Professional Documents
Culture Documents
Development and Piloting Application of A Performance Measurement Model in Greek Public Sector Hospitals
Development and Piloting Application of A Performance Measurement Model in Greek Public Sector Hospitals
(University of London)
by
Lambros Georgopoulos
September 2004
Contents
Page No
Chapter 1 – Introduction
1.1 Introduction 6
1.2 Background 6
1.3 Objectives & Scope 7
1.4 Hospital Definition 8
1.5 Hospital Performance 8
1.6 Importance of Hospital Performance Measurement 9
1.7 Hospital Performance Measurement Systems 11
1.8 Terminology 18
Chapter 2 – Methodology
2.1 Secondary Research 19
2.1.1 Books & Journals 19
2.1.2 Internet Research 20
2.2 Primary Research 21
2
5.3 Excluded indicators 39
5.4 Suggested indicators for adoption in Greece 41
Chapter 7 – Discussion
7.1 Conclusion 70
7.2 Recommendations 72
References 74
URL’s 79
Appendix 1 Outpatients’ Questionnaire 81
Appendix 2 A&E Questionnaire 83
Appendix 3 Inpatients’ Questionnaire 85
Appendix 4 ‘Trust the Doctors’ Questionnaire 87
Appendix 5 A&E Questionnaire Results 88
Appendix 6 Outpatients’ Questionnaire Results 89
Appendix 7 Inpatients’ Questionnaire Results 90
Appendix 8 ‘Trust the Doctors’ Questionnaire Results 91
Appendix 9 Health Statistics – Greece & Albania 92
Appendix 10 Concentrated table of results 93
Appendix 11 Staff breakdown & percentage of organic positions covered 94
Appendix 12 Outpatients’ waiting times 95
Appendix 13 UHP’s website 96
3
This work is dedicated to the memory of two deceased members of my family; my
father Spiros and my brother Pavlos. Their ever love, encouragement and support is
still alive.
4
Synopsis
Since 1984, public hospitals in Greece operate under the Greek NHS umbrella.
Despite the 20 years of NHS existence, the system is characterised by the
absence of a hospital performance measurement model.
The piloting application of the model is attempted at UHI and UHP, two of the
most reputable hospitals in Greece. Despite omissions of standards,
frameworks and guidelines, the current absence of business like managerial
approach and a poorly implemented health system, the application of the
model proves the feasibility of a performance measurement system within
Greek public sector hospitals.
The application results further prove that through the use of the model,
valuable information are produced, enabling the identification of better
performing organizations and the detection of performance variations.
5
Chapter 1: Introduction
1.1. Introduction
1.2. Background:
The 1980’s was a decade where the life of Greek citizens has changed dramatically.
Entrance to the European Union in 1981 and the clear support of the Greek
population to the socialist party - which had as a keynote address “change” - in the
national election later that year, were the first signs of a decade where the Greek
population would come across new situations. Health issues were a top priority -like
in most countries of the word- in the agenda of the newly elected government. The
population had voted for ‘change’ and health issues could not be exempted from this
change. The main change within health was the introduction of the NHS (ESY) in
1983 1 -in an attempt to benchmark Western European health systems- as well as the
creation of newly built public hospitals.
Few years after their election, as part of the health agenda, the socialist government
allocated a contract to a German company to build three university hospitals in the
administration capitals of three different provinces. Patras -the capital of
Peloponnesus region-, Iraklion -island of Crete’s capital - and Ioannina -the capital of
1
Hellenic Republic - Health, Health Care & Welfare in Greece Report, 2003
6
Epirus region- were the three cities honored. Using similar building plans, materials
and construction methodology, the German consortium delivered three modern
hospitals before the end of the decade (1989) and a year later (1990) 2 the hospitals
opened their doors to the public. Fifteen years after their opening, the three hospitals
are considered as leading public hospitals amongst the Greek public opinion. This is
a remarkable achievement, especially considering the fact that Greece -where half of
the population lives in the capital city- is a highly centralized country in all areas of
public administration and service.
Regardless of their reputation -mainly arising from some specialized treatments they
offer-, there are serious concerns regarding their overall performance. The majority -if
not all- public hospitals in Greece show negative financial figures while at the same
time they have huge debts. A highly reputable Greek newspaper, “To Vima” 3 claims
that the debts of Greek hospitals are approximately around €2 billion whereas €1.5
billion is due to uncollected debts from insurance funds. According to CASPE 4, there
have been no government initiatives to introduce quality assurance or any
accreditation system for health care services. Furthermore, there are no examples of
fully operational accreditation programs, nor of organizational standards by which
hospital facilities may be evaluated. Hospitals’ performance in Greece is regarded as
synonymous with reputation.
As this study attempts to develop a model for the measurement of public hospitals’
performance, the above hospitals are an excellent case to examine due to their
similarities. The hospitals were built at the same period; as it will be seen later in this
study, they have similar facilities, staffing pattern, capacity and organizational
structure. As part of the NHS -like all public hospitals- , they use the same
recruitment processes, they operate within the same legal framework and external
context and finally, they are funded from the same resources. Homogeneity is a
critical factor when comparing performance as it increases the significance of the
outcomes.
2
University Hospital of Ioannina – www.uhi.gr
3
To Vima, 09.05.2004
4
CASPE – www.caspe.co.uk
7
• How can we measure public hospital performance within the Greek health
care arena?
Hospitals are complex organisations which may vary in terms of size, operations,
facilities, equipment, quality, staff, organisational structure, infrastructure,
ownership, and many more parameters. There are different definitions of hospitals,
ranging from “a health facility where patients receive treatment” -according to
various dictionaries- to the World Health Organisation’s (WHO) definition 5 of “an
organized effort to provide a specific set of medical services, usually physically
located in one or several buildings, and related to specialized cure (diagnosis and
treatment) and care (as opposed to the primary care level) with the input of health
professionals, technologies and facilities”. For the purpose of this study, the
WHO’s definition is used.
5
WHO, Barcelona 2003
8
WHO’s 6 definition of performance: “the best that could be achieved with the same
resources”.
6
WHO, World Health Report 2000
9
invest when it comes to the private sector. The use of performance indicators attracts
media publicity and raises public awareness. Identification of performance variations
may also help guide quality improvement efforts, whereas high-performing
management teams can be identified and high-performance drivers can be
determined.
An example from the United States can give a clear picture of the importance of
performance measurement: Solutient 7 estimated that if all hospitals in the United
States performed at the level of the 2002 Top 100 hospitals, the impact for the
population insured under the Medicare system would be:
The newest use of performance measurement systems has been identified in the
UK’s NHS where foundation hospitals 8 -the best performing hospitals- are allowed
increased autonomy. The NHS plan involves less inspection and monitoring from the
Department of Health for foundation hospitals, with some more benefits including:
Griffith & King 9 further claim that a variety of stakeholder groups are putting
increased pressure on providers for measured performance; they are demanding
data on quality and patient satisfaction, although simultaneously pressing for lower
costs. Finally, Hearnshaw et al 10, state that if desirable performance measures are
set according to appropriate criteria, the attainment of these targets should result in
improved care. In contrast, Cluzeau et al11 support that if quality of care is assessed
against inappropriate criteria, attainment of targets may not effect any improvement
in care and resources may be wasted in ineffective quality improvement activities.
7
Solutient, National Report 2002
8
The Guardian, 31.03.2004
9
Griffith J.R., King J.G. 2000
10
Hearnshaw et al, 2002
11
Cluzeau et al, 1995
10
At this point it is necessary to define the term “indicators”. Based on Boyce’s 12
definition of indicators for health care systems, the term ‘indicators’ or ‘measures’
is used in this project for: statistics or other units of information which reflect,
directly or indirectly, the performance of the hospital.
The Greek health system is not the only European health system lacking on hospital
performance measurements. Despite the fact that hospitals are an important part of
any health care system, as well as the fact that a large proportion of the health
budgets are allocated to hospitals, according to the 2002 WHO report 13 “systems of
performance measurement are poorly developed throughout Europe”. Due to the
language barrier, this study has not examined systems of other European Countries.
Our research identified at least five European countries were such systems exist
(Sweden, Denmark, Spain, Norway and the Netherlands) but -as previously
mentioned- the study will emphasize in the Anglo-Saxon systems (US, Canada,
Australia and the USA).
1 Financial
2 Customer
3 Internal Business Processes
4 Learning & Growth
12
Boyce N., 2002
13
WHO, World Health Report 2000
11
Voelker et al 14 claim that the BSCs are particularly appropriate for organizations in
turbulent industries such as healthcare. The box below originates from the same
source and represents the adaptation of the BSC in healthcare organisations.
The Association used an approach based on the work of Baker & Pirk who adapted
Kaplan’s ‘balance scorecard’ to the Canadian hospital setting. Baker & Pirk’s
approach describes performance across four dimensions or “quadrants” critical to the
strategic success of any health care organization as seen in the following box. These
four dimensions remain separate, not integrated as in the balanced scorecard.
14
Voelker et al, 2001
12
According to the OHA’s 2003 report 15, thirty five indicators are selected over the four
dimensions, based on their scientific soundness, relevance and feasibility.
Researchers restricted the number of indicators to a manageable level, balancing the
wide scope of the study with the need for conciseness.
Chris Ferao, a finance manager at William Osler Health Center in Canada, has
developed an excellent set of hospital performance indicators. But, the problem with
the ‘Global Hospital Indicators’ lies with the fact that the indicators are evaluating
hospital performance exclusively from the financial prism. He separated the
indicators in four categories:
15
CIHI, Hospital Report 2003
13
Although his measurement system can not be used to examine holistic performance,
some indicators of Ferao’s model can be partially used in the construction of
developed models.
16
Dr Foster - www.drfoster.co.uk
14
within the United Kingdom public sector. The indicators are designed by CHI and the
Department of Health (DoH) to reflect a wide range of performance issues, following
consultations with many stakeholders. The model uses eight key targets in
conjunction with performance indicators developed by the ‘balanced scorecard
methodology’. The key targets represent the most significant factors in determining
overall performance. Performance against targets is assessed in terms of the degree
the target has been achieved. The 2004/05 targets are:
1 12hrs wait for emergency admission via A&E post decision to admit
2 All cancers: 2 week wait
3 Elective patients waiting longer than the standard
4 Financial management
5 Hospital cleanliness
6 Outpatient and elective (inpatient and daycare) booking
7 Outpatients waiting longer than the standard
8 Total time in A&E: 4hrs or less
The performance indicators used by the model are focused towards the coverage of
three areas:
1 Clinical
2 Patient
3 Capacity & capability
15
Contrary to Dr Foster’s model characterised as ‘the elephant’, Solucient’s model can
be metaphorically described as a ‘fox’ model by being small, fast and clever. The
model is designed for hospitals operating within the American health system, thus
making few of the measures impossible to apply in the ‘non-for-profit’ European
public hospital sector. Solutient’s performance measures are listed below.
1 Volume Indicators
2 Utilization Indicators
3 Mortality for procedures
4 Mortality for conditions
It should be mentioned that this model applies for inpatients, ignoring outpatients and
day patients.
The method used by the agency is a set of quality indicators (QIs) organised in three
‘modules’. The first module is Prevention QIs which attempt to identify hospital
admissions which could have been avoided. The second module is similar with
Wisconsin’s methodology, reflecting quality of care inside the hospital whereas the
16
third module is the patient’s safety QIs, reflecting quality of care by focusing on
iatrogenic events.
The Institute, based on the principle that the elements for reviewing the performance
of health care delivery are efficiency (cost/unit of output) and effectiveness (quality,
appropriateness etc) developed a set of indicators. The measures were developed
under a framework by the National Ministers Benchmarking Working Group, ignoring
including access and equity indicators
A research on quality and outcome indicators for acute healthcare services on behalf
of the above department of the Australian Government has sought for indicators
across eight dimensions:
1 Access
2 Efficiency
3 Safety
4 Effectiveness
5 Acceptability
6 Continuity
7 Technical Proficiency
8 Appropriateness
17
Productivity Commission, 1999
17
1.8 Terminology
Evening surgeries: a term used in the Greek health arena to describe private
surgeries operating by hospital doctors within public hospitals. The government
prohibited academics and NHS doctors from running private surgeries. In exchange,
they have been offered the option to examine outpatients and operate their surgeries
within public hospitals during the evenings, sharing the income with the hospital.
ESY: Greek National Health System
Casemix: tool for improving and managing health systems through the categorization
of patient episodes. It is been used for output measurement; resource allocation;
(funding, paying, charging); quality improvement; comparative analyses and the
monitoring of trends over time 18.
Diagnosis Related Groups(DRGs): the best known casemix classification. It is a
patient classification system which provides the types of patients a hospitals treats.
Junior Doctors: Unspecialized doctors. Those medical professionals who work in
hospitals on a trainee programme in order to gain their specialization qualification.
Health Centres: Primary care health units of the Greek NHS. They are usually
located in areas with a small population and aim to provide basic health care
treatment.
Hospital: an organized effort to provide a specific set of medical services, usually
physically located in one or several buildings, and related to specialized cure
(diagnosis and treatment) and care (as opposed to the primary care level), with the
input of health professionals, technologies and facilities 19
Hospital Acquired (or nosocomial) Infections: An infection caught while hospitalized.
The medical term often used is "nosocomial", coming from the Greek word nosus
(disease) and komio (taking care of). “Nosocomio” is the Greek word for hospital.
Hospital Performance Indicators: statistics or other units of information which reflect,
directly or indirectly, the performance of the hospital in maintaining or increasing the
well-being of its target population 20
Organic Positions: staffing positions necessary for the smooth operation of a public
organization, decided by the Greek government through enactment.
Performance: the best that could be achieved with the same resources 21
PESY: Regional Health Authority
18
Australian Department of Health & Ageing - www.health.gov.au
19
WHO
20
Boyce, 2002
21
WHO
18
Chapter 2: Methodology
The first stage of the research was the secondary research, involving internet search
and literature review of the subject. The secondary research attempted to identify
literature related to performance evaluation theory, systems, models and tools
currently in place in advanced health systems such as the United States, Canada,
Australia and UK. The research focused on bibliography which could assist towards
the evaluation of overall hospital performance from various perspectives. Despite the
fact that there were a variety of resources related to performance measurement,
there was extremely limited bibliography on evaluating performance within hospitals.
In common with other researchers 22, we found the literature search to be ineffective
and inefficient tool for accessing the body of knowledge of healthcare performance
indicators applications.
Books were searched through Imperial College library’s database by alternating and
combining relevant keywords such as:
22
Kazandjian et al, 1995 / Boyce N., 2002 / Centre for Health Policy studies, 1994
19
• performance measurement
• performance evaluation / evaluating performance
• quality management
• hospital performance
• quality/performance indicators
• quality/performance measures
• benchmarking
• performance management system
Journals and online articles were searched through the library’s e-journal service.
The most useful electronic resources found to be
• EBSCO
• Factiva
• Science Direct
• Blackwell Synergy
• Cochrane
Furthermore, as the area investigated is very specific with limited resources, the
‘snowball’ method has been used. Every time a useful article or a relevant report was
identified, its references were filtered, with the most relevant references becoming
the references of this study.
International
International Organization for Standardization (ISO)
World Health Organization (WHO)
International Society for Quality in Health Care (ISQua – based in Australia)
20
Australia
Australian Council for Health Care Standards
Australian Institute of Health & Welfare
USA
Wisconsin Department of Health and Family Services
Agency for Health Care Reseacrh & Quality
Department of Health
Joint Commission on Accreditation of Healthcare Organizations
The National Quality Forum
The Association of Maryland Hospitals & Health Systems – Quality Indicator Project
Canada
Ontario Hospital Association
Ontario Ministry of Health
Canadian Institute for Health Information
Canadian Council for Health Services Accreditation
UK
Commission for Health Improvement
Department of Health
Dr Foster – Delivering Intelligent Health Care Information
Greece
Ministry of Health & Welfare
University Hospital of Ioannina
University Hospital of Patras
Numerous websites of private and public hospitals
Following the analysis of the various performance systems and the initial
development of a model which adapts the most appropriate dimensions in the Greek
Hospitals’ case, primary research was carried in order to apply the new model. The
primary research involved the collection of ‘raw’ data with the performance indicators
identified during the secondary research ‘guiding’ the primary research. The variety
of performance measures lead to collection of both quantitative data (financial figures,
length of stay, number of patients etc) as well as qualitative data (patient satisfaction,
reputation, etc). Primary research involved a combination of structured and semi-
structured interviews, patient surveys through the use of questionnaires and analysis
of existing data and reports, either produced by the two hospitals or by statistic
21
organisations. The sample for our questionnaires was consisted of local citizens,
inpatients and outpatients of various departments as well as visitors of the A&E
department. Several interviews were conducted with key staff, involving hospital and
departmental managers, doctors, academics, nurses, administrative personnel and
health authorities. Initial contacts were established well in advance, whereas the
‘snowball’ and ‘networking’ methods were successfully attempted.
Primary research was crucial for the successful completion of the study. The
development of the model, its dimensions and indicators were highly affected by
primary research. The higher the level, relevance and accuracy of available hospital
data, the easier was to construct the indicators. In many cases, as it will be
mentioned at a later stage of the project, the lack of sufficient and relevant data led to
the exclusion of a number of widely used indicators from use within the Greek health
system.
As the study examines the holistic hospital performance, primary research attempted
to involve as many stakeholders as possible. The study further attempted to involve
as many questionnaire respondents as possible, but the time constraint narrowed the
project’s potential.
22
Chapter 3: The operational environment
3.1 Introduction
The purpose of this chapter is to examine the operational environment of the Greek
public hospitals. An analysis of the political, economic, socio-demographic and
administrative environment is attempted in the ‘country facts’ section. The chapter
further analyzes the health status of the country and describes the National Health
System. Finally, it explains some particular characteristics of the Greek public sector
hospitals, most of them identified through the research.
Greece is a country with a population of 11 millions, with almost half of them living in
the capital city of Athens. The official unemployment rate for 2001 was at 10.5% 23 but
various newspapers claim that the real figure is much higher than the official
published figure. Contrary to the impression of most people, Greece is a
mountainous country with 80% 24 of the land being mountainous or hilly. The country
is a member of the European Union since 1981 and -as a member of the European
Monetary Union- in 2002 the country replaced its currency (Drachma) with the Euro
(€).
Since 1975 the country’s political system is parliamentary democracy with the two
major political parties -Socialists & Democrats- gaining between 80% and 90% in
every single election over the last 25 years 25. The administrative system is highly
centralized with decision-making taking place in the capital city of Athens. The 2004
Olympics hosted by Athens further worsened the problem of centralization, with huge
amounts of money from the government budget allocated to Athens for infrastructure
development. Over the last 5 years, Athens acquired a new Airport, a new
underground system, new road networks, circular roads and numerous stadiums.
There is a lot of debate on whether the Olympics would affect positive or negative the
country’s development but most people seem to agree that the gap between Athens
and the rest of the country has now increased dramatically.
23
National Observatory on Employment – www.epa.gr
24
Greece Now! – www.greece.gr
25
Office of the Prime Minister www.primeminister.gr
23
13 Administrative Regions 52 Prefectures
The country’s figures in life expectancy reach 76 years for males -3rd amongst EU
countries-, whereas female life expectancy is at 81 years 27. Eighteen percent 28 of the
population was over 65 years old in 2002, a figure which ranks Greece in the 2nd
position amongst EU countries, with Italy leading the ranking table. At the same time,
the country has the largest percentage of regular smokers within the EU, with 45% of
male and 35% of female being regular smokers according to the WHO 29. The total
expenditure on health per capita amount to $1814 in 2002, and the ratio of acute
care beds per 1000 citizens was 4 in 2000 30. The public funding as a percentage of
the total pharmaceutical expenditure is at 71.5% 31 . According to the National
Statistics organisation the ratio of population per doctor was one for every 250
citizens in 2000 whereas the ratio of population per dentist was one for every 850
citizens for the same year. These ratios are far lower than the average EU ratios,
26
Ministry of Internal Affairs & Public Administration www.ypes.gr
27
National Statistics Organisation www.statistics.gr
28
OECD Health Data 2004,
29
WHO – 2002 Database
30
OECD Health Data 2004
31
OECD Health Data 2004
24
especially when compared with UK, where the 2004 dentist ratio for example is one
for every 3.500 citizens, as there are only 18000 dentists registered with the General
Dental Council to practice in the UK 32. Contrary to most European countries, Greece
seems to have an oversupply of medical professionals. In most reports researched,
numerous health status indicators record rather high figures; however, there is
substantial concern from the authors of the reports over the reliability of Greek data
due to measurement and registration problems in Greece.
The Greek National Health System (ESY) was established in 1983 aiming to improve
access to care, slow down the growth of private care, and provide a free, equitable
and comprehensive health care coverage. Initially, ESY was based on the principles
of equity, development of the primary health care, government control over the
provision of health care and decentralization. Following the implementation of the
new legislation, the government focused on the acquisition of significant capital
infrastructure within the public health sector. Over the 21 years of ESY existence,
numerous new hospitals were built, including several university hospitals. Mossialos
claimed back in 1997 33 that ESY had only been partially implemented and previous
structures remained unchanged. He states that structural changes are hindered due
to a lack of consensus amongst political, social and medical-professional groups.
Since the early 1990s, emphasis for reform has shifted towards a more managerial
and market-oriented approach.
Almost all the population is covered by one of the few hundred social insurance funds
with the government studying the integration of the numerous insurance funds to a
single social insurance fund. The health system is financed through a mixture of
taxation and social insurance, with direct private payments and private health
insurance representing only a small supplementary system of financing. The country
spends 5% of its GDP for health 34 and public hospitals were funded by the state
(70%) and sickness funds (30%) back in 1998 35 . Nowadays, taxation and social
insurance remain the main source of funding but recent legislation has offered
32
General Dental Council - www.gdc-uk.org
33
Mossialos, 1997
34
OECD Health Data 2004
35
European Parliament, 1998
25
hospitals the chance to gain some extra income from the ‘evening surgeries’ as well
as from the rental of shops and patients’ contributions.
Medical professionals have the choice to operate privately, with the majority of them
having contracts with the social insurance funds. The NHS does not operate with
part-time staff and only recently the government banned NHS doctors and academics
to run private surgeries parallel with their NHS duties.
Part of the shift to the more managerial and market oriented approach mentioned
above was the introduction of the Regional Health Authorities (PESYs), a couple of
years ago. In addition, the socialist government introduced health managers to
operate public hospitals. The hospital managers -in the majority of the cases- are
politically appointed persons. Previously, the hospital managers were politically
appointed persons with the difference that they had complete lack of knowledge in
any aspect of hospital administration –hospitals were managed even by civil
engineers or lawyers!-. Following the defeat of the socialist government in March
2004 elections, the latest reforms of the system seem to be pointless since the
intentions of the new government is to cease the PESYs and constrain the role of the
hospital manager.
Through the primary research we identified that the system, from an administrative
point of view is extremely problematic. The only hospital performance figures found
at PESY offices were dated back in 2000! Following visits at the University Hospital
of Patras as well as the University Hospital of Ioannina, we realised that each
hospital –although they are the leading and larger hospitals in a big geographic area-
had only one person each, responsible for the statistics. At the same time, in most
European hospitals the organisational structures include a fully employed statistics
department, with numerous clinical coders, several administration staff and a few
managers.
26
more problematic area. Human resource planning is carried centrally by the ministry
of health with the human resource departments of the hospitals having minimum
autonomy. The ministry decides on the number of professionals needed by each
hospital -“organic positions”- and appoints accordingly, but this is a process which is
not carried over regularly. The consequences are: out of date planning and
understaffing. Hospitals’ staffing needs may increase over the years but the ‘organic
positions’ allocated by the ministry of health remain the same as possible
amendment involves a large amount of bureaucracy.
More than that, hospitals operate with less staff than the organic positions. As it can
be seen in the graph below (detailed figures can also be found in Appendix 11), both
hospitals examined operate with major staff shortages, thus -according to
organisational behaviour and human resource management theories- leading to staff
dissatisfaction, lack of motivation, poor productivity and overall poor organisational
performance.
Employee wages are similar for all public hospitals as they are determined centrally
by the government. The variances are subject to the type of profession and years of
27
service, with the primary research indicating no signs of performance related pay and
variances across different organisations.
The primary research also indicated that Greek public hospitals -despite a
government legislative adjustment of debts in May 2001 36- have huge debts, mainly
arising due to uncollected debts from the insurance funds. This can either be seen as
an ineffectiveness of public hospitals’ finance departments, mismanagement or even
an improper health system. The Greek Health Minister, right after his appointment
earlier this year stated 37 in a daily newspaper: “even the health authorities are not
aware of the amount they owe. We have to estimate by ….asking the suppliers for
the amounts owed!”
The range of the services offered by each hospital varies but all hospitals have
similar departments from a patient’s perspective. These departments are:
36
Athens News Agency, 27/4/2004
37
Kathimerini, 25/3/2004
28
Greek public sector hospitals are not using any casemix classification to monitor
quality of care, utilization of services and control costs. The research identified no
national hospital information programme or any Diagnostic Related Group framework.
Several clinics within the public hospitals use casemix methodology, but this is a
result of individual initiatives rather than an organised effort.
29
Chapter 4 - Development of the performance measurement model
4.1 Introduction
Quality of care and heath outcomes is an issue in the majority of the systems
examined with patient experience also possessing an important element. The
Canadian systems appear to be the most ‘balanced’, as they attempt to view and
judge hospital performance from different prisms.
30
The Australian models have produced some of the most unique indicators, such as
emergency readmissions, geographic variation in admission, cost per casemix
adjusted discharge and many more. The use of casemix methodology within the
Australian health system, allows Australian models greater flexibility for indicator
development. The fact that such indicators can be produced implies a well organised
health system with emphasis on data collection and utility. Overall, the Australian
models focus mainly on issues such as patient experience and waiting times, costs
and clinical effectiveness.
38
Voelker, 2001
31
organisational success. Three dimensions are inspired from the BSC methodology
for the construction of the model; internal, patient & financial. It should be mentioned
that those dimensions are also found in other models; e.g. the patient dimension is
common across the majority of the performance measurement systems. The fourth
dimension of the BSC (growth and development) was excluded for two main reasons:
In addition to the three dimensions borrowed from Kaplan’s BSC, two further
dimensions are included; clinical and volume/complexity. The clinical dimension is
used as we consider it crucial for the evaluation of hospital performance. Hospitals
are health care provision organisations, with the clinical activity constituting the main
activity of any hospital of the world. In addition, the dimension is used in the majority
of the models reviewed. The rationale behind the selection of the final dimension -
volume and complexity- lies with the fact that the model intends to be used for
hospitals comparisons. The volume and complexity dimension, acting as the
throughput element of the model, allows the development of indicators which will
enable reliable comparisons.
Numerous medical articles and journals have suggested that, outcomes for patients
may be better in hospitals where doctors perform such procedures regularly, rather
than occasionally. According to a report on behalf of Texas Health Care Information
Council 39 , better quality may be associated with greater volume; however, low-
volume providers may have excellent outcomes. Since volume alone is not an
outcome (result) measure, where possible, volume indicators are evaluated along
with other indicators. Contrary to the Texas Health Care Information Council,
Michigan Hospital Association claims the number of cases treated is not an indicator
39
Texas Health Care Information Council, 2002
32
of the quality of care you might receive in a hospital. However, they acknowledge that
volume helps you understand the size of the hospital.
There have been numbers of studies about the impacts of volume on hospital
effectiveness, offering somewhat conflicting views 40. Following the literature review,
this study considers that volume indicators are indicators for quality for non-
emergency procedures, such as outpatients and evening surgeries visits. People
who seek emergency treatment do not choose health care provider, but still, there
might be a link in this area between volume and quality as ‘practise makes perfect’. It
also considers that –taking into account the given resources- volume indicators can
represent efficiency. The literature review examined, further indicated that hospitals’
volume and range of services have a profound effect on the type of patients the
organisation treats.
The hypothesis is a hospital A with an occupancy rate of 99% and a hospital B with
an occupancy rate of 50% for the same time period. Obviously, due to the lack of
volume indicators hospital B seems more efficient as it operates in maximum
occupancy. Examining the same case from an antithesis prism -which is the use of
volume indicators- we get the following results: Hospital A has 40 beds whereas
hospital B has 1000 beds. Hospital A has only one paediatric clinic whereas hospital
B meets all patients’ needs.
40
PHC4, 2002
33
Patient Focus:
The patient focused dimension indents to examine how well hospitals satisfy their
customers’ needs. Patients’ expectations should be met to the higher level possible,
as they are those who finance public hospitals either through taxation, insurance
funds contributions or direct payments.
In UK, improving the experience of every single individual patient was at the centre of
NHS Plan reforms whereas patient experience is a major issue in the majority –if not
all- performance measurement systems of western world countries.
Based on UK’s Commission for Health improvement view of patient care 41 , our
dimension is covering five key elements:
Overall, hospitals are built with the purpose of providing medical service to patients.
Patients can not be excluded from any hospital performance measurement system
as they are the key stakeholders of any health system.
41
Commission for Health Improvement - www.chi.nhs.uk
34
Under the patient dimension, the model is finally using 7 indicators. Three of the
indicators refer to patient experience, two measures of patient waiting time, a
measure of hospital cleanliness and a measure of food quality.
Internal Focus:
Internal focus is concerned with how hospitals are managed and operating. The
intention of this dimension is to develop indicators capable of examining the staffing
patterns, the shortage of staff as well as indicators emphasising more on the
operating efficiency of the hospitals.
Bloom 42
found that experienced staff reduced personnel costs, while staffing
decisions have a significant impact not only on personnel costs, but operating costs
of hospitals. The literature reviewed indicates that occupancy rates and average
length of stay are critical for the evaluation of hospital performance as they are highly
associated with the allocation of resources. It should be mentioned that ALOS was
the only indicator used by all models examined. In addition, the dimension attempts
to examine the use of the new technologies, such as multifunction websites, by the
hospital administration.
Seven indicators are constructed under the internal dimension of the model. The lack
of business like management approach in public hospital, prevents the development
of further measures, such as LOS for selected DRGs, staff absence rates, patient
complaints etc.
Clinical Focus:
The clinical component of the model is concerned with the clinical effectiveness and
the health outcomes of the hospital. The UK NHS defines clinical effectiveness as
doing the right thing, at the right time, in the right way, for the right patient.
Regarding outcomes, there is some debate on its definition. Baumberg 43 gives the
simplest definition: “The results of health care process”. As hospitals are
organisations operating with their main purpose being the provision of health care,
the measurement of the effectiveness and health outcomes can not be excluded by
any hospital performance measurement system.
42
Bloom J. et al,1993
43
Baumberg L. et al, 1995
35
As the study examines holistic hospital performance, not departmental or sectional
performance, the clinical approach constitutes the most difficult to apply. The
difficulty on the measurement of health outcomes and clinical effectiveness is
relevant to the capacity, capability and complexity of the organisation. Specialized
hospitals (such as a paediatric or an oncology hospital) can easier measure their
health outcomes and clinical effectiveness as the symptoms, processes and
treatment methods do not vary significantly. The difficulty in the construction of
performance indicators with the purpose of measuring clinical excellence is further
increased by the lack of patient categorisation systems –as in the case of Greek
hospitals-.
Financial Focus:
The financial dimension is included in our model as hospitals are funded from
citizens directly or indirectly and the financial health of organisations such as
hospitals is of major importance to a variety of stakeholders. The government,
taxpayers, the opposition parties and suppliers are some of the key interesting
parties of a hospitals’ financial health, each of them for different reasons.
All the ratios derive from the organisations financial statements, with the problem in
the case of Greek public sector hospitals being that they do not construct and publish
36
financial statements such as balance sheets, profit and loss accounts and cash flow
statements. More than that, the hospitals do not valuate their fixed assets such as
land, buildings and equipment. The research indicated that the accounting system
currently in use is misty and allows great flexibility for manipulation of results
between different organisations. The Greek government intentions are to change the
current procedures and implement international accounting standards within public
sector organisations, in line with the trend of implementing private sector
management elements within public sector organisations. Until then, the Greek
public hospitals’ financial health analysis can be performed based on figures selected
from a small range of options.
The financial dimension of the model is the component with the most indicators.
Totally, 8 indicators are found to be applicable; assets (excluding fixed assets),
liabilities, income, debtors, expenses, evening surgeries’ profit, cost/patient and
cost/inpatient day. Despite the incompatibility of the majority of financial ratios, the
financial figures included in the model are still of the purest measures of hospital
performance, as they are clear measures of hospitals financial health.
37
Chapter 5 - Construction of the model’s performance indicators
5.1 Introduction
Good indicators should be easy to understand and use by intended audience. For
comparisons to be made, Boyce claims 44 that common indicator definitions must
exist and be systematically applied in data generation, with common data collection
methodologies and results that are risk-adjusted. He further states that lack of
adjustment for relative urgency or disease severity (at a minimum), reports on waiting
times are of limited utility. Thompson -a professor of epidemiology- has a similar
approach with Boyce, as he states 45 that reliability requires explicit and detailed
indicator definition, inclusion and exclusion criteria and sub-indicator risk stratification
where appropriate. Our literature review indicated that there are no standard, widely
accepted, universal supported indicators. Studying numerous quality indicators, we
have selected a set of 35 indicators, in an attempt to match the following criteria:
44
Boyce N., 2002
45
Thompson R., 2002
38
5.3 Excluded indicators
Some widely used indicators were excluded from the set of the 35 indicators, mainly
due to their disharmony with the abovementioned criteria and the operational
characteristics of Greek public hospitals. The ‘popular’ indicators excluded are listed
below:
Time between decision for admission and hospitalization: Contrary to the rest of
the indicators, this measurement can not be applied for a completely different reason;
Greek hospitals prove to be extremely efficient in this area. Interviews with hospitals’
staff and inpatients indicated that the time between the decision for a patient’s
admission (taken either at A&E or outpatients department) and actual hospitalization
is minimal. Following a decision for admission, patients are directly sent to the
relevant clinics, with further tests and examinations occurring in the clinics.
39
use of such data will lead to unrealistic results. Public sector hospitals’ staff often
enjoy small periods of vacations without informing personnel department as they are
‘covered’ by colleagues.
Length of Stay (LOS) for selected DRGs: A performance indicator used in several
performance measurement systems. LOS is heavily affected by severity. When LOS
is calculated for selected DRGs, the risk of significance failure is minimized. Public
hospitals in Greece do not use casemix methodology, with LOS failing to be risk and
severity adjusted.
Mortality Rates for DRGs with low mortality rates: a risk-adjusted measure of
clinical effectiveness, which -like in the case of the previous indicator- can not be
applied due to the lack of casemix methodology usage.
Average Salary: The average salary of the hospital is calculated by the sum of
wages divided by the total number of staff. It is a measure used in models where the
operational environment is either private or gives hospitals the autonomy to pay
higher wages and attract better staff -like the foundation hospitals in UK’s NHS case-.
The average salary figure may indicate a higher staff satisfaction and higher quality
of staff. In case the indicator is applied at Greek public sector hospitals, the results
will be misleading. Wages are determined centrally by the Ministry of Health, with
figures being relatively consistent at a national level. If a large variance is observed, it
can only be a result of a variance in the synthesis of staff rather than an indicator of a
better paying organisation. A commissioning manager or a cardiologist in a small
general hospital will be paid the same amount with a commissioning manager or a
cardiologist in a large teaching hospital. Variances between similar positions are
minimal, thus invalidating the use of the average salary ratio.
40
resources. In managerial terms, pre-booking is equivalent to demand prediction
which enables the organisation to plan accordingly. In addition, the indicator is
excluded as the reliability of the data is in doubt. Although appointments are entered
into a computerized system, the total number of appointments is carried out manually,
thus decreasing the reliability of the data. Further than that, the research found that a
few departments do record non pre-booked appointments whereas others do not.
Given that the study examines overall hospital performance, if the current data is
used, the results will be unrealistic.
The above graph shows the model, including the 35 suggested indicators for
adoption in Greece, categorised according to the 5 dimensions described in the
previous chapter.
41
Patient Experience
A&E wait: an indicator measuring the time it takes for a patient since its arrival at the
A&E department till he/she actually sees the doctor. Ideally, an electronic ticketing
system could be in place and waiting time would be easily known. In Greek reality,
there is no such a system in any of the Greek hospitals’ A&E departments visited.
Waiting time is extremely difficult to estimate, especially since it is subject to factors
such as severity and unpredicted demand. As a result, the indicator is calculated
through systematic observations. Regular visits at A&E departments, at different
dates and times are needed. In addition, since the exact estimation of the figure is
currently unrealistic, the waiting time (wt) is divided into three categories:
Outpatients wait: similarly with the A&E wait indicator, outpatients wait is used to
identify the time patients need to book an appointment with the hospitals’ outpatients
department. Outpatients department includes numerous specialities. In an attempt to
identify the outpatient surgeries with the higher demand, the research examined
hospital reports where the volume per outpatient speciality was recorded. According
to the internal reports, the outpatient surgeries reflecting the higher demand are listed
below:
• Paediatrics
• Pathology
• Cardiologic
• Dermatology
42
• Surgical
• Orthopaedics
• Urology
• ENT
• Gynaecology
In order to measure the outpatients wait from a holistic approach, the wait for each
surgery is identified and the average of the nine specialities constitutes the average
waiting time of the outpatients department. As the indicator would be applied in
hospitals varying in size and range of services, some hospitals may not offer some of
the abovementioned surgeries. In that case, the sum of waiting times is divided with
the total number of surgeries available. As a result, the reliability of the measure is
irrespective of the complexity of the outpatients department.
Hospital food: Food is another major issue for hospitals. Within hospitals, food is not
only a mean of satisfying hunger. Nutrition is critical for patients’ recovery and poor
quality of food may lead to undesirable outcomes. Patients often find food
unpalatable and large quantities are thrown away at huge costs for the hospitals. The
indicator measures hospital food quality using patients’ own experiences through the
inpatients’ questionnaire (Appendix 3).
46
Butler P., 2001
43
Volume & Complexity
Beds: The total number of available beds within the hospital. The indicator is used to
measure capacity. In addition, due to the particular characteristics of Greek public
hospitals (eg: doctors’ specialization positions are allocated according to the bed size
of each hospital), the indicator is also a measure of complexity and staffing volume.
Hospital Clinics: The figure is measured by the sum of all the clinics operating within
the hospital. It is an indicator of organizational complexity and range of services
offered. Because a broad range of clinical services increases market potential, the
number of clinical services has a positive correlation with hospital performance 47.
A&E visits: The number of patients seen in the A&E department for a specific period
of time. A volume indicator with data gained from the hospitals’ statistics department.
Outpatients’ visits: The number of patients visited the outpatients department for a
specific period of time. The figure represents a volume indicator. Taking into account
the given resources, the indicator can indirectly constitute an efficiency measure.
Evening Surgeries: The third and less important volume measure. It shows the
number of patients registered for an evening surgery appointment. Data are collected
from the evening surgeries secretariat as all appointments are recorded through a
computerized system. More than a volume indicator, the figure can also –indirectly-
be an indicator of medical staff quality. The larger the number of visits, the higher the
demand of a hospital’s medical staff.
Inpatients: The final and one of the most important volume indicators. It represents
the number of patients stayed within the hospital for more than 24hours. The data is
gained from the hospitals’ statistics departments.
Employees: The total number of staff working in the hospital. The level of an
organisation’s human resources is an indicator of its volume, especially in the service
industry. The indicator does not distinguish between different professions and
experience, as this is examined on the next set of indicators which relates to
efficiency and capability. The figure does not include medical or nursing students and
47
Friedman B. & Shortell S. 1988
44
staff working within the hospital premises but not directly linked with hospital
operations. Such staff might be canteen, private restaurant and gift shop employees.
Surgical Operations: The sum of the surgical operations occurred in the hospital
premises. The results are considered as volume indicators, but -taking into account
the given resources- it can be also seen as an efficiency indicator.
Financial Management:
Evening Surgeries Profit: Income oriented from the operation of the evening
surgeries. It measures whether the operation of the ‘evening patients surgeries’
generates sufficient revenue.
Liabilities: All of the hospitals’ financial obligations that have a negative value. As in
the case of the assets measure, the figure is indicative of the financial health of any
organisation.
Income: Money received from sales and any other income of the hospitals. The
higher the income the better for the hospital operation as it increases the capability of
meeting its obligations
45
Expenses: Goods or services purchased directly for the running of the hospitals. The
figure includes drugs, medical materials, food, wages of academic staff, construction
works, consumables, cleaning, security etc. The measure is indicative of the
hospitals’ financial requirements.
Cost/Inpatient Day: The cost for each day a patient is hospitalized. The ratio is
calculated by the sum of Expenses, divided by the sum of: LOS multiplied with
Inpatients figure.
Cost/Inpatient: The total cost for each patient hospitalized calculated by the sum of
expenses divided by the number of inpatients. In other worlds, it is the amount of
money spent to treat each patient. McNamee 48 claims that you always want your
income to exceed costs; otherwise you are paying patients for the privilege of treating
them.
Cost per Inpatient = Expenses / Inpatients
The cost per inpatient and inpatient day figures are not only indicators of financial
performance. Within the public sector, they can be used to determine pricing policy
and patient co-payment amounts. It can be claimed that the lower the figure, the
more efficient use of financial resources, but when comparing between different
hospitals, this is valid only when factors such us severity and health outcomes are
taken into consideration.
Efficiency, Capacity
Occupancy: the average daily census divided by the average number of hospital
beds during a reporting period 49. It indicates the degree to which a hospital is full. In
the private sector it is an indicator of financial health as high occupancy rates
increase the likelihood of profitability and further indicate a demand for service. In
public sector hospitals, especially within health systems which aim to provide free
48
McNamee K., 2001
49
American Hospital Association – www.aha.org
46
health care treatment and hospitals operate on a non-for-profit status, the rate is
more a sign of volume, efficiency and demand for service.
Junior Doctors’ Percentage: Studies done in a number of countries 50, including the
U.S., Australia, Israel and the U.K., indicated that most medical errors involve junior
doctors. The same study claims that junior doctors are not the only to blame but
numerous medical errors are a result of junior doctors in over their heads and
50
Medical Post, 25/09/2001
47
improperly supervised. The indicator, expressed as a percentage, shows the
proportion of junior doctors to the total number of doctors.
Doctors/bed: The total number of doctors working in the hospital divided by the total
number of beds available in the hospital. The indicator is used to represent the
operational efficiency of the hospital. A large amount of doctors per bed possibly
results in operational inefficiency whereas a small amount of doctors per bed results
in lower access and quality of care. The literature reviewed shows that there is no
ideal number of doctors per bed. Nevertheless, when comparing between hospitals
the measure is indicative of operational efficiency, access and quality of care.
Nurses/bed: The total number of nurses working in the hospital divided by the total
number of beds available at the hospital. As in the previous indicator, there is no
ideal number of nursed per bed. Nursing personnel represent the majority of Greek
public hospitals’ staff, consisting a vital component for the smooth operation of any
hospital.
51
Elges M., 2002
48
hospitals’ websites, based on Elges’s 52 suggestions as well as on the navigation and
study of web pages of various hospitals around the world. Each criterion counts 1
scoring point with the maximum a hospital can reach being at 12 points, in the case it
meets all 12 criteria.
Mortality Rates:
Patient survival is a universally accepted measure of hospital quality 53 and acute
care systems typically look at intrahospital mortality as one indicator of health
outcome 54. The lower the mortality rate, the greater the survival of the patients in the
hospital. The mortality rate indicator could increase its validity if factors such as risk,
severity, patients’ lifestyle choices and socioeconomic issues are considered. This
study ignores risk and severity elements as these are impossible to be applied,
considering the current situation of the Greek public sector hospitals. Experts say 55
mortality is just one indicator of hospital performance. It cannot be used as the
primary measure of the overall quality of care provided by a hospital. The rate is
calculated by dividing the total number of deaths occurred in the hospital over a
certain period of time by the sum of inpatients at the same period of time.
52
Elges M., 2002
53
Solutient, 2002
54
Wolff A.M., 194
55
Michigan Hospital Association, 2002
49
Mortality Rate: Deaths / Inpatients
Legal Actions: The number of legal actions against the hospital -per year- as a result
of clinical negligence. The figure is an indicator of clinical errors and omissions. The
American Institute of Medicine 56 states that over 100.000 patients die every year as a
result of medical errors and omissions which could be avoided. Despite the indicator
being found only in one out of all the reports studied, it is a performance measure
which perfectly corresponds to the Greek case. The main reasons are:
• the interest the media show on pain, death, human error and disaster –
especially when all those are combined together, as it happens on clinical
negligence-
• the increased number of legal actions against hospitals and medical staff as a
result of patient awareness and access to information
• the financial impact, when hospitals are forced to pay huge amounts on
compensations
Trust the doctors: a measurement of the perception of local citizens to the doctors of
the hospital. The indicator is used in a few of the reports reviewed. It is a measure of
health outcomes, as public opinion regarding the credibility of the doctors shapes
from the experience they have. This experience might be through their own use of
the services, their friends or relatives visit to the hospital or even from the experience
of people who don’t know personally but they know their hospital experience through
the media. A general public questionnaire is used to gather the data for this figure.
The questionnaire (Appendix 4) is distributed to local citizens, aiming to gather –for
the purpose of this study- 30 respondents for each hospital. Respondents should be
over 18years old and live within 60km distance of the hospital. The sample excludes
citizens who live in the area for less than a year, as the indented sample needs to
have some direct or indirect experience from the hospital. The interview includes
only one simple question. Citizens are asked to rate how much they trust the doctors
of the particular hospital in a scale of 1 to 10, with 1 being “not at all” and 10 being “I
fully trust them”.
56
Institute of Medicine - www.iom.org
50
prolonged hospital stay and increase of treatment costs 57. In the United Kingdom, a
study on the effects of surgical-site infections has shown an increase on the average
LOS of 8.2 days, at a cost of £1,041 58. HAIs are used in this model as a clinical
effectiveness indicator but they also indirectly constitute an indicator of efficiency in
terms of LOS, allocation of human and financial resources.
The existence of a HAI Committee is obligatory -by law-in each Greek public sector
hospital. The Committee’s roles include the planning of HAIs prevention and control.
On a regular basis, the committee has to record the HAIs rate of the hospital. At the
same time, the Hellenic Centre of Infectious Diseases Control (KEEL) operates at a
national level to monitor HAIs. KEEL is currently running a national piloting
programme of internal infections indicators.
The HAI rate can be either obtained from the hospital’s HAI Committee or the KEEL
piloting program. At this point, a dilemma exists on which is the proper rate. The
answer is given by the recency factor and the rate of the most recent record is used.
57
Brachman P.S., 1981
58
Coello R. et al, 1993
51
Chapter 6 - Application of the model
6.1 Introduction
Ioannina is 450km away from the capital city of Athens, with the whole region being
isolated from the rest of the country. A mountainous area with poor road networks,
extremely high unemployment rates, few industries and various European Union
reports listing the region as the poorest amongst all European Union Regions.
Despite these factors, the region has a tradition in culture and education, with a
highly reputable university.
59
Greek National Statistics Organisation, 2001
52
The UHI is one of the largest employers Region of Epirus
locally and one of the very few large
organisations in the area. Opened in
1989, the hospital is one of the largest
Greek Hospitals covering the health
needs of the whole region of Epirus, the
islands of Lefkas and Corfu, as well as
the southern district of the neighbouring
Albania.
View of UHI
Within the Epirus region, there are 3 more hospitals, one in each prefecture but the
research indicated that those hospitals do not offer a full range of services and a
large proportion of patients are either redirected -depending on the severity- to
Ioannina or choose themselves on beforehand to visit the University Hospital. This is
also the case with the islands of Lefkas and Corfu, two very popular holiday
destinations. Especially in Corfu, the only hospital has a very poor reputation and
fails to meet the increased demand of the summer period due to the influx of tourists.
The hospital has attracted the media attention in numerous cases. Locals prefer to
visit Ioannina as a result of dissatisfaction on the quality of care, the range of
services, access and waiting time of Corfu’s hospital. Further than that, the hospital
attracts a large amount of neighbouring Albanians who see UHI as the most
convenient place for secondary care. It is only 60km away from the Albanian borders;
the majority of the services offered are not available in Albania, whereas the quality
of care is much better than in the poorly developed Albanian Health System.
53
As a major teaching hospital, UHI works closely with the medical school of the
University of Ioannina. It provides a full range of hospital services and several of the
hospital’s clinics have built a strong reputation for excellence and innovation.
Orthopaedic, ophthalmology and urology clinics’ reputation has passed beyond the
Greek borders, with patients including celebrities from around the world according to
the local and national press.
Its organisational structure seen in the diagram below is a typical structure of a public
sector hospital. As previously mentioned, from a managerial perspective, the hospital
is divided in four areas: medical, technical, nursing and administration. Medical
sector is further categorised in 3 main categories: pathology, surgical and laboratory
whereas 2 more smaller divisions exist: psychiatric and social medicine.
The majority of the employees in the hospital (44%) are nurses with doctors
representing one fifth of the total employees. Administrative staff represents another
fifth, with technical and scientific employees sharing the remaining 15%.
54
UHI’s staff breakdown
UHI’s view
Covering an area of 61.135m², the
hospital is one of the largest hospitals in
Greece, employing approximately 2000
employees. As a teaching hospital, more
than a hundred of the hospital’s
employees are academic staff of Patras
Medical School, one of the seven medical
55
schools of the country. Each year, 120 60 new students are accepted in Patras
Medical School, through tough examinations at a national level. Considering the
duration of the study for a medical degree, which is at least 6 years in Greece, there
is a minimum of 700 medical students practising in the hospital each year.
The hospital shares the health care demand of the city with one more public hospital,
“St Andreas” General Hospital, whilst there is not any private hospital operating in the
region. In addition to “St Andreas”, there are two smaller, specialized hospitals:
Thorax Diseases Hospital and “Karamandanio” Hospital of Paediatric Diseases.
Part of Western Greece’s Regional Health Authority, the hospital shares the health
needs of 741.282 61 citizens of the Region with 8 more public general hospitals and
20 health centres spread amongst the 3 abovementioned prefectures. University
Hospital of Patras is the largest hospital in the Region of Western Greece.
Additionally, it is the only hospital of the area to be regarded by the Greek
Government and the Organising Committee of the 2004 Olympic Games as an
“Olympic Hospital”.
The majority of the patients visiting the hospital come from the three Prefectures of
Western Greece. A small proportion of 8% of inpatients -according to the hospital
database- visit the hospital from other parts of Greece (8%). It should be mentioned
that 15 years after its opening, some clinics of the hospital are now very reputable,
thus attracting patients from various parts of Greece for selected operations.
UHP has an identical organisational structure with UHI. As it can be seen in the
organisational structure diagram, UHP is managed by a board with the hospital
manager being one level lower in terms of hierarchy. In fact, the hospital manager’s
role is the most important as he chairs the board. The board consists of 6 “tactic” and
4 “acting” members: the assistant manager, the president of the hospital’s scientific
committee as well as the directors of the medical, nursing and administrative
divisions are the tactic members.
60
Ministry of National Education & Religious Affairs – www.ypepth.gr
61
Greek National Statistics Organisation, 2001
56
UHP’s organisational structure
Staff composition of UHP has negligible differences with that of UHI. As it can be
seen in the following graph, nursing staff represent the majority of the employees,
with administrative and medical employees following with 22% and 21% respectively.
The remaining percentage is covered by technical (5%) and scientific staff (9%).
Comparisons require reliable data 62 and this study attempted to achieve maximum
reliability. Interviews were contacted as part of the data gathering process, as well as
during the performance analysis stage. Direct observations were carried parallel to
other data gathering methods such us informal surveys. A large amount of valuable
62
Thompson R., 2002
57
and reliable information was gathered through the organisations’ own records and
reports. Due to the particular characteristics of the study and the sensitivity of
medical data, access to medical records was not possible.
The reliability of the data is heavily dependent on the initial data entry process of the
hospitals. Overall, the research identified that the computerized processes of public
hospitals produce reliable data, whereas the non-computerised processes are
problematic in terms of data reliability. In many cases, the research was diminished
in order to exclude unreliable elements. A typical example was with an efficiency
indicator referring to the percentage of pre-booked appointments. The indicator was
excluded during the development of the indicators phase as half of the process is
carried manually.
6.5 Results
The piloting application of the model at UHI and UHP has produced the following
results, categorised according to the 5 dimensions of the model;
In terms of beds, Patras’ hospital capacity is at 83% of Ioannina’s hospital which has
691 beds. Paradoxically, the research indicated that Ioannina operates with less staff
than Patras, as at July 2004, there were 1908 and 1965 employees respectively. The
research indicated that UHI and UHP show similar volumes of operations and
laboratory examinations. In the first case 1.300 more patients were operated in UHI
than UHP -a difference of 15%- within 2003, whereas UHP has performed 200.000
more laboratory examinations than UHI within the same year -a difference of 5%-.
As it can be seen in the graph below, in terms of complexity, UHI appears to operate
a slightly larger range of clinics. We found 29 clinics within UHI, 2 more than UHP.
The above figure does not mean that UHI covers the full range of services UHP
offers and vice versa. A few of the 27 clinics operating in Patras -such as
paedosurgery-, do not exist in Ioannina while at the same time, some of the clinics
UHI operates -such us child orthopedics- do not exist in Patras.
58
Volume & Complexity Dimension: Results (I)
The University Hospital of Patras served 84% of the outpatients that Ioannina served
(135,289 patients), a difference of approximately 35,000 patients for 2003. The
evening surgeries volume has a similar ratio with capacity, with Patras reaching 73%
of Ioannina’s volume (21,073 patients). The higher variance in the volume indicators
exists in the inpatients as well as in the A&E figures. In the first case Patras served
only 60% of the inpatients that Ioannina served (53,927 inpatients), whereas in the
second case Patras A&E department has received 2.22 times more patients than
Ioannina.
The explanation of the low volume of inpatients in Patras lies with the fact that Patras
claim to have approximately 30,000 day-patients. At the same time, Ioannina’s
hospital has recorded only 500 day-patients for the same year! The research
indicated that the irregularity exists due to the lack of a national standard of patients’
categorisation system. Some hospitals treat haemodialysis patients and other
categories of regular visitors as day-patients whereas other hospitals, like in the case
of Ioannina, do not record them at all.
There is no clear explanation regarding the irregularity of the A&E indicators, where
Patras shows 2.22 times more visits than Ioannina. Both organisations share the
health care demand of their local areas by being on duty in a 1:1 ratio. Consequently,
both A&E departments operate approximately 185 days a year. Following
discussions with staff of both hospitals, what seems to be the most logical
59
explanation for such a discrepancy is either a different patient recording system at
A&E department or a failure of UHI to record all A&E visits.
Efficiency, Capability
University Hospital of Ioannina found to work at a lower occupancy rate for 2003 than
University Hospital of Patras, despite its inpatients figure being 1.66 times more than
the Patras’ figure (53927 Vs 32433). The explanation behind the occupancy level lies
behind two reasons:
60
the staff end up with a two months holiday period each year as they are ‘covered’ by
fellow colleagues”.
University Hospital of Patras is also ahead on the number of doctors per bed with its
figure very close at a 1:1 ratio while University Hospital of Ioannina operates at a
rate of 0.67 doctors/bed. UHP further overrides UHI in the nurses/bed ratio where it
reaches a 1.54 figure. At the same time UHI operates with a ratio of 1.29 as it can be
seen in the graph below.
Patras’s website meets 58% of the criteria set by reaching 7 out of the 12 website
criteria, whereas Ioannina’s hospital does not currently have a website thus meeting
0% on this indicator. Out of the 12 functions, Patras’s website (Appendix 13) is
missing a staff directory, login to accounts system, online appointment scheduling,
newsrooms/discussion boards and a careers section. It should be mentioned that in
the beginning of the study, Ioannina’s hospital did have a website but it was of poor
content and last updated a couple of years ago. Administrative staff of the hospital
claims that a new website is under construction and this might be the reason behind
the “page can not be found” message on the internet explorer window when the
www.uhi.gr address is entered.
Contrary to the occupancy and LOS results which could not be up to date and
represent the last year (2003) the rest of the indicators are as at July 2004.
61
Patient Experience
The patients’ experience data collection process was the most time consuming part
of the research. The questionnaires were distributed between July and August 2004.
The survey involved: 60 outpatients, 60 inpatients of randomly selected clinics, 60
A&E patients and 60 local citizens for the ‘trust the doctors’ questionnaire. All
respondents were randomly selected. The level of respondents was selected to be
30 persons per questionnaire per hospital due to the time constraint of the study.
The low number of respondents limits the ability of the study to elicit accurate
conclusions. Nevertheless, the main purpose of the survey was not to compare the
two hospitals but to test the applicability of the selected indicators.
The analysis of the questionnaire results indicates that both hospitals achieved a
greater patient satisfaction rate on their emergencies departments. In a scale of 1 to
10, the average score of the 30 respondents of each hospital is 7.43 for Ioannina and
7.37 for Patras (Appendix 5). The worst scoring for both hospitals is found in their
outpatients department with 66% satisfaction for Ioannina and 64% for Patras
(Appendix 6). The questionnaire results indicated that both hospitals have similar
achievements in terms of patient experience as it can be seen in the following graph.
UHI achieved a slightly higher score in the A&E & outpatients departments with UHP
overcoming UHI at the inpatients experience results (7.17 Vs 7.03 – Appendix 7).
The analytical results can be found in the appendices section of this study.
The patient experience set of indicators involves two more measures; cleanliness &
food quality. In the first measure, both hospitals reached high levels. The average
score of the 90 respondents for each hospital was 8.32 for UHI and 7.94 for UHP.
The high levels achieved in the cleanliness results were not repeated with the food
quality measure where 30 inpatients of UHI rated the food with an average score of
6.03. At the same time, a similar number of inpatients at Patras rated hospital food
with an average of 7.07 out of 10.
The patient component of the model further involves two ‘waiting’ measures’;
outpatients wait and A&E wait. Outpatient wait found to be between 3 and 4 days in
both hospitals as it can be seen in the graph below. The detailed waiting time for
each outpatient surgery for both hospitals can be seen in Appendix 12. Numerous
62
visits were needed, various times were selected and long hours were spent in the
A&E department of the hospitals, in order to accurately identify the A&E waiting time.
The findings are similar for both organisations. The large majority of A&E visitors see
a doctor within the first sixty minutes.
Financial
If patients’ experience results were the most time consuming part of this study,
financial data gathering was the most difficult part of the research. As Greek public
hospitals have huge debts, the financial performance of public hospital often attracts
media attention. As a result, hospital staff was very suspicious when we asked for
financial data. In the case of Ioannina, despite that these data -by law- should be
publicly available, following an effort of 3 weeks, we finally had to fill an application
form and get the written authorisation of the hospital manager in order to access the
data.
The poor financial performance of Greek public sector hospitals can be also seen in
the following graphs. UHI is performing better than UHP, but both hospitals show
negative results. In both cases, the liabilities figures are larger than the assets figures,
resulting in a net loss. UHI’s loss is almost half of Patras’, as the assets figure is
larger (€74m Vs €64m), while at the same time the liabilities figure is smaller (€82m
63
Vs €85m) than Patras. UHI financial figures are better than UHP, despite that UHI
serves a large amount of neighbouring Albanias, who are treated as ‘poverty stricken’
and contribute a minimal amount towards their health costs.
The next graph shows a more detailed view of the financial performance of both
hospitals. The first indicator -evening surgeries’ profits- is larger in Ioannina
(€570,000 Vs €440,000) because UHI treated a larger amount of patients in its
evening surgeries as it was mentioned in the volume indicators (21073 patients Vs
17715 patients). According to the above results, UHI had a profit of €27 per patient
while UHP had approximately a profit of €24, representing €3 less per patient. The
income indicator shows Ioannina to gain 12.5% more income than Patras, a figure
which can be characterised logical, as Ioanina treats a larger number of patients. The
debtors’ amount is similar for the two hospitals, at a level of €50m. Considering the
higher volume of UHI as well as the larger amount of its assets, normally UHPs
debtors should have been lower than UHI’s level.
Finally, the wider variation is found at the expenses figures. Within 2003, UHP shows
an expense figure of €71m, 2.15 times or €38m more than the expenses of UHI for
the same year! A more detailed research on the reasons of such a variance indicated
that UHP spent €3.5m in medical equipment within 2003. At the same time, the
64
hospital implemented a construction project of €7m for refurbishments, renovations
and expansion. A further reason lies with the fact that UHP was named an ‘Olympic
Hospital’ by the Greek government. This actually meant that the hospital would be in
duty during the 2004 Summer OIympic Games and large amounts were spent to
ensure that the hospital had the adequate facilities to meet the criteria set by the
International Olympic Committee. Despite the above factors, the research indicated
some irregularities in the expenses of UHP when compared with UHI. Out of the
€71m expense figure of Patras for the 2003financial year, €20.8m were spent on
drugs and €18.3m were spent on medical material. At the same period, UHI -a
hospital which treats a larger volume of patients and has a higher capacity despite
operating with less staff- spent €9.2m and €8.6m for drugs and medical material
respectively, less than half the amount UHP spent!
The expense figures are also affecting the cost/inpatient and cost per/inpatient day
indicators. As it can be seen in the following graph, the two hospitals show a huge
variance in the abovementioned indicators over the 2003 period. The cost per
65
inpatient day in Patras is 2.3 times more than in Ioannina (€406 Vs €176) and
similarly, the cost per inpatient is 3.5 times more than in Ioannina (€2177 Vs €616).
Considering the distance between Patras and Ioannina (200km), the fact that the two
hospitals are funded from the same resources as well as the fact that UHI has spare
capacity, and –most of all- the cost/inpatient and cost/inpatient day figures,
theoretically, it would be more cost-effective to transfer hospitalized patients from
UHP to the University Hospital of Ioannina! Each hospitalized patient costs €1,561
(£1,115) more in Patras than in Ioannina. Alternatively, the difference between a
patient day in Patras and a patient day in Ioannina is €227 (£162).
Clinical
Legas actions: According to the legal advisor of the university Hospital of Ioannina,
since January 2003, there was only one legal action for medical errors and omissions
against the hospital. Mr Georgoulis also added that the court decision was in favour
of the hospital and no compensation has to be paid. During the same period, there
were three legal actions against the University Hospital of Patras. Mr Nikoletatos,
informed us that all cases are still pending. Medical professionals of the two hospitals
who have also worked in other hospitals of the country, informed us that this is a very
66
low figure for both hospitals, as legal actions against the hospitals are an often
phenomenon in most hospitals of the country.
Mortality Rates: There is a significant difference between the two hospitals in the
mortality rate figure. Ioannina appears to have 9.6 deaths per 1000 patients, less
than half of Patras’s rate (20.2 deaths per 1000 patients). The indicator is not risk
adjusted, but taking into account that Ioannina’s hospital serves a large proportion of
neighbouring Albanians whose health status is much lower than the Greek
population 63 -as it can be seen in Appendix 9-, this is a contradicting result.
Trust the doctors: Local citizens seem to trust the doctors of both hospitals at a high
level. Both hospitals achieved high scores in the survey carried between July and
August 2004 as part of this study. Ioannina reached 94% and Patras 92% (Appendix
8). Following discussions with many questionnaire respondents, the reasons for such
high results are:
Internal Infections:
Two prevalence studies for the estimation of hospital acquired infections (HAIs) were
carried out in the University of Ioannina, during an effective nosocomial infection
program 64 . The first study was carried in September 1999 and the second on
December 2000. According to an interview with the president of Ioannina’s Internal
63
WHO – World Health Report 2003
64
Levidiotou S., Barba V., et al 2002
67
Infections Committee, as the Committee Members are engaged in many roles within
the hospital and overloaded in their initial positions, since December 2000, there are
no further data. The first study involved a sample of 488 patients with the second
study recording 448 patients. Both studies reported an HAI rate of 9.1% 65, a figure
which is in line with rates from surveys carried out in other European countries 66.
Despite our efforts, none of Patras’ Internal Infection Committee members could be
contacted, but the most recent figures on the HAIs rate of the two hospitals are found
in KEEL’s National Piloting Programme of Internal Infections Indicators. Two studies
were carried out by KEEL, the first on June 2002 and the second on June 2003.
According to the results of the study, Ioannina were ranked at the top of the list, with
a rate of 0% -the lower HAIs rate nationally-, while Patras scored a rate of 3.14%. It
should be mentioned that the samples where 136 and 542 patients for UHI and UHP
respectively, whereas the average rate of the 24 hospitals involved in the program
was at 6.59% 67.
65
Levidiotou S. et al, 2002
66
Gikas A. et al, 2004
67
Hellenic Centre for Infectious Diseases Control - www.keel.org.gr
68
6.6 Overall Performance
The fully concentrated version of results -including indicators of all dimensions- for
each hospital is available in Appendix 10. In addition an overall comparison graph
using a scoring system of 1 to 10 for each indicator is shown below. Due to the time
constrain, the study does not attempt to apply weighting methods at the indicators.
The development of the following graph -which is only used to graphically represent
the variances between the two organisations-, is a result of a very simplified
approach. Each of the 35 figures was converted to a 1 to 10 scale, with 1
representing the lowest possible and 10 indicating the maximum possible result. The
simplified approach also turned indicators were the larger value signaled worst
performance (eg: mortality rates and legal actions figures), in such a way that the
larger the value, the lower the scoring at the 1 to 10 scale.
The graph shows that UHI covers a wider area within the 35 indicators circle, which
actually means that the hospital’s overall performance is better than UHP. This is a
consequence of UHI achieving better performance figures in more indicators than
UHP. Nevertheless, we can not yet conclude that UHI is –in overall terms- a better
performing organisation than UHP. This can only be claimed when the health system
enables the application of severity and risk factors.
69
Chapter 7 - Conclusions and recommendations
7.1 Conclusions
• Despite the demand for public reporting clearly increasing, Greek public
sector hospitals have done little to embrace the opportunities this trend brings.
The science of healthcare quality indicators is not even in its infancy. To the
best of our knowledge, this study is the first ever work approaching
performance measurement within the Greek NHS!
• Greek Health System and public sector hospitals are well behind in terms of
data/info gathering and usage, performance management and quality
assurance initiatives. Decision makers of the Greek NHS seem to focus their
attention on the daily running of the organisations (micro-approach) rather
than on the future of the organisations, their development and growth (macro-
approach).
70
• The current accounting system obstructs the implementation of a
performance measurement model which includes a financial focus. The
implementation of International Accounting Standards within public sector
organisations will facilitate financial performance analysis and comparisons
between different organisations through the use of financial ratios.
• The application of the model and its results further proved that comparisons
can be made, useful information can be produced, variances can be detected
and performance can be judged. The results provided evidence of similar
organisations –in terms of volume- having huge variances in performance, as
in the case of the financial dimension of the two hospitals examined.
71
matters regarding health management and strong professional groups
resisting to change.
It is important to mention that the results produced in this study might seem a result
of an easy data gathering process, as similar European establishments might be
publishing such figures in their website. In reality, the data collection process was the
most difficult part of the study -an effect of the mismanagement of Greek hospitals
and absence of available information regarding their operation-.
7.2 Recommendations
72
collection for usage in decision making. The current situation on patient and
hospital activity data can be described as chaotic, with no interest from any
stakeholder to take advantage of this huge amount of information. It is
characteristic that the sum of employees in the statistic departments of the
two hospitals selected, (which cover together an area of more than 1 million
people - 10% of the country’s population), is 2. Alternatively, one person for
half a million of population!
73
References
Baker G.R., G.H. Pink (1995). A Balanced Scorecard for Canadian Hospitals.
Healthcare Management Forum. 8(4): 7–13
Butler P., NHS Quality & Performance, Q&A: Hospital Cleanliness, Guardian,
10/04/2001
Canadian Institute for Health Information, Hospital Report: Acute Care, 2003
Centre for Health Policy Studies, Columbia – MD, Understanding and Choosing
Clinical Performance Measures for Quality Improvement: Development of a
typology, Report for the Department of Health and Human Services, 31/01/1994
Cluzeau F., Littlejohns P., Grimshaw J., Hopkins A., Appraising Clinical Guidelines
and Development of Criteria – a pilot study, Journal of Interprofessional Care
1995; 9:227–9
Coello R., Glenister H., Fereres J. et al, The Cost of Infection in Surgical Patients:
a case control study, Journal of Hospital Infections, 25:239-250, 1993
74
Comarow A., Higher Volume, Fewer Deaths, US News & World Report Online, July
17, 2000
Davies H.T.O., Lampel J., Trust in Performance Indicators? Quality in Health Care,
7:159-162, 1998
Department of Health & Family Services, Quality & Outcome Indicators for Acute
Healthcare Services, Australian Government Publishing Service, 1997
Elges M., Driving traffic to your Web site - Web Technology in Focus, Healthcare
Review, Jan 8, 2002
Ferrao C., Global Hospital Indicators, Ontario Joint Policy & Planning Committee
Guy’s & St Thoma’s Hospital, Working Together to Improve Patient Care, Annual
Report 2002-03
75
Health Care Quality Steering Committee, Consumer & Provider Views on Key
Dimensions of Quality Hospital Care: A review of the Literature, Rhode Island
Department of Health
Hearnshaw H., Harker R., Cheater F., Baker R., Grimshaw G., A Study of the
Methods used to Select Review Criteria for Clinical Audit. Health Technol
Assess; 6(1), 2002
Hellenic Republic – Ministry of Health & Welfare, Health, Health Care & Welfare in
Greece, 2003
Kaplan R., Norton D., The Balanced Scorecard – Measures that drive
performance, Harvard Business Review, Feb p.301, 1992
Kazandjian V., Wood P., Lawthers J., Balancing Science and Practice in Indicator
Development: The Maryland Hospital Association Quality Indicator Project,
International Journal of Quality in Health Care 7 (1): 39 – 46, 1995
Luft H.S. & Hunt S.S., Evaluation of Individual Hospital Quality through Outcome
Statistics. JAMA; 255: 2780, 1986
McNamee K., Business & Health Sense, Acupuncture Today, Vol 2, Issue 1,
January 2001
Medical Post, Medical Errors: the result of recurring system failures, Vol. 37,
Issue 32, 25.09.2001
Mossialos E., Citizens Views on Health Care Systems in the 15 Member States
of the European Union, Health Economics 6(2) 109-116, 1997
76
Productivity Commission, Private Hospitals in Australia, Commission Research
Paper, Aus Info, Canberra 1999
Solucient, 100 Top Hospitals: Benchmark for Success, National Report, 2002
To Vima, Public Hospitals in the …Acute Care Department: Reasons lead public
hospitals having huge debts, 09.05.2004
Wilson R. C., The UK Approach to Better Hospital Food, King’s College Hospital,
Denmark Hill, London, United Kingdom
77
World Health Organisation, World Health Report: Shaping the Future, 2003
78
URL’s
79
The National Quality Forum (USA) www.qualityforum.org
University Hospital of Ioannina (GR) www.uhi.gr
University Hospital of Patras (GR) www.pgnp.gr
Wisconsin Department of Health and Family Services (USA) www.dhfs.state.wi.us
World Health Organization (INT) www.who.int
80
Appendix 1 - Outpatients Questionnaire
In the next set of questions, please circle your answer. The scale in the first question
is from 1 to 10, with 1 being “very dirty” and 10 being “very clean”. In the last
question, the scale remains the same with 1 being “very poor” and 10 being
“excellent”. Before answering the last question, please consider:
81
Patient Initials: ______
Age: ______
1 2 3 4 5 6 7 8 9 10
How would you rate your overall experience in the outpatients department?
1 2 3 4 5 6 7 8 9 10
82
Appendix 2 - A&E Questionnaire
In the next set of questions, please circle your answer. The scale in the first question
is from 1 to 10, with 1 being “very dirty” and 10 being “very clean”. In the second
question, the scale remains the same with 1 being “very poor” and 10 being
“excellent”. Before answering the last question, please consider:
• How long did you wait to be seen by a doctor
• Whether different members of staff gave you conflicting information
• Whether doctors explained the reasons for any actions/treatments
• Whether you were given the right amount of information
• Whether staff explained risks/benefits of any treatment needed
• Whether staff explained medication side effects
• Whether you had enough time to discuss your problem with the doctor
• Whether medical professionals were listening what you had to say
• Whether your questions were answered clearly
• Whether you were treated with respect and dignity
• How clean was the A&E department
• Whether you had enough privacy when treated/examined
83
Patient Initials: ______
Age: ______
1 2 3 4 5 6 7 8 9 10
How would you rate your overall experience in the A&E department?
1 2 3 4 5 6 7 8 9 10
84
Appendix 3 - Inpatients’ Questionnaire
In the next set of questions, please circle your answer. The scale is from1 to 10, with
1 being “very poor” and 10 being “excellent”, except in the first question where the
scale remains the same but 1 represents “very dirty” and 10 “very clean”. Before
answering the last question please consider:
85
Patient Initials: ________
Age: ____________
1 2 3 4 5 6 7 8 9 10
1 2 3 4 5 6 7 8 9 10
1 2 3 4 5 6 7 8 9 10
86
Appendix 4 - Trust the Doctors Questionnaire
In the one and only question, please circle your answer. The scale is from1 to 10,
with one being “not at all” and 10 representing “fully trust them”.
Initials: ________
Age: __________
87
Appendix 5 - A&E Questionnaire - Results
88
Appendix 6 - Outpatients’ Questionnaire – Results
89
Appendix 7 - Inpatients’ Questionnaire – Results
90
Appendix 8 - Trust the Doctors Questionnaire – Results
91
Appendix 9 - Health Statistics: Greece & Albania
Greece Albania
92
Appendix 10 – Concentrated table of results
income 36 32 €M 2003
debtors 50 50 €M 2003
expenses 33 71 €M 2003
cost/inpatient 616 2177 € 2003
cost/inpat day 176 403 € 2003
mortality 9.6 20.2 per 1000 2003
legal actions 1 3 2003
clinical
93
Appendix 11 - Staff Breakdown & Percentage of Organic Positions Covered by
Occupational Group
94
Appendix 12 - Outpatients’ wait by surgery
95
Appendix 13 - University Hospital of Patras’s website www.pgnp.gr
A well designed website which is updated regularly. Available only in Greek language.
96