Contact
Join
Login

The value of cause of death data

Medical certification of cause of death

Coding causes of death to statistical categories
The International Classification of Diseases

Cause of death: where there is no physician
Verbal autopsy diagnostic algorithms

Automated verbal autopsy
What is automated verbal autopsy and how does it differ from medical certification of cause of death?

Incorporating verbal autopsy into the civil registration and vital statistics system

Learn how to use the Learning Centre

Move your way through the CRVS system or simply click on a topic to dive into a specific subject.

An account lets you:

  • Save resources from our Library
  • Track your progress through the Learning Centre
  • Sign-up for our free newsletter

Medical certification of cause of death

Avoiding ill-defined and unusable underlying causes

A common error is to report the ‘mode of dying’ as the COD. For example, ‘heart failure’, ‘cardiac arrest’, ‘shock’, ‘brain failure’, ‘hepatic failure’ and ‘renal failure’ should not be reported as the COD, because organ failures do not usually occur without a precipitating cause. Therefore, if organ failure is reported on the death certificate, the underlying condition responsible for the failure should also be reported. For example, ‘acute renal failure due to diabetes mellitus’ or ‘liver failure due to hepatitis B infection’. 

Where only the mode of dying is reported on the death certificates (for example, ‘heart failure’) then a potentially wide range of underlying causes will be missed. The health measures needed to prevent these underlying causes are often very different, so key information needed to inform national and local planning and health policy development will be missed. The true underlying cause of a death certified as being due to ‘heart failure’ could be as diverse as a major noncommunicable disease, a stab injury or post-partum haemorrhage. Perpetuating such confusion through COD statistics will greatly reduce their health value.

There will always be cases for which the physician does not have enough evidence to certify the COD with certainty. However, it is extremely important to try to avoid using vague and ill-defined terms for the UCOD in the MCCOD. Ill-defined (often called unusable) causes of death do not provide the kind of information needed guide health policy and planning. Studies have shown that it is possible to reduce the percentage of these unusable codes by gathering all available information related to a death and following death certification guidelines.

In general, the proportion of deaths coded to ill-defined categories should not exceed 10% of all deaths in the age group 65 years and over, and should account for <5% of deaths occurring in those under 65 years. Ill-defined and unusable conditions reported from most countries show very consistent patterns, and the following conditions are the most commonly reported:

  1. Septicaemia/sepsis
  2. Heart failure
  3. Pneumonia, unspecified
  4. Senility
  5. Depression
  6. Essential primary hypertension
  7. Cardiac arrest.

The following sections describe the main conditions that are commonly erroneously described as the UCOD, and explain the reasons why these conditions are considered ill-defined and unusable for health purposes.


Sepsis

When reporting deaths due to infections, it is important to specify the cause of the infection, site of the infection and responsible organism (where available). If the UCOD was reported just as sepsis or septicaemia, the important information regarding the UCOD, and subsequent value for public policy, is lost as sepsis can be due to many conditions.

download
Heart failure

The diagram below shows that there are many causes that could contribute to heart failure. It is important to specify the underlying cause so the data have utility for public health purposes. Deaths from stabbings have much different public health implications than deaths from myocardial infarctions.

heart_failure_figure
Pneumonia

Pneumonia appears in leading causes of death lists published by many countries. Pneumonia can be a UCOD in some instances (for example, community acquired pneumonia); however, most of the time pneumonia is the immediate COD resulting from a variety of other underlying causes of death.
The true underlying cause of a death attributed to ‘pneumonia’, which public health policy is attempting to prevent, could be as diverse as a terminal cancer, an injury or an infection such as HIV. The public health prevention strategies for each would be vastly different, and hence it is critically important to avoid using such vague and ill-defined diagnoses when the true UCOD is something else.

pneumonia_figure
Senility

In many countries, ‘old age’ or ‘senility’ is commonly reported as the UCOD.  Senility is not a valid cause of death . Age of the patient is reported in the demographic section of the death certificate. We understand that in patients with comorbidities it may sometimes be hard to select one UCOD. But it is important to look for evidence for the correct UCOD that resulted in the death of the elderly person. Any elderly person can die from any acute or chronic condition.

Depression

Patients suffering from severe depression sometimes commit suicide. In such instances the injury and circumstance lead to death (suicide by hanging) should be reported as the UCOD. Depression should be reported as a contributory condition in Part 2.

Hypertension

In most instances, primary hypertension is a contributory cause for conditions like stroke or acute myocardial infarction. Therefore, primary hypertension should be reported in Part 2 of the death certificate as the contributory cause, and not as the underlying cause. Secondary hypertension can be reported as an intermediate COD; however, the condition  leading to  the secondary hypertension should be reported as the UCOD.



Read more 

Naghavi M et al (2010).  Algorithms for enhancing health utility of national causes-of-death data. Population Health Metrics.

Rampatige R et al (2014).  Systematic review of statistics on causes of deaths in hospitals: Strengthening the evidence for policy-makers. Bulletin of the World Health Organization.


Guidance for assessing and interpreting the quality of mortality data using ANACONDA
Guidance for assessing and interpreting the quality of mortality data using ANACONDA

For users of ANACONDA (statisticians and/or analysts in health and statistics departments, researchers, or other experts working with mortality data).

Download resource (1.94MB)
More Info

Authors: Mikkelsen L, Lopez AD

Publication date: October 2017

Resource type: CRVS resources and tools

Related resources: Course prospectus: ANACONDA

img-video-icon
Brazil's health data: How have they improved so much and so quickly?

Interview with Dr Mohsen Naghavi, Professor of Global Health at the Institute of Health Metrics and Evaluation at the University of Washington. Filmed in October 2017 during the Meeting on Improving the Quality of Information on Causes of Death in Brazil in Recife, Brazil.

Open video

© University of Melbourne 2018   For more information on copyright visit our website terms