**3. Results**

#### *3.1. Coverage of Measles-Containing Vaccine in the Study Population*

There were 795,734 individuals aged 2 to 25 years of age, alive and resident in Wales as of 31 August 2021. Of these, 35,254 did not have a record in the NCCHD and a further 111,585 were not registered with a GP who submits data to the SAIL Databank. Using NCCHD data only, coverage of one dose of measles-containing vaccine in these 648,895 remaining individuals was 96.2% and coverage of two doses in those aged 4 to 25 years of age was 92.0%. After reconciling with GP data, coverage increased to 97.1% for one dose and 93.8% for two doses (Figure 1). Of those who were vaccinated, 1620 had received measles-containing vaccines other than MMR for their first dose and 2781 had received measles-containing vaccines other than MMR for their second dose. The majority of non-MMR measles vaccines were given to those aged 15 to 21 years (with the highest proportion received by 20-year-olds, 1.3%). The proportion of all measles vaccines received that were non-MMR was under 0.5% in all other age groups.

**Figure 1.** Coverage of one (**a**) and two (**b**) doses of measles-containing vaccine in those aged 2 to 25 years of age, alive and resident in Wales as of 31 August 2021. The improvement in coverage from reconciling the National Community Child Health Database and primary care GP data is also shown.

### *3.2. Determinants of Measles Vaccination Coverage*

After exclusion of 4688 individuals with vaccine refusal codes, there were 644,207 individuals aged 2 to 25 years in the equity study population. In a univariable analyses, month of birth was the only variable that was not significantly associated with vaccination uptake of either dose. Having had previous vaccinations was strongly associated with having had at least one dose of measles-containing vaccine; OR 177.45 (95% CI 162.99–193.60) for pneumococcal vaccine, OR 100.25 (95% CI 96.52–104.14) for three doses of pertussis vaccine and OR 27.60 (95% CI 25.82–29.50) for two doses of rotavirus vaccine.

Age first registered with a primary care GP in Wales was most strongly associated with vaccine coverage, with those first registering at secondary school age (12 to 16 years of age) least likely to be recorded as vaccinated, compared to those born in Wales. Those born outside of the UK were less likely to be vaccinated, OR 0.07 (95% CI 0.06–0.07) for one dose. For groups with at least 100 persons, coverage of one dose was under 80% in those born in Romania, Bulgaria, Syria, Lithuania, Turkey, Slovakia, Czech Republic, Nigeria, Zimbabwe, Iraq, South Africa and Asia (not otherwise specified). Coverage was also higher in those who had English or Welsh recorded as a language, OR 8.45 (95% CI 7.88–9.06) for one dose. For groups with at least 100 persons, coverage of one dose was under 80% in those recorded as speaking Bulgarian, Romanian, Lithuanian, Russian, Hungarian, Slovak, Italian and Spanish. There was also association with ethnicity and coverage, with those who were in a combined Black, Asian, Mixed or other ethnic group having lower coverage than those in the combined White ethnic group. In the univariable analysis those with a recorded religion of Buddhism, Islam, Pagan or other religions were less likely to be vaccinated than those who stated they had no religion.

Females were more likely to be vaccinated than males, OR 1.09 (95% CI 1.05–1.12) for one dose. Mothers who were older (36 years and over) and younger (under 17) when giving birth were less likely to have children who were vaccinated, as well as those born in to families with more children (OR 0.16 95% CI 0.14–0.18 if sixth or greater compared to first born). There was variation by health board and deprivation quintile of residence, with vaccination less likely in more deprived areas. Coverage was also lower in urban areas, compared to rural areas, OR 0.74 (95% CI 0.71–0.77) for one dose. Those who have ever been eligible for free school meals were less likely to be vaccinated, OR 0.86 (95% CI 0.82–0.89) as well as those who were born to mothers who smoked during pregnancy (OR 0.76 95% CI 0.69–0.84) and mothers who had no qualifications compared to those with at least GCSE qualifications. People were less likely to be vaccinated with two doses if they had a school exclusion record (OR 0.76 95% CI 0.72–0.80), although this association was not seen with one dose. This association was stronger for those with a permanent exclusion record compared to a temporary exclusion record.

There was no association with vaccination and premature birth for dose one but coverage of two doses was significantly lower, OR 0.92 (95% CI 0.87–0.97). Those who consulted with their GP at least once between 1 September 2020 and 31 August 2021 were more likely to be vaccinated, and coverage was significantly higher in those with recorded comorbidities. Those with chronic pulmonary disease, renal disease and uncomplicated diabetes were significantly more likely to be vaccinated, and those with liver disease, peptic ulcer and rheumatic disease were significantly less likely to be vaccinated. Coverage of at least one dose in those with hearing loss or sight loss was higher than the rest of the study population; coverage of two doses in those with sight loss was not significantly different. Coverage of two doses in those with a learning disability was lower than the general population, OR 0.62 (95% CI 0.55–0.71), although there was no difference for one dose. Those who attend, or have attended, a special school had lower coverage of one and two doses.

The univariable analyses including the full cohort can be found in Supplementary Table S1.

The variables included in the final model are presented in Table 1. The cohort was restricted to those aged 4 to 25 years without missing information across all variables, producing a study population of 419,405. This restricted cohort had higher vaccine coverage overall, and led to some notably different estimates in the univariable analyses. Although having a comorbidity score of one was associated with higher vaccination coverage compared to those with no comorbidities, those with a score of three or more were less likely to be vaccinated in the restricted cohort. Also, younger mothers (aged under 17) in the restricted cohort were more likely to have vaccinated children compared to the univariable analysis in the full cohort, which showed they were less likely.

In the multivariable analysis, after controlling for other factors, the strongest association with vaccination uptake was birth order (OR 0.21 95% CI 0.17–0.26 for one dose if sixth or greater compared to first born) and being born outside of the UK (OR 0.21 95% CI 0.18–0.25 for one dose). Living in a more deprived area of residence was still associated with lower coverage but the association was not as strong. The association with recorded language, free school meal eligibility and mothers' highest qualification was also slightly reduced. Having a comorbidity score of 3 or more was no longer significant. However, the biggest difference was seen in ethnicity, where those in the combined Asian ethnic group were more likely to be vaccinated with at least one and two doses after controlling for other factors (OR 1.58 (95% CI 1.27–1.97) for at least one dose), and those in the combined Black ethnic group were no longer significantly less likely to be vaccinated. Differences were also seen in those who had recorded religion of Islam, who were more likely to be vaccinated with one and two doses after controlling for other factors (OR 1.45 (95% CI 1.13–1.87) for at least one dose).

*Vaccines* **2023**, *11*, 680

**Table 1.** Uptake of one or two doses of measles-containing vaccine in those aged 4 to 25 years alive and resident in Wales as of 31 August 2021, without a vaccine refusal code, by individual characteristics. Odds Ratios and 95% Confidence Intervals are also presented. Analysis restricted to those with complete information across all variables.








\*Datasuppressedtocomplywithstatisticaldisclosurepolicy; Groupswithuptakeunder95%areindicatedwithboldtext.
