1. PERFECT = PERFormance, Effectiveness and Cost of Treatment episodes

2. GENERAL AIM OF PERFECT
To develop methods for register-based measurement of the cost-effectiveness of treatment and to create a comparative database that allows the treatments given and their costs and outcomes to be compared between hospitals, hospital districts, regions and population groups.

3. Costs and Outcomes Health care system Determinants of a disease
Pre-stage
of a disease
Acute stage of
a disease
Chronic stage of
a disease
Costs and
Outcomes
Health care system
Economic
resources
Incentives
Medical
knowledge

4. PERFECT
Produces comparative information on treatments and their costs and outcomes for treatment monitoring and development.
Creates indicators and models for monitoring the content, quality and cost-effectiveness of treatment episodes in specialised medical care.
Assesses factors that influence costs and outcomes.
Develops methods for the register-based measurement of cost-effectiveness, and comes up with proposals concerning the data content of national level registers in order to improve the continuous monitoring of cost-effectiveness.
Develops an approach and methodology that can be subsequently applied to other disease groups as well.
Compares costs and outcomes at an international level

5. PERFECT - disease groups
The focus will be on selected disease groups with sufficient significance in terms of costs and burden of illness:
Acute myocardial infarctions, extended later to revascular procedures (CABG, PTCA)
Hip fracture
Breast cancer
Hip and knee replacements
Very low birth weight infants
Schizophrenia
Stroke

6. Organisation of the Project
Each subprojects has own expert group, together 50 clinical experts
Develop the disease/health problem specific protocols
Define the content of data
Is responsible for basic reports
Aims to do international comparison
THL / CHESS is responsible for overall coordination of the project, gathering and analysing the data and for health economic, health service research and statistical expertise

7. Description of PERFECT-Project
BASIC REPORTS
THL
Hospital discharge register,
Hospital productivity
(Benchmarking)
database
SOCIAL INSURANCE
INSTITUTION
Register on Health and
Social Benefits
STATISTICS OF FINLAND
Cause-of-Death Register
OTHER
REGISTERS
Implant Register on
Orthopaedic Endoprostheses,
Hospitals patient registers
RESEARCH
PERFECT DATA BASE
FEEDBACK

8. Content of basic reports http://www. thl
Levels
Hospital Districts (responsible for providing specialist care in Finland) based on the municipality of the patient)
Hospitals (over 50 patients), based on patients treated in a hospital
Indicators
Basic information on patients such as number of patients, age structure, co-morbidity (about 40 indicators)
Process indicators describing length of stay, outpatient visits, use of procedures, drugs, cost of care (about 140 indicators) .
Indicators describing outcomes of patients (about 60 indicators)

9. Current status of Perfect
Acute myocardial infarction (regional level and hospital data available from the years /2013)
Hip Fracture (regional and hospital level data available from years /2013)
Hip and knee replacements (regional and hospital level data available from years )
Very low birth weight infants (regional and hospital level data available from years )
Schizophrenia (regional and hospital level data available from years cohorts)
Stroke (regional and hospital level data available from years /2013)

10. Key questions
Definition of an episode: When it starts and when it finishes (follow up time)?
Balancing: what can be done on routine basis with scientific/methodological aspects

11. Definitions of the episodes
Admission to ward A
Procedure/treatment in ward A
Admission to ward B
Discharge to another hospital
Outpatient visit
Medication purchase
Total episode of care
First hospital episode
time
Discharge home or nursinghome

12. Solutions used in basic reports
1) A definition of patient group so that that they are as comparable as possible
Examples
New hospitalised AMI (ICD-10 I21-I22) aged 40-85:Patients were excluded if
i) they were discharged alive and had a length of stay, including transfers, of less than 3 days,
ii) they had been hospitalised for AMI during the previous year (365 days)
iii) if the were institutionalised before hospital admission for AMI
Stroke. First ever stroke and not institutionalised
2) Risk adjustment for co morbidity using information on previous use of hospital inpatient care (since 1987), registered individuals suffering from certain specified chronic conditions (Social insurance institution) and purchases of prescribed medicines
3) Standardising by modelling and calculation of confidence intervals

13. Measurement of outcome
So far mostly risk adjusted traditional measures (7 day, 30 day, 1 and 5 year) mortality, infections, readmissions, reoperations, days until return to home, days spent at home in during follow up (e.g. one year) placement in long-term care

15. Measurement of cost
Estimation of standard (relative) cost for each outpatient visit and inpatient discharge in the database Based on patient level cost data in one hospital district.
Assumes that unit cost of producing specific procedures or DRGs are same all over country in each year
Cost of prescribed drugs based on their retail prices

16. Measurement of cost in basic reports
Cost and utilisation data for all inpatient care, outpatient visits in specialised care and private doctors’ visits and prescribed medicines
Standardisation modelled using GLM (log link with gamma distribution). Two or three part model for cost of prescribed medicines. Deaths or survival time included in standardisation

21. Next steps Extending the approach to primary and social care
Pilot project in capital area and municipality of Kouvola (about 1.2 million parsons)
Data 2006 
Sub projects
Children
Psychiatric patients
Elderly
Substance abusers.
Comparison with Oslo and Stockholm

22. One year cost (EUR) of stroke patients, 2009-2010

23. Challenges Reporting of secondary diagnosis and procedures
Time lag, reliable comparison need pooled 2-3 years data
Development of quality registers? Under discussion
Development of Patient Data Repository system, under current legislation not possible to use for research and statistics

24. Extension of Perfect: EuroHOPE

25. Partners
Centre for Health and Social Economics (CHESS),National Institute for Health and Welfare, Finland
Centre for Research on Health and Social Care Management, Universita Commerciale Luigi Bocconi, Milano, Italy
Health Services Management Training Centre, Semmelweis University, Budapest, Hungary
National Institute of Public Health and the Environment, the Netherlands
University of Oslo, Department of Health Management and Health Economics, Norway
Ragnar Frisch Centre for Economic Research, Oslo, Norway
University of Edinburgh, Scotland
Medical Management Centre (MMC), Karolinska Institutet, Stockholm, Sweden

26. Patient group specific work in EuroHope (I)
Five patient groups subject to
acute myocardial infarction (AMI)
stroke
hip fracture
breast cancer
very low birth weight infants
Clinical experts from each of the participating countries
The protocols define
inclusion/exclusion criteria
definition of cycle of care (when it starts, follow-up etc.)
comorbidities (used in risk adjustment)
specification of outcome measures

27. data used for comparative research
EuroHOPE Data
Comparison of countries, regions and hospitals
EuroHOPE research
National EuroHOPE database
National EuroHOPE compari-son data
International
EuroHOPE comparison data
National discharge register
National mortality register
Other national registers
Protocols
Anonymous
individual level
data used for comparative research
National research and bencmarking

29. Regional variation in mortality, AMI
Age- and sex-adjusted one-year mortality by regions, AMI in 2008 (2009 in Norway).
AMI: most of the Italian and Swedish and all the Norwegian regions performed better than average regions for all countries in one year mortality, whereas some Finnish, most of the Scottish and all the Hungarian regions performed poorer than average
The use of PCI had a negative but not statistically significant effect at the regional level.

30. Regional variation in mortality, hip fracture
Age- and sex-adjusted one-year mortality by regions, hip fracture in 2008 (Norway 2009)
In hip fracture, well-performing regions were found—in addition to Italy and Sweden—from Norway
Regarding ischaemic stroke and hip fracture patients, the regional differences in length of stay and mortality were not related to any of the analysed regional level factors.

31. Measurement of hospital quality performance (30-day survival)
Empirical Bayes estimates of hospital effects for quality obtained from a model, where age, gender, comorbidities and transfers to a higher level hospital are taken into account
The effects do not as such have exact practical interpretation but we can estimate that survival difference between the lowest and highest hospital was 30 percentage points (min 67.5, max 97.5) in the care of AMI patients
Stockholm

32. Measurement of hospital quality performance (30-day survival)
Empirical Bayes estimates of hospital effects for quality obtained from a model, where age, gender, comorbidities and transfers to a higher level hospital are taken into account
The effects do not as such have exact practical interpretation but we can estimate that survival difference between the lowest and highest hospital was 30 percentage points (min 67.5, max 97.5) in the care of AMI patients
Stockholm

33. Hospital-level comparison: Measurement of hospital quality performance (30-day survival)
Empirical Bayes estimates of hospital effects for quality obtained from a model, where age, gender, comorbidities and transfers to a higher level hospital are taken into account
The effects do not as such have exact practical interpretation but we can estimate that survival difference between the lowest and highest hospital was 30 percentage points (min 67.5, max 97.5) in the care of AMI patients

34. Hospitals cost performance in care of AMI patients based on empirical Bayes estimates of random coefficient

35. EuroHOPE now and future
Maintains national and regional indicators at
Provides audience with scientific and policy relevant results
Health Policy articles – 2 pieces on the air already!
Health Economics Supplement – Beginning of 2015
Variety of clinical articles – 4 papers submitted
Stream of publications in EuroHOPE Discussion Papers Series at
Continues the performance evaluation and extends the activity to other countries and other patient groups.
BRIDGE project: updating the data from Finland, Sweden, Norway, Italy, Hungry and Denmark.
CanHOPE.
Nordic comparison including socioeconomic data (Finland, Norway)