Testing the robustness of the adapted SDG indicator 3.b.3 methodology to determine access to essential medicines for children

Abstract

Background Within Sustainable Development Goal 3, a target is to provide access to affordable essential medicines and vaccines for the entire population on a sustainable basis. Therefore, the SDG indicator 3.b.3 is created to measure the access to essential medicines for the general population. However, there is currently no standardized way to determine the access to essential medicines for children. By adapting the original SDG indicator 3.b.3 to a child specific indicator, a method for assessing access to essential medicine for all children was developed. The purpose of this study is to perform multiple sensitivity analyses on different input parameters that influence the child-specific SDG indicator 3.b.3 methodology to test the robustness of the adapted method.

Method The historical World Health Organization/Health Action International data from 10 different countries were combined into one dataset with 25 hypothetical facilities which together formed a hypothetical country. The data on 19 of the 22 medicines in the young children medicine basket was matched. In addition to this base case set, another dataset has been created in which more medicine prices were included. With these datasets multiple sensitivity analyses were performed on the input parameters: the national poverty line (NPL), number of units needed for treatment (NUNT) and the burden of disease. The average facility score was calculated to compare outcomes of different sensitivity analysis.

Results The average facility scores of the multiple sensitivity analyses were compared to the average facility scores of 35,25% (base case dataset) and 68,18% (price dataset). Changing the NPL did not result in a relevant shift (range: 33,62 – 36,62%) of the average facility score. The results showed that using the NUNT instead of units per treatment had little impact on the average facility score (range: 33,33 – 41,21%) and it did not matter whether the minimum, maximum or average NUNT was used. Also, both the base case dataset and the price dataset showed little influence on the average facility score when changing the burden of the disease.

Conclusion The average facility scores in the sensitivity analyses on the different parameters (NPL, NUNT and burden of disease) do not show a relevant change, so it can be concluded that robustness of the adapted SDG indicator 3.b.3 methodology is proved.