RESUMO
BACKGROUND: Sustainable Development Goal (SDG) indicator 3.b.3 monitors progress in medicines' accessibility for adults and has significant limitations when applying to medicines for children. An adapted indicator methodology was developed to fill this gap, but no proof of its robustness exists. We provide this evidence through sensitivity analyses. METHODS: Data on availability and prices of child medicines from ten historical datasets were combined to create datasets for analysis: Dataset 1 (medicines selected at random) and Dataset 2 (preference given to available medicines, to better capture affordability of medicines). A base case scenario and univariate sensitivity analyses were performed to test critical components of the methodology, including the new variable of number of units needed for treatment (NUNT), disease burden (DB) weighting, and the National Poverty Line (NPL) limits. Additional analyses were run on a continuously smaller basket of medicines to explore the minimum number of medicines required. Mean facility scores for access were calculated and compared. RESULTS: The mean facility score for Dataset 1 and Dataset 2 within the base case scenario was 35.5% (range 8.0-58.8%) and 76.3% (range 57.2-90.6%). Different NUNT scenarios led to limited variations in mean facility scores of + 0.1% and -0.2%, or differences of + 4.4% and -2.1% at the more critical NPL of $5.50 (Dataset 1). For Dataset 2, variations to the NUNT generated differences of + 0.0% and -0.6%, at an NPL of $5.50 the differences were + 5.0 and -2.0%. Different approaches for weighting for DB induced considerable fluctuations of 9.0% and 11.2% respectively. Stable outcomes with less than 5% change in mean facility score were observed for a medicine basket down to 12 medicines. For smaller baskets, scores increased more rapidly with a widening range. CONCLUSION: This study has confirmed that the proposed adaptations to make SDG indicator 3.b.3 appropriate for children are robust, indicating that they could be an important addition to the official Global Indicator Framework. At least 12 child-appropriate medicines should be surveyed to obtain meaningful outcomes. General concerns that remain about the weighting of medicines for DB and the NPL should be considered at the 2025 planned review of this framework.