*2.3. Statistical Analyses*

Statistical analyses followed the WHO tetanus serosurvey guidance published in 2018 [14]. Estimates of tetanus and diphtheria seroprotection (at ≥0.01, ≥0.1 and ≥1.0 IU/mL antibody levels) and Wilson 95% confidence intervals were calculated using sample weights and a Taylor series linearization method to account for survey design. To assess risk factors, seroprotection was calculated by sociodemographic characteristics (age, sex, zone, state, urban/rural, and wealth quintile) and associations were assessed using Rao-Scott chi-square tests. A *p*-value of <0.05 was considered statistically significant. To assess waning immunity and the need for tetanus–diphtheria booster dose introduction, estimates of the proportions of children by age and different antibody level categories for tetanus and diphtheria (<0.01, 0.01 to 0.099, 0.1 to 0.99, and ≥1.0 IU/mL) were calculated; geometric mean antibody levels and 95% confidence intervals (95%CI) were also calculated accounting for the survey design. We assessed the contribution of routine immunization, waning immunity, and natural exposure (for diphtheria) by assessing percentage of children seroprotected (≥0.01 IU/mL) against diphtheria (including through natural infection) but not against tetanus (<0.01 IU/mL) by age group (4–11 months, 12–23 months, 24–35 months, 36–47 months, and 48–59 months; 0–4, 5–9 and 10–14 years). All statistical analyses were conducted using SAS v9.4 (SAS Institute, Inc., Cary, NC, USA).

Finally, we triangulated data on the proportion of children with minimal and full seroprotection to tetanus by age with vaccination coverage to assess comparability between serosurvey results and vaccination coverage results in those age groups given that tetanus immunity can only be achieved by vaccination. For this purpose, we used DTP3 coverage among children 12–23 months of age reported from three data sources that covered the birth cohorts included in this survey (born during 2004–2018): (1) WHO/UNICEF estimates of national immunization coverage (WUENIC); (2) country-reported official estimates for DTP3 coverage for each birth cohort [27]; and (3) when available for the birth cohorts included in this survey, we also used vaccination coverage survey estimates of DTP3 coverage among children 12–23 months from Demographic and Health Surveys (DHS) (for 2018, 2013, and 2008) [28] and Multiple Indicator Cluster Surveys (MICS) (for 2016, 2011, and 2007) [29].

### **3. Results**
