On-time vaccination was defined as vaccine receipt within 30 days of the recommended age. MMR indicates measles-mumps-rubella; MMR2, dose 2 of MMR-containing vaccine in cohort born between July 1 and December 31, 2012 (A) and cohort born between January 1 and June 30, 2009 (B).
Customize your JAMA Network experience by selecting one or more topics from the list below.
Macartney K, Gidding HF, Trinh L, et al. Evaluation of Combination Measles-Mumps-Rubella-Varicella Vaccine Introduction in Australia. JAMA Pediatr. 2017;171(10):992–998. doi:10.1001/jamapediatrics.2017.1965
What is the effect of introduction of combination measles-mumps-rubella-varicella vaccine at age 18 months as the second dose of measles-containing vaccine on vaccine coverage and risk of vaccine-associated febrile seizures in Australia?
A national cohort study of vaccine coverage before and after measles-mumps-rubella-varicella vaccine introduction showed improvement in uptake and timeliness for all 4 vaccine components. Despite the peak incidence of all-cause febrile seizures occurring at age 18 months and a known increased risk of febrile seizures following the first dose, in a self-controlled case series analysis including 1471 children, use of measles-mumps-rubella-varicella vaccine at 18 months was not associated with an increased risk of febrile seizures.
Measles-mumps-rubella-varicella combination vaccine was safely incorporated into the Australian National Immunisation Program schedule and improved population-level protection against these serious viral diseases.
Incorporating combination vaccines, such as the measles-mumps-rubella-varicella (MMRV) vaccine, into immunization schedules should be evaluated from a benefit-risk perspective. Use of MMRV vaccine poses challenges due to a recognized increased risk of febrile seizures (FSs) when used as the first dose in the second year of life. Conversely, completion by age 2 years of measles, mumps, rubella, and varicella immunization may offer improved disease control.
To evaluate the effect on safety and coverage of earlier (age 18 months) scheduling of MMRV vaccine as the second dose of measles-containing vaccine (MCV) in Australia.
Design, Setting, and Participants
Prospective active sentinel safety surveillance comparing the relative incidence (RI) of FSs in toddlers given MMRV and measles-mumps-rubella (MMR) and a national cohort study of vaccine coverage rates and timeliness before and after MMRV vaccine introduction were conducted. All Australian children aged 11 to 72 months were included in the coverage analysis, and 1471 Australian children aged 11 to 59 months were included in the FS analysis, with a focus on those aged 11 to 23 months.
Main Outcomes and Measures
MMRV vaccine safety, specifically, the RI of FSs after MMRV vaccine at age 18 months, compared with risk following MMR vaccine and vaccine uptake for 2-dose MCV and single-dose varicella vaccine, focusing on timeliness.
Of the 1471 children, the median age at first FS was 21 months (interquartile range [IQR], 14-31 months). Three hundred ninety-one children were aged 11 to 23 months and had at least 1 FS included in the analysis; of these, 207 (52.9%) were male. A total of 278 children (71.1%) had received MMR followed by MMRV vaccine, 97 (24.8%) had received MMR vaccine only, and 16 (4.1%) had received neither vaccine. There was no increased risk of FSs (RI, 1.08; 95% CI, 0.55-2.13) in the 5 to 12 days following MMRV vaccine given as the second MCV to toddlers. Febrile seizures occurred after dose 1 of MMR vaccine at a known low increased risk (RI, 2.71; 95% CI, 1.71- 4.29). Following program implementation, 2-dose MCV coverage at age 36 months exceeded that obtained at age 60 months in historical cohorts recommended to receive MMR vaccine before school entry, and on-time vaccination increased by 13.5% (from 58.9% to 72.4%). Despite no change in the scheduled age of varicella vaccine, use of MMRV vaccine was associated with a 4.0% increase in 1-dose varicella vaccine coverage.
Conclusions and Relevance
To our knowledge, this is the first study to provide evidence of the absence of an association between use of MMRV vaccine as the second dose of MCV in toddlers and an increased risk of FSs. Incorporation of MMRV vaccine has facilitated improvements in vaccine coverage that will potentially improve disease control.
Some parents and health care workers are finding decision making regarding immunization increasingly complex, presenting a barrier to timely vaccine uptake.1,2 A commonly reported concern is the number of injections given to children.1,2 Combination vaccines reduce the number of injections needed and may improve vaccine acceptance, coverage, and, ultimately, disease control. However, various other factors surrounding use of combination vaccines, including cost-effectiveness, safety, availability, and country- or region- specific disease epidemiology, require consideration.3 In the past decade, several countries have faced challenges in incorporating the combination measles-mumps-rubella-varicella (MMRV) vaccine into their immunization schedules. Although both available MMRV vaccines (Priorix-Tetra [GlaxoSmithKline Biologicals SA] and ProQuad [Merck & Co Inc]) offer the advantage of a single injection against 4 diseases, prelicensure studies showed an increased risk of fever in first-dose recipients aged 12 to 23 months compared with children who received measles-mumps-rubella (MMR) and varicella vaccines separately.4,5 This reaction was presumed to be related to potentiation of the immune response to the measles virus component. Postlicensure studies subsequently reported an approximately 2-fold increased risk of febrile seizures (FSs) following MMRV compared with giving separate MMR and varicella vaccines.6,7 This finding prompted a withdrawal of a preferential recommendation for use of MMRV as the first measles-containing vaccine (MCV) in the United States and Germany.8,9
Before July 2013, MMRV vaccine was not used in Australia. Two doses of MMR vaccine were scheduled on the National Immunisation Program (NIP) and spaced 3 years apart at ages 12 months and 4 years, similar to the US and UK schedules. However, data from the national Australian Childhood Immunisation Register (ACIR) in 2012 showed that vaccine uptake was suboptimal; approximately 92% of children had received 2 MCVs by age 5 years,10 and modeling demonstrated an increased risk of measles outbreaks associated with low 2-dose immunity in younger children. Disease outbreaks arising from measles importations11,12 demonstrated the need to improve 2-dose coverage at all ages, but especially in the young. A single dose of monovalent varicella vaccine had been scheduled under NIP at age 18 months since November 2005,13 but coverage by age 2 years was only 86%, although it increased to 92% by 5 years. Declines in varicella-related morbidity and mortality had occurred,14-17 but modeling18 suggested that improved 1-dose coverage was needed to decrease the risk of shifting disease to older age groups where higher disease severity occurs.
To address these challenges, the decision was made to include MMRV vaccine on the Australian NIP at age 18 months as the second MCV dose from July 2013 onward, as reported in Table 1. The risk-benefit assessment that underpinned this change was based on 2 hypotheses: (1) higher and earlier population-level vaccine coverage of 2 doses of MCV and 1 dose of varicella vaccine would be achieved by bringing forward the scheduled age for the second MCV dose to 18 months and replacing MMR with MMRV vaccine, and (2) when used as the second instead of the first dose of MCV, MMRV vaccine would not be associated with an increased risk of FSs, even though it would be provided to children aged 18 months, when the incidence of FSs peaks. The vaccine safety and evaluation plan for MMRV vaccine introduction included active, prospective sentinel FS surveillance using the Paediatric Active Enhanced Disease Surveillance (PAEDS) network19,20 and analysis of vaccine uptake using the ACIR. We aim to present the findings of this evaluation, examining the effect of the program change on (1) vaccine safety, specifically, the risk of MCV-associated FS and (2) vaccine uptake and timeliness.
Active, prospective sentinel FS surveillance was conducted from May 1, 2013 (2 months before MMRV vaccine introduction), to June 30, 2014, by the PAEDS Network at 5 Australian tertiary pediatric hospitals, as previously described.19,21 At each site, emergency department and inpatient databases were scanned daily by PAEDS surveillance nurses to ascertain possible FS presentations in all children younger than 5 years. Periodic review of all International Statistical Classification of Diseases and Related Health Problems, Tenth Revision, Australian Modification–coded FS encounters (code R56.0) was also conducted to capture additional cases. Clinical and demographic data were collected from the medical records and caregiver interviews, and all FS diagnoses were confirmed. All children had immunization records obtained from the ACIR, both at FS presentation and at study end (to identify all vaccine exposures).
In Australia, the timing of vaccine administration is highly associated with NIP-recommended schedule points. Therefore, our analysis cohort was restricted to children who were aged 11 to 23 months. Analysis was further restricted to include only children who had (1) 1 dose of MMR vaccine followed by 1 dose of MMRV vaccine at least 27 days later (consistent with NIP recommendations), (2) 1 dose of MMR vaccine (as some had not yet received MMRV vaccine), or (3) no MMR or MMRV vaccine (unvaccinated children, who contribute to the age-specific relative incidence [RI]).22 Children who received MMRV vaccine as their first MCV were excluded because this schedule was not consistent with NIP recommendations and occurred rarely.
The main study outcome was the RI of FS in the 5 to 12 days after the first and second MCV doses compared with nonrisk periods (baseline) within the same person. An additional risk period of 13 to 30 days after each MCV was also included to identify any longer-term risk.
Relative incidences were calculated using the self-controlled case series (SCCS) method of analysis.3,4 The SCCS method requires FS cases only and compares the FS rate during biologically plausible, predetermined risk periods with nonrisk periods (baseline) within the same person using conditional Poisson regression models; thus, all fixed confounders (eg, sex) are automatically adjusted for.22 Because age is a strong predictor of FS and is time varying, all models were adjusted for the effect of age (using 3 age groups in the base case: 11-14, 15-18, and 19-23 months). We removed the −1- to −13-day period before vaccination from the baseline time because it may be associated with a lower FS risk (an FS occurrence may delay receipt of scheduled vaccines).19,22
The primary analysis included children who had both first and subsequent FS episodes (considered unique episodes), in which the subsequent FS was separated by at least 7 days from a previous episode.3,5 Two sensitivity analyses were conducted: (1) adjustment for age using finer intervals (1-month age groups) and (2) restriction of the analysis to first FS episodes.
The ACIR is a nearly complete electronic population register. It includes approximately 99% of all children registered with the national public health insurance scheme, Medicare, that covers all citizens, permanent residents, and select visa holders. During the study period, the ACIR recorded receipt of all vaccines provided up to age 7 years.23 Doses recorded on the ACIR (renamed the Australian Immunisation Register [AIR] from November 2016 and including all aged persons) are linked to financial incentives for families and health care professionals, providing a basis for complete reporting.7,24
The study outcome was immunization coverage of consecutive, 3-month national cohorts of children born between January 1, 2009, and December 31, 2012, who had reached the ages of 24, 36, 48, and 72 months, respectively, for receipt of MMR, varicella, and/or MMRV vaccine by December 2015 (Table 1).
Coverage estimates for receipt of a second MCV dose and single varicella vaccine dose, either on time (within 30 days of the recommended age) or at scheduled assessments dates, were compared between the pre-MMRV and post-MMRV periods. Data were analyzed in SAS, version 9.3 (SAS Institute Inc); Stata, version 12 (StataCorp); and Excel 2007 (Microsoft Corp).
Each PAEDS hospital obtained ethical approval to conduct the FS safety study: Sydney Children’s Hospital Network Human Research Ethics Committee; Princess Margaret Hospital Human Research Ethics Committee; Women’s and Children’s Hospital Network Human Research Ethics Committee; QLD Children’s Health Services (Royal Children’s Hospital) Human Research Ethics Committee; and the Royal Children’s Hospital Human Research Ethics Committee (Melbourne). Specific ethics approval was not required for vaccine coverage analysis, as we conducted our study using deidentified ACIR data supplied by the Australian Government Department of Human Services for the purposes of program evaluation.
During the study analysis period, 1668 unique FS episodes were identified in 1471 children younger than 5 years. Of these children, 1335 (90.8%) had only 1 episode and 136 (9.2%) had 2 or more episodes separated by at least 7 days. The median age at the time of the first FS was 21 months (interquartile range [IQR], 14-31 months), similar to the median age at receipt of MMRV vaccine of 18 months (IQR, 18-19 months) and the peak age at FSs shown previously.19 After restriction to age 11 to 23 months and the recommended vaccine sequence, there were 465 FS episodes in 391 children. Ten children with 12 FSs (10 of 401 cases [2.4%]) were excluded because the recommended schedule was not followed (only 4 had MMRV as dose 1, which was an insufficient sample to analyze FS risk). Of the 391 children included, 278 (71.1%) had received MMR followed by MMRV vaccine, 97 (24.8%) had received MMR vaccine only, and 16 (4.1%) had received neither vaccine. Further data are provided in Table 2.
Table 3 provides the results of the primary and sensitivity self-controlled case series analyses. In the primary analysis, which adjusted for age using 3 age groups, there was no significantly increased risk of FSs within the 5- to 12-day risk period following MMRV, the prevaccination period, or the 13- to 30-day postvaccination period. The RI of FSs was raised in the 5 to 12 days following MMR vaccine (MCV dose 1) (RI, 2.71; 95% CI, 1.71-4.29), and there was a significantly lower risk in the 2 weeks before vaccination. The results of the sensitivity analyses were similar to those of the primary analyses.
As reported in Table 4, within 2.5 years following MMRV introduction, 2-dose MCV coverage increased to 93.8% at age 36 months, which exceeded the most recent preprogram historical coverage level of 92% at age 60 months (1 year after the previous age 48-month schedule point). Coverage with varicella-containing vaccine, consistently recommended at 18 months and assessed at age 24 months, increased by 4% after the change (Table 4). Overall, on-time immunization with the second MCV (defined as vaccine receipt within 30 days of the recommended age) improved by 13.5% (from 58.9% to 72.4%) (Figure). During this time, there was virtually no change in the coverage of MMR dose 1 (recommended at age 12 months and measured at age 18 months), which increased by only 0.5% (Table 4).
We present a comprehensive evaluation of the effect of 2 simultaneous changes to the Australian NIP that are relevant to child immunization programs worldwide: introduction of combination MMRV vaccine and bringing forward the scheduled age for provision of the second MCV. Our evaluation demonstrates that MMRV vaccine introduction in Australia has been associated with improved coverage and timeliness of protection against all 4 diseases, with on-time vaccination increasing by 13.5%. This change has been done with no effect on the overall safety profile of the program; use of MMRV vaccine as dose 2 of MCV at the age of 18 months was not associated with an increased risk of FSs.
Global efforts to control measles rely on achieving and maintaining high 2-dose vaccine coverage (preferably >95%) at a country and subnational (district) level.25 Australia was 1 of the first 4 countries in the World Health Organization Western Pacific Region to reach measles elimination status, declared in March 2014.12 However, before 2014 and despite an overall reduction in the incidence of measles, notification rates were highest in infants and children aged 1 to 4 years, outbreaks often involved young children, and measles vaccine coverage for 1 and 2 vaccine doses was suboptimal at 92% nationally.11 This figure also masked small areas of lower coverage and lack of timely uptake.26 Within 2.5 years of implementing our compressed schedule at ages 12 and 18 months, we have demonstrated that more children were fully protected against measles at an earlier age. In the United States, which has recently experienced a resurgence of measles,25 MCV dose 2 is recommended at ages 4 to 6 years, and uptake of dose 1 at age 12 months could be more timely. The 2014 US National Immunization Survey estimated that 92% of children aged 19 to 35 months had received 1 dose of MMR vaccine (range, 84%-97%).27
Each country needs to assess its own unique disease epidemiology, immunization program characteristics, and barriers to vaccine uptake to determine the optimal timing of MCV doses. However, for children in whom vaccination is delayed whether due to missed opportunities, access issues, or vaccine hesitancy, earlier scheduled measles vaccination offers more opportunities to provide catch-up vaccination, particularly before school entry. One potential downside of earlier second-dose vaccination is the potential for waning immunity. Modeling the effect of this schedule change on population immunity to measles in Australia was sensitive to assumptions regarding the extent of waning of vaccine-derived immunity.28 Waning immunity may also be an issue for the less-efficacious mumps component of the vaccine; ongoing disease surveillance will be important to monitor for this potential outcome and, if needed, adjust policy recommendations accordingly. However, several European countries, Canadian provinces, and low- to middle-income countries under the Expanded Program on Immunization use a similarly compressed MCV schedule.
Australia has had a 1-dose varicella vaccination program for children aged 18 months since late 2005.13,16 We show that MMRV introduction has rapidly been associated with improvements in the absolute level and timeliness of coverage against varicella over that achieved with the single-antigen vaccine. While our study design cannot confirm a direct cause-and-effect relationship, reasons for this increase in coverage may include (1) reduced prior attendance at the 18-month schedule point due to parental (and clinician) perceptions of varicella as a mild disease for which an appointment for the immunization was not considered sufficiently important (no other vaccine was recommended at this schedule point between 2003 and 2016); (2) increased encouragement for children to attend the 18-month immunization visit due to the inclusion of other antigens, particularly measles, in the vaccine; and/or (3) reduction of the overall number of scheduled injections, which was appealing to caregivers and clinicians. Although a routine 2-dose varicella immunization schedule, as adopted in the United States in 2007, would offer improved protection against varicella, the addition of a second varicella dose was previously rejected for NIP inclusion in Australia on the basis of inadequate incremental cost-effectiveness.16
To our knowledge, this study is the first to demonstrate that administration of MMRV vaccine as dose 2 of the MCV at age 18 months is not associated with an increased risk of FSs despite peak FS incidence at this age.29 A US assessment showed no increased FS risk when MMRV dose 2 was given at age 4 to 6 years30; however, overall FS incidence is much lower in that age group. These results are also consistent with those of the previous PAEDS study showing no increased risk of FSs after monovalent varicella vaccine at age 18 months and confirming the well-described fold increase in FSs after MCV dose 1.19,29 Six postmarketing studies6,7,31-34 and a meta-analysis35 have consistently shown a 2-fold increase in FS risk in the risk window of approximately 5 to 12 days after MMRV dose 1 in toddlers compared with giving MMR or MMR and the varicella vaccine separately. Although this finding equates to a relatively low absolute excess of 4.3 FSs per 10 000 doses,7 even a low risk has been viewed as concerning. In Australia, an unexpected high rate of FSs occurred from 1 seasonal influenza vaccine brand (Fluvax [Afluria in the United States]) in children younger than 5 years in 2010. Although most FS cases resolved without sequelae, permanent neurologic damage occurred in 2 children.36,37 In the United States, primary care clinicians are reported as being unlikely to recommend MMRV,38 and in Germany, there has been a decline in varicella vaccination uptake.9
This study has a number of limitations and strengths. Although we showed no statistically significant association between FSs and MMRV vaccine, the point estimate was above 1 and CIs were wide (RI, 1.08; 95% CI, 0.55-2.13), thereby not excluding a very low level of risk. In addition, FS case capture was taken from sentinel pediatric hospital surveillance and may not be representative of all Australian children with FSs. However, each site also functions as a community-based hospital, most children resided nearby and had simple FSs, and our analysis was robust in demonstrating the known association with MCV dose 1. Our ecologic cohort design to assess vaccine coverage changes is subject to unrecognized biases or confounding factors and does not prove that the schedule change was the necessary or only causative factor in improving vaccine uptake. However, no other major programmatic or procedural changes occurred during the study period that would otherwise have increased coverage, and, notably, MCV dose 1 uptake did not change substantially over time. In Australia, all NIP vaccines are commonwealth government procured, and our 8 state and territory health departments undertake oversight program delivery, resulting in more prescriptive use of vaccine combinations and brands than in other countries where vaccine choice is influenced by individual immunization clinicians or insurers. Together with our comprehensive national vaccine register, this control enables accurate evaluation of changes in coverage in response to new vaccine introduction. Although our study primarily reports on 1 brand of MMRV vaccine (Priorix-Tetra), based on first principles, we believe that these results would likely be similar for the other registered MMRV vaccine (ProQuad).
Data from this study help clinicians to better understand the link between measles and varicella virus–containing live vaccines and the risk of FSs—a common but serious early childhood condition that occurs in response to fever from any source. We present a comprehensive evaluation of the incorporation of MMRV vaccine into the Australian NIP, demonstrating an association with improved vaccine uptake and timeliness while maintaining overall program safety. Our findings should help to inform childhood immunization policy decision making regarding use of these vaccines in other countries.
Accepted for Publication: May 10, 2017.
Corresponding Author: Kristine Macartney, MD, National Centre for Immunisation Research and Surveillance, The Children's Hospital at Westmead, Locked Bag 4001, Westmead, NSW 2145, Australia (firstname.lastname@example.org)
Published Online: August 14, 2017. doi:10.1001/jamapediatrics.2017.1965
Author Contributions: Drs Macartney and Wood had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Macartney, Gidding, Dey, Richmond, Gold, Crawford, McIntyre, Wood.
Acquisition, analysis, or interpretation of data: Macartney, Gidding, Trinh, Wang, Dey, Hull, Orr, McRae, Richmond, Crawford, Kynaston, McIntyre.
Drafting of the manuscript: Macartney, Gidding, Trinh, Dey, Richmond, Gold, McIntyre.
Critical revision of the manuscript for important intellectual content: Macartney, Gidding, Wang, Dey, Hull, Orr, McRae, Richmond, Crawford, Kynaston, McIntyre, Wood.
Statistical analysis: Macartney, Gidding, Trinh, Wang, Dey, Hull.
Obtained funding: Macartney, Wood.
Administrative, technical, or material support: Macartney, Dey, Orr, McRae, Richmond.
Study supervision: Macartney, Gidding, Richmond, Crawford.
Conflict of Interest Disclosures: None reported.
Funding/Support: Drs Gidding and Wood are supported by Australian National Health and Medical Research Council Career Development Fellowships. Funding from the Australian Government Department of Health and the NHMRC Project Grant ID number 1049557 supported the conduct of the study.
Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Group Information: Paediatric Active Enhanced Disease Surveillance (PAEDS) Network members include Kristine Macartney, MD, Karen Orr, RN, Jocelynne McRae, MPH, Peter McIntyre, PhD, Nicholas Wood, PhD, and Laura Rost, RN (National Centre for Immunisation Research and Surveillance); Peter Richmond, MBBS, and Christopher Blyth, PhD (School of Paediatrics and Child Health, University of Western Australia); Michael Gold, MB, CHB (Department of Paediatrics, University of Adelaide); Nigel Crawford, PhD (Royal Children's Hospital); Jennifer A. Kynaston, MBBS, Julia E. Clark, BM, BS, and Sonia Dougherty, RN (Lady Cilento Children's Hospital); Robert Booy, PhD, and Elizabeth Elliott, PhD (School of Child and Adolescent Health, University of Sydney); Jim Buttery, MD (School of Public Health and Preventive Medicine, Monash University); Helen Marshall, PhD (Robinson Research Institute, The University of Adelaide); Thomas Snelling, PhD (Wesfarmers Centre for Vaccines and Infectious Diseases, Telethon Kids Institute, University of Western Australia); Michael Nissen, PhD (previously Royal Children's Hospital, Brisbane); Alissa McMinn, RN, and Donna Lee, RN (Monash Children's Hospital); Carolyn Finucane, RN, Christine Robins, RN, Carol Orr, RN, and Jacki Connell, RN (Princess Margaret Hospital); Christine Heath, RN, and Mary Walker, RN (Women's and Children's Hospital); Sharon Tan, RN, Helen Knight, RN, and Jennifer Murphy, RN (The Children's Hospital at Westmead).
Additional Contributions: Alexandra Hendry, PhD, of the National Centre for Immunisation Research and Surveillance, assisted with manuscript preparation. No compensation was received.
Create a personal account or sign in to: