Friday, September 08, 2006
Dying, the beloved country
HIV/Aids, malaria, diabetes and the December holidays: they'll all be the death of us, says Stats SA communications manager Trevor Oosterwyk in his analysis of the organisation's Mortality Report, released today
September 7, 2006
By Trevor Oosterwyk
For any country, the health and survival of its citizens is one of its main concerns. Knowing a particular age group's chances of dying for either sex, how the risks have changed over time, and understanding the causes of death are crucial in building an accurate picture of mortality and forming the basis for effective policy planning to improve survival rates.
In South Africa, the objective of reducing mortality is reflected in the Government Programme of Action.
This programme identifies key areas of intervention and focuses on research and initiatives to reduce deaths from malaria, tuberculosis, non-communicable diseases such as diabetes, asthma and hypertension, and unnatural causes of death, as well as the ongoing concern of HIV.
Today Statistics SA released data on deaths registered in the country from 1997 to 2004. The data is organised by age, sex and year of death.
Not all deaths are registered, but there was a higher percentage of deaths registered between 1997 and 2004 than in previous periods. Statistics SA estimates that, in 2000, more than 80% of the deaths of those aged 15 and over were registered.
The release of the statistics on registered deaths between 1997 and 2004 marks a high point in the quality of official data on mortality in Africa.
Statistics SA has also released an extensive report on adult deaths in the country in the same period. This report analyses death rates by age, sex, year of death and cause of death for people aged 15 to 64.
The data on deaths of children from birth to the age of 14 is much more incomplete.
It is estimated that less than 60% of the deaths of children were registered for any year between 1997 and 2004, while the data for the deaths of those above 65 were distorted by age exaggeration.
The report's data on deaths by age, sex and year are adjusted to take into account the incompleteness of death registration.
The methodology used to estimate the percentage of deaths registered is the same as that used by Statistics SA to estimate the mid-year population and to estimate the prevalence of HIV. The same approach is used by UNAids and the Department of Health.
The application of this methodology is part of Statistics SA's efforts to become the country's preferred provider of quality statistics.
A key part of this effort has been to apply consistent and defensible procedures to produce mid-year population estimates, to estimate the percentage of deaths registered, and to estimate the prevalence of HIV by age and sex.
Starting with the mid-year population estimates for 2004, this new procedure was applied to estimate the population by age, sex and population group, and to estimate the population by age and sex for every province. This approach uses the United Nations Spectrum programme with the UN East Asian mortality pattern.
The values of the input parameters were determined from the start by consultation between Statistics SA, the Department of Health and UNAids.
Values for earlier years were recalculated after the production of the 2004 mid-year population estimates, using the same methodology. For the estimate of the percentage of deaths registered used in the mortality report, a consistent methodology was applied for the years 1997 to 2004.
The Statistics SA mid-year population estimates have been adopted by the departments of health, housing, social development, water affairs and forestry, and the National Treasury. The population estimates are important for assessing the state of service delivery and for projections of future needs and demands.
Beginning with estimates for 2005 ANC (Antenatal Clinic) data, the Department of Health has used the same methodology to produce estimates of HIV prevalence by age and sex.
The department and Statistics SA have consulted extensively with each other, and the input assumptions for a given year were the same whether applied by the department or Statistics SA.
Beginning with estimates for 2006, mid-year population estimates have been made by Statistics SA at the district council level. With the 2007 Community Survey, it is planned that mid-year population estimates will be made at municipality level.
There is an annual workshop to discuss the input assumptions for these estimates. All the estimates are published and available to the public, government and business.
The co-operation between Statistics SA, the Department of Health and UNAids, and the application of a consistent methodology across an increasing number of areas, helps ensure that analyses and conclusions based on the generated data are the result of real patterns in the data, rather than the result of differences in the assumptions used to estimate results in different areas or for different years.
This facilitates the development of policy based on actual levels and trends rather than on ungrounded assumptions.
One of the main findings of the mortality report is that death rates in South Africa rose between 1997 and 2004 for every five-year age group for each sex, except for males aged 15 to 19. Some of these increases are large. The death rates more than tripled for females aged 20 to 39 and more than doubled for males aged 30 to 44.
Increases in death rates were smaller for the young and old. For each sex, for those aged 15 to 19 and 55 to 64, death rates between 1997 and 2004 increased by 20% or less.
Overall death rates from infectious diseases more than tripled for males and increased almost five-fold for females between 1997 and 2004. A large part of this was due to HIV.
HIV death rates have a distinctive age pattern: there is an increase in a given age and then a rapid decline at older ages. This peak occurs between the ages of 30 and 34 for females and between 35 and 39 for males.
According to the report, many HIV deaths are registered with some other cause of death. Based on the age pattern of death rates by sex, it is likely that a high proportion of deaths registered as death by parasitic diseases, parasitic opportunistic infections, certain disorders of the immune mechanism and maternal conditions (females only) are actually caused by HIV.
The report said this non-reporting of HIV deaths probably concealed how many actual deaths resulted from HIV given the age pattern of death.
Some registered causes of death rise to a peak with age and then decline at older ages more slowly than HIV, especially for males. For these causes of death, the report says, some of the deaths are likely to be from HIV. Other causes of death include infectious diseases, tuberculosis, malaria and nutritional deficiencies.
A major challenge for researchers is to find a reliable method of determining what deaths with a registered cause of something other than HIV are actually caused by HIV. Until that has been accomplished, the report says, it is not possible to count the number of HIV deaths by age, sex and year of death.
There was virtually no change in overall cancer death rates for either sex between 1997 and 2004. Cancer death rates declined for those aged between 50 and 64.
Non-communicable diseases caused a large portion of the deaths of older people. Increases in deaths from non-communicable diseases are a concern throughout the world.
Overall, death rates from non-communicable diseases increased by 45% for males and by 60% for females between 1997 and 2004. For each sex, the death rate from cancer, stroke and other circulatory causes combined rose by 12% between 1997 and 2004.
There is concern over malaria deaths, especially as drug-resistant strains of the disease have become common.
Overall malaria death rates rose between 1997 and 1999 and declined between 1999 and 2004. The overall death rate from malaria in 2004 was 45% higher than the 1997 value for males, however, and the overall death rate from malaria in 2004 was 93% higher than the 1997 rate for females.
Diabetes is also a great concern, the report says. The adoption of a Western diet, with large amounts of fat and few fruits and vegetables, led to overweight and obesity.
This increased the risk of Type II diabetes, which can be fatal if not effectively treated. Death rates from diabetes and obesity rose for each sex between 1997 and 2004.
Female death rates were always higher than male death rates, but the gap between the sexes narrowed over time.
Death rates from unnatural causes changed little between 1997 and 2004, the report finds, although these declined at the ages of 15 to 19 and at older ages.
Unnatural causes of deaths are homicide, suicide or accidents. Unnatural cause deaths are sometimes called external cause deaths, violent deaths or injury deaths.
The risk of unnatural deaths had large variations across the seasons, said the report. The risk of an unnatural death is 25% higher on any given day in December than on a day in other months. The holiday period is especially risky for unnatural causes of death, including deaths from firearms, knives, transport, drowning and falls.
The number of homicides has declined since the late 1990s, says the report. Homicide rates, however, remain high for males aged between 35 and 39.
Homicide accounts for most unnatural deaths of males aged between 35 and 39. Male homicide rates are about six times higher than female homicide rates.
The findings in the Statistics SA adult mortality report give a more complete picture of the level and causes of adult deaths than has previously been available.
The report also draws attention to a number of questions needing further research.
It highlights particular threats to the South African population, including the effects of diet (as shown in increasing diabetes death rates) and the risks of unnatural death in the December holiday period.
HIV/Aids, malaria, diabetes and the December holidays: they'll all be the death of us, says Stats SA communications manager Trevor Oosterwyk in his analysis of the organisation's Mortality Report, released today
September 7, 2006
By Trevor Oosterwyk
For any country, the health and survival of its citizens is one of its main concerns. Knowing a particular age group's chances of dying for either sex, how the risks have changed over time, and understanding the causes of death are crucial in building an accurate picture of mortality and forming the basis for effective policy planning to improve survival rates.
In South Africa, the objective of reducing mortality is reflected in the Government Programme of Action.
This programme identifies key areas of intervention and focuses on research and initiatives to reduce deaths from malaria, tuberculosis, non-communicable diseases such as diabetes, asthma and hypertension, and unnatural causes of death, as well as the ongoing concern of HIV.
Today Statistics SA released data on deaths registered in the country from 1997 to 2004. The data is organised by age, sex and year of death.
Not all deaths are registered, but there was a higher percentage of deaths registered between 1997 and 2004 than in previous periods. Statistics SA estimates that, in 2000, more than 80% of the deaths of those aged 15 and over were registered.
The release of the statistics on registered deaths between 1997 and 2004 marks a high point in the quality of official data on mortality in Africa.
Statistics SA has also released an extensive report on adult deaths in the country in the same period. This report analyses death rates by age, sex, year of death and cause of death for people aged 15 to 64.
The data on deaths of children from birth to the age of 14 is much more incomplete.
It is estimated that less than 60% of the deaths of children were registered for any year between 1997 and 2004, while the data for the deaths of those above 65 were distorted by age exaggeration.
The report's data on deaths by age, sex and year are adjusted to take into account the incompleteness of death registration.
The methodology used to estimate the percentage of deaths registered is the same as that used by Statistics SA to estimate the mid-year population and to estimate the prevalence of HIV. The same approach is used by UNAids and the Department of Health.
The application of this methodology is part of Statistics SA's efforts to become the country's preferred provider of quality statistics.
A key part of this effort has been to apply consistent and defensible procedures to produce mid-year population estimates, to estimate the percentage of deaths registered, and to estimate the prevalence of HIV by age and sex.
Starting with the mid-year population estimates for 2004, this new procedure was applied to estimate the population by age, sex and population group, and to estimate the population by age and sex for every province. This approach uses the United Nations Spectrum programme with the UN East Asian mortality pattern.
The values of the input parameters were determined from the start by consultation between Statistics SA, the Department of Health and UNAids.
Values for earlier years were recalculated after the production of the 2004 mid-year population estimates, using the same methodology. For the estimate of the percentage of deaths registered used in the mortality report, a consistent methodology was applied for the years 1997 to 2004.
The Statistics SA mid-year population estimates have been adopted by the departments of health, housing, social development, water affairs and forestry, and the National Treasury. The population estimates are important for assessing the state of service delivery and for projections of future needs and demands.
Beginning with estimates for 2005 ANC (Antenatal Clinic) data, the Department of Health has used the same methodology to produce estimates of HIV prevalence by age and sex.
The department and Statistics SA have consulted extensively with each other, and the input assumptions for a given year were the same whether applied by the department or Statistics SA.
Beginning with estimates for 2006, mid-year population estimates have been made by Statistics SA at the district council level. With the 2007 Community Survey, it is planned that mid-year population estimates will be made at municipality level.
There is an annual workshop to discuss the input assumptions for these estimates. All the estimates are published and available to the public, government and business.
The co-operation between Statistics SA, the Department of Health and UNAids, and the application of a consistent methodology across an increasing number of areas, helps ensure that analyses and conclusions based on the generated data are the result of real patterns in the data, rather than the result of differences in the assumptions used to estimate results in different areas or for different years.
This facilitates the development of policy based on actual levels and trends rather than on ungrounded assumptions.
One of the main findings of the mortality report is that death rates in South Africa rose between 1997 and 2004 for every five-year age group for each sex, except for males aged 15 to 19. Some of these increases are large. The death rates more than tripled for females aged 20 to 39 and more than doubled for males aged 30 to 44.
Increases in death rates were smaller for the young and old. For each sex, for those aged 15 to 19 and 55 to 64, death rates between 1997 and 2004 increased by 20% or less.
Overall death rates from infectious diseases more than tripled for males and increased almost five-fold for females between 1997 and 2004. A large part of this was due to HIV.
HIV death rates have a distinctive age pattern: there is an increase in a given age and then a rapid decline at older ages. This peak occurs between the ages of 30 and 34 for females and between 35 and 39 for males.
According to the report, many HIV deaths are registered with some other cause of death. Based on the age pattern of death rates by sex, it is likely that a high proportion of deaths registered as death by parasitic diseases, parasitic opportunistic infections, certain disorders of the immune mechanism and maternal conditions (females only) are actually caused by HIV.
The report said this non-reporting of HIV deaths probably concealed how many actual deaths resulted from HIV given the age pattern of death.
Some registered causes of death rise to a peak with age and then decline at older ages more slowly than HIV, especially for males. For these causes of death, the report says, some of the deaths are likely to be from HIV. Other causes of death include infectious diseases, tuberculosis, malaria and nutritional deficiencies.
A major challenge for researchers is to find a reliable method of determining what deaths with a registered cause of something other than HIV are actually caused by HIV. Until that has been accomplished, the report says, it is not possible to count the number of HIV deaths by age, sex and year of death.
There was virtually no change in overall cancer death rates for either sex between 1997 and 2004. Cancer death rates declined for those aged between 50 and 64.
Non-communicable diseases caused a large portion of the deaths of older people. Increases in deaths from non-communicable diseases are a concern throughout the world.
Overall, death rates from non-communicable diseases increased by 45% for males and by 60% for females between 1997 and 2004. For each sex, the death rate from cancer, stroke and other circulatory causes combined rose by 12% between 1997 and 2004.
There is concern over malaria deaths, especially as drug-resistant strains of the disease have become common.
Overall malaria death rates rose between 1997 and 1999 and declined between 1999 and 2004. The overall death rate from malaria in 2004 was 45% higher than the 1997 value for males, however, and the overall death rate from malaria in 2004 was 93% higher than the 1997 rate for females.
Diabetes is also a great concern, the report says. The adoption of a Western diet, with large amounts of fat and few fruits and vegetables, led to overweight and obesity.
This increased the risk of Type II diabetes, which can be fatal if not effectively treated. Death rates from diabetes and obesity rose for each sex between 1997 and 2004.
Female death rates were always higher than male death rates, but the gap between the sexes narrowed over time.
Death rates from unnatural causes changed little between 1997 and 2004, the report finds, although these declined at the ages of 15 to 19 and at older ages.
Unnatural causes of deaths are homicide, suicide or accidents. Unnatural cause deaths are sometimes called external cause deaths, violent deaths or injury deaths.
The risk of unnatural deaths had large variations across the seasons, said the report. The risk of an unnatural death is 25% higher on any given day in December than on a day in other months. The holiday period is especially risky for unnatural causes of death, including deaths from firearms, knives, transport, drowning and falls.
The number of homicides has declined since the late 1990s, says the report. Homicide rates, however, remain high for males aged between 35 and 39.
Homicide accounts for most unnatural deaths of males aged between 35 and 39. Male homicide rates are about six times higher than female homicide rates.
The findings in the Statistics SA adult mortality report give a more complete picture of the level and causes of adult deaths than has previously been available.
The report also draws attention to a number of questions needing further research.
It highlights particular threats to the South African population, including the effects of diet (as shown in increasing diabetes death rates) and the risks of unnatural death in the December holiday period.
