*3.2. Identifying Potential Exemplars in Reducing Zero-Dose Children since 2000*

*Absolute progress.* The DRC, Ethiopia, and India registered the largest absolute reductions in national no-DTP prevalence and subnational gaps from 2000 to 2019 (Figure 2; Table 2). Supplementary Figure S1A–C show both national and subnational no-DTP trends over time for each country.

In 2000, 51.9% of under-one children had no doses of DTP in the DRC nationally, with the country experiencing a 46.8 percentage-point gap between the territories with 5th and 95th percentiles for no-DTP prevalence (i.e., 28.5% to 75.3%). By 2019, national no-DTP prevalence levels fell to 6.9%, a 45.0 percentage-point decline. The DRC's subnational gaps narrowed by 30.8 percentage-points by 2019, decreasing to a total of 16.0 percentage points across the 5th and 95th percentiles of territories (i.e., 1.5% to 17.5%). As highlighted by Figrues 2 and S1A, subnational gaps started narrowing faster from 2015–2019 than in previous time periods.

For Ethiopia nationally, 63.4% of under-one children lacked any doses of DTP in 2000, but no-DTP prevalence fell 51.6 percentage points to 11.7% by 2019. Across its zones, Ethiopia had a 47.9 percentage-point gap between the 5th and 95th percentile levels of no-DTP prevalence in 2000, spanning from 39.6% to 87.5%. This subnational gap decreased by 17.2 percentage points by 2019, to a 30.8 percentage-point difference between the 5th and 95th percentile no-DTP levels across zones (i.e., 1.6% to 32.4%). However, amid such marked gains over the last 19 years, Ethiopia's reductions in subnational gaps have stagnated from 2016–2019 (Figure 2 and Figure S1B).

In India, national no-DTP prevalence was 30.9% in 2000 with a 52.5 percentage-point gap between the 5th and 95th percentile for no-DTP levels across districts (i.e., 7.2% to 59.7%). By 2019, 7.2% of under-one children had no doses of DTP in India nationally, a 23.6 percentage-point decline. Subnational gaps in India decreased 32.7 percentage points between 2000 and 2019, falling to 19.7 percentage-point difference in 2019 (i.e., 1.6% to 21.3%). Although overall subnational gaps have narrowed (Figure 2 and Figure S1C), several districts still exceeded 30% of under-one children with no doses of DTP in 2019.

1DWLRQDOQRí'73SUHYDOHQFH

**Figure 2.** Comparing no-DTP trajectories for potential exemplars in reducing zero-dose children, 2000 to 2019. National no-DTP prevalence is represented on the x-axis and the subnational gap (as measured by the difference between the 5th and 95th percentile no-DTP prevalence across secondlevel administrative units) is represented on the y-axis. Each corner represents an extreme for each of these no-DTP metrics, with the lower right-hand corner—low national no-DTP prevalence and low subnational inequality—being the direction in which every location should strive to reach to equitably reduce no-DTP prevalence. Trends in the two no-DTP metrics for the potential exemplars are highlighted in black, with each circle representing a year from 2000 to 2019 that is color-coded from orange (2000) to blue (2019). The light gray trajectories represent the other 51 countries in this analysis.

*Relative progress*. As measured by the percentage change in no-DTP metrics between 2000 and 2019, Bangladesh and Burundi achieved the largest relative reductions in no-DTP prevalence (Figure 2; Table 2). National and subnational no-DTP trends are illustrated in Supplementary Figure S1D,E.

In 2000, estimated national prevalence of no-DTP was 8.8% in Bangladesh, already below the 10% target set forth by the Global Vaccine Action Plan for 2020 [44]. However, by 2019, the percentage of under-one children with no doses of DTP fell to 0.8% in Bangladesh, a 98.0% decline since 2000. Subnational gaps in Bangladesh fell by 99.9% since 2000, narrowing from a 14.4 percentage-point difference for the 5th and 95th percentiles across districts in 2000 (i.e., 3.3% to 17.7%) to approximately 0.01 percentage-points in 2019. Absolute subnational gaps began narrowing faster after about 2012 (Supplementary Figure S1D).

For Burundi, national no-DTP estimates were 18.4% in 2000, but decreased to 2.3% in 2019—an 87.3% reduction. Across Burundi's communes in 2000, there was a 17.1 percentagepoint difference for the 5th and 95th percentiles in no-DTP prevalence (i.e., 10.2% to 27.3%). By 2019, this subnational gap fell to 2.4 percentage points, representing an 85.9% decline

across communes at the 5th and 95th percentile (i.e., 1.5% to 3.9%). Progress accelerated after 2005 (Figure 2, Supplementary Figure S1E), when the country's 13-year civil war ended [45].

Table 2 details these estimates for 2000, 2019, and across change metrics for all 56 countries, while Figure 2 depicts trajectories for no-DTP across national levels and subnational gaps for potential exemplars in reducing zero-dose children for each year between 2000 to 2019.

#### *3.3. Comparing Divergent No-DTP Trajectories since 2000 for Select Locations*

Focusing on the four Gavi Learning Hub countries—Nigeria, Mali, Uganda, and Bangladesh—we mapped their no-DTP trajectories from 2000 to 2019 against potential exemplars in reducing zero-dose children (Figure 3); the exception was Bangladesh, which achieved among the largest relative reductions in national no-DTP prevalence and subnational gaps since 2000. Accordingly, Figure 3 excludes Bangladesh.

**Figure 3.** Comparing Gavi Learning Hub country no-DTP trajectories since 2000 to potential exemplars in reducing zero-dose children. Bangladesh, a Gavi Learning Hub country, was identified as potential exemplar based on its marked progress on relative no-DTP metrics of change (Figure 2). Accordingly, we focus on Nigeria, Mali, and Uganda here. National no-DTP prevalence is represented on the x-axis and the subnational gap (as measured by the difference between the 5th and 95th percentile no-DTP prevalence across second-level administrative units) is represented on the y-axis. Each corner represents an extreme for each of these no-DTP metrics, with the lower right-hand corner—low national no-DTP prevalence and low subnational inequality—being the direction in which every location should strive to reach to equitably reduce no-DTP prevalence. Trends in the two no-DTP metrics for the Gavi Learning Hub countries are highlighted in black, with each circle representing a year from 2000 to 2019 that is color-coded from orange (2000) to blue (2019). The teal trends represent trajectories for the potential exemplars in reducing zero-dose children based on their absolute or relative progress since 2000. The light gray trajectories represent the other countries in this analysis.

For Nigeria, Ethiopia was its closest 'neighbor' in terms of national no-DTP prevalence in 2000—55.5% in Nigeria and 63.4% in Ethiopia—with diverging no-DTP trajectories through 2019 (i.e., 28.9% in Nigeria and 11.7% in Ethiopia). From 2000 to 2019, Nigeria consistently had among the highest subnational no-DTP disparities in the world; even in 2000, when Ethiopia had the fourth highest subnational gap in no-DTP among included countries (47.9 percentage points; Table 2), Nigeria's subnational gap was more than 20 percentage points higher (71.7; Table 2). Nonetheless, given how Ethiopia markedly reduced no-DTP subnational gaps at the same time trends in Nigeria's subnational disparities more or less stagnated, they may be well-aligned for cross-country learning.

For Mali, the DRC was its closest 'neighbor' for subnational no-DTP prevalence gaps in 2000, with Mali experiencing a 47.8 percentage-point gap and the DRC having a 46.8 percentage-point disparity. By 2019, Mali still had a subnational gap exceeding 40 percentage points (40.5, Table 2) while the DRC reduced its subnational gap to 16.0 (Table 2). National no-DTP prevalence was more variable for Mali and the DRC, with Mali's national no-DTP levels registering far lower than the DRC's in 2000 (36.4% and 51.9%, respectively) but then only moderately declining to 17.1% by 2019. In contrast, the DRC's national no-DTP prevalence decreased to 6.9% in 2019. Yet the DRC's no-DTP metrics from 2010–2015—the time before the country accelerated no-DTP reductions—parallel Mali's 2019 no-DTP measures. Accordingly, this more recent time period may support optimal cross-country learning for Mali.

For Uganda, Burundi aligned most closely to its 2000 no-DTP measures but showed divergences by 2019. In 2000, national no-DTP prevalence was 21.9% in Uganda and 18.4% in Burundi; by 2019, their no-DTP estimates were 6.7% and 2.3%, respectively (Table 2). Subnational gap trends were less similar for these two countries, with Burundi's no-DTP subnational gap in 2000 being narrower (17.1 percentage points) than that of Uganda's (29.2 percentage points). Each country recorded sizeable declines in subnational no-DTP gaps, with Uganda's falling to 6.3 percentage points and Burundi's to 2.4. In many ways, both Uganda and Burundi could offer meaningful lessons around reducing subnational disparities among unvaccinated children.
