MAJOR ARTICLE
Influenza A/Subtype and B/Lineage Effectiveness Estimates for the 2011–2012 Trivalent Vaccine: Cross-Season and Cross-Lineage Protection With Unchanged Vaccine
1
British Columbia Centre for Disease Control, Vancouver, 2University of British Columbia, Vancouver, British Columbia, 3Institut National De Santé Publique Du Québec, Québec, 4Public Health Ontario, Toronto, 5University of Toronto, Ontario, 6University of Calgary, Alberta, 7Alberta Provincial Laboratory, Calgary, Alberta; 8National Microbiology Laboratory, Winnipeg, 9University of Manitoba, Winnipeg, 10Winnipeg Regional Health Authority, Winnipeg and 11 Cadham Provincial Laboratory, Winnipeg, Manitoba, Canada
Background. We estimate vaccine effectiveness (VE) against both influenza A/subtypes and B/lineages in Canada for the 2011–2012 trivalent inactivated influenza vaccine (TIV) with components entirely unchanged from the 2010–2011 TIV and in the context of phenotypic and genotypic characterization of circulating viruses. Methods. In a test-negative case-control study VE was estimated as [1-adjustedOddsRatio] × 100 for RT-PCRconfirmed influenza in vaccinated vs nonvaccinated participants. Viruses were characterized by hemagglutination inhibition (HI) and sequencing of antigenic sites of the hemagglutinin (HA) gene. Results. There were 1507 participants. VE against A(H1N1)pdm09 was 80% (95% confidence interval [CI], 52%– 92%): circulating viruses were HI-characterized as vaccine-matched and bore just 2 aminoacid (AA) differences from vaccine. VE against A/H3N2 was 51% (95% CI, 10%–73%): circulating viruses were HI-characterized as vaccinerelated but bore ≥11AA differences from vaccine. VE against influenza B was 51% (95% CI, 26%–67%) in total: 71% (95% CI, 40%–86%) for lineage-matched B/Victoria and 27% (95% CI, −21% to 56%) for lineagemismatched B/Yamagata. For both influenza A and B types, VE was similar among recipients of either 2010–2011 or 2011–2012 TIV alone, higher when vaccinated both seasons. Conclusions. Phenotypic and genotypic characterization of circulating and vaccine viruses enhances understanding of TIV performance, shown in 2011–2012 to be substantial against well-conserved A(H1N1)pdm09 and lineagematched influenza B, suboptimal against genetic-variants of A/H3N2, and further reduced against lineage-mismatched influenza B. With unchanged vaccine components, protection may extend beyond a single season. Keywords. influenza; sentinel surveillance; vaccines and immunisation; trivalent influenza vaccine; vaccine effectiveness; influenza A subtype; influenza B lineage.
Since 2004, effectiveness of the annually reformulated trivalent inactivated influenza vaccine (TIV) against
Received 3 November 2013; accepted 23 December 2013; electronically published 19 January 2014. Presentations at meetings: Results from this study have been presented in part at: 10th Canadian Immunization Conference, Vancouver, BC, Canada, 3–5 December 2012. Correspondence: Danuta M. Skowronski, MD, 655 West 12th Ave, Vancouver, British Columbia V5Z 4R4, Canada ([email protected]). The Journal of Infectious Diseases 2014;210:126–37 © The Author 2014. Published by Oxford University Press on behalf of the Infectious Diseases Society of America. All rights reserved. For Permissions, please e-mail: [email protected]. DOI: 10.1093/infdis/jiu048
126
•
JID 2014:210 (1 July)
•
Skowronski et al
medically-attended, laboratory-confirmed influenza has been estimated in Canada through a test-negative case-control design [1–7]. Component-specific vaccine effectiveness (VE) estimation is embedded within the routine sentinel surveillance network, enabling direct linkage of clinical data with detailed phenotypic and genotypic characterization of contributing, circulating viruses for context. During the 2010–2011 season, this network identified suboptimal VE against the A/H3N2 subtype, which composed 60% of all influenza detections that season [7]. Suboptimal VE was associated with
Downloaded from http://jid.oxfordjournals.org/ at University of South Dakota and School of Medicine on May 8, 2015
Danuta M. Skowronski,1,2 Naveed Z. Janjua,1,2 Suzana Sabaiduc,1 Gaston De Serres,3 Anne-Luise Winter,4 Jonathan B. Gubbay,4,5 James A. Dickinson,6 Kevin Fonseca,6,7 Hugues Charest,3 Nathalie Bastien,8 Yan Li,8 Trijntje L. Kwindt,2 Salaheddin M. Mahmud,9,10 Paul Van Caeseele,9,11 Mel Krajden,1,2 and Martin Petric1,2
METHODS Epidemiologic
Several hundred community-based practitioners contribute to annual VE estimation in Canada from the 5 most populous provinces, including British Columbia (BC), Alberta, Manitoba (newly joined 2011–2012), Ontario, and Quebec. As previously described [1–7], practitioners at designated sentinel sites may collect nasal/nasopharyngeal swabs from all patients presenting within 7 days of influenza-like illness (ILI) onset. ILI is defined as the acute onset of fever and cough and ≥1 of sore throat, arthralgia, myalgia, or prostration; for 2011–2012, fever was not required in the elderly ≥65 years old. At the time of specimen collection, the attending practitioner obtains epidemiologic information from consenting patients/
parents/guardians using a standardized questionnaire affixed to the laboratory requisition. Information includes date of symptom onset, current immunization status, and month/year of vaccine receipt, as well as influenza immunization during the prior 2 seasons and during the fall 2009 pandemic immunization campaign. Details related to special pediatric immunization dosing requirements are not sought. Comorbidity is recorded on the questionnaire as ‘yes’/‘no’/‘unknown’ to any of the chronic medical conditions defined by Canada’s National Advisory Committee on Immunization as increasing the risk of influenza complications, without specifying the condition [8]. Ethics review boards in each participating province approved this study. Immunization
Immunized participants primarily receive vaccine during the regular autumn immunization campaign. Influenza vaccine is provided free-of-charge to all citizens ≥6 months old in Alberta, Ontario, and Manitoba. In BC and Quebec, vaccine is provided free-of-charge to individuals in designated high-risk groups and their close contacts [8]. Others are encouraged to receive vaccine but must purchase it. For the 2011–2012 season, live attenuated vaccine was available for those 2–59 years old but was not publicly funded in participating provinces. Similarly, an adjuvanted-subunit TIV was also approved for use in the elderly but only a few jurisdictions in Ontario and BC funded its use, primarily for long-term-care residents. Per WHO recommendation, 2011–2012 TIV components were unchanged from 2010–2011 and included A/California/ 7/2009(H1N1)-like (hereafter A(H1N1)pdm09), A/Perth/16/ 2009(H3N2)-like and B/Brisbane/60/2008(Victoria-lineage) [8, 9]. As in 2010–2011, manufacturers substituted A/California/7/2009(H1N1) NYMC-X-179A or -X-181 and A/Victoria/ 210/2009(H3N2)-NYMC-X-187 as considered to be antigenically equivalent strains. Laboratory
Specimens were tested for influenza at provincial reference laboratories by real-time reverse-transcription polymerasechain reaction (RT-PCR) according to standard provincial procedures. All RT-PCR positive specimens were inoculated into cell culture for virus isolation. An aliquot of successfully cultivated virus was then submitted to the National Microbiology Laboratory (Canada’s influenza reference laboratory) for HI characterization, performed using post-infection ferret antisera against WHO-recommended reference strains (eg, A/ Perth/16/2009(H3N2)) and turkey erythrocytes; reduced titers for H3N2 were confirmed with guinea pig erythrocytes [10]. Isolates were identified as antigenically most-similar to the prototype reference strain according to the reciprocal of the highest HI titer. Previously a ≥4-fold reduction in HI titer was considered signal of potential antigenic drift; this has more recently been revised to a ≥8-fold reduction [10]. 2011–2012 Influenza Vaccine Effectiveness
•
JID 2014:210 (1 July)
•
127
Downloaded from http://jid.oxfordjournals.org/ at University of South Dakota and School of Medicine on May 8, 2015
divergent mutations in both the H3N2 vaccine component and circulating H3N2 viruses and with surveillance reports of increased long-term care facility outbreaks involving highly vaccinated residents and healthcare workers [7]. The vaccine component recommended by the World Health Organization (WHO) for the 2010–2011 TIV was A/Perth/16/2009, but the egg-passaged reassortant strain upon which manufacturers based the vaccine component was A/Victoria/210/2009 [NYMC-X-187] (hereafter “X-187”). X-187 was considered antigenically equivalent to A/Perth/16/ 2009 but later shown on genetic analysis to differ by 6 amino acids (AA) at antigenic sites [7]. Although conventional hemagglutination inhibition (HI) assay also characterized all H3N2 isolates as well matched to A/Perth/16/2009, phylogenetic analysis revealed 2 variants circulating through the sentinel network, represented by the clade 5 A/HongKong/2121/2010-like strain (comprising approximately 90% of sequenced H3N2), and the clade 7 A/Victoria/208/2009-like strain (comprising 10%) [7]. These circulating H3N2 variants differed from X-187 by 12AA and 8AA mutations, respectively, at antigenic sites resulting in reduced pairwise AA identity with vaccine of 90.8% and 93.9%, respectively. Consistent with genomic findings of vaccine-virus divergence, epidemiologic findings showed suboptimal VE of approximately 40% for the 2010–2011 H3N2 component. All 3 TIV strains recommended in 2010–2011 by the WHO remained unchanged in 2011–2012 [8]. The 2011–2012 season was further noteworthy as both influenza A/subtypes (H1N1, H3N2) and B/lineages (Victoria, Yamagata) cocirculated in Canada. Here we report VE estimates against both A/subtypes and B/lineages for a single season’s trivalent vaccine, providing detailed phenotypic and genotypic characterization of directly contributing circulating viruses in relation to corresponding vaccine components. Given shared cross-season vaccine components we also report prior 2010–2011 TIV effects on 2011–2012 protection and for A(H1N1)pdm09, assess residual effects of 2009 monovalent AS03-adjuvanted pandemic vaccine receipt.
A subset of sentinel A(H1N1)pdm09 HA1/HA2 and H3N2 HA1 genes from viruses detected across the season and contributing to VE analysis were sequenced for phylogenetic and pair-wise AA identity based on antigenic maps for H1 (50AA residues across antigenic sites Cb/Sa/Ca2/Ca1/Sb) and H3 (131AA residues across antigenic sites A-E; Supplementary 1). Influenza B viruses that could not be characterized at the lineage-level by HI were identified through phylogenetic analysis (Ontario/Alberta) or by an influenza B-lineage-specific 1-step conventional RT-PCR assay (BC/Manitoba/Quebec) [11]. Because antigenic maps for influenza B have not yet been established, its further genotypic analysis was not undertaken.
Specimens collected from sentinel sites between November 1 (week 44, 2011) and April 30 (week 18, 2012) are eligible for inclusion in primary VE analysis each year with exclusion criteria as shown in Figure 1 [1–5, 7]. A specimen testing positive for influenza is considered a case and a specimen testing negative for all influenza types/subtypes is considered a control
RESULTS Participant Profile
There were 1507 participants included in 2011–2012 VE analysis (Figure 1). As previously, adults 20–49 years old
Figure 1. Specimen exclusion for influenza vaccine effectiveness analysis, Canada, 2011–2012 sentinel surveillance system. Exclusions shown here in stepwise fashion to arrive at total case and control tally (ie, those meeting multiple exclusion criteria are counted on the basis of the first exclusion criterion met in the list shown). Abbreviation: PCR, polymerase chain reaction. 128
•
JID 2014:210 (1 July)
•
Skowronski et al
Downloaded from http://jid.oxfordjournals.org/ at University of South Dakota and School of Medicine on May 8, 2015
VE Analysis
[1–7]. The odds ratio (OR) for medically attended, laboratoryconfirmed influenza in vaccinated vs nonvaccinated participants is estimated by logistic regression with adjustment for clinically relevant confounders including age, comorbidity, province, week of specimen collection, and interval between ILI onset and specimen collection [1–7]. VE is calculated as [1-adjustedOR] × 100. In sensitivity analyses, 2011–2012 VE estimates were assessed with restriction to reflect delayed influenza circulation (January– May, 2012), specimen collection interval ≤4 days, those without comorbidity, and by age. The effects of prior 2010–2011 TIV and 2009 monovalent-adjuvanted vaccine receipt, alone and/ or in combination with 2011–2012 TIV were explored through indicator-variable analysis.
contributed most (702/1507; 47%) and the elderly least (130/ 1507; 9%) (Table 1) [1–5, 7]. Overall, the proportion of participants who had received vaccine was 15% (70/454) among cases and 30% (334/1096) among controls [4, 5, 7]. Most participants immunized in 2011–2012 also reported receiving the prior season’s 2010–2011 TIV (293/367; 80%). Receipt of 2009 monovalent pandemic vaccine was reported by 42% (596/1427) overall and among 72% (267/368) of those immunized in 2011–2012.
Table 1.
After excluding those immunized
Influenza A/Subtype and B/Lineage Effectiveness Estimates for the 2011–2012 Trivalent Vaccine: Cross-Season and Cross-Lineage Protection With Unchanged Vaccine
1
British Columbia Centre for Disease Control, Vancouver, 2University of British Columbia, Vancouver, British Columbia, 3Institut National De Santé Publique Du Québec, Québec, 4Public Health Ontario, Toronto, 5University of Toronto, Ontario, 6University of Calgary, Alberta, 7Alberta Provincial Laboratory, Calgary, Alberta; 8National Microbiology Laboratory, Winnipeg, 9University of Manitoba, Winnipeg, 10Winnipeg Regional Health Authority, Winnipeg and 11 Cadham Provincial Laboratory, Winnipeg, Manitoba, Canada
Background. We estimate vaccine effectiveness (VE) against both influenza A/subtypes and B/lineages in Canada for the 2011–2012 trivalent inactivated influenza vaccine (TIV) with components entirely unchanged from the 2010–2011 TIV and in the context of phenotypic and genotypic characterization of circulating viruses. Methods. In a test-negative case-control study VE was estimated as [1-adjustedOddsRatio] × 100 for RT-PCRconfirmed influenza in vaccinated vs nonvaccinated participants. Viruses were characterized by hemagglutination inhibition (HI) and sequencing of antigenic sites of the hemagglutinin (HA) gene. Results. There were 1507 participants. VE against A(H1N1)pdm09 was 80% (95% confidence interval [CI], 52%– 92%): circulating viruses were HI-characterized as vaccine-matched and bore just 2 aminoacid (AA) differences from vaccine. VE against A/H3N2 was 51% (95% CI, 10%–73%): circulating viruses were HI-characterized as vaccinerelated but bore ≥11AA differences from vaccine. VE against influenza B was 51% (95% CI, 26%–67%) in total: 71% (95% CI, 40%–86%) for lineage-matched B/Victoria and 27% (95% CI, −21% to 56%) for lineagemismatched B/Yamagata. For both influenza A and B types, VE was similar among recipients of either 2010–2011 or 2011–2012 TIV alone, higher when vaccinated both seasons. Conclusions. Phenotypic and genotypic characterization of circulating and vaccine viruses enhances understanding of TIV performance, shown in 2011–2012 to be substantial against well-conserved A(H1N1)pdm09 and lineagematched influenza B, suboptimal against genetic-variants of A/H3N2, and further reduced against lineage-mismatched influenza B. With unchanged vaccine components, protection may extend beyond a single season. Keywords. influenza; sentinel surveillance; vaccines and immunisation; trivalent influenza vaccine; vaccine effectiveness; influenza A subtype; influenza B lineage.
Since 2004, effectiveness of the annually reformulated trivalent inactivated influenza vaccine (TIV) against
Received 3 November 2013; accepted 23 December 2013; electronically published 19 January 2014. Presentations at meetings: Results from this study have been presented in part at: 10th Canadian Immunization Conference, Vancouver, BC, Canada, 3–5 December 2012. Correspondence: Danuta M. Skowronski, MD, 655 West 12th Ave, Vancouver, British Columbia V5Z 4R4, Canada ([email protected]). The Journal of Infectious Diseases 2014;210:126–37 © The Author 2014. Published by Oxford University Press on behalf of the Infectious Diseases Society of America. All rights reserved. For Permissions, please e-mail: [email protected]. DOI: 10.1093/infdis/jiu048
126
•
JID 2014:210 (1 July)
•
Skowronski et al
medically-attended, laboratory-confirmed influenza has been estimated in Canada through a test-negative case-control design [1–7]. Component-specific vaccine effectiveness (VE) estimation is embedded within the routine sentinel surveillance network, enabling direct linkage of clinical data with detailed phenotypic and genotypic characterization of contributing, circulating viruses for context. During the 2010–2011 season, this network identified suboptimal VE against the A/H3N2 subtype, which composed 60% of all influenza detections that season [7]. Suboptimal VE was associated with
Downloaded from http://jid.oxfordjournals.org/ at University of South Dakota and School of Medicine on May 8, 2015
Danuta M. Skowronski,1,2 Naveed Z. Janjua,1,2 Suzana Sabaiduc,1 Gaston De Serres,3 Anne-Luise Winter,4 Jonathan B. Gubbay,4,5 James A. Dickinson,6 Kevin Fonseca,6,7 Hugues Charest,3 Nathalie Bastien,8 Yan Li,8 Trijntje L. Kwindt,2 Salaheddin M. Mahmud,9,10 Paul Van Caeseele,9,11 Mel Krajden,1,2 and Martin Petric1,2
METHODS Epidemiologic
Several hundred community-based practitioners contribute to annual VE estimation in Canada from the 5 most populous provinces, including British Columbia (BC), Alberta, Manitoba (newly joined 2011–2012), Ontario, and Quebec. As previously described [1–7], practitioners at designated sentinel sites may collect nasal/nasopharyngeal swabs from all patients presenting within 7 days of influenza-like illness (ILI) onset. ILI is defined as the acute onset of fever and cough and ≥1 of sore throat, arthralgia, myalgia, or prostration; for 2011–2012, fever was not required in the elderly ≥65 years old. At the time of specimen collection, the attending practitioner obtains epidemiologic information from consenting patients/
parents/guardians using a standardized questionnaire affixed to the laboratory requisition. Information includes date of symptom onset, current immunization status, and month/year of vaccine receipt, as well as influenza immunization during the prior 2 seasons and during the fall 2009 pandemic immunization campaign. Details related to special pediatric immunization dosing requirements are not sought. Comorbidity is recorded on the questionnaire as ‘yes’/‘no’/‘unknown’ to any of the chronic medical conditions defined by Canada’s National Advisory Committee on Immunization as increasing the risk of influenza complications, without specifying the condition [8]. Ethics review boards in each participating province approved this study. Immunization
Immunized participants primarily receive vaccine during the regular autumn immunization campaign. Influenza vaccine is provided free-of-charge to all citizens ≥6 months old in Alberta, Ontario, and Manitoba. In BC and Quebec, vaccine is provided free-of-charge to individuals in designated high-risk groups and their close contacts [8]. Others are encouraged to receive vaccine but must purchase it. For the 2011–2012 season, live attenuated vaccine was available for those 2–59 years old but was not publicly funded in participating provinces. Similarly, an adjuvanted-subunit TIV was also approved for use in the elderly but only a few jurisdictions in Ontario and BC funded its use, primarily for long-term-care residents. Per WHO recommendation, 2011–2012 TIV components were unchanged from 2010–2011 and included A/California/ 7/2009(H1N1)-like (hereafter A(H1N1)pdm09), A/Perth/16/ 2009(H3N2)-like and B/Brisbane/60/2008(Victoria-lineage) [8, 9]. As in 2010–2011, manufacturers substituted A/California/7/2009(H1N1) NYMC-X-179A or -X-181 and A/Victoria/ 210/2009(H3N2)-NYMC-X-187 as considered to be antigenically equivalent strains. Laboratory
Specimens were tested for influenza at provincial reference laboratories by real-time reverse-transcription polymerasechain reaction (RT-PCR) according to standard provincial procedures. All RT-PCR positive specimens were inoculated into cell culture for virus isolation. An aliquot of successfully cultivated virus was then submitted to the National Microbiology Laboratory (Canada’s influenza reference laboratory) for HI characterization, performed using post-infection ferret antisera against WHO-recommended reference strains (eg, A/ Perth/16/2009(H3N2)) and turkey erythrocytes; reduced titers for H3N2 were confirmed with guinea pig erythrocytes [10]. Isolates were identified as antigenically most-similar to the prototype reference strain according to the reciprocal of the highest HI titer. Previously a ≥4-fold reduction in HI titer was considered signal of potential antigenic drift; this has more recently been revised to a ≥8-fold reduction [10]. 2011–2012 Influenza Vaccine Effectiveness
•
JID 2014:210 (1 July)
•
127
Downloaded from http://jid.oxfordjournals.org/ at University of South Dakota and School of Medicine on May 8, 2015
divergent mutations in both the H3N2 vaccine component and circulating H3N2 viruses and with surveillance reports of increased long-term care facility outbreaks involving highly vaccinated residents and healthcare workers [7]. The vaccine component recommended by the World Health Organization (WHO) for the 2010–2011 TIV was A/Perth/16/2009, but the egg-passaged reassortant strain upon which manufacturers based the vaccine component was A/Victoria/210/2009 [NYMC-X-187] (hereafter “X-187”). X-187 was considered antigenically equivalent to A/Perth/16/ 2009 but later shown on genetic analysis to differ by 6 amino acids (AA) at antigenic sites [7]. Although conventional hemagglutination inhibition (HI) assay also characterized all H3N2 isolates as well matched to A/Perth/16/2009, phylogenetic analysis revealed 2 variants circulating through the sentinel network, represented by the clade 5 A/HongKong/2121/2010-like strain (comprising approximately 90% of sequenced H3N2), and the clade 7 A/Victoria/208/2009-like strain (comprising 10%) [7]. These circulating H3N2 variants differed from X-187 by 12AA and 8AA mutations, respectively, at antigenic sites resulting in reduced pairwise AA identity with vaccine of 90.8% and 93.9%, respectively. Consistent with genomic findings of vaccine-virus divergence, epidemiologic findings showed suboptimal VE of approximately 40% for the 2010–2011 H3N2 component. All 3 TIV strains recommended in 2010–2011 by the WHO remained unchanged in 2011–2012 [8]. The 2011–2012 season was further noteworthy as both influenza A/subtypes (H1N1, H3N2) and B/lineages (Victoria, Yamagata) cocirculated in Canada. Here we report VE estimates against both A/subtypes and B/lineages for a single season’s trivalent vaccine, providing detailed phenotypic and genotypic characterization of directly contributing circulating viruses in relation to corresponding vaccine components. Given shared cross-season vaccine components we also report prior 2010–2011 TIV effects on 2011–2012 protection and for A(H1N1)pdm09, assess residual effects of 2009 monovalent AS03-adjuvanted pandemic vaccine receipt.
A subset of sentinel A(H1N1)pdm09 HA1/HA2 and H3N2 HA1 genes from viruses detected across the season and contributing to VE analysis were sequenced for phylogenetic and pair-wise AA identity based on antigenic maps for H1 (50AA residues across antigenic sites Cb/Sa/Ca2/Ca1/Sb) and H3 (131AA residues across antigenic sites A-E; Supplementary 1). Influenza B viruses that could not be characterized at the lineage-level by HI were identified through phylogenetic analysis (Ontario/Alberta) or by an influenza B-lineage-specific 1-step conventional RT-PCR assay (BC/Manitoba/Quebec) [11]. Because antigenic maps for influenza B have not yet been established, its further genotypic analysis was not undertaken.
Specimens collected from sentinel sites between November 1 (week 44, 2011) and April 30 (week 18, 2012) are eligible for inclusion in primary VE analysis each year with exclusion criteria as shown in Figure 1 [1–5, 7]. A specimen testing positive for influenza is considered a case and a specimen testing negative for all influenza types/subtypes is considered a control
RESULTS Participant Profile
There were 1507 participants included in 2011–2012 VE analysis (Figure 1). As previously, adults 20–49 years old
Figure 1. Specimen exclusion for influenza vaccine effectiveness analysis, Canada, 2011–2012 sentinel surveillance system. Exclusions shown here in stepwise fashion to arrive at total case and control tally (ie, those meeting multiple exclusion criteria are counted on the basis of the first exclusion criterion met in the list shown). Abbreviation: PCR, polymerase chain reaction. 128
•
JID 2014:210 (1 July)
•
Skowronski et al
Downloaded from http://jid.oxfordjournals.org/ at University of South Dakota and School of Medicine on May 8, 2015
VE Analysis
[1–7]. The odds ratio (OR) for medically attended, laboratoryconfirmed influenza in vaccinated vs nonvaccinated participants is estimated by logistic regression with adjustment for clinically relevant confounders including age, comorbidity, province, week of specimen collection, and interval between ILI onset and specimen collection [1–7]. VE is calculated as [1-adjustedOR] × 100. In sensitivity analyses, 2011–2012 VE estimates were assessed with restriction to reflect delayed influenza circulation (January– May, 2012), specimen collection interval ≤4 days, those without comorbidity, and by age. The effects of prior 2010–2011 TIV and 2009 monovalent-adjuvanted vaccine receipt, alone and/ or in combination with 2011–2012 TIV were explored through indicator-variable analysis.
contributed most (702/1507; 47%) and the elderly least (130/ 1507; 9%) (Table 1) [1–5, 7]. Overall, the proportion of participants who had received vaccine was 15% (70/454) among cases and 30% (334/1096) among controls [4, 5, 7]. Most participants immunized in 2011–2012 also reported receiving the prior season’s 2010–2011 TIV (293/367; 80%). Receipt of 2009 monovalent pandemic vaccine was reported by 42% (596/1427) overall and among 72% (267/368) of those immunized in 2011–2012.
Table 1.
After excluding those immunized
