Health impact of routine measles vaccination and 1 supplementary immunisation activities in 14 high burden 2 countries: a DynaMICE modelling study

14 Background: WHO recommends ≥95% population coverage with two doses of measles- 15 containing vaccine (MCV). Most countries used routine services to offer MCV1 and later, 16 MCV2. Many countries conducted supplementary immunisation activities (SIAs), offering 17 vaccination to all persons in a given age range irrespective of prior vaccination history. We 18 estimated the relative impact of each dose and delivery route in 14 high burden countries. Methods: We used an age-structured dynamic model (DynaMICE), to estimate the health 21 impact of different vaccination strategies on measles susceptibility and burden over 22 2000 ─ 2020. We estimated their incremental impact using averted cases and deaths and their 23 efficiency using number needed to vaccinate (NNV) to avert an additional measles case. 25 Findings: Compared to no vaccination, MCV1 implementation averted 823 million cases and 26 9.5 million deaths, with a median NNV of 1.41. Adding routine MCV2 to MCV1 further averted 27 108 million cases and 0.4 million deaths, while adding SIAs to MCV1 led to 249 million averted 28 cases and 4 million deaths. Despite a larger incremental impact, adding SIAs to MCV1 showed 29 reduced efficiency compared to adding routine MCV2, with median NNVs of 6.15 and 5.41, 30 respectively.


Introduction 39
Between 2000-2020, measles deaths were estimated to have decreased by 94% globally [1], 40 achieved mostly through routine immunisation (RI) and supplementary immunisation activities 41 (SIAs) with measles-containing vaccines (MCV) [2][3][4][5]. MCV1 is defined by the World Health 42 Organization (WHO) as the first routine dose of MCV given during the first year of life 43 (recommended at 9 or 12 months of age) while MCV2 is defined as the second routine dose 44 of MCV (recommended at 15-18 months of age). SIAs refer to vaccination campaigns that 45 deliver vaccine doses using strategies going beyond routine services and have usually been 46 non-selective, that is offering vaccine irrespective of past vaccination history. Throughout this 47 paper, the term SIA indicates non-selective SIAs. 48 49 Since the introduction of measles vaccination in low-and middle-income countries (LMICs), 50 recommendations around measles vaccination strategies have been revised. Historically, 51 LMICs relied on MCV1 with SIAs to interrupt transmission and reach unvaccinated children. 52 In 2009, WHO recommended introducing MCV2 once a country reached 80% MCV1 53 coverage, retaining an emphasis on aiming for high coverage with MCV1 as soon as possible 54 after a child loses maternal antibodies. In 2017, this policy was revised to recommend that 55 countries include MCV2 in RI schedules regardless of MCV1 coverage alongside operational 56 support to strengthen RI infrastructure when incorporating MCV2. In part due to concerns 57 about the sustainability of funding for nationwide non-selective SIAs and their potential to 58 disrupt routine services [6,7], WHO has proposed that eventually such SIAs can be phased 59 out once countries achieve over 95% coverage of both routine doses [8]. Implementation of SIAs over time has been motivated by different goals and needs. For 66 example, SIAs have been used to increase population immunity in countries with low MCV1 67 or MCV2 coverage. In such settings, SIAs have been cited as a highly effective and equitable 68 strategy for protecting hard-to-reach communities with children who would otherwise be 69 missed by RI [11,12], although the relative reach of SIAs versus RI varies between and within 70 countries [13]. Importantly, SIAs were a major component of the measles elimination strategy 71 in the region of the Americas, with high routine MCV1 coverage and occasional follow-up SIAs 72 sustaining elimination for many years [14]. To prevent measles transmission and subsequent 73 outbreaks, a commonly used (rule-of-thumb) criterion is that a follow-up SIA should be 74 conducted before the number of susceptible children under 5 years of age approaches the 75 size of one birth cohort [15,16]. Historically, this rule-of-thumb has been influential in informing 76 the timing of SIAs so the accumulation of susceptibles remains below the size of one birth 77 cohort and measles transmission can be interrupted and elimination achieved [8]. In practice, 78 even if countries recognise that a follow-up SIA is due and correctly identify the age groups 79 with the highest prevalence of susceptibility, delays in obtaining funding or competing priorities 80 may lead to delayed implementation of an SIA or a narrower than ideal age range, which 81 reduces its impact [16]. 82 83 In 2012 is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 12, 2022. coverage was either 10% lower than the country's MCV1 coverage or equal to the country's 167 MCV2 coverage in that year, whichever was larger (dashed lines in appendix figure S1). In 168 addition, we modelled two alternative assumptions about the likelihood of receipt of an SIA 169 dose according to past vaccination history (appendix figure S2). In this analysis, we specifically 170 describe 'zero-dose' population as those receive no MCV doses. One assumption is that SIA 171 doses preferentially reached already-vaccinated children, and any remaining doses after all 172 already-vaccinated children are reached are then given to zero-dose children, while the other 173 assumes a strategy that reaches zero-dose children first, and the remaining doses are given 174 to already-vaccinated children. While these two distribution assumptions are hypothetical, 175 these analyses strengthen our understanding on the potential ranges of incremental health 176 impact and efficiency that SIAs can provide when added to MCV1 only strategy. 177 Role of the funding source 178 The funders had no role in study design, data collection, data analysis, data interpretation, or 179 writing of the report. All authors had full access to all data in the study and had final 180 responsibility for the decision to submit for publication. 14 high-burden countries presented a substantial decline in incidence rates in scenarios with 188 MCV1 only. In the scenario where MCV1 and MCV2 were used without SIAs, the annual 189 burden of measles declined slowly over time, and endemic transmission continued. With 190 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 12, 2022. ; https://doi.org/10.1101/2022.07.11.22277494 doi: medRxiv preprint MCV1 and SIAs, there was a more rapid decline in measles burden, but large-scale outbreaks 191 were predicted. Overall, the greatest absolute burden reduction attributable to MCV1, MCV2,192 and SIAs in comparison to no vaccination over 2000-2020 was in India, China, and Nigeria 193 respectively (appendix figure S3), which are countries with the highest IHME measles 194 incidence estimates and largest population sizes. 195 196 The estimated total number of susceptible children under 5 years of age shows varying trends 197 by vaccination delivery strategy (figure 3 trend being 10% lower than MCV1 (appendix figure S1). Compared to MCV1 alone, early 242 introduction of MCV2 would have averted more cases than the historical introduction, resulting 243 in a further reduction of 97 million measles cases across study countries. In terms of efficiency, 244 only slight improvement was seen from early MCV2 introduction, with a median NNV dropping 245 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 12, 2022. Using 20 years of MCV coverage data, we investigated the relative impact and efficiency of 262 different MCV strategies in 14 high measles burden countries that included a wide range of 263 socio-economic, demographic and routine immunisation settings. We describe health impact 264 in terms of cases, deaths, and DALYs averted, and the proportion of years during which 265 vaccination scenarios kept the susceptible population below one birth cohort, while efficiency 266 is described in terms of NNV to avert a measles case. 267 268 The use of MCV1 resulted in the highest relative health impact of any dose, and the best 269 efficiency in reducing measles burden. The vaccination scenario using MCV1 and SIAs 270 together can more effectively keep the size of the susceptible population below the size of one 271 birth cohort and had a bigger effect on predicted incidence in comparison to using MCV1 and 272 MCV2 together. Furthermore, SIAs averted a greater proportion of measles burden and 273 reduced the size of the susceptible population more than MCV2, whereas efficiency of SIAs 274 as assessed by NNV to avert a measles case was lower than that for MCV2. However, there 275 was marked variation between countries in the relative efficiency of each incremental strategy. 276 The strategies used between 2000-2020 in these countries greatly reduced measles burden 277 compared to a no-vaccination scenario but, other than in China, were not predicted to prevent 278 large outbreaks. This is consistent with other analyses in LMICs [33]. 279 280 The relatively high impact but reduced efficiency of SIAs could also be interpreted from the 281 viewpoint of dose deliverywhile SIAs could be delivered to more people, many doses were 282 predicted to reach previously vaccinated children (appendix figure S4). Even though 283 prioritisation of zero-dose children for measles vaccination during SIAs may not always be the 284 most efficient strategy as it is operationally challenging and costly, successfully reaching zero-285 dose children first results in an effective strategy to avert measles burden. 286 287 Compared to SIAs, MCV2 had relatively less impact in countries that had not yet met or 288 exceeded 90% routine coverage of MCV1, even under the assumption of early MCV2 289 introduction. Accompanied by better reach of zero-dose children by SIAs, early introduction of 290 MCV2 could have substantially reduced incidence over time. In countries that implemented 291 routine MCV2 early on and maintained high levels of coverage, such as China and India, there 292 was little difference in estimated incidence rate between historical scenarios and "optimal" 293 assumptions for MCV delivery that MCV2 was introduced in 2000 and SIA doses were given 294 first to zero-dose children (appendix figure S5). This finding suggests that in the future SIAs 295 may be needed less often if high coverage of MCV2 can be successfully attained and 296 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 12, 2022. an immunisation programme is acceptable [39]. Further data, such as costs for vaccine 341 procurement and delivery, will be useful in understanding cost-effectiveness of immunisation 342 programmes. 343 344 The resources required for SIAs, including economic and human resources as well as 345 logistical challenges, can be major deterrents to their implementation. Despite several 346 unknowns regarding interpretation of estimated NNVs, our results show that routine MCV2 is 347 not always more efficient than SIAs. Furthermore, the current trend towards including multiple 348 interventions in a single SIA or integrating many of the components of SIA planning across 349 different interventions may increase efficiency, although it will be important to monitor the 350 effectiveness of "integrated campaigns" [40,41]. 351 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

352
We critically assessed the incremental impact of different measles vaccination strategies to 353 inform future decisions about vaccination planning and policies. Understanding the relative 354 impact and efficiency of the first routine dose, second routine dose, and SIAs of MCV will 355 assist stakeholders in assessing the value of measles vaccination programmes, and further 356 identify improved pathways towards measles elimination. 357 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 12, 2022. The authors have no conflicts of interests to declare. 380 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

398
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

412
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

450
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 12, 2022.