Death certificate collection and coding: Current practices in Europe and approprieteness for rare di

1
Death certificate collection and coding Current
practices in Europe and approprieteness for rare
diseases
G Pavillon E Jougla INSERM CépiDc - IFR 69 -
France Epidemiological Center on Medical Causes
of Death
Rare diseases Task Force Working group on health
indicators Monday 30 January 2006
2
Summary

• Mortality indicators
• Causes of death coding
• Automated coding systems
• Electronic death registration
• International collaborations
• Application to rare diseases

3
Mortality indicators
4

• Use of mortality indicators
• Trend studies (time evolution)
• Regional and international comparisons (space)

• Data comparability ?
• Certification level (which cause of death ?)
• Codification level (which classification ?)

5
London 1632 mortality tables Abortive and
stillborn 445 Affrighted 1 Aged 628 Ague 43 Ap
oplex and meagrom 17 Bit by a mad dog 1 .....
6
International co-ordination
? 1853 First international congress
? 1855 International list (139 groups)
? 1900 International Classification of Diseases
Revision 1 (ICD1) (179 groups)
? 1948 WHO - ICD-6
7

• International regulations on
• Classification
• Underlying cause definition
• Death certificate form
• Rules and guidelines
• Statistical definitions
• Tabulation

8
Causes of death coding
9

• ICD-10 (1992-1994)
• 12 000 classes
• 3 volumes
• - Tabular lists
• - Instruction manual
• - Alphabetical index

10

• Underlying cause
• a) the disease or injury which initiated the
  train of morbid events leading directly to death,
  or
• b) the circumstances of the accident or violence
  which produced the fatal injury.

11
(No Transcript)
12
International form of death certificate Part I
a) .. b) .. c)
.. d) .. Part II
...
13
ICD Classification Rules

• ? Selection Rules
• To identify the underlying cause of death

? Modification Rules To improve precision and
usefulness
14
Part I a) Respiratory
failure b) Lung metastasis c) Cancer of
bladder d) Part II Alcoholism
15
Part I a) Respiratory failure b)
Influenza c) Cardiac insufficiency
d) Part II
Part I a) Acute pancreatitis b)
anorexia, dehydration c) cardio-respiratory
arrest d) Part II
16
(No Transcript)
17

• Data are comparable if
• Use the ICD rules
• Use the same medical knowledge

18

• How to avoid coder differences?
• Improve training
• Use the same medical knowledge
• Use automated coding systems

19
Automated coding systems
20

• An Automated Coding System (ACS)
• Codes each condition reported on the death
  certificate
• Selects the underlying cause of death

21

• An ACS includes
• a dictionary
• Acute Myocardial infarction I219
• ICD rules
• Medical knowledge

22
MMDS Coding system (NCHS)
First implementation tests in 1968 Dictionary
(160 000 entries) Medical knowledge base
(decision tables) Coding modules (SuperMicar,
Micar) Underlying cause module (ACME)

23
MMDS decision tables are a standard but Coding
modules are difficult to transpose to
non-English- speaking countries
Countries using MMDS US, UK, Australia
... or parts of it Sweden, Brazil, Hungary,
France ...
24
Styx Coding system (CépiDc Inserm)
First implementation in 2000 Dictionary (100
000 entries) Medical knowledge base (MMDS
decision tables)

Countries using Styx France, Canada (Québec),
Tunisia, Luxembourg
25
Styx architecture
Medical causes of death text
STYX
26
(No Transcript)
27
Electronic death registration
28
Current system
29

• Drawbacks
• Time
• Confidentiality
• Interactivity

30
Electronic death registration
31

• Advantages
• Quick assessment of mortality level
• Elimination of paper documents
• Improvement of confidentiality
• Improvement of quality (interactive system)

32

• Schedule
• Preliminary study (end 2005)
• Experimentation 2006
• Deployment 2007 (hospitals)

33
International collaborations
34

• International Collaborations
• World Health Organisation (WHO)
• European Community (Eurostat)
• International Collaborative Effort (NCHS)

35

• World Health Organisation
• Mortality Forum
• Mortality Reference Group
• Mortality statistics

36

• European Community (Eurostat)
• Report on Automated Coding Systems
• Report on Comparability and Quality Improvement
• Report on Certification training tools
• Quality control tool
• Mortality statistics (Regional level)
• Atlas on Mortality

37

• ICE (NCHS)
• International meetings (1996, 1999, 2003)
• Recommendations
• User's group on decision tables

38
? Usefulness of mortality statistics implies the
production of high quality data
? Automated coding systems contribute to quality
? International collaborations are necessary to
develop such systems
39
Application to rare diseases
40
Classification ICD 9
41
Classification ICD 10
42
Mean annual number of deaths (19912002)
total Underlying cause
43
Mean annual number of deaths (19912002) by
sex Underlying cause
44
Mean annual number of deaths (1999) by
age Underlying cause
45
Mean annual number of deaths (19912002)
Total Underlying and multiple causes
46
Mean annual number of deaths (19911999) by
sex Underlying and multiple causes
47
Mean annual number of deaths (2002)
Total Analysis of causes of death text Code
E75.2 (Other sphingolipidosis)
48
Analysis of rare disease mortality
? Underlying cause
? Multiple causes
? Text of causes
49
Death certificate collection and coding Current
practices in Europe and approprieteness for rare
diseases
G Pavillon E Jougla INSERM CépiDc - IFR 69 -
France Epidemiological Center on Medical Causes
of Death
Rare diseases Task Force Working group on health
indicators Monday 30 January 2006
50
(No Transcript)