.. _dw_section:

Calculating disability weights
************************************
Disability weights (DWs) are associated with health states and not diseases. 
For every disease, an average DW, taking into account the different health states and 
possible disease sequelae as well as a severity distribution, have been computed. 
This methodology is primarily based on the approach proposed by IHME in :cite:p:`GBD2016DiseaseandInjuryIncidenceandPrevalenceCollaborators2017`.

A full list of DWs can be found here :numref:`dw_table`. Specific information for each disease 
can be found using the links below. 

  * :ref:`section_dw_diab`
  * :ref:`section_dw_stroke`
  * :ref:`section_dw_ihd`
  * :ref:`section_dw_copd`
  * :ref:`section_dw_arthritis`
  * :ref:`section_dw_backpain`
  * :ref:`section_dw_cirrhosis`
  * :ref:`section_dw_ckd`
  * :ref:`section_dw_depression`
  * :ref:`section_dw_dementia`
  * :ref:`section_dw_cancers`
  * :ref:`section_dw_injuries`
  
.. _section_dw_diab:

Diabetes
========
In the IHME 2015 dataset we have access to the prevalence of different sequelae ("Diabetic foot", "Amputation due to diabetes mellitus", 
"Diabetic neuropathy", "Vision impairment due to diabetes mellitus") available by country, gender 
and age-group, which have been used to compute the average DW associated 
with diabetes. 

It was assumed that people could only have one sequela. Therefore, where sequela data was available, it was assigned with 
the remaining lumped into 'Uncomplicated diabetes mellitus'.  

.. _diab_dw:
.. csv-table:: Disability weight for diabetes 
   :file: tables/diabetes_DW.csv
   :widths: 25, 25, 10

.. _section_dw_stroke:

Stroke
=======

The mid-level of severity: “Stroke, long‐term consequences, moderate plus cognition problems” is used 
as the average DW associated with long-term disability due to long-term consequences 
following an acute episode of stroke. 

.. _cerebro_dw:
.. csv-table:: Disability weight for stroke by severity level
   :file: tables/cerebro_dw.csv
   :widths: 25, 25, 10

.. _section_dw_ihd:

Ischaemic heart disease
=======================

During the acute phase of myocardial infarction, a weighted average between the two DWs associated with 
"Acute myocardial infarction, days 1-2” and “Acute myocardial infarction, days 3-28” is used. 

The disability weight associated with “Angina pectoris: moderate “is used as the average DW associated 
with long-term disability due to long-term consequences associated with acute myocardial infarction. 

.. _ihd_dw_csv:
.. csv-table:: Disability weight for IHD by severity level 
   :file: tables/ihd_dw.csv
   :widths: 25, 25, 10

.. _section_dw_copd:

Chronic obstructive pulmonary disease
=====================================

The 2015 prevalence rate of COPD by stage (III-V) are available by country, gender and age-group in the 
IHME dataset. These data have been used to compute the average disability weight associated with COPD. 

.. _copd_dw: 
.. csv-table:: Disability weight for COPD by severity level 
   :file: tables/copd_dw.csv
   :widths: 25, 25, 10

.. _section_dw_arthritis:

Rheumatoid arthritis
====================

Using the severity split and the associated DWs, the weighted average is 0.24717 and was used for rheumatoid arthritis disability weight.

.. _arthr_dw:
.. csv-table:: Disability weight for rheumatoid arthritis by severity level
   :file: tables/arthr_dw.csv
   :widths: 15, 30, 10, 30

.. _section_dw_backpain:

Low back pain
=============
The proportion of low back pain and leg pain was assumed to be 30%, which is a conservative estimated 
based on IHME (GBD) estimates :cite:p:`GBD2016DiseaseandInjuryIncidenceandPrevalenceCollaborators2017`. The severity distribution and the associated DWs were 
combined to compute an average DW for low back pain of 0.125857.

.. _back_dw:
.. csv-table:: Disability weight for back pain by severity level
   :file: tables/1_backpain_dw.csv
   :widths: 30, 40, 10

.. _back_dw2: 
.. csv-table:: Back pain severity prevalences
   :file: tables/2_backpain_dw.csv
   :widths: 20, 40, 40

.. _section_dw_cirrhosis:

Cirrhosis
=========
The DW for Cirrhosis is shown in the table below. 

.. _cirrdw: 
.. csv-table:: Disability weight for cirrhosis by severity and health status
   :file: tables/cirr_dw.csv
   :widths: 25, 25, 10

.. _section_dw_ckd:

Chronic kidney disease
======================
As stage IV and V of chronic kidney disease are modelled, the two following disability weights were directly used in the simulation

.. _ckd_dw: 
.. csv-table:: Disability weight for chronic kidney disease by severity level
   :file: tables/ckd_dw.csv
   :widths: 25, 25, 10


.. _section_dw_depression:

Depression
==========

The severity distribution and the associated DWs were combined to compute an 
average DW for major depressive disorder of 0.219.

.. _depression_dw:
.. csv-table:: Disability weight for major depressive disorder by severity 
   :file: tables/depression_dw.csv
   :widths: 15, 35, 10, 20
   :header-rows: 1


.. _section_dw_dementia:

Dementia
========

The severity distribution and associated DWs for dementia were combined to compute an average 
DW for the following three age groups: 

 * <70 years: 0.218
 * 70-79 years: 0.240 
 * 80+ years: 0.267 


.. _dem_dw: 

.. table:: Disability weight for dementia by severity level and age group 
   :widths: 15, 15, 5, 30, 30, 30

   +----------------+-----------------------+--------+--------------+--------------+-------------+
   | **Severity**   | **Health State**      | **DW** | **Severity distribution by age**          |
   +                +                       +        +--------------+--------------+-------------+
   |                |                       |        |**<70**       | **70-80**    |**80+**      |
   +----------------+-----------------------+--------+--------------+--------------+-------------+
   | Mild           | Dementia: mild        | 0.165  | 79% (71-86%) | 71% (63-78%) | 61% (53-68%)|
   +----------------+-----------------------+--------+--------------+--------------+-------------+
   | Moderate       | Dementia: moderate    | 0.388  | 17% (11-23%) | 19% (14-24%) | 26% (22-30%)|
   +----------------+-----------------------+--------+--------------+--------------+-------------+
   | Severe         | Dementia: severe	    | 0.545  | 4% (2-7%)    | 9% (5-13%)   | 12% (7-17%) |
   +----------------+-----------------------+--------+--------------+--------------+-------------+


.. _section_dw_cancers:

Cancers
=======

DWs for cancers are linked to the various stages of cancer, of which there are three.  
As the stages in cancer are not modelled, the DW associated with “diagnosis and primary 
therapy” is used for all cancers. 

.. _cancer_dw: 
.. csv-table:: Disability weight for cancer by severity level
   :file: tables/cancer_dw.csv
   :widths: 25, 25, 10

.. _section_dw_injuries:

Injuries
========
For injuries, average disability weights were computed using the framework developed by IHME and explained 
in the supplementary appendix of :cite:p:`Vos2012` using two specific matrixes. 

The first matrix (see :numref:`dw_injuries_GBD`) maps ICD10 codes to “23 nature of injuries”. Each nature of injury refers 
to specific health states (long- and short-term disability weights). The second matrix (see :numref:`inj_dw`) 
describes the proportion of the different natures of injuries in each cause of injury 
(Falls, Road accident, etc.). 

These two matrices were combined to compute an average short/long term disability weight for every injury 
modelled. When an individual is injured, the short-term disability weight applies until recovery or death. 
The individual is then attached to an injury sequelae and the long-term disability weight is applied. 

.. _dw_injuries_GBD:
.. figure:: Injuries.png
   :alt: injuries_GBD
   :width: 800px
   :align: center
   :figclass: align-center

   Relationship between GBD Cause Names and Injury N-codes

.. _inj_dw:
.. csv-table:: Disability weight for injuries by severity level
   :file: tables/injuries_dw.csv