Question #1. Should Harmonia conduct an EPI survey early next year? What are your considerations?

I strongly believe that we need to implement a coverage survey in Harmonia. The reasons are the following:

• DHS coverage results (3 years ago) indicate sizeable differences between administrative coverage and actual coverage (8% of difference for Penta1 and up to 16 % of difference for Penta3). This is supported by the fact that the 15-year old denominator used to estimate administrative coverage is very likely to be inaccurate. Overall, administrative coverage data are not very reliable and this warrants a new survey.
• The coverage data available came from the DHS. While these data may have some validity, we are not sure that the methodology used for DHS is the same as the updated WHO survey methodology. This means that even the coverage results as per the DHS may not be as valid as that of the updated survey methodology.
• PCV was introduced 3 years ago, and since then, no coverage survey has been conducted to assess the uptake of this important but costly vaccine.
• 3 years ago, when the DHS was implemented, the availability of cards was 80%. This means that in 1 child out of 5, the vaccination status could not be ascertained. This problem can be dealt with in a new survey that will also collect written records available at the health facility level.

Question #2. The steering committee includes representatives from the national Ministry of Health, National Statistics Office, WHO, UNICEF, and a donor. Given what you have already learned about Harmonia, what do you think will be their primary question of interest for the survey (estimation, comparison, or classification)? Why?

I think this survey should be an estimation survey, owing to the following reasons:

• It is difficult to aim for comparisons because we are not sure to have detailed results for each district (if for instance sampling unit is not the district) and for each population subgroup. To add to that, the method used previously (DHS) is not comparable to the one that will be used now; consequently, it is not appropriate to compare the results coming from two surveys that used different methodologies.
• For the same reasons (lack of district-specific results), we may not be able to do classification among the different districts and subgroups based on vaccine coverage performance

Question #3. Who will comprise the target population and why?

In the current routine immunization schedule in Harmonia, MCV2 is administered at the age of 18 months. To account for that, the target population of this survey will be the children in the age-group 24-35 months. Trying to accommodate childbearing women (for tetanus coverage) and pre-adolescent girls (for HPV coverage) in this same survey will be cumbersome, cost and time-consuming.