Looking beneath the surface of leading causes of death

Leading causes of death provide a simplified way of organising and reporting mortality data to enable broad changes in patterns of mortality to be monitored over time. Leading causes are compiled according to the underlying cause of death – the disease or condition which started the chain of morbid events leading to death – and are represented as counts rather than population based mortality rates.

While tabulations of leading causes are therefore easily understood, large amounts of detail available in the mortality dataset are not apparent when looking at leading cause groupings. This article provides examples of three key areas which are not represented in leading cause data:

• Details of individual conditions that comprise leading cause groupings.
• Changes in the age structure or distribution of deaths within leading cause groupings.
• Information on the full range of diseases or conditions which are recorded on death certificates (multiple causes of death).

Diseases that comprise underlying cause categories

Most leading cause categories comprise multiple diseases or conditions. Two simple examples are Australia’s first and second leading causes of death, Ischaemic heart disease and Dementia, including Alzheimer’s disease.

Ischaemic heart disease comprises the International Classification of Diseases (ICD) categories I20–I25. There are two very common causes of death within these six diseases, Acute myocardial infarction (I21) (commonly known as heart attacks), and Chronic ischaemic heart disease (I25) (e.g. coronary atherosclerosis) which is characterised by a reduced oxygen supply to the heart.

While the death rate from Ischaemic heart disease has been decreasing steadily for many years, the death rates for acute and chronic ischaemic heart diseases have declined at different rates. Over time, slower reductions in the death rate for chronic ischaemic heart disease have seen it become the predominant cause within the Ischaemic heart disease grouping as shown in the graph below.



The reduction in mortality rates for Ischaemic heart diseases is a critically important change when considering patterns of mortality in Australia. However, the more rapid reduction in mortality rates for acute ischaemic heart diseases compared to chronic heart diseases is also important. It provides further insight into how advancements in medicine are reducing ischaemic heart disease mortality, and also highlights the importance of chronic ischaemic heart disease as a public health issue into the future.

Dementia, including Alzheimer’s disease (F01, F03, G30) is Australia’s second leading cause of death and it is expected to become the leading cause within the next few years. The diseases which comprise this leading cause group are Alzheimer’s disease (G30), Vascular dementia (F01) and Unspecified dementia (F03).

While it is estimated that up to 70% of dementia cases in Australia are caused by Alzheimer’s disease (Dementia Australia), the proportion of dementia deaths certified as Alzheimer’s disease is far lower, accounting for less than one third of deaths in this leading cause category.

While both Alzheimer’s disease and Vascular dementia provide an insight into the cause of dementia symptoms, deaths certified only as dementia and classified as Unspecified dementia record only the symptom and not the cause. Given that the principle of underlying cause is to identify the disease or condition that led to the symptom, these deaths would be considered to be poorly defined.

In total, there were 26,030 deaths where dementia was mentioned on the death certificate. The table below shows the total count of mentions of each type of dementia to appear on a death certificate in 2017. While 13,729 of deaths due to dementia are certified as the underlying cause of death, an additional 12,301 deaths had dementia recorded as an associated cause of death. Common underlying causes for these 12,301 deaths include Ischaemic heart disease, stroke and cancer.



Those deaths with an underlying cause of Unspecified dementia (8,076 deaths) have an unknown cause of the dementia symptoms. Had the cause of the dementia been certified, this would have been chosen as the underlying cause. For most of these deaths it is likely the underlying cause would have been Alzheimer’s disease, Vascular dementia, Ischaemic heart disease or stroke, but the proportion of each is unknown.

Dementia is clearly an important health issue as life expectancy and the number of very elderly people increases. Therefore there would be great benefit in obtaining better insights into the causes of more dementia deaths in Australia to enable better targeting of future medical research and health interventions. If the cause of dementia symptoms were known in all cases this would change the composition of this leading cause category and likely the relevant ranking within leading cause tabulations.

Changes in the age structure of deaths

Changes in leading causes of death over a period of time highlight one aspect of changing patterns of mortality. Looking at changes in the ages at which deaths occur provides a further insight into the ability to prevent and treat diseases, regardless of changes in the number or rate at which deaths occur.

When considering deaths from breast cancer, the standardised death rate has declined over the past 20 years (1998-2017) from 14.5 deaths per 100,000 people to 10.1. In addition to this decrease in the death rate, there has also been an approximate increase of 4.5 years in the median age at death.

It is important to consider changes in the age structure of the population when interpreting changes in the age distribution of deaths, hence the importance of age-standardised death rates. However, changes can highlight an enhanced ability to diagnose and treat disease among younger people. Enhanced understanding of the disease and more advanced treatment options are all factors which are contributing to decreased breast cancer death rates among younger women.



Multiple causes of death

Most deaths in Australia are certified by medical practitioners using the Medical Certificate of Cause of Death (MCCD). Certifying doctors are asked to record those diseases or conditions which contributed to the death in a sequence ending with the underlying cause. On average, there were 3.4 causes of death mentioned on medical certificates in 2017.

When tabulating leading causes of death, it is important that each death is only counted once. The selection of an underlying cause of death allows each death to be assigned to a single cause, but also means that other conditions listed on the death certificate are not taken into account in underlying cause (and leading cause) tabulations.

Multiple cause tabulations provide an opportunity to better understand the full contribution of a disease or condition to mortality in Australia. For instance, there were 13,729 deaths where the underlying cause of death was recorded as Dementia, including Alzheimer’s disease, but in total it was mentioned on 26,030 death certificates. Similarly there were 4,839 deaths with an underlying cause of Diabetes, but diabetes was mentioned on 17,020 death certificates.

Multiple causes of death can provide an important insight into the overall prevalence and impact of a disease from a mortality perspective. Multiple causes are also becoming more important to consider as people live to an older age and are likely to suffer multiple conditions (often chronic) at the time of death. An example of this would be the large number of Ischaemic heart disease deaths where hypertension and diabetes are also mentioned on the death certificate.

While fully understanding patterns within the multiple cause data can be complex, the ABS is working with partners to seek better ways to utilise this important data resource in future mortality tabulations.

Conclusion

Presentation of mortality by leading causes provides a simplified way of understanding key trends and changes in patterns of mortality over time. Far more detail than that presented in leading cause tabulations is available within the full dataset. Much of that detail is used in key national health reports, including tabulations by individual causes, groups of causes that relate to particular health issues (i.e. drug related deaths or preventable deaths), age-specific and age-standardised death rates and more detailed geographic breakdowns.

Where mortality data lacks specificity (i.e. Unspecified dementia) it is likely that dedicated studies would be required to fill in missing details. The use of linked health datasets and engagement with certifying medical practitioners could help in bridging that particular data gap.

The use of multiple causes of death data is currently mostly targeted at certain diseases and conditions. Building frameworks for better use of multiple cause data has been the subject of international efforts over recent years and progress is being made. There will always be a need to tabulate mortality data according to single underlying cause of death, but as deaths from multiple co-morbidities increase in number, the need to fully describe the circumstances relating to a death will also increase.


Reference

1. Dementia Australia, Alzheimer's disease. Viewed 24 September 2018.
https://www.dementia.org.au/about-dementia/types-of-dementia/alzheimers-disease