*2.2. Study Outcome*

We examine maternal tetanus immunization coverage via the standard measure Protection at Birth (PAB), the proportion of women whose most recent live birth was protected against neonatal tetanus [32,33]. A birth is considered protected from tetanus if the mother (a) received at least two doses of TTCV during the pregnancy for her most recent live birth; (b) received at least two doses of TTCV, the last one within 3 years of the most recent live birth; (c) received at least 3 doses of TTCV, the last one within 5 years of the most recent live birth; (d) received at least 4 doses of TTCV, the last one within 10 years of the most recent live birth; or (e) ever received at least 5 doses of TTCV at any time prior to the most recent live birth. This measure is based on women whose most recent live birth occurred in the 59 or 23 months prior to the survey for DHS or MICS, respectively. This difference in time frame is due to the data collection methodologies of the two survey families. Additionally, maternal tetanus vaccination is ascertained via recall in DHS, while maternal vaccination cards are requested for confirmation in MICS and recall is used only if no card is available.

### *2.3. Dimensions of Inequality*

Based on drivers of inequality identified in previous publications on childhood vaccination as well as data availability in DHS and MICS, we examined the following four dimensions of inequality: household wealth (country-specific wealth quintiles) [6–9], maternal age (15–19, 20–49) [10], maternal education (none, primary, secondary or higher) [8,9,11], and place of residence (urban, rural) [9,11–13].

For a set of sub-analyses, we classified countries based on World Bank 2022 income groups: low-income, lower-middle income, or upper-middle income [34]. Only two highincome countries (Uruguay and Trinidad and Tobago) had available data, so they were excluded from these sub-analyses.

#### *2.4. Statistical Analyses*

We first present the latest situation of inequality in MNT vaccination coverage for each country (using the most recent survey available from 2011 to 2020) via disaggregated data and summary measures of inequality. For each of the four dimensions of inequality (household wealth, maternal age, maternal education, and place of residence), we calculated the following, based on the country-specific estimates:


For each median value estimated, we also present the 95% confidence interval (CI) of the median, calculated using the centile Stata command with default specifications. This uses a binomial-based method described in Mood and Graybill 1963 that makes no assumptions on the distribution of the coverage variable [35,36].

We also examined changes over time in coverage levels and coverage inequalities. For this analysis, we included countries with at least one survey in the period 2011–2020 and one in the period 2001–2010, where the two surveys were at least 5 years apart. When multiple surveys in a time range were available, the most recent survey that maintained a 5-year gap was used. To assess changes in inequality over time, we first examined annual absolute change in national coverage levels, calculated as the national coverage in the more recent survey minus the national coverage in the older survey divided by the number of years between surveys. We then calculated annual absolute excess change in coverage, which compares the annual rate of change in the least and most advantaged subgroups. This is calculated as (absolute annual change for least advantaged subgroup) minus (absolute annual change for most advantaged subgroup). Several patterns in coverage can lead to positive (pro-disadvantaged) or negative (pro-advantaged) excess change. For example, a positive excess change in coverage value can arise when both groups have increasing coverage but the increase in the disadvantaged group is faster than the increase in the advantaged group; when both groups have decreasing coverage but the decrease in the disadvantaged group is slower than the decrease in the advantage group; or when the disadvantage group increased (or had no change in) coverage while the advantage group had a decrease (or had no change in) coverage. Excess change in coverage has been previously used to portray change in inequality over time in diphtheria-tetanus-pertussis (DTP) immunization coverage and in other maternal health outcomes [7,37].

As an additional post hoc analysis, we examined whether trends in inequality differed substantially based on MNTE achievement status, assessing inequality metrics separately for countries who have achieved MNTE vs. those who have not achieved MNTE.

Relevant survey sampling designs were taken into account when calculating point estimates of disaggregated data and corresponding 95% CIs at the country level. Statistical significance was set at *p* < 0.05 for all comparisons, and 95% CIs are reported throughout.

We conducted all analyses in Stata 17, and we developed data visualizations using Tableau version 2022.1.1.

#### **3. Results**
