yess GEAR - Ask an Expert
You are here Ask an Expert
Amanuel Yigezu
Health Economist  |  Ethiopia
Hello there. I am conducting a thesis on CEA. my outcome report is in cost per HIV positive client identified. The WHO threshold recommend GDP per capita compared with cost per DALY averted. I saw one study report their finding in cost per positive client identified and compare it with the country's GDP per capita. my question is is it possible to compare the cost per positive clients with countries GDP per capita instead of using DALY averted? If not how can I use threshold value for reports which are not in DALY aversion? Thanks

Expert Replies:

Mark Jit

Professor  |  United Kingdom  |   Replied: 22 Aug 2018 at 00:58
Dear Amanuel,

WHO does not actually recommend the automatic use of GDP per capita-based thresholds for use in country economic evaluations. Instead, countries are encouraged to develop their own thresholds, and to use them in the context of other factors including budget impact and feasibility.

Further information can be found in Bertram et al. Cost-effectiveness thresholds: pros and cons. Bull World Health Organ. 2016 Dec 1;94(12):925-930.

Regardless of what threshold is ultimately used, a threshold for a cost-effectiveness ratio with DALYs as the denominator should not be used for a cost-effectiveness ratio in natural units (eg. positive client identified), unless you think that 1 positive client identified is exactly equivalent to 1 DALY averted.

You have two options:
(i) Find out from the decision maker or budget holder what their willingness to pay to identify one HIV positive client is.
(ii) Construct a decision analytic model of HIV natural history, diagnosis and treatment to work out how many DALYs are averted per HIV positive client identified.


Amanuel Yigezu

Health Economist  |  Ethiopia  |   Replied: 18 Apr 2019 at 19:11
Dear Mark,
Thank you for the previous comments. I am now conducting economic evaluation and I wanted to know a method to construct HIV natural history and know DALY averted per HIV positive individual.
I also have another question,
Is it possible to construct a decision analysis model for 18 package? (Decision tree model or Markov model?)

Leave your reply:

Please sign in before leaving a reply.

Expert Profiles

Mark Jit

London School of Hygiene & Tropical Medicine, UK