The GiveWell Blog

Trachoma: An example of the need for long-term monitoring

When is a measured program-impact not a real impact? When it doesn’t last.

A study published today in PloS NTDs evaluated the impact of four doses of azithromycin (one every 6 months), and monitored trachoma prevalence throughout the drug administration period and for 2 years after the last dose.

In the first 24 months (from the start of the program through 6 months after the last treatment) prevalence decreased from 63.5% to 2.6%. In the 18 months after the last treatment, prevalence increased to 25.2%.

There was wide village-to-village variation in the prevalence before, during, and after treatment. While all villages ended up with lower infection rates of trachoma than they started with, in some (in particular village numbers 10 and 12 in this table), the infection rate 18 months post-treatment was close to pre-treatment levels.

Another reason to think that more monitoring is probably worth the cost.