The Western Pacific Regional Plan of Action for Measles Elimination (2003) aims to ‘achieve and maintain 95% population immunity to measles in each birth cohort within each district of each country in the Region’ as one of its three main strategies towards measles elimination [20]. This is the first study that aims to do so by evaluating immunity conferred from measles vaccination in Malaysian birth cohorts throughout the life course. Our findings show that the Effective Vaccination Coverage (EVC) throughout the life course of Malaysian birth cohorts since 1982 is below the 95% immunity target threshold required to achieve measles elimination, contributing to the persistence of measles in Malaysia.
This study provides an assessment of national-level population immunity of measles, and does not consider subnational heterogeneity or vaccine waning. Additionally, there is no published literature as yet on national-level serological assessments of measles immunity. However, district and national laboratory serological assessments are available for comparison. One subnational study collected results of measles seroassays from September 2014 to January 2015 from government health clinics in a Malaysian district, r; here, individuals aged 15 to 24 years old showed a 74% positive titre (95% CI 70–78%) and children aged 6 to 9 years old showed 90% positive titre (95% CI 87–94%) [21]. A prior study conducted in the year 2008 using measles samples from Malaysian hospitals showed that 82.8% of individuals above the age of seven were seropositive for measles [22]. Malaysia has consistently reported more than 95% coverage of MCV1 and MCV2 in most years since 1982, yet the observations in this study and previous serological assessments show WUENIC estimates may not be the most accurate representation of measles population immunity in the country.
In 2013, updated guidelines for measles elimination were published by WPRO to recommend that countries ‘achieve and maintain > 95% vaccination coverage of two doses of MCV through routine immunisation, adding SIAs when required’ [23]. We find that the Malaysian population immunity conferred through measles vaccination increased significantly through the introduction of MCV2 and national immunisation campaigns. As noted in this study, a larger proportion of individuals receive MCV2 beginning in 2004; given that efficacy of measles vaccination improves with subsequent doses, EVC improves in these birth cohorts. Our findings corroborate this, showing that in 2.5% of samples, i.e. the best-case scenario, EVC crosses the 95% threshold. However, the study also finds that there is still a significant proportion of individuals in Malaysia who have not received any measles vaccination or only received MCV1.
These ‘best-case scenario’ findings, driven by higher vaccine efficacy, between dose coverage (BdC), proportion of individuals with MCV2, and a lower proportion of zero-dose individuals, represents the strategy required for immunisation programmes to meet the 95% threshold for measles control. Achieving these outcomes, however, requires immunisation programme capacity to monitor BdC. Since 1982, Malaysia has made significant strides towards the elimination of measles, implementing evidence-based vaccination policies that have resulted in a decreased burden. However, Malaysia is reliant on paper-based mechanisms for immunisation coverage reporting [24]. Paper-based systems depend on the aggregated facility or district-level data, and may be subject to delays or lags. This results in overreporting of estimates and increases the risk of unreadable, missing, or inconsistent vaccination coverage data. Electronic immunisation registries (EIR) circumvent these issues, especially ones that are interoperable with other electronic systems that handle patient records and civil registration systems for more accurate population and vaccine dose data [25, 26]. For Malaysia, a country aiming for measles elimination, the objective is to maintain high coverage and BdC through routine immunisation with periodic SIAs to cover immunity gaps. Conversely, in a country with a high burden of measles and low MCV1 coverage looking to achieve disease control, decorrelating measles vaccination doses provides a greater opportunity for zero-dose individuals to receive MCV1.
However, the implementation of a mechanism for monitoring BdC must consider contextual health system factors and feasibility. Developing health information management systems is costly and laborious, butalleviate user workload when compared to traditional paper-based systems [27, 28]. In this regard, there have been examples of innovative mixed paper and electronic immunisation registries that have shown promise in low-resource settings. Given the budgetary, infrastructural, and technical considerations at the national level in most digital health projects, it remains to be seen if this technology can be scalable [29].
There are several methodological strengths in this study. The use of probability theory to proportionate the population by the number of doses provides an opportunity to account for BdC in settings that do not have a mechanism for its monitoring. It also estimates the zero-dose population and gives insight into the effectiveness of SIAs in vaccinating susceptible populations. This is important as relying solely on vaccination coverage data may underestimate the proportion of susceptible individuals, as individuals with MCV1 may be subject to re-vaccination in these campaigns [30]. Future research should explore the age distribution of measles cases within the Malaysian population and correlate the findings of EVC by birth cohorts at the district level. This will better inform programme managers on the need for age-targeted immunisation campaigns as well as serological studies to evaluate population immunity against measles. This methodology provides flexibility in application, as it can be used for other multi-dose vaccination programmes.
Depending on the type of vaccine and efficacy data available, incorporating primary vaccination failure as a function of the methodology may need to be considered in subsequent research. This study uses the ‘all-or-nothing’ approach to define measles vaccine efficacy, and does not consider the waning of immunity conferred through vaccination as it does not play a significant role in measles transmission [31]. Future research should build upon the aforementioned aspects in order to provide more reliable estimates on the effectiveness of measles vaccination in a given population.
Uncertainty within our findings may stem from our use of UNWPP estimates of population demography and vaccination coverage estimates, given that the study methodology matched these estimates to immunisation data to allow for temporal analysis of population immunity. This may result in an overestimation or underestimation of the target population given that these are external sources of population estimates, and not in-country demographic statistics used by measles vaccine programme managers in Malaysia. Target population estimates influence the precision of vaccination coverage figures significantly, and increasingly so in settings with higher coverage levels [32]. Further, the temporal EVC changes seen within the population are based on vaccination delivery occurring as per the routine immunisation schedule. This may not necessarily reflect the situation in real life [33, 34]. Delays in vaccination administration may result in a child being part of the susceptible group for a longer period, shifting changes in EVC values in birth cohorts to higher age groups. Future research should explore the timeliness of measles vaccination to determine the proportion of children who receive measles immunisation at the recommended age.
Infants have protection from maternal antibodies against measles infection which wanes around 6 months from birth. The first dose of routine measles vaccination (MCV1) is usually given to infants at 9 months old. The methodology proposed in this paper does not account for maternal protection and the gap between its waning and the first dose of MCV vaccination, because this paper is primarily interested in vaccine induced immunity. Immunity induced by natural infection is also out of the scope in this paper. To be able to account for disease induced immunity, additional input of age specific measles cases and deaths are required. We consider this as a future research topic.
Add Comment