a relatively flat nursing hierarchy with few total of the ways in which it divides its labour
supervisors; into distinct tasks and then achieves coordina-
+ the chief nurse had a strong position in the tion among them.”
management structure of the hospital; Most standard texts in management studies
+ nurses had autonomy to make clinical deci- have at least one chapter on organisational
sions in their own areas of competence and structures. Dawson,21 for example, in a chapter
had control over their own practice; entitled “Coordination and control: structure
+ decision making was decentralised at the and organisational design” defines organisa-
level of the unit; tional structure as “. . . the socially created pat-
+ staYng was adequate and limits were placed tern of rules, roles and relationships that exist
on the number of new nursing graduates; within [the organisation].” In contrast, the cul-
+ methods to facilitate communication be- ture of an organisation refers to the collection
tween nurses and physicians were estab- of values and beliefs within it. Mintzberg
lished; implies that there is a strong relationship
+ the organisation of nurses’ work promoted between culture and structure. His classifi-
accountability and continuity of care—for cation of organisational configurations suggests,
example, primary nursing care; for example, that organisations with relatively
+ the institution demonstrated the value it non-hierarchical structures such as universities
attached to nurses—for example, by invest- are likely to have very diVerent cultures from
ing in their education. organisations such as the army that have a
Aiken and colleagues at the University of strong hierarchical structure. One of the most
Pennsylvania have since shown in a series of interesting features of an organisational struc-
studies that cardinal features of the “magnet” ture is the extent to which it is centralised or
hospitals are related to lower mortality rates,6 decentralised.
increased patient satisfaction,16 and lower According to Simon et al22: “An administra-
burnout rates17 and needle stick injuries among tive organisation is centralised to the extent
nursing staV.18 These methodologically sophis- that decisions are made at relatively high levels
ticated studies use a research strategy whereby in the organisation; decentralised to the extent
data gathered from individuals about their that discretion and authority to make impor-
sense of autonomy, control over their own tant decisions are delegated by top manage-
work, and quality of communication are aggre- ment to lower levels of executive authority”.
gated to describe important characteristics of The two concepts are not mirror images; some
the organisation. This enables the researchers empirical work suggests that they are, in fact,
to estimate the relationship between structural weakly correlated.23 Studies also suggest that

www.qualityhealthcare.com
Management matters 45

there are at least two separate notions embed-

ded in the concept of decentralisation, the first + Participation in work decisions: perma-
referring to the hierarchical levels at which nent programmes where workers take a
decisions are made (or influenced) and the formal direct role in decisions about their
second referring to the extent to which work
diVerent levels participate in the decision mak- + Consultative participation: long term
ing process. These arguments suggest that both interventions such as quality circles
centralisation and decentralisation should be where employees are consulted as man-
seen as multidimensional concepts. agers make decisions
Traditionally, centralisation is measured in + Short term participation: brief but formal
two main ways. Pugh et al,24 who pioneered the exercises in participatory decision mak-
examination of organisational structures, col- ing about job issues
lected data on, for example, the Chief Execu- + Informal participation: managers and
tive’s span of control and the ratio of workers to subordinates engage in informal influ-
supervisors, mainly from documentary evi- ence sharing despite the absence of a for-
dence collected from organisations. Hage and mal programme
Aiken25 focused on participation in decision + Employee ownership
making and hierarchy of authority. In their + Representative participation: employees
landmark study of centralisation these authors are elected as council or board members
interviewed staV in 16 social welfare and health
organisations in the USA where they focused Box 1 DiVerent types of participation.28
on behaviour—particularly participation in
hiring, promotion, policies, and feature of the employment relation rather than
programmes—as well as the extent to which sporadic or exclusive. This would explain why
each individual felt he or she had to defer to a consultative, short term, and representative
supervisor. Although both of these measures participation—which are either episodic or
have been well used in organisational research, involve only selected individuals rather than all
we might question whether either would employees—appear to have less impact on per-
provide an adequate measure of decentralisa- formance and job satisfaction than other more
tion in the NHS in the UK. consistent forms of participation.
Acorn et al26 used Hage and Aiken’s25 instru- The significance and cost eVectiveness of
ment to survey acute care nurse managers in a participative management have been the sub-
recent US study. The dependent variables of jects of some debate in management studies.
interest were autonomy, job satisfaction, and The literature seems to suggest that participa-
commitment to the organisation. Scores on tion has some beneficial eVects, but is it worth
decentralisation were not normally distributed the costs of reorganisation and training of staV
because, the authors argue, most hospitals in that would be involved in implementation? To
the USA are decentralised to some degree, and establish the current state of knowledge
were recoded as a trichotomous variable. Path Wagner27 examined 10 meta-analyses which
analysis showed that decentralisation produced focused on the eVects of participation on job
significant positive eVects on autonomy, job satisfaction and performance. The author
satisfaction, and organisational commitment, excluded from the review studies of delegation
and influenced commitment through auton- where managers relinquished all influence to
omy and job satisfaction. their subordinates, studies of consultation
Decentralisation is related to the notion of where subordinates were involved in idea gen-
“participative management” which is widely eration but were not involved in selecting the
used in organisational behaviour and manage- final idea, and more comprehensive and exten-
ment studies. Wagner27 defines participation as sive programmes, such as job enrichment
“. . . a process in which influence is shared interventions and quality of work life pro-
among individuals who are otherwise hierar- grammes. He concluded that current evidence
chical unequals”. Participative management is consistent with the claim that participation
practices mean the involvement of managers has a statistically significant positive eVect on
and subordinates in information processing, job performance and satisfaction but that the
decision making, and problem solving. Cotton actual eVects are limited in size. In practical
et al28 identified 68 studies of participation but terms, this leaves unanswered the question of
found that there were some important diVer- whether the high costs associated with intro-
ences in the way researchers defined the key ducing participation in management are justi-
term. In order to analyse the eVects of diVerent fied. If, as Wagner suggests might be the case,
forms of participation they classified studies the eVect of participation is cumulative, with
into six groups depending on the focus of the “small episodic influences” building over time,
study, as shown in box 1. then it may indeed be a good strategy for an
Cotton et al28 then showed that not all forms organisation to pursue. Current research,
of participation are equally eVective. The which is predominantly cross sectional, may
“winners” appear to be participation in work miss changes that occur over time.
decisions, informal participation (which was Decentralisation and participative manage-
associated primarily with enhanced job per- ment are related to a number of other “innova-
formance), and employee ownership (associ- tive work practices” which have been reviewed
ated with enhanced job satisfaction). We might by Ichniowski et al.29 Within this broad term
speculate that participation appears to be most they include eVorts to improve workers’
eVective when it is a permanent and inclusive involvement (such as profit sharing, flexible

www.qualityhealthcare.com
46 West

and broadly defined work assignments), im-

proved communication and dispute resolution + Leadership and management skills
mechanisms, and worker participation in + Clear organisational objectives and
decision making. These can be contrasted with strategies for achieving them
traditional work practices where jobs have clear + The “fit” between organisational objec-
boundaries and associated rates of pay, where tives, external environment, and strategy
there are clear lines between workers and + Change management processes
+ An organisational culture that is receptive
supervisors, decisions are made almost exclu-
to change and good working relationships
sively by managers, and communication flows
among key actors
through the formal chain of command. They + Good relationships between separate
concluded that: “Innovative human resource parts of the NHS network with clarity
management practices can improve business over “who does what”
productivity, primarily through the use of + Use of performance management, sup-
systems of related work practices designed to ported by good information systems, to
enhance worker participation and flexibility in drive change
the design of work and decentralisation of + Recognition that “good performance” is
management tasks and responsibilities”. They multidimensional and that the compo-
also suggested that there are potentially large nents of performance, such as quality of
payoVs—that is, the consequences of adopting care and financial success, are not neces-
participative work practices can have economi- sarily competitive
cally important eVects on the performance of
firms that adopt them. Perhaps the most Box 2 Organisational and managerial factors related to
organisational performance.33
important finding is that the specific work prac-
tice is less eVective than the co-existence of a could find no quality studies of the determi-
number of similar practices that improve nants of performance in trusts. They con-
productivity and attitudes as well as decrease cluded that the most comprehensive, illumi-
turnover and accidents. This is the phenom- nating, and useful research on performance
enon of “bundling”, which is used to describe determinants in healthcare settings has been
the combination of high involvement work carried out in the USA by Shortell and
practices and supporting management prac- colleagues.31 This work, which was mainly
tices. “Workers cannot make good decisions conducted on managed care organisations,
without suYcient information and training, raises important findings and questions for the
and they are unlikely to make suggestions if implementation of primary care groups.
they feel that this will cost them their jobs or There is some evidence—for example, in the
reduce their pay”.30 It is tempting to conclude work by Shortell et al,31 Pettigrew,32 and Collins
that some underlying cultural shift in the rela- and Porras33—for the impact of a number of
tionship between workers and managers is a organisational and managerial factors that are
necessary prerequisite for beneficial changes in related to organisational performance in both
the structure and functioning of the organis- the public and private sectors (box 2).
ation. In other words, tinkering with one or two Pettigrew et al32 criticised this work for the
organisational innovations is not enough. The historical tendency to focus on one determi-
question of the extent to which high involve- nant of quality such as human resource
management practices, rather than attempting
ment work practices and supporting manage-
to construct and estimate multivariate models.
ment practices have been adopted in the NHS
In some ways the idea of “bundling” can be
has still to be determined. However, if trusts do seen as an attempt to redress the balance in
vary on these dimensions, it makes an empiri- favour of more complex models.
cal test of the relationship between work prac- These authors identified the recent theoreti-
tices and quality of care at least theoretically cal writings of industrial economists Milgrom
possible. and Roberts34 as an important impetus to
Similar conclusions emerged from a review future work in this area. The complementari-
of the literature on the determinants of ties approach argues that sets of factors can be
organisational performance commissioned by mutually reinforcing in their eVects on per-
the National Health Service Executive and formance. Their recommendation is that fu-
conducted by Pettigrew and colleagues at ture research on the performance of healthcare
Warwick and Aston Business Schools.30 They institutions should, at least in part, use the idea
were asked to identify and synthesise what is of complementarities.
known and not known about the determinants
of performance in private and public sector Towards a model of the organisational
organisations, and about the practices and impacts on quality of patient care
techniques of performance management. The aim of this paper has been to identify, from
They found there is more literature on the literature, some of the organisational
performance measurement, less on perform- variables that might belong in a model of qual-
ance management, and least on the determi- ity of care. These appear to fall into a number
nants of performance. Relative to research on of categories which are summarised in box 3.
the private sector, research on the determi- Structural features of the organisation might
nants of public sector performance is very lim- include the extent to which the organisation is
ited in quality as well as quantity. In fact, they centralised or decentralised, which can be

www.qualityhealthcare.com
Management matters 47

important. Both of these categories are at the

Independent variables organisational level of analysis because neither
+ Organisational structures, e.g. decentrali- set of variables is reducible to the behaviour of
sation individuals. They are therefore logically prior
+ Organisational processes, e.g. human to the social psychological variables that describe
resource management practices, coordi- the experience of working in a particular place.
nation of care, interprofessional relation- In this category the literature stresses the rela-
ships tionship between participation in decision
+ Environmental variables, e.g. quality of making, sense of involvement in the organis-
relationships with other organisations ation, and sense of autonomy and control.
+ Social psychology of work, e.g. individu- Taken together, the structural features of the
als’ experience of participation in deci- organisation and the processes it employs par-
sion making, sense of autonomy, and tially determine the subjective experiences of
control workers (staV outcomes) such as job satisfaction
+ “Fit” among strategy, structure, and and morale, and contribute to organisational
environment outcomes such as diYculties in recruiting and
retaining staV.
Intermediate variables The main variable that we want to explain is
+ StaV outcomes, e.g. job satisfaction quality of patient care which could be estimated,
+ Organisational outcomes, e.g. rate of in the first instance, by using the NHS
sickness and absenteeism performance indicators published for England
in 1999 and 2000. These could also be used to
Control variables calculate more theoretically defensible depend-
+ Hospital characteristics, e.g. size, spe- ent variables such as the “failure to rescue”
cialisation, teaching status, number of measure described by Silber et al.14 We also
staV need to consider a number of control variables.
+ Patient characteristics, e.g. severity of ill- These are variables which have been shown to
ness, multiple diagnoses have a significant association with quality—for
+ Characteristics of work, e.g. predictabil- example, size, teaching status, extent of spe-
ity of admission patterns, volume cialisation, staV number and skill mix, and the
+ Socioeconomic factors, e.g. social class volume and case mix of patients. This group of
characteristics of local population, urban/ variables would also include human capital
rural location variables such as the training, education and
+ Economic variables, e.g. financial state of experience of staV, or tenure in the case of the
the organisation top management team. Finally, having criti-
cised previous research for omitting environ-
Dependent variables mental variables, we could include some meas-
+ Clinical indicators, e.g. deaths in hospital ures of the extent to which the organisation is
within 30 days of admission influenced or controlled by external forces and
+ Adverse events, e.g. medication errors, the quality of relationships they enjoy with
falls other organisations. A diagrammatic represen-
+ Complications, e.g. hospital acquired tation of the model is shown in fig 1.
infections
+ Constructed indicators, e.g. failure to Taking this work forward
rescue Understanding how the organisation and man-
+ Administrative targets, e.g. state of wait- agement of hospitals aVects the quality of
ing lists, financial viability patient care is no mean task. Previous research
+ Patients and carers’ experiences, e.g. can help us to identify some of the variables
complaints, response to surveys that appear to be relevant, but we do not yet
have a theory that reflects the complexity of the
Box 3 Types of variables that might be included in relationships involved. This paper, by drawing
modelling organisational outcomes on a number of diVerent fields of literature, has
sought to identify variables at diVerent levels of
measured either as the level at which decisions analysis—individual, organisational, and
are taken or by the number of levels in the environmental—that might be linked. The next
hierarchy. The ratio of supervisory to non- step will be to articulate how they might be
supervisory positions is a crude measure of related to each other and to build simple mod-
centralisation. NHS hospitals also diVer in the els that will guide empirical investigation. This
extent to which departments such as finance or will entail dealing with issues such as causal
personnel are devolved out of their own profes- ordering, identification of mechanisms, and
sional departments into management teams. specification of temporal sequences that have
The extent to which clinicians are involved in dogged this tradition of research for many
management also seems to be an important years. Ideally, future research will be more
distinguishing feature of some current NHS theoretically informed, will use longitudinal
trusts, which might have important implica- rather than cross sectional research designs (so
tions for sharing power and responsibility. that the problem of causal ordering can be
Organisational processes, such as innovative addressed), and will use statistical methods
human resource management practices and such as multilevel modelling which allow for
procedures to facilitate communication, con- the inclusion of variables at diVerent levels of
flict resolution, and participation are also analysis.

www.qualityhealthcare.com
48 West

Independent Intervening Dependent 6 Aiken LH, Smith HL, Lake ET. Lower Medicare mortality
among a set of hospitals known for good nursing care. Med
variables variables variables Care 1994;32:771–87.
7 Secretary of State for Health. A first class service. London:
Department of Health, 1998.
8 Buetow SA, Roland M. Clinical governance: bridging the
Organisational gap between managerial and clinical approaches to quality
structure of care. Quality in Health Care 1999;8:184–90.
9 Department of Health. Clinical governance: quality in the new
NHS. London: The Stationery OYce, 1999.
10 McKee M, Black N. Does the current use of junior doctors
Staff outcomes in the United Kingdom aVect the quality of medical care?
Soc Sci Med 1992;34:549–58.
11 Flood AB. The impact of organisational and managerial
Organisational Indicators of factors on the quality of care in health care organisations.
Med Care Rev 1994;51:381–429.
processes quality of care 12 Hospital volume and health care outcomes, costs and
patient access. EVective Health Care 1996; 2(8).
Organisational 13 Mitchell PH, Shortell SM. Adverse outcomes and variations
in organisation of care delivery. Med Care 1997;35(11
outcomes Suppl).
14 Silber JH, Williams SV, Krakaeur H, et al. Hospital and
patient characteristics associated with death after surgery.
Relationships to A study of adverse events and failure to rescue. Med Care
the environment 1992;30:615–29.
15 McClure A, Polin M, Sovie M, et al. Magnet hospitals: attrac-
tion and retention of professional nurses. Kansas City,
Control variables Missouri: American Academy of Nursing, 1983.
16 Aiken LH, Sloane DM, Lake ET. Satisfaction with inpatient
AIDS care: a national comparison of dedicated units and
Figure 1 Proposed model of the organisational impacts on quality of patient care scattered beds. Paper presented at the Annual Meeting of
the ASA, New York, 1996.
17 Aiken LH, Sloane DM. EVects of organizational innovations
A great deal of research is currently under- in AIDS care on burnout among urban hospital nurses.
way that will strengthen the evidence on which Work and Occupations 1997;24:453–77.
18 Aiken LH, Sloane DM, Klocinski J. Hospital nurses’ risk of
recommendations about the organisation and occupational exposure to blood: prospective, retrospective
management of hospitals can be based. How- and institutional reports. Am J Public Health 1997;87:103–
7.
ever, the process of producing good quality 19 Aiken LH, Sloane DM, Sochalski J. Hospital organisation
research can be prolonged. In the meantime, it and outcomes. Quality in Health Care 1998;7:222–6.
20 Mintzberg H. The structure of organizations. Englewood
is important to communicate the importance CliVs, NJ: Prentice-Hall, 1971.
of organisational factors to clinicians, to whom 21 Dawson S. Analysing organisations. Basingstoke: Macmillan
they may be relatively unknown. Medical and Business, 1996.
22 Simon HA, Guetzkow H, Kozmetsky G, et al. Centralisation
nursing education tends to focus, quite rightly, vs. decentralisation in organising the Controller’s Department.
on individual patient care, and an awareness of New York: Controllership Foundation, 1954.
23 Khandwalla PN. Viable and eVective organisational design
how each clinical encounter is constrained or of firms. Acad Management J 1973;16:481–95.
enabled by the system within which it is 24 Pugh DS, Hickson DJ, Hinings CR, et al. Dimensions of
organisational structure. Admin Sci Quarterly 1968;14:115–
embedded can take many years of clinical 26.
practice. We all need to become much more 25 Hage J, Aiken M. Relationship of centralisation to other
structural properties. Admin Sci Quarterly 1967;12:72–92.
conscious of how the way we work together, 26 Acorn S, Ratner PA, Crawford M. Decentralisation as a
and the way that care is organised, aVects determinant of autonomy, job satisfaction, and organisa-
tional commitment among nurse managers. Nursing Res
patients’ experience of the healthcare system. 1997;46:52–8.
27 Wagner JA III. Participation’s eVects on performance and
satisfaction: a reconsideration of research evidence. Acad
The author would like to thank David Barron, Alison Kitson, Management Rev 1994;19:312–30.
Juliet McDonnell, Jan Savage, and Cherill Scott who made use- 28 Cotton JL, Vollrath DA, Froggatt KL. Employee
ful comments and suggestions on previous versions of this participation: diverse forms and diVerent outcomes. Acad
paper. Management Rev 1998;13:8–22.
29 Ichniowski CT, Kochan A, Levine D, et al. What works at
1 Penrose ET. Biological analogies in the theory of the firm. work: overview and assessment. Industrial Relations 1996;
Am Econ Rev 1952;42:804–19. 35:299–333.
2 Perrow C. The analysis of goals in complex organisations. 30 Pettigrew A, Brignall TJS, Harvey J, et al. The determinants of
Am Sociol Rev 1961;26:854–66. hospital performance: a review of the literature. Review
3 Meyer MW. Measuring performance in economic organisa- conducted for the NHSE and available from Warwick
tions. In: Smelser NJ, Swedberg R, eds. The handbook of Business School, 1999.
economic sociology. Princeton, NJ: Princeton University 31 Shortell S, Zimmerman J, Rousseau D, et al. The perform-
Press, 1994: 556–78. ance of intensive care units: does good management make
4 Barron DN, West E, Hannan MT. A time to grow and a time a diVerence? Med Care 1998;32:508–25.
to die: growth and mortality of credit unions in New York 32 Pettigrew AM. Organising to improve performance. Hot topics.
City, 1914–1990. Am J Sociol 1994;100:381–421. Warwick Business School, 1999: 1, 5.
5 Scott WR, Forrest WH Jr, Brown BW Jr. Hospital structure 33 Collins JC, Porras JI. Built to last: successful habits of visionary
and postoperative mortality and morbidity. In: Shortell companies. New York: Harper Business, 1994.
SM, Brown M, eds. Organisational research in hospitals. 34 Milgrom P, Roberts J. Complementarities and fit: strategy,
Chicago: Blue Cross Association, Inquiry Book. 1976: structure and organisational change in manufacturing. J
72–89. Account Econ 1995;19:179–208.

www.qualityhealthcare.com