RESEARCH ARTICLE Open Access
The impact of repeated vaccination on influenza vaccine effectiveness: a systematic review and meta-analysis Lauren C. Ramsay1, Sarah A. Buchan2, Robert G. Stirling2,3, Benjamin J. Cowling4, Shuo Feng4, Jeffrey C. Kwong1,2,5,6,7 and Bryna F. Warshawsky1,8*
Background: Conflicting results regarding the impact of repeated vaccination on influenza vaccine effectiveness (VE) may cause confusion regarding the benefits of receiving the current season’s vaccine.
Methods: We systematically searched MEDLINE, Embase, PubMed, and Cumulative Index to Nursing and Allied Health Literature from database inception to August 17, 2016, for observational studies published in English that reported VE against laboratory-confirmed influenza for four vaccination groups, namely current season only, prior season only, both seasons, and neither season. We pooled differences in VE (ΔVE) between vaccination groups by influenza season and type/subtype using a random effects model. The study protocol is registered with PROSPERO (registration number: CRD42016037241).
Results: We identified 3435 unique articles, reviewed the full text of 634, and included 20 for meta-analysis. Compared to prior season vaccination only, vaccination in both seasons was associated with greater protection against influenza H1N1 (ΔVE = 26%; 95% CI, 15% to 36%) and B (ΔVE = 24%; 95% CI, 7% to 42%), but not H3N2 (ΔVE = 10%; 95% CI, –6% to 25%). Compared to no vaccination for either season, individuals who received the current season’s vaccine had greater protection against H1N1 (ΔVE = 61%; 95% CI, 50% to 70%), H3N2 (ΔVE = 41%; 95% CI, 33% to 48%), and B (ΔVE = 62%; 95% CI, 54% to 68%). We observed no differences in VE between vaccination in both seasons and the current season only for H1N1 (ΔVE = 4%; 95% CI, –7% to 15%), H3N2 (ΔVE = –12%; 95% CI, –27% to 4%), or B (ΔVE = –8%; 95% CI, –17% to 1%).
Conclusions: From the patient perspective, our results support current season vaccination regardless of prior season vaccination. We found no overall evidence that prior season vaccination negatively impacts current season VE. It is important that future VE studies include vaccination history over multiple seasons in order to evaluate repeated vaccination in more detail.
Keywords: Influenza, Vaccine effectiveness, Repeated vaccination
Background Seasonal influenza vaccination is the predominant strat- egy for preventing influenza-related morbidity and mor- tality. Annual vaccination is recommended because of waning immunity and because influenza strains undergo
antigenic drift, necessitating reviewing and, in most sea- sons, changing of the vaccine to better match the up- coming season’s strains . Because of the frequently changing vaccine, influenza vaccine effectiveness (VE) is assessed annually. With increasing numbers of people being immunized
against influenza annually, the impact of repeated vac- cination has gained significant interest. Of particular concern are older adults (65 years and above), who tend to have more comorbidities as they age, as both age and co-morbidities increase their risk of influenza-associated complications . If repeated vaccination negatively
* Correspondence: email@example.com Lauren C Ramsay and Sarah A Buchan are joint first authors. Jeffrey C Kwong and Bryna F Warshawsky are joint senior authors. 1Public Health Ontario, 480 University Avenue Suite 300, Toronto, Ontario M5G 1V2, Canada 8Department of Epidemiology and Biostatistics, Western University, 1151 Richmond St, London, Ontario N6A 3K7, Canada Full list of author information is available at the end of the article
© The Author(s). 2017 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Ramsay et al. BMC Medicine (2017) 15:159 DOI 10.1186/s12916-017-0919-0
impacts current VE, then having been repeatedly vacci- nated in earlier years may be detrimental to the protec- tion of older adults when they need it most. Studies from the 1970s and 1980s found inconsistent results re- garding the impact of repeated vaccination [3, 4]. In 1999, a systematic review and meta-analysis of field studies, trials, and serologic studies found no evidence of negative impacts of repeated vaccination . More recently, some studies have found VE to be reduced in those who received repeated prior influenza vaccina- tions [6–8]. Since most VE studies now report estimates taking
into account vaccination status for both current and prior seasons, we sought to evaluate the impact of repeated vaccination on VE through a systematic review and meta-analysis. We aimed to assess the impact of repeated vaccination to provide evidence to support patient and clinician decision-making about receiving the current season’s influenza vaccine. We considered two patient-relevant scenarios, (1) for those who received last season’s vaccine, should they also receive this season’s vaccine? (vaccination in both seasons versus prior season only) and (2) for those who did not receive last season’s vaccine, should they receive this season’s vaccine? (vaccination in current season only versus neither season). We also considered a policy-relevant scenario, comparing VE for vaccination in both seasons versus the current season only. This latter scenario is not relevant to patients because they cannot alter their vaccination history; however, the findings may influence policy decisions regarding whether or not to offer annual vaccination to the entire population if there was evidence suggesting that repeated vaccination could negatively impact future VE.
Methods Search strategy and selection criteria We searched MEDLINE, Embase, PubMed, and Cumu- lative Index to Nursing and Allied Health Literature (CINAHL) databases from inception to August 17, 2016. We developed a unique search strategy for each database with the assistance of a scientific librarian; across all databases, the search terms included “influenza”, “immunization”, “vaccine”, and “effectiveness”, and articles were restricted to those published in English (Additional file 1). Two reviewers (SB, LR) independ- ently screened titles and abstracts, and hand-searched the references of the included articles. Eligible studies used observational study designs (e.g.,
prospective cohort, test-negative case-control) and reported VE against medically attended, laboratory- confirmed influenza for four mutually exclusive vaccin- ation groups, namely current season only, prior season only, both current and prior seasons, and neither season
(reference group). Prior season vaccination referred pri- marily to vaccination status in the year immediately prior to the season being examined. Studies with other definitions of prior season (e.g., any dose in the prior two seasons) were excluded from the meta-analysis, but were described in a qualitative synthesis. We excluded interim VE reports that were superseded by end-of- season reports, and conference abstracts and proceed- ings. We followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines for reporting results .
Risk of bias assessment We used the Newcastle–Ottawa scale (NOS) to assess the risk of bias of included case-control and cohort stud- ies . Two reviewers (SB, LR) independently evaluated the quality of each study based on the domains of selec- tion, comparability, and either exposure (for case- control studies) or outcome (for cohort studies). For studies using the test-negative design, we determined whether calendar time had been included in the adjusted analyses . Studies were categorized as being at low, moderate, or high risk of bias if they were missing one or less items, two to three items, or more than three items on the NOS, respectively . Any disagreements between the two reviewers were resolved by consensus.
Data analysis Two reviewers (SB, LR) abstracted the data using a struc- tured electronic data extraction form, extracting study characteristics (e.g., study design, recruitment setting, case definition) and VE estimates for the four vaccination groups, with discrepancies adjudicated by consensus. Whenever possible, we extracted VE reports by influenza type/subtype and age group and only included the most specific results reported (e.g., by age group or influenza type/subtype) in the meta-analysis. Because specific lineage information for influenza B was often unavailable, we used overall estimates for influenza B. For each study included in the meta-analysis, VE esti-
mates for current season only, prior season only, and both current and prior seasons were assessed against the reference group who were not vaccinated in either season. In the present study, VE estimates from each study were compared for those vaccinated in both the current and prior seasons to those vaccinated in the prior season only and to those vaccinated in the current season only by subtracting the VE estimates. The abso- lute differences in VE (ΔVE) were stratified by influ- enza type/subtype and season and calculated as (1) vaccinated in both seasons compared to the prior sea- son only (ΔVE = VEboth – VEprior only), and (2) vacci- nated in both seasons compared to the current season only (ΔVE = VEboth – VEcurrent only). In both of the
Ramsay et al. BMC Medicine (2017) 15:159 Page 2 of 18
above scenarios, a ΔVE greater than zero implies a higher VE estimate when vaccinated in both seasons than in either the current or the prior season only. We also assessed the VE of those vaccinated in the current season only compared to those vaccinated in neither season (pooled VEcurrent only). We calculated confidence intervals for ΔVE by boot-
strapping using 1000 samples . Similar to previous work , we took 1000 samples from VEcurrent only, VEprior only, and VEboth. We then estimated 1000 mea- sures of ΔVE for both ΔVE = VEboth – VEcurrent only and ΔVE = VEboth – VEprior only; the 2.5% and 97.5% percen- tiles for ΔVE were computed as the confidence intervals. We used a random effects model to pool ΔVE estimates to compare the overall difference between vaccination in both seasons with vaccination in either the prior season only or the current season only. To compare VE for those vaccinated in the current season versus those
vaccinated in neither season, we use