Effectiveness of Seasonal Trivalent Influenza Vaccine for Preventing Influenza Virus Illness Among Pregnant Women: A Population-Based Case-Control Study During the 2010–2011 and 2011–2012 Influenza Seasons (original) (raw)

Journal Article

,

1

Influenza Division

,

Centers for Disease Control and Prevention (CDC)

,

Atlanta, Georgia

Correspondence: Mark G. Thompson, PhD, Influenza Division, Centers for Disease Control and Prevention, 1600 Clifton Rd NE, MS A-32, Atlanta, GA 30333 (isq8@cdc.gov).

Search for other works by this author on:

,

2

Division of Research

,

Kaiser Foundation Research Institute

,

Oakland, California

3

Department of Health Research and Policy, School of Medicine

,

Stanford University, California

Search for other works by this author on:

,

4

Abt Associates, Inc, Cambridge, Massachusetts

Search for other works by this author on:

,

1

Influenza Division

,

Centers for Disease Control and Prevention (CDC)

,

Atlanta, Georgia

5

Battelle Memorial Institute

,

Atlanta, Georgia

Search for other works by this author on:

,

2

Division of Research

,

Kaiser Foundation Research Institute

,

Oakland, California

Search for other works by this author on:

,

6

Center for Health Research

,

Kaiser Permanente Northwest

,

Portland, Oregon

Search for other works by this author on:

,

4

Abt Associates, Inc, Cambridge, Massachusetts

Search for other works by this author on:

,

1

Influenza Division

,

Centers for Disease Control and Prevention (CDC)

,

Atlanta, Georgia

Search for other works by this author on:

,

2

Division of Research

,

Kaiser Foundation Research Institute

,

Oakland, California

Search for other works by this author on:

,

6

Center for Health Research

,

Kaiser Permanente Northwest

,

Portland, Oregon

Search for other works by this author on:

... Show more

Accepted:

04 November 2013

Published:

26 November 2013

• PDF
• Split View
• • Article contents
  • Figures & tables
  • Video
  • Audio
  • Supplementary Data
• Cite

Cite

Mark G. Thompson, De-Kun Li, Pat Shifflett, Leslie Z. Sokolow, Jeannette R. Ferber, Samantha Kurosky, Sam Bozeman, Sue B. Reynolds, Roxana Odouli, Michelle L. Henninger, Tia L. Kauffman, Lyndsay A. Avalos, Sarah Ball, Jennifer L. Williams, Stephanie A. Irving, David K. Shay, Allison L. Naleway, for the Pregnancy and Influenza Project Workgroup, Susan Chu, Janet Cragan, Anne McIntyre, Julie Villanueva, Alicia Fry, Joe Bresee, Jerome Tokars, Jane Seward, Effectiveness of Seasonal Trivalent Influenza Vaccine for Preventing Influenza Virus Illness Among Pregnant Women: A Population-Based Case-Control Study During the 2010–2011 and 2011–2012 Influenza Seasons, Clinical Infectious Diseases, Volume 58, Issue 4, 15 February 2014, Pages 449–457, https://doi.org/10.1093/cid/cit750
Close

Navbar Search Filter Mobile Enter search term Search

Abstract

Background. Although vaccination with trivalent inactivated influenza vaccine (TIV) is recommended for all pregnant women, no vaccine effectiveness (VE) studies of TIV in pregnant women have assessed laboratory-confirmed influenza outcomes.

Methods. We conducted a case-control study over 2 influenza seasons (2010–2011 and 2011–2012) among Kaiser Permanente health plan members in 2 metropolitan areas in California and Oregon. We compared the proportion vaccinated among 100 influenza cases (confirmed by reverse transcription polymerase chain reaction) with the proportions vaccinated among 192 controls with acute respiratory illness (ARI) who tested negative for influenza and 200 controls without ARI (matched by season, site, and trimester).

Results. Among influenza cases, 42% were vaccinated during the study season compared to 58% and 63% vaccinated among influenza-negative controls and matched ARI-negative controls, respectively. The adjusted VE of the current season vaccine against influenza A and B was 44% (95% confidence interval [CI], 5%–67%) using the influenza-negative controls and 53% (95% CI, 24%–72%) using the ARI-negative controls. Receipt of the prior season's vaccine, however, had an effect similar to receipt of the current season's vaccine. As such, vaccination in either or both seasons had statistically similar adjusted VE using influenza-negative controls (VE point estimates range = 51%–76%) and ARI-negative controls (48%–76%).

Conclusions. Influenza vaccination reduced the risk of ARI associated with laboratory-confirmed influenza among pregnant women by about one-half, similar to VE observed among all adults during these seasons.

Because pregnant women appear to be vulnerable to severe disease and secondary complications from influenza [1], vaccination with trivalent inactivated influenza vaccine (TIV) is recommended for all pregnant women in the United States [2]. TIV coverage in this population has risen to about 50% in the United States [3], but no vaccine effectiveness (VE) studies of TIV among pregnant women have assessed laboratory-confirmed influenza outcomes [4, 5]. Although several studies have found a similar antibody response to the vaccine among pregnant and nonpregnant women [6, 7], previous studies have only compared rates of nonspecific respiratory illness among vaccinated and unvaccinated pregnant women, with mixed results [6, 8–11].

To address this gap in the literature, we estimated influenza vaccine effectiveness in preventing laboratory-confirmed influenza illness among pregnant women in 2 metropolitan areas during 2 influenza seasons (2010–2011 and 2011–2012). We compared the proportion of women who were vaccinated among influenza cases with the proportions vaccinated among controls with acute respiratory illness (ARI) who tested negative for influenza. This approach, often referred to as the test-negative design, is believed to minimize potential bias and confounding by the propensity to seek healthcare [12, 13] and has been employed in other studies of VE among adults during the same seasons [14–17]. We also compared the vaccination status of cases with another control group as an additional estimate of VE: women who did not have ARI matched by site, season, and trimester. Given previous reports of differences in VE depending on prior vaccine exposure [18], including a recent report from one of our study seasons [19], as a secondary objective we also examined whether VE among pregnant women varied depending on vaccination history during the previous season.1

Figure 1.

Number of pregnant women with febrile acute respiratory illness who tested positive (cases) or negative (controls) for influenza viruses by week of testing during the 2010–2011 and 2011–2012 seasons. The influenza season is defined by the illness onset date of the first and last case identified at each study site. For the 2010–2011 season, Kaiser Permanente Northern California (KPNC) surveillance (from 3 January 2011 to 24 April 2011) identified 32 influenza positives and 58 influenza negatives; Kaiser Permanente Northwest (KPNW) surveillance (from 12 December 2010 to 24 April 2011) identified 37 influenza positives and 53 influenza negatives. For the 2011–2012 season, KPNC surveillance (from 10 January 2012 to 18 May 2012) identified 21 influenza positives and 57 influenza negatives; KPNW surveillance (from 22 January 2012 to 1 May 2012) identified 10 influenza positives and 24 influenza negatives.

METHODS

Study Design

We have previously published the detailed methods of our population-based case-control study, the Pregnancy and Influenza Project (PIP) [20]. In brief, the study took place at 2 study sites during the 2010–2011 and 2011–2012 influenza seasons. Participants were members of Kaiser Permanente who had at least 1 prenatal visit in the Northwest (KPNW) region (Portland, Oregon, metropolitan area) or the Northern California (KPNC) region (San Francisco Bay Area). Participants were offered small incentives in the form of gift cards. The institutional review boards at both sites approved study instruments, procedures, and written consent documents.

Participant characteristics were assessed as part of an enrollment interview [20]. We examined 3 measures of health status. Self-rated health was assessed with a standard 5-level rating of overall health from poor (1) to excellent (5) [21]. We ascertained the presence of a high-risk medical condition (associated with increased risk of influenza complications [22]; codes available from the authors) during the year prior to conception and pregnancy complications (using a subset of the International Classification of Diseases, Ninth Revision, Clinical Modification [_ICD-9 CM_] codes 640–649 related to adverse pregnancy outcomes) from medical records.

Influenza Vaccination

We obtained influenza vaccination status during the study year and the prior year from medical records and self-report (if the participant was vaccinated outside the health plan). Vaccine components for the study years and prior seasons are listed in Supplementary Table A; the vast majority of tested circulating viruses in the United States were antigenically similar to vaccine components during both seasons [23, 24]. Current season vaccination was defined by receipt of seasonal influenza vaccine ≥14 days before illness onset; for matched controls, the illness onset date for the corresponding case was used.

Surveillance

During both study seasons, we identified potential ARIs using daily surveillance of electronic medical records for medically attended acute respiratory illness (MAARI) (using ICD-9-CM codes 460–466 and 480–488). In addition, during the first season, weekly Internet- or telephone-based surveillance monitored the occurrence of non–medically attended ARI among the participants at both sites. First-season participants were encouraged to contact staff directly if they became ill, and those who did not complete a weekly surveillance report received a reminder email or telephone call from project staff to assess current ARI status. For both influenza seasons, trained study staff collected respiratory specimens at participants' homes for ARIs that included fever and cough within 8 days of illness onset.

Case and Control Definitions

Cases were pregnant women with ARI identified through surveillance with real-time reverse transcription polymerase chain reaction (rRT-PCR)–confirmed influenza infections. Influenza-negative controls were pregnant women with ARI who tested negative for influenza. We included only 1 ARI per participant: either the influenza-positive illness or the first influenza-negative illness for those with >1 ARI and no positive test result.

As a secondary control group, we identified ARI-negative controls who had no record of a medical visit for ARI and no self-report of an illness with fever and cough since the start of the local influenza season through the date of illness onset for the index case. We matched 2 ARI-negative controls to each case by season, study site, and trimester of the case at the time of illness. The method of identifying these ARI-negative controls differed by season. During the second season, we recruited ARI-negative controls prospectively within 2 weeks of identifying an influenza case from randomly selected sets of all currently pregnant women in the same trimester who were members of the KPNC or KPNW health plans. ARI-negative controls for the first season were identified retrospectively at random from a previously recruited cohort of 1432 pregnant women, with 803 (56%) who completed at least 4 weekly surveillance reports eligible to be controls [20]. For potential ARI-negative controls in both seasons, the absence of MAARI was confirmed by electronic medical record review. The absence of non–medically attended ARI was determined by review of weekly surveillance reports during the first season and by a screening interview in the second season.

Laboratory Methods

Respiratory specimens were collected using nasopharyngeal (NP) swabs and stored with transport medium in cryovials at −70°C prior to overnight shipping on dry ice to the Marshfield Clinic Research Foundation laboratory (Marshfield, Wisconsin). Real-time RT-PCR was performed using a LightCycler Real-Time PCR System (Roche Diagnostics, Indianapolis, Indiana) and Invitrogen SuperScript III Platinum One-Step Quant RT-PCR chemistry (Life Technologies, Grand Island, New York). Protocols, primer, probes, and proficiency panel were provided by the Influenza Division of the Centers for Disease Control and Prevention.

Statistical Analysis

VE was estimated as 100% X (1 – odds ratio [ratio of odds of vaccination among cases to the odds of vaccination among controls]) using unconditional logistic regression for the influenza-negative control analyses and using conditional logistic regression for matched ARI-negative control analyses. A 95% confidence interval (CI) was calculated for each estimate. We estimated that 89 cases were needed to achieve 80% power (α = .05) to detect a VE of 50% with 2 controls per case and a vaccination rate of 50% among controls.

Both VE models adjusted for age, race, ethnicity (Hispanic), and high-risk medical conditions. Conditional logistic regression models using ARI-negative controls included strata for trimester, season, and site to account for matching. Analyses with influenza-negative controls adjusted for site, season, trimester, whether or not the illness was MAARI, and days between illness onset and NP swab collection. These covariates were chosen to reflect other recent influenza VE studies and to facilitate comparisons with VE estimates for adults during the same influenza seasons [14, 15]. Potential confounding, as indicated by significant associations with both vaccination and influenza outcome or a change of >10% in the VE estimate [15, 16], was also examined for characteristics in Table 1; however, we observed no other indications of confounding. Potential effect modification by prior vaccination status was examined by including main effects for current and prior season vaccination status plus an interaction term into each model. Analyses were conducted using SAS version 9.1 software (SAS Institute).

Table 1.

Characteristics of 492 Pregnant Women

Abbreviations: ARI, acute respiratory illness; BMI, body mass index; IQR, interquartile range; KPNC, Kaiser Permanente Northern California; KPNW, Kaiser Permanente Northwest; NS, not statistically significant.

a Two ARI-negative controls were matched to each case by season, study site, and trimester. ARI-negative controls had no record of a medical visit for ARI and no self-report of an illness with fever and cough since the start of local influenza season through the date of illness onset for the index case.

b Test of difference between cases and ARI-negative controls. Comparisons with categorical variables used χ2 tests. Comparisons with continuous variables used Mann-Whitney U tests.

c Test of difference between cases and influenza-negative controls.

d High-risk medical condition prior to pregnancy was indicated by 1 or more medical visits in the year prior to conception for a condition associated with increased risk of influenza complications, including cancer, diabetes, and neurological disorders as well as lung, heart, immune, and kidney disease (codes available upon request).

e Pregnancy complication was indicated by a medical visit associated with a subset of International Classification of Diseases, Ninth Revision codes 640–649 related to adverse pregnancy outcomes.

f Trimester at the time of the illness onset for the case or influenza-negative control; for ARI-negative controls the index date is the illness onset of their matched case.

g Current season vaccinated defined as receipt of vaccine at least 14 days prior to index illness onset; for ARI-negative controls the index illness is the illness of their matched case. Current season vaccination status and date were confirmed by medical record for 91% of vaccines and by self-report for 9% of vaccines.

h Prior season vaccination status was confirmed by medical record for 83% of vaccinees and by self-report for 17% of vaccines.

Table 1.

Characteristics of 492 Pregnant Women

Abbreviations: ARI, acute respiratory illness; BMI, body mass index; IQR, interquartile range; KPNC, Kaiser Permanente Northern California; KPNW, Kaiser Permanente Northwest; NS, not statistically significant.

a Two ARI-negative controls were matched to each case by season, study site, and trimester. ARI-negative controls had no record of a medical visit for ARI and no self-report of an illness with fever and cough since the start of local influenza season through the date of illness onset for the index case.

b Test of difference between cases and ARI-negative controls. Comparisons with categorical variables used χ2 tests. Comparisons with continuous variables used Mann-Whitney U tests.

c Test of difference between cases and influenza-negative controls.

d High-risk medical condition prior to pregnancy was indicated by 1 or more medical visits in the year prior to conception for a condition associated with increased risk of influenza complications, including cancer, diabetes, and neurological disorders as well as lung, heart, immune, and kidney disease (codes available upon request).

e Pregnancy complication was indicated by a medical visit associated with a subset of International Classification of Diseases, Ninth Revision codes 640–649 related to adverse pregnancy outcomes.

f Trimester at the time of the illness onset for the case or influenza-negative control; for ARI-negative controls the index date is the illness onset of their matched case.

g Current season vaccinated defined as receipt of vaccine at least 14 days prior to index illness onset; for ARI-negative controls the index illness is the illness of their matched case. Current season vaccination status and date were confirmed by medical record for 91% of vaccines and by self-report for 9% of vaccines.

h Prior season vaccination status was confirmed by medical record for 83% of vaccinees and by self-report for 17% of vaccines.

RESULTS

Study Population Characteristics

A total of 105 women who were rRT-PCR-positive for influenza were identified (Figure 1). We excluded 5 women because their respiratory specimens were collected >8 days after illness onset. Of the 100 influenza cases included in analysis, 45 were infected with influenza A(H1N1)pdm09 virus, 33 with A(H3N2), 1 with an unsubtyped A virus, and 21 with influenza B viruses. Of 217 women with ARI who tested influenza negative, 25 were excluded because respiratory specimens were collected >8 days after illness onset, leaving 192 influenza-negative controls. We identified 200 ARI-negative controls. Most of the illnesses among the influenza cases (83/100 [83%]) and the influenza-negative controls (151/192 [79%]) were medically attended.

Characteristics of the 100 influenza cases, 200 matched ARI-negative controls, and 192 influenza-negative controls are presented in Table 1. Most cases and controls were in their second or third trimester (82%), and most did not have a preexisting high-risk medical condition (80%) or have pregnancy complications (79%). Cases and both groups of controls were similar, with the exception of the following characteristics: ARI-negative controls were more likely to be white and not Hispanic; more ARI-negative controls described their health as excellent (81%) compared to cases (68%), and influenza-negative controls were more likely to have a high-risk medical condition (30%) than cases (17%) (all P < .05). The percentage of women who were obese prior to their pregnancy was highest among influenza-negative controls (30%), followed by influenza cases (15%), and lowest among ARI-negative controls (9%) (P < .05). Respiratory specimens were collected from cases about 1 day earlier than influenza-negative controls (median = 4 vs 5 days since illness onset; P < .05). We identified no consistent differences between vaccinated and unvaccinated participants in characteristics listed in Table 1 (Supplementary Table B). However, among all participants and notably among ARI-negative controls, vaccination rates were lower among women in their first trimester and among women with parity of 3 or more.

Effect of Current Season Vaccination

Among influenza cases, 42% were vaccinated compared to 58% vaccinated among influenza-negative controls and 63% vaccinated among matched ARI-negative controls. The unadjusted and adjusted VE estimates (with 95% CIs) using the influenza-negative controls were 48% (16%–68%) and 44% (5%–67%) and using the matched ARI-negative controls were 54% (26%–71%) and 53% (24%–72%).

Precision of VE estimates by season and (sub)type were limited by the small number of cases. Using influenza-negative controls, the adjusted VE point estimate was higher in the 2010–2011 season (57% [12%–79%]) than in the 2011–2012 season (27% [−67 to 68%]). After combining seasons, type- and subtype-specific VE estimates using influenza-negative controls were similar for A(H1N1)pdm09 virus infection (57% [16%–78%]), A(H3N2) (41% [−25 to 72%]), and influenza B (46% [−33 to 79%]) infections. Similar findings were observed for season- and (sub)type –specific VE estimates using the matched ARI-negative controls (data not shown). As an additional sensitivity analysis, adjusted VE was calculated excluding illnesses swabbed >4 days after illness onset, but resulted in little change in VE point estimates using influenza-negative controls (39% [−26% to 71%]) or matched ARI-negative controls (57% [17%–78%]). Similarly, VE estimates were also unchanged when the small number of participants with vaccination status confirmed by self-report only were excluded (data not shown).

Effect of Current and Prior Season Vaccination

In secondary analyses with both groups of controls, we found that VE for receipt of the current season's TIV differed significantly by receipt of prior season's vaccine (all P < .01 for interaction effects in both models), indicating that the effect of current season vaccination may not be independent of prior seasonal vaccination (Supplementary Table C). We did not observe effect modification by receipt of monovalent A(H1N1)pdm09 vaccination (data not shown). Therefore, we estimated VE for 3 categories of vaccine exposure (ie, vaccination in prior season only, current season only, or both seasons); women who received neither vaccine served as the referent group. Significant VE was observed for all 3 vaccine exposure categories in both models. The point estimates for VE against influenza A and B ranged from 51% to 76% using the influenza-negative controls and 48% to 76% in the ARI-negative control model (Table 2). A similar pattern of VE estimates were noted when stratified by season or virus (sub)type (Supplementary Table D). No consistent differences in participant characteristics were identified when comparing women with different combinations of prior and current season vaccinations (Supplementary Table E).

Table 2.

Combinations of Current and Prior Seasonal Trivalent Influenza Vaccine Exposure and Estimates of Unadjusted and Adjusted Vaccine Effectiveness for Each Vaccine Exposure

Abbreviations: ARI, acute respiratory illness; CI, confidence interval; VE, vaccine effectiveness.

a Vaccine effectiveness was estimated as 100% X (1– odds ratio [ratio of odds of being vaccinated among the cases to the odds of being vaccinated among the controls]) using conditional logistic regression for matched case-control analyses and unconditional logistic regression for the influenza-negative control model. Cases were matched with 2 ARI-negative controls by study site, season, and trimester. Both multivariate models were adjusted for age, race, ethnicity (Hispanic), and high-risk medical condition. VE calculated with influenza-negative controls was also adjusted for site, season, trimester, whether or not the illness was medically attended, and days between illness onset and respiratory specimen collection.

b Referent group = no prior vaccination and no current (study season) vaccination.

Table 2.

Combinations of Current and Prior Seasonal Trivalent Influenza Vaccine Exposure and Estimates of Unadjusted and Adjusted Vaccine Effectiveness for Each Vaccine Exposure

Abbreviations: ARI, acute respiratory illness; CI, confidence interval; VE, vaccine effectiveness.

a Vaccine effectiveness was estimated as 100% X (1– odds ratio [ratio of odds of being vaccinated among the cases to the odds of being vaccinated among the controls]) using conditional logistic regression for matched case-control analyses and unconditional logistic regression for the influenza-negative control model. Cases were matched with 2 ARI-negative controls by study site, season, and trimester. Both multivariate models were adjusted for age, race, ethnicity (Hispanic), and high-risk medical condition. VE calculated with influenza-negative controls was also adjusted for site, season, trimester, whether or not the illness was medically attended, and days between illness onset and respiratory specimen collection.

b Referent group = no prior vaccination and no current (study season) vaccination.

In summary, the VE estimates of prior (76%), current (58%–75%), and both (48%–51%) season vaccinations did not differ significantly as evidenced by overlapping confidence intervals (Table 2). Women who received either or both vaccines had similarly reduced likelihoods of being influenza positive. To illustrate, 115 of 292 pregnant women with ARI were not vaccinated in either the prior or current season; 47% (54/115) of these women tested positive for influenza compared to 27% (26/96) of women who were vaccinated in both seasons and 17% (4/23) and 28% (16/58) of women vaccinated during the prior or current season only, respectively (Supplementary Figure 1).

DISCUSSION

We observed substantial and significant vaccine effects. When combined with mounting evidence for the vaccine's safety [11, 25], these results support the importance of vaccination among pregnant women in the United States and internationally [26]. The effectiveness of vaccination with the current season's vaccine that we observed among pregnant women using the test-negative design (44% [95% CI, 5%–67%]) was within the range of VE estimates observed among all adults by other VE studies with similar controls during the same seasons [14–17], and consistent with VE point estimates reported in a recent meta-analysis [27]. We observed similar VE using a second group of matched controls without ARI. As one would expect, the effectiveness we observed for influenza infection confirmed by rRT-PCR was higher than the efficacy found for febrile ARIs without influenza confirmation (36% [95% CI, 4%–57%]) in a randomized trial with pregnant women in Bangladesh [10].

In our population-based study of pregnant women over 2 influenza seasons, vaccination in the prior season modified the effect of vaccination during pregnancy. Seasonal influenza vaccine received during the current or the prior season reduced the risk of ARI with laboratory-confirmed influenza by more than one-half. Specifically, contingent on whether the woman received the vaccine in only 1 or both seasons, we observed an adjusted VE of 51%–76% comparing cases to controls who had ARI but tested negative for influenza and observed a similar range of VE (48%–76%) with our ARI-negative controls.

We observed this protective effect for vaccination during 2 seasons when the vaccine components remained the same, the influenza B strain was consistent going back an additional year, and all 3 vaccine components were antigenically similar to circulating viruses [23, 24]. The similar protection conferred by influenza vaccination in either the current or prior season in our study should not be interpreted as indicating that a pregnant woman can forgo vaccination if she were vaccinated in the previous season. Our study took place during unusual seasons in which most vaccine components and circulating strains remained stable. Although our findings are consistent with some other studies that have observed sustained influenza antibody seroprotection across seasons [28, 29] and the potential for cross-season [17] and cross-strain protection [30, 31] from prior vaccination, our study is limited by the lack of serologic data, which might give insight into the mechanism behind the effects we observed.

Other researchers have also observed that VE can be modified by prior vaccination [19, 32, 33], especially during seasons with homologous vaccine viruses [18], but additional research is needed to confirm and explain this effect. Indeed, our study was not designed to evaluate the effect of vaccine exposure over consecutive years. As a result, our study was limited by imperfect records for prior vaccinations, especially for the 2009 pandemic vaccine that was often administered outside of traditional care settings [34]. Although receipt of pandemic vaccine did not modify the effect of current season vaccination in our study, receipt of pandemic and seasonal vaccinations were highly correlated in our study and others [34, 35], and thus pandemic vaccination may have contributed to the effects we observed in a way that we could not disentangle.

Strengths of this evaluation of the effectiveness of TIV among pregnant women include confirmation of influenza virus infection by rRT-PCR, documentation of vaccination status and medical conditions from medical records, and the use of 2 complementary control groups. Other strengths include our prospective surveillance and recruitment of women in 2 geographic regions over 2 seasons when all 3 influenza strains included in the vaccine were in circulation.

Our evaluation also has at least 5 limitations. First, we do not know the respiratory pathogens associated with the ARIs of influenza-negative controls, although recent research [36] and modeling studies [12] suggest this is not likely to be a significant source of bias in VE estimates among adults. Second, given the challenges of studying laboratory-confirmed influenza among pregnant women [20], we had to conduct the study over 2 seasons with modest changes in study design to enroll a sufficient number of cases to calculate a VE estimate. Specifically, while all ARIs in the second season were medically attended, about one-third of the influenza and noninfluenza ARIs during the first season were not medically attended. During the second season, we recruited matched ARI-negative controls from all eligible pregnant women in the participating health plans, but in the first season, ARI-negative controls were selected from a previously recruited prospective cohort. We adjusted for season and whether the illness was medically attended in our multivariate model. In addition, similar influenza strains circulated in both seasons [23, 24], and we found that VE was similar when stratified by season and virus. These factors suggest that differences in control enrollment by season did not substantially affect our results. Third, it is possible that residual or unmeasured confounding may have biased our results in unknown ways, although we found little evidence of confounding by participant characteristics. Additionally, the influenza-negative control model also adjusted for days from illness onset to testing and whether or not the ARI was medically attended. Fourth, conclusions about VE among pregnant women from our study may be limited to this population: women mostly in their second or third trimester, who tended to be in good health, and received prenatal care within a managed care organization. Similarly, our conclusions are limited to the measured influenza illness outcomes and do not address the prevention of asymptomatic infections or atypical manifestations.

In conclusion, influenza vaccination reduced the risk of ARI associated with laboratory-confirmed influenza among pregnant women by about one-half, which is similar to VE observed among all adults during these seasons. Thus, taken together with other recent studies, influenza vaccines appear to be not only safe for mothers and fetuses [11, 25] but equally as effective as in nonpregnant adults. As pregnant women are at risk for serious influenza-related outcomes, our results support the recent World Health Organization recommendation that all countries with influenza vaccination programs vaccinate pregnant women [26]. Further research is needed to confirm the direct benefits of influenza vaccination for pregnant women across all trimesters and establish the extent to which maternal vaccination confers secondary benefits to their infants, as suggested by several recent studies [10, 37–40].

Notes

Acknowledgments. Members of the PIP Workgroup at the Centers for Disease Control and Prevention (CDC), in addition to the authors, included: Susan Chu, PhD, Janet Cragan, MD, Anne McIntyre, PhD, and Julie Villanueva, PhD. We thank Alicia Fry, MD, MPH, Joe Bresee, MD, Jerome Tokars, MD, and Jane Seward, MBBS, MPH, for their feedback on early versions of this manuscript.

Disclaimer. The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the CDC, Abt Associates, Inc, or the Kaiser Foundation Research Institute.

Financial support. This work was supported by the CDC (contract 200-2010-F-33132 to Abt Associates Inc).

Potential conflicts of interest. All authors: No reported conflicts.

All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.

References

1

Effects of influenza on pregnant women and infants

,

Am J Obstet Gynecol

,

2012

, vol.

207

3 suppl

(pg.

S3

-

8

)

2

Prevention and control of influenza: recommendations of the Advisory Committee on Immunization Practices (ACIP)

,

MMWR Recomm Rep

,

2004

, vol.

53

RR-6

(pg.

1

-

40

)

3

Monitoring seasonal influenza vaccination coverage among pregnant women in the United States

,

Am J Obstet Gynecol

,

2012

, vol.

207

3 suppl

(pg.

S9

-

16

)

4

Inactivated influenza vaccines: methods, policies, and politics

,

J Clin Epidemiol

,

2009

, vol.

62

(pg.

677

-

86

)

5

Is routine influenza immunization warranted in early pregnancy?

,

Vaccine

,

2009

, vol.

27

(pg.

4754

-

70

)

6

Effectiveness of polyvalent influenza vaccine in pregnancy. Report of a controlled study during an outbreak of Asian influenza

,

Obstet Gynecol

,

1964

, vol.

23

(pg.

830

-

7

)

7

et al.

Influenza immunization in pregnancy—antibody responses in mothers and infants

,

N Engl J Med

,

2010

, vol.

362

(pg.

1644

-

6

)

8

Effectiveness of influenza vaccine during pregnancy in preventing hospitalizations and outpatient visits for respiratory illness in pregnant women and their infants

,

Am J Perinatol

,

2004

, vol.

21

(pg.

333

-

9

)

9

et al.

Safety of influenza vaccination during pregnancy

,

Am J Obstet Gynecol

,

2005

, vol.

192

(pg.

1098

-

106

)

10

et al.

Effectiveness of maternal influenza immunization in mothers and infants

,

N Engl J Med

,

2008

, vol.

359

(pg.

1555

-

64

)

11

et al.

Risk of fetal death after pandemic influenza virus infection or vaccination

,

N Engl J Med

,

2013

, vol.

368

(pg.

333

-

40

)

12

The case test-negative control design for studies of the effectiveness of seasonal influenza vaccine

,

Vaccine

,

2013

, vol.

31

(pg.

3104

-

9

)

13

The test-negative design for estimating influenza vaccine effectiveness

,

Vaccine

,

2013

, vol.

31

(pg.

2165

-

8

)

14

Team IMC-CS

Early estimates of seasonal influenza vaccine effectiveness in Europe among target groups for vaccination: results from the I-MOVE multicentre case-control study, 2011/12

,

Euro Surveill

,

2012

, vol.

17

pii:20146

15

et al.

Effectiveness of seasonal influenza vaccines in the United States during a season with circulation of all three vaccine strains

,

Clin Infect Dis

,

2012

, vol.

55

(pg.

951

-

9

)

16

et al.

Effectiveness of the 2010–11 seasonal trivalent influenza vaccine in Spain: cycEVA study

,

Vaccine

,

2012

, vol.

30

(pg.

3595

-

602

)

17

et al.

A sentinel platform to evaluate influenza vaccine effectiveness and new variant circulation, Canada 2010–2011 season

,

Clin Infect Dis

,

2012

, vol.

55

(pg.

332

-

42

)

18

Variable efficacy of repeated annual influenza vaccination

,

Proc Natl Acad Sci U S A

,

1999

, vol.

96

(pg.

14001

-

6

)

19

Influenza vaccine effectiveness in the community and the household

,

Clin Infect Dis

,

2013

, vol.

56

(pg.

1363

-

9

)

20

et al.

The Pregnancy and Influenza Project: design of an observational case-cohort study to evaluate influenza burden and vaccine effectiveness among pregnant women and their infants

,

Am J Obstet Gynecol

,

2011

, vol.

204

6 suppl 1

(pg.

S69

-

76

)

21

Self-rated health and mortality: a review of twenty-seven community studies

,

J Health Soc Behav

,

1997

, vol.

38

(pg.

21

-

37

)

22

et al.

Prevention and control of influenza with vaccines: recommendations of the Advisory Committee on Immunization Practices (ACIP), 2010

,

MMWR Recomm Rep

,

2010

, vol.

59

RR-8

(pg.

1

-

62

)

23

Centers for Disease Control and Prevention

,

Antigenic characterization

,

2012

24

Centers for Disease Control and Prevention

,

Was last season's vaccine a good match for circulating viruses?

,

2011

25

Safety of influenza immunization during pregnancy for the fetus and the neonate

,

Am J Obstet Gynecol

,

2012

, vol.

207

3 suppl

(pg.

S38

-

46

)

26

Vaccines against influenza. WHO position paper—November 2012

,

2012

, vol.

87

(pg.

461

-

76

)

Releve epidemiologique hebdomadaire/Section d'hygiene du Secretariat de la Societe des Nations [Weekly epidemiological record/Health Section of the Secretariat of the League of Nations]

27

Efficacy and effectiveness of influenza vaccines: a systematic review and meta-analysis

,

Lancet Infect Dis

,

2012

, vol.

12

(pg.

36

-

44

)

28

et al.

Influence of prior influenza vaccination on antibody and B-cell responses

,

PLoS One

,

2008

, vol.

3

pg.

e2975

29

et al.

Immunogenicity, boostability, and sustainability of the immune response after vaccination against influenza A virus (H1N1) 2009 in a healthy population

,

Clin Vaccine Immunol

,

2011

, vol.

18

(pg.

1401

-

5

)

30

et al.

HI responses induced by seasonal influenza vaccination are associated with clinical protection and with seroprotection against non-homologous strains

,

Vaccine

,

2012

, vol.

30

(pg.

5262

-

9

)

31

et al.

Impacts on influenza A(H1N1)pdm09 infection from cross-protection of seasonal trivalent influenza vaccines and A(H1N1)pdm09 vaccines: systematic review and meta-analyses

,

Vaccine

,

2012

, vol.

30

(pg.

3209

-

22

)

32

Protection against influenza after annually repeated vaccination: a meta-analysis of serologic and field studies

,

Arch Intern Med

,

1999

, vol.

159

(pg.

182

-

8

)

33

Efficacy of sequential annual vaccination with inactivated influenza virus vaccine

,

Am J Epidemiol

,

1988

, vol.

127

(pg.

353

-

64

)

34

Centers for Disease C, Prevention

Seasonal influenza and 2009 H1N1 influenza vaccination coverage among pregnant women—10 states, 2009–10 influenza season

,

MMWR Morb Mortal Wkly Rep

,

2010

, vol.

59

(pg.

1541

-

5

)

35

Centers for Disease C, Prevention

Influenza vaccination coverage among pregnant women—United States, 2010–11 influenza season

,

MMWR Morb Mortal Wkly Rep

,

2011

, vol.

60

(pg.

1078

-

82

)

36

Influenza vaccination is not associated with detection of non-influenza respiratory viruses in seasonal studies of influenza vaccine effectiveness

,

Clin Infect Dis

,

2013

, vol.

57

(pg.

789

-

93

)

37

The association between influenza vaccine in pregnancy and adverse neonatal outcomes

,

J Obstet Gynaecol Can

,

2012

, vol.

34

(pg.

714

-

20

)

38

Influenza vaccine given to pregnant women reduces hospitalization due to influenza in their infants

,

Clin Infect Dis

,

2010

, vol.

51

(pg.

1355

-

61

)

39

et al.

Neonatal outcomes after influenza immunization during pregnancy: a randomized controlled trial

,

CMAJ

,

2012

, vol.

184

(pg.

645

-

53

)

40

et al.

Maternal influenza immunization and reduced likelihood of prematurity and small for gestational age births: a retrospective cohort study

,

PLoS Med

,

2011

, vol.

8

pg.

e1000441

Published by Oxford University Press on behalf of the Infectious Diseases Society of America 2013. This work is written by (a) US Government employee(s) and is in the public domain in the US.

Topic:

• pregnancy
• influenza
• adult
• influenza vaccines
• orthomyxoviridae
• respiratory tract infections
• vaccination
• vaccines

Supplementary data

I agree to the terms and conditions. You must accept the terms and conditions.

Submit a comment

Name

Affiliations

Comment title

Comment

You have entered an invalid code

Thank you for submitting a comment on this article. Your comment will be reviewed and published at the journal's discretion. Please check for further notifications by email.

Citations

Views

Altmetric

Metrics

Total Views 10,450

9,140 Pageviews

1,310 PDF Downloads

Since 12/1/2016

Citations

131 Web of Science

×

Email alerts

More on this topic

Related articles in PubMed

Citing articles via

• Latest

• Most Read

• Most Cited

More from Oxford Academic