Burden of Disease - DALY (YLD & YLL) worksheets

Page content: About DALYs | Estimation of YLD | Disability weights | Excel and DisMod files | Feedback | References

About disability-adjusted life years (DALYs)

The Victorian BOD studies (Mortality, and Morbidity) use the DALY as a population measure of incident lost years of healthy life due to a wide range of diseases, injuries and selected risk factors.

The DALY extends the concept of potential years of life lost due to premature death (PYLL) to include equivalent years of "healthy" life lost by virtue of being in states of ill-health. DALYs for a disease or health condition are calculated as the sum of the years of life lost due to premature mortality (YLL) in the population and the "years lived with disability" (YLD) for incident cases of the health condition:

DALY = YLL + YLD

The loss of healthy life due to non-fatal health conditions (YLD) requires estimation of the incidence of the health condition (disease or injury) in the specified time period. For each new case, the number of years of healthy life lost is obtained by multiplying the average duration of the condition (to remission or death) by a severity weight that measures the loss of healthy life using an average QALY weight for the disease state. The DALY is described in detail in Murray and Lopez (1996).

The Australian YLD calculations depart from the GBD methodology in the following three areas:

• The GBD discounted YLDs at 3% per annum and applied age weights that gave higher weight to a year of life in young and mid adult years, and lower weight to a year of life at very young and older years. The Australian projects do not use age weights but are using a 3% discount rate in line with the GBD.
  
  
• The Australian studies use a set of Dutch weights for conditions common in developed countries, supplemented by weights used in the GBD study for other conditions. In general, the Dutch and GBD weights are reasonably consistent, but in the longer term, it would be desirable to carry out weighting exercises in Australia to examine how appropriate the weights are in the Australian context.
  
  
• The GBD did not attempt to deal with the effects of comorbidities on YLD estimates for individual diseases. The Australian project adjusts YLD estimates for comorbidities between mental disorders, between congenital malformations and between physical disorders at older ages.

General method for estimation of YLD

Years lived with disability are essentially calculated as follows (ignoring the complications of discounting):

YLD = No. incident cases x Average duration (years) x Disability weight

In order to make a consistent and meaningful estimate of YLD for a condition, it is crucial to clearly define the condition under consideration in terms of case or episode, and severity level or disease stage. It is then necessary to ensure that the disability weight and the population incidence/prevalence data relate to the same case definition. The most difficult step in estimating YLD for most diseases is matching existing population data to the disease stage/severity categories for which the weights are available. Getting this wrong can result in substantial error in the YLD estimate. Disability weights are discussed further below.

For some conditions, numbers of incident cases are available directly from disease registers or epidemiological studies, but for most conditions, only prevalence data are available. In these cases, a software program called Dismod is used to model incidence and duration from estimates of prevalence, remission, case fatality and background mortality. The underlying model is as follows:

Figure 1. DisMod modelling of incidence, prevalence and duration of disease

Where remission rates and/or case fatality rates are not known, they are usually estimated from available evidence. While this affects the age distribution of incident cases and YLD, total YLD are quite insensitive to these assumptions, since the YLD are effectively determined by incidence x duration = prevalence. In other words, it does not make a whole lot of difference what combination of incidence and remission rates (and thus derived durations) you choose. If you aim at the same prevalence figures, the combination of even quite different incidence and duration figures will still lead to more or less the same YLD calculations.

Disability weights

The DALY uses explicit preference weights for health states. These preference weights are derived using a Person Trade Off method with small groups of health experts who are asked to determine weights for a set of indicator health conditions using a deliberative process (Murray and Lopez 1996). The Netherlands has carried out a project to measure weights for 53 diseases of public health importance, involving the estimation of weights for 175 disease stages and/or severity levels (Stouthard et al 1997).

The Dutch weights have the advantage of greater detail in terms of severity levels and disease stages, allowing Australian information on severity distributions and disease progression to be used. This goes some way towards allowing the use of population data on disability and health state distributions in the estimation of YLD. The Dutch study also described each disease stage in terms of a standardised health state description using a variant of the EuroQol classification of health states, the EQ-5D+, which included a sixth dimension for cognitive functioning.

There are some disease categories included in the Australian study for which there are no weights available in either the Dutch or GBD studies. To assist in estimating provisional weights for these, we have fitted a multiplicative regression model for the single attribute states defined by the six dimensions of the EQ-5D+. The regression model is fitted to the Dutch weights using the specified EQ-5D+ descriptions to define the independent variables and the estimated average weight as the dependent variable. This enables us to estimate provisional weights for conditions for which we can specify a description of the resulting health outcomes using the EQ-5D+.

Table 1. Some examples of disability weights from the Dutch study (Stouthard et al 1997)

top of page

Excel and DisMod files

All calculations of DALYs (YLD and YLL) presented in the Victorian Burden of Disease Study: Morbidity, can be downloaded as self-extracting Zipped Excel and DisMod files.

DisMod is a shareware software program developed at Harvard University that allows one to find a set of incidence rates by age that match observed prevalences, given estimates of remission rates and cause-specific mortality risk derived from population data or epidemiological studies.

Download the DisMod software from the Harvard University DisMod website

The self-extracting Zip (.exe) files, below, each contain a number of Excel and DisMod files. After downloading, double-click the .exe file to extract the compressed files.

top of page

Feedback

These worksheets represent the first attempt in Australia to carry out a systematic and comprehensive analysis of the incidence, prevalence, remission, case fatality and severity of diseases ensuring internal consistency and using a common currency for years of healthy life lost due to premature mortality and non-fatal health states. YLD estimates have been made for 176 disease and injury categories involving over 1000 stages, severity levels and/or sequelae.

While every attempt has been made to identify the best available information in relation to each disease and injury category, and to consult as widely as possible, it must be emphasised that these YLD estimates should be seen as developmental. In some cases, it was not possible to go beyond simple models and assumptions about some key parameters, in the time frame available.

The worksheets are being made available to any interested persons in the hope that we will receive constructive criticism and feedback on data, assumptions, models and methods. We also hope that these worksheets will enable others to further develop and refine specific YLD estimates for subsequent analyses of burden of disease in Australia, or for specific disease or injury analyses.

Feedback is an essential component of the development process for burden of disease analyses so that future estimates are as useful as possible to epidemiologists, researchers, health planners and economic evaluation and health priority setting processes.

If you have questions or comments relating to the worksheets, please contact:

AIHW Burden of Disease project: bod@aihw.gov.au

or Colin Mathers, Australian Institute of Health and Welfare:
Email: colin.mathers@aihw.gov.au

or Theo Vos, Victorian Department of Human Services:
Email: theo.vos@dhs.vic.gov.au

References

Murray CJL, Lopez AD, eds. (1996) The Global Burden of Disease: a comprehensive assessment of mortality and disability from diseases, injuries and risk factors in 1990 and projected to 2020. Harvard: Harvard School of Public Health.

Stouthard MEA, Essink-Bot ML, Bonsel GJ, Barendregt JJ, Kramer PG, van de Water HPA, Gunning-Schepers LJ, van der Maas PJ (1997). Disability Weights for Diseases in the Netherlands. Rotterdam: Erasmus University.

top of page