If you’re trying to compare very different health problems, say for example, malaria vs. maternal health vs. road injuries, you need a common unit.
That’s what a DALY is.
In simple terms, a DALY represents one year of healthy life lost. It combines two things:
Years of life lost (YLL) due to early death, and Years lived with illness or disability (YLD)
So, DALY = YLL + YLD
As a welfare metric, the Cost/DALY averted is a good #outcome proxy, much more than output data on reach and coverage, i.e., number of patients, number of clinics, and all.
This is because with DALYs, you are not just counting the number of patients treated; you are measuring actual improvement in health and life expectancy.
For example, a school feeding program in Kenya reaches about 600,000 students a day. A very impressive output, but is it a good proxy for impact? Not quite.
A good outcome proxy would be DALYs from anemia averted as a result of the feeding program. Anemia is notoriously common in U5s and 5-9s in Africa. So to construct our DALYs, we collect data:
1. Prevalence rates of anemia in primary schoolers in Kenya
2. Effectiveness rates of SFPs on iron stores. For this, RCTs/meta-analyses are the standard.
3. GBD weighting of mild/moderate iron deficiency anemia
4. Mortality rates of anemia in this cohort
5. Disease duration
6. Life expectancy in Kenya
With these, we calculate total DALYs averted and then discount based on certain assumptions: attribution, dead weight, external validity, and so on.
The #WorldBank, WHO, and global health grant makers use this metric to compare multiple health interventions across countries. Interventions that avert the most DALYs per dollar are naturally ranked higher as best buys.
However, like any modelling, the Cost/DALY averted has its blind spots, more on that later.
