Burden of Disease Research Unit Cause of death statistics from vital registration Debbie Bradshaw WHO-FIC Collaborating Centre (Under designation)

Burden of Disease Research Unit

Cause of death statistics

from vital registration

Debbie Bradshaw

WHO-FIC Collaborating Centre (Under designation)

Outline of presentation

• Illustrate the public health value of cause of death statistics – City of Cape Town

• Process of collecting cause of death statistics• ascertaining the cause of death• underlying cause of death• ICD coding• cause of death statistics

• Quality of cause of death certification

Cause of death profile in Cape Town, 2004

Source: Western Cape BOD project using City of Cape Town mortality data

Age specific HIV death rates, Cape Town 2001 - 2006

HIV Males

0

50

100

150

200

0-4 5-14 15-24 25-34 35-44 45-54 55-64 65-74 75+

HIV Females

0

50

100

150

200

0-4 5-14 15-24 25-34 35-44 45-54 55-64 65-74 75+

2001

2004

2006


Age-standardised death rates (per 100 000 population)

Cape Town, 2001 – 2006

0

20

40

60

80

100

120

2001 2002 2003 2004 2005 2006

Ischaemic heart disease Stroke

Diabetes mellitus Hypertensive heart disease


0

200

400

600

800

1000

1200

1400

1600

1800

2000

Kha

yelit

sha

Eas

tern

Mitc

hels

Pla

in

Klip

fon

tein

Tyg

erb

erg

We

ste

rn

Nor

ther

n

So

uth

ern

Ca

pe T

ow

n

Dea

ths

pe

r 1

00 0

00

HIV/AIDS Other Communicable/mat/peri/nutrition Non-communicable diseases Injuries

Age-standardised death rates (per 100 000 population)

by broad cause for sub-districts of Cape Town, 2006

Source: Groenewald et al, 2008

Leading causes of premature mortality, Cape Town 2006

2006

16.1

14.4

8.4

5.3

4.1

3.7

3.4

3.4

2.9

2.5

0 5 10 15 20

HIV/AIDS

Homicide

Tuberculosis

Road traffic

Lower Respiratory infections

Ischaemic heart disease

Stroke

Diabetes Mellitus

Low birth weight and RDS

Diarrhoeal Diseases

Percentage

Percentage of total Years of Life Lost

(YLLs)


Public health importance of mortality data

• Monitor the health of the population– Leading causes of death– Mortality rates (geographic variations)– Trends over time

• Inform decisions about health policy and strategy– Prevent premature deaths– Service provision – Health budgets

• Evaluate health service outcomes

Attending doctor: • Establish diagnosis • Complete death certificate (International form – WHO)

Coding by Statistical Office:• Code causes of death (ICD code for each cause listed)• Classify cause of death (select a single underlying cause of death for

stats according to ICD selection rules)• Check validity, query

Analysis by Statistical Office:• Tabulate and disseminate data

Source: Adapted from Johansson LA, 2008

Main stages in production of cause of death statistics

G MEDICAL CERTIFICATE OF CAUSE OF DEATH Part 1 Enter the diseases, injuries or complications that caused the death. Do not enter the mode of dying, such as cardiac or respiratory arrest, shock, or heart failure. List only one cause on each line. IMMEDIATE CAUSE (a) …………………………...……… (Final disease or condition Due to, (or as a consequence of) resulting in death Sequentially list condition, if any, leading to (b) ………………………….………. immediate cause. Enter Due to, (or as a consequence of) UNDERLYING CAUSE last (c) …………………………………. Due to, (or as a consequence of) (d) ……………………………….. Part 2 Other significant conditions contributing to death but not resulting in underlying cause given in Part 1 .……………………………… __________________________________________________________________________________

Approximate interval between onset and death (Days/Months/Years) ………………. ……………….. ………………. ………………. ……………….

FOR OFFICE USE ONLY ICD-10

Immediate cause on top line

Any causal sequence with underlying cause

at the bottom

Contributing cause but not in causal sequence

in Part 2

According to ICD-10:

• The Immediate Cause is the final disease, injury or complication directly causing the death. It should be noted that the mechanism of death or terminal event (for example, cardiac arrest or respiratory arrest) is not considered to be a cause of death. The mechanism of death should not be reported as the immediate cause of death as it is a statement not specifically related to the disease process, and it merely attests to the fact of death.

• The Underlying Cause of Death is the disease or injury that started the sequence of events leading directly to death or the circumstances of the accident or violence that produced the fatal injury. In the case of a violent death, the form of external violence or accident is antecedent to the injury entered, although the two events may be almost simultaneous.

• Part II is for reporting all other significant diseases, conditions, or injuries that contributed to death but which did not result in the underlying cause of death given in Part I.

A 75-year-old female had a 15-year history of non-insulin-dependent diabetes mellitus, a 13-year history of mild hypertension treated with thiazide diuretics, and an uncomplicated myocardial infarction 6 years prior to the present illness. She was found disoriented at her home and brought to hospital. On admission she was noted to be unresponsive, without focal neurologic signs, and severely dehydrated with a blood pressure of 90/60. Initial laboratory tests disclosed severe hyperglycemia, hyperosmolarity, azotemia, and mild ketosis without acidosis. A diagnosis of hyperosmolar nonketotic coma was made. The patient was treated with fluids, electrolytes, insulin and broad-spread antibiotics. Within 72 hours, the patient’s hypersomolar, hyperglycemic state was resolved. However, she remained anuric with progressive azotemia. The patient died on the 8th hospital day in severe renal failure.




Acute renal failure

Hyperosmolar nonketotic coma

Diabetes mellitus, non-insulin dependent

Hypertension, Previous myocardial infarction

5 Days

8 Days

15 Years

Female aged 77 years, stumbled and fell over while cleaning the house and sustained a fracture of the neck of the left femur. She had an operation for insertion of a pin the following day. Four weeks later her condition deteriorated, she developed hypostatic pneumonia and died two days later.




Terminal Hypostatic Pneumonia

Fractured Left Neck of Femur

Accidental fall while cleaning at home

2 Days

4 weeks

4 weeks




Cancer

Diabetes




Stroke

Pneumonia




Cardiac arrest

Terminology

Terminology that should be avoided

• Ill-defined / non-specific conditions– Old age– Headache– “Natural causes”

• Mechanisms of death– Heart failure– Kidney failure– Dehydration– Hypoxia– Sepsis

Things that should not be on the certificate

• Abbreviations– DM II– MI– MS– HONK

• Stories– The patient presented three days ago with severe

abdominal pain, but the family says it’s been going on for a long time. At surgery extensive peritoneal sepsis of unknown cause was found.

Things that should not be on the death certificate

In the case of Human Immunodeficiency Virus

– Check hospital notes for HIV-tests or treatment

– Underlying COD: Human immunodeficiency virus– Intermediate COD: Acquired Immunodeficiency syndrome– Immediate COD: Tuberculosis / Cryptococcal Meningitis, etc.

Processing information from the medical certificate

• Code each cause according to the International Statistical Classification of Diseases and Related Health Problems (ICD-10)

• Apply ICD-10 rules to establish the Underlying cause of death

• Check the validity of the information – and query or correct

• Check consistency of data eg prostate cancer for female

http://www.who.int/classifications/icd/en/

ICD – 10

A00 CholeraA00.0 Cholera due to Vibrio cholerae 01, biovar choleraeA00.1 Cholera due to Vibrio cholerae 01, biovar eltorA00.9 Cholera, unspecified

Z99 Dependence on enabling machines and devices, not elsewhere classifiedZ99.0 Dependence on aspiratorZ99.1 Dependence on respiratorZ99.2 Dependence on renal dialysisZ99.3 Dependence on wheelchairZ99.8 Dependence on other enabling machines and devicesZ99.9 Dependence on unspecified enabling machine and device

ICD – 10 ChaptersA00 – B99 Certain infectious and parasitic diseasesC00 – D48 NeoplasmsD50 – D89 Diseases of the blood and immune mechanismE00 – E90 Endocrine, nutritional and metabolic disordersF00 – F99 Mental and behavioural disordersG00 – G99 Diseases of the nervous systemH00 – H59 Diseases of the eye and adnexaH60 – H95 Diseases of the ear and mastoid processI00 – I99 Diseases of the circulatory J00 – J99 Diseases of the respiratory diseasesK00 – K93 Diseases of the digestive systemL00 – L99 Diseases of the skin and subcutaneous tissueM00 – M99 Diseases of the musculoskeletal system N00 – N99 Diseases of the genitourinary systemO00 – O99 Pregnancy, children and puerperiumP00 – P99 Perinatal conditionsQ00 – Q99 Congenital malformationsR00 – R99 Symptoms and signs not elsewhere definedS00 – T99 Nature of injuryV01 – Y98 External causes of morbidity and mortalityZ00 – Z99 Reasons for encounter with health service

Problems with cause of death statistics

Errors can occur at all stages of production:– Diagnosis – Death certification

• Errors (mechanism of death/no UC, competing causes, incorrect sequence)

• Insufficient information

– Coding – Classification (incorrect or inconsistent application of ICD selection rules)

– Analysis

TB - male

0

100

200

300

400

500

600

7000

5-9

15-1

9

25-2

9

35-3

9

45-4

9

55-5

9

65-6

9

75-7

9

85+

1996

2000/01 sample

TB - female

0

50

100

150

200

250

300

0

5-9

15-1

9

25-2

9

35-3

9

45-4

9

55-5

9

65-6

9

75-7

9

85+

1996

2000/01 sample

HIV - males

0

200

400

600

800

1000

0

5-9

15-1

9

25-2

9

35-3

9

45-4

9

55-5

9

65-6

9

75-7

9

85+

1996

2000/01 sample

HIV - females

0

200

400

600

800

1000

0

5-9

15-1

9

25-2

9

35-3

9

45-4

9

55-5

9

65-6

9

75-7

9

85+

1996

2000/01 sample

HIV - males

0

200

400

600

800

1000

0

5-9

15-1

9

25-2

9

35-3

9

45-4

9

55-5

9

65-6

9

75-7

9

85+

1996

2000/01 sample

HIV - females

0

200

400

600

800

1000

0

5-9

15-1

9

25-2

9

35-3

9

45-4

9

55-5

9

65-6

9

75-7

9

85+

1996

2000/01 sample

Source: Groenewald et al, 2005

Death certification quality in Cape Winelands and Overberg:

Percentage ill-defined causes of death by sub-district

0

5

10

15

20

25

30

Per

cen

tag

e

2004 26 10 7 14 12 15 10

2005 27 10 7 12 8 14 7

2006 20 13 9 9 17 13 10 10 10

Breede River Winelands

Breede Valey Drakenstein Stellenbosch Witzenberg Cape Agulhas Overstrand Swellen Theewaters

Review of certificates in Vangaurd: Major errors

• Incorrect sequencing in Part 1

• Competing causes of death in Part 1

• Lack of a proper underlying cause of death

• A mechanism of death, without an underlying cause of death

• One or more major errors in DNF

28.7%

15.3%

14.8%

13.5%

43.4%

Source: Burger et al., 2007

Review of certificates in Vangaurd : Minor errors

• Absence of a time estimate between onset of disease and death

• Use of abbreviations

• Recording of inappropriate information

• Illegible handwriting

• One or more minor errors in DNF

81.5%

23.7%

13.0%

2.5%

86.1%Source: Burger et al., 2007

In total 91,7% of cases had at least one error!!

Characteristic OR 95% CI p-value

Sex 0.8017

Male 1.0 -

Female 1.0 0.8-1.4

Age 0.1260

Neonates 1.0 -

1 mth – 4 yrs 1.4 0.1-21.4

5-14 2.5 0.1-122.3

15-24 7.3 0.3-203.4

25-34 3.2 0.1-78.7

35-44 4.5 0.2-108.7

45-54 5.3 0.2-124.9

55-64 3.8 0.2-90.4

65-74 6.2 0.3-147.3

75+ 7.0 0.3-168.0

Serious major errors and associated factors, Academic Hospital in Cape Town

Characteristic OR 95% CI p-value

Department 0.5375

Casualty 1.0 -

Medical general 1.8 1.2-2.8

Medical specialised 2.0 1.2-3,2

Surgical 1.4 0.9-2.2

Neonates 3.3 0.2-62.8

Cause of death<0.00

1

Neoplasms 1.0 -

Circulatory 3.4 2.1-5.5

Infectious/parasitic 4.3 2.3-8.0

Respiratory 4.7 2.5-9.0

Endocrine/nutrtional/metab

17.2 8.7-34.0

Digestive 6.3 3.1-12.9

Perinatal conditions 3.4 0.8-15.0

Genitourinary 17.

3 7.8-38.2

Other 5.8 2.9-11.5

Source: Nojilana et al., 2008

not recorded on death certificate

recorded on death certificate

Tested for HIV

not tested for HIV

HIV +ve

HIV -ve

recorded on death certificate

not reported on death certificate

71

33 38

6 32

15 17

Potentially HIV/AIDS related causes*

0 6

242HIV test information from medical record for sub-sample

Brief intervention for internsProportion with adequate score

N Pre-test%

Post-test%

McNemar’sp-value

Overall group 24 13 88 <0.0001

Intervention group

13 15 84 0.0027

Control group 11 9 91 0.0027

Source: Pieterse et al., 2008

http://who.int/bookorders

www.healthmetricsnetwork.org

[email protected]

http://who.int/bookorders

http://www.healthmetricsnetwork.org/

Determinants of HealthSocio-economic and demographic factorsEnvironmental and behavioural risk factors

Service seeking behaviours

Health System Inputs Outputs OutcomesPolicy Information Coverage Financing Service Utilisation

Human resources (availability Organisation and quality)

Health status

Mortality

Morbidity/Disability

Well-being

Domains of Measurement

Source: Health Metrics Network

http://www.who.int/classifications/icd/en/

AcknowledgementBOD Unit and collaborators

Dr Pam GroenewaldDr Lene BurgerMs Desiree PieterseMs Beatrice Nojilana

WHO-FIC Collaborating CentreMs Lyn HanmerDr Pam GroenewaldMr Malute TshivaseMs Margie SchneiderProf Jenny JelsmaMr David BourneMe Sedick Isaacs

Local, provincial and national governmentCity of Cape TownWestern Cape Department of HealthDepartment of Home Affairs Statistics South Africa

Burden of Disease Research Unit Cause of death statistics from vital registration Debbie Bradshaw WHO-FIC Collaborating Centre (Under designation)

Documents

immediate cause of death

cause of death profile

mechanism of death

violent death

fact of death

death certification

broad cause

production of cause