Skip Navigation

Effectiveness of Inactivated Influenza Vaccines Varied Substantially with Antigenic Match from the 2004–2005 Season to the 2006–2007 Season

  1. Edward A. Belongia1,
  2. Burney A. Kieke1,
  3. James G. Donahue1,
  4. Robert T. Greenlee1,
  5. Amanda Balish2,
  6. Angie Foust2,
  7. Stephen Lindstrom2 and
  8. David K. Shay2

    for the Marshfield Influenza Study Groupa

  1. 1 Marshfield Clinic Research Foundation, Marshfield, Wisconsin
  2. 2 Influenza Division, Centers for Disease Control and Prevention, Atlanta, Georgia
  1. Reprints or correspondence: Edward A. Belongia, MD, Epidemiology Research Center (ML2), Marshfield Clinic Research Foundation, 1000 North Oak Ave., Marshfield, WI 54449 (belongia.edward{at}marshfieldclinic.org).

  1. Presented in part: American Society for Microbiology Conference on Tuberculosis: Past, Present and Future, New York, New York, 20–24 June 2000 (abstract 153).

 
Next Section

Abstract

Background. We estimated the effectiveness of inactivated influenza vaccines for the prevention of laboratory-confirmed, medically attended influenza during 3 seasons with variable antigenic match between vaccine and patient strains.

Methods. Patients were enrolled during or after a clinical encounter for acute respiratory illness. Influenza infection was confirmed by culture or reverse-transcriptase polymerase chain reaction. Case-control analyses were performed that used data from patients who were ill without influenza (hereafter, “test-negative control subjects”) and data from asymptomatic control subjects from the population (hereafter, “traditional control subjects”). Vaccine effectiveness (VE) was estimated as [100 × (1 − adjusted odds ratio)]. Influenza isolates were antigenically characterized.

Results. Influenza was detected in 167 (20%) of 818 patients in 2004–2005, in 51 (14%) of 356 in 2005–2006, and in 102 (11%) of 932 in 2006–2007. Analyses that used data from test-negative control subjects showed that VE was 10% (95% confidence interval [CI], −36% to 40%) in 2004–2005, 21% (95% CI,−52% to 59%) in 2005–2006, and 52% (95% CI, 22% to 70%) in 2006–2007. Using data from traditional control subjects, VE for those seasons was estimated to be 5% (95% CI, −52% to 40%), 11% (95% CI, −96% to 59%), and 37% (95% CI, −10% to 64%), respectively; confidence intervals included 0. The percentage of viruses that were antigenically matched to vaccine strains was 5% (3 of 62) in 2004–2005, 5% (2 of 42) in 2005–2006, and 91% (85 of 93) in 2006–2007.

Conclusions. Influenza VE varied substantially across 3 seasons and was highest when antigenic match was optimal. VE estimates that used data from test-negative control subjects were consistently higher than those that used data from traditional control subjects.

Influenza is an important cause of death and serious illness in the United States. As a result, annual influenza vaccination is recommended for young children, elderly people, other individuals at high risk for serious influenza-related complications, and close contacts of these groups [1]. Antigenic drift necessitates frequent changes in the composition of influenza vaccines, and these changes must be specified 7–9 months in advance of the influenza season to allow for the production and distribution of vaccines [2]. Vaccine effectiveness (VE) in the field can vary from season to season, but several observational studies have suggested that influenza vaccine retains some effectiveness when antigenic drift hasoccurred between vaccine and circulating strains. However, it is difficult to generalize these results because influenza infection was not laboratory confirmed in most studies, nor was antigenic match determined for viruses recovered from study participants [3–8]. We conducted population-based case-control studies to estimate the annual effectiveness of trivalent inactivated influenza vaccine for preventing medically attended, laboratoryconfirmed influenza illness during 3 consecutive influenza seasons. Influenza viruses recovered from study participants were characterized each season to assess antigenic match.

Previous SectionNext Section

methods

Study population. The source population included residents of the Marshfield Epidemiologic Study Area (MESA), a dynamic, population-based cohort of approximately 54,000 residents living in 14 zip code areas surrounding Marshfield, Wisconsin. In this area, nearly all residents receive their inpatient and outpatient care from Marshfield Clinic facilities, which use an electronic medical record that captures ≥90% of outpatient visits, 99% of deaths, and 95% of hospital discharges for the population [9–12]. Prior to each influenza season, we identified MESA residents for whom influenza vaccination was recommended by the Advisory Committee on Immunization Practices (ACIP) on the basis of age or individual risk of influenza complications. These individuals were eligible to be enrolled and tested for influenza during or after a clinical encounter for acute respiratory illness during the influenza season. For the 2004– 2005 season, community-dwelling residents of MESA were eligible to be enrolled with acute respiratory illness if they were 6–23 months old or ≥65 years old on 1 November 1 2004, or if they were 2–64 years old and had been diagnosed with a high-risk medical condition [13]. For the 2005–2006 influenza season, eligibility was expanded to include all adults≥50 years old, on the basis of ACIP recommendations for that season [14]. For the 2006–2007 season, eligibility was further expanded to include all children 6–59 months old, on the basis of the then-current ACIP recommendations [15].

Each season, individuals were classified as having a medical condition that placed them at high risk for complications of influenza infection if they had ≥2 visits to the Marshfield Clinic during the preceding 12 months that involved an International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) diagnosis code for a high-risk condition. The ICD-9 codes included diagnoses in the following chronic disease categories: cardiac disease, pulmonary disease, renal disease, liver disease, diabetes mellitus, immunosuppressive disorders, malignancies, neurological/musculoskeletal disease, metabolic disease, cerebrovascular disease, and circulatory system disease.

Influenza vaccination and immunization status. MarshfieldClinic providers and local vaccination clinics used only trivalent inactivated influenza vaccine (Sanofi-Pasteur) during each influenzaseason. Despite concerns about vaccine availability in fall of 2004, Marshfield Clinic received sufficient doses to meet demand during all 3 seasons. Influenza vaccination status was determined by a real-time, internet-based vaccination registry used by all publicand private vaccination providers serving the population (http://www.recin.org). A validation study of the registry demonstrated that it correctly identified 96% to 98% of all influenza vaccinationsthat were delivered to the study population (E.A.B., unpublished data).

Adults and children were classified as immunized beginning14 days after receipt of influenza vaccination. Children <9 yearsold were classified as fully immunized only if they had received ≥2 doses of influenza vaccine. Children in this age group wereclassified as partially immunized if they had received only 1 of 2recommended doses. The number of partially immunized childrenwas insufficient to generate a separate VE estimate, andthey were excluded from all analyses.

Ascertainment of influenza illness. Members of the studycohort were recruited by trained research coordinators duringor after an inpatient or outpatient medical encounter for acute respiratory illness of <10 days duration. Potential participants whose illness duration was≥10 days (at the time of the encounter) were excluded because they were unlikely to test positive for influenza [16, 17]. Research coordinators used an electronic appointment system to identify and recruit eligible patients in all primary care clinics and in urgent care on weekdays, evenings, and weekends. Appointments made within the previous 3 days were selectively identified, which reduced the need to evaluate patients who were seen for a previously diagnosed condition. Patients with acute respiratory or febrile illness were identified by reviewing comments in the electronic appointment system or by querying the clinical staff. Eligible patients were also recruited at an acute care hospital that is contiguous with Marshfield Clinic (St. Joseph's Hospital). Most ill patients who were not approached during a clinical encounter were identified on the following day by use of electronic diagnosis codes entered by attending physicians (ICD-9-CM codes 382.0, 382.4, 382.9, 460–466, 480, 483–486, 487, 490, 780.6, and 786.2). These individuals were contacted by telephone, and a swab sample was obtained at home from those who were eligible and consented. A nasopharyngeal swab sample (from adults and children⩾13 years old) or nasal swab sample (from children 6 months–12 years old) was obtained from all participants [18, 19]. Most swab samples were obtained by research coordinators immediately after the clinical encounter, but samples obtained by clinical staff were used when available to avoid duplicate swabs. Each participant (or parent) completed a short interview to assess illness symptoms and onset date. Patients with laboratory-confirmed influenza were contacted again ⩾14 days after enrollment to ascertain hospital admissions related to that illness. Ill patients with a negative influenza test could be reenrolled with a new respiratory illness after a 28-day exclusion period to allow time for recovery from the first illness episode. The date that study enrollment was initiated and the duration of recruitment varied by season. The initiation of recruitment was based on isolation of influenza at Marshfield Laboratories and evidence of increasing influenza detection at the Wisconsin State Laboratory of Hygiene. The start date and duration of study recruitment for each of the 3 seasons was as follows: 3 January 2005 (12 weeks), 8 March 2006 (6 weeks), 22 January 2007 (10 weeks).

Control groups. Two control groups were used: traditional control subjects and test-negative control subjects. Traditional control subjects were randomly sampled with replacement each week from individuals in the source population who did not have a clinical encounter for acute respiratory illness prior to that week. These individuals were contacted by telephone and screened for the occurrence of any acute respiratory illness during the enrollment period. Traditional control subjects were enrolled in a 2:1 ratio to case subjects, with matching by week of case identification and age group (6–23 months, 2–8 years, 9–17 years, 18–49 years, 50–64 years, or ⩾65 years). For the 2006– 2007 season, the age groups used for matching in children were 2–4 years and 5–17 years, based on expanded vaccine recommendations. The test-negative control group was made up of study participants with acute respiratory illness who tested negative for influenza.

Influenza virus isolation and characterization. Viral cultures and real-time reverse-transcriptase polymerase chain reaction (RT-PCR) were performed at the Marshfield Clinic Research Foundation. Both culture and RT-PCR were performed in 2005–2006 and 2006–2007, but only culture was performed during the 2004–2005 season. Nasopharyngeal and nasal swab samples were placed in viral transport medium (M4-RT; Remel) and refrigerated or stored on ice until delivered to the laboratory on the same day that they were collected. Cultures were set up within 1 working day after receipt of samples. For the 2004–2005 season, monolayered rhesus monkey kidney cells (Diagnostic Hybrid) were inoculated and incubated at 35°–37°C on a rotating roller drum. Cultures were examined for cytopathic effect, and those without cytopathic effect were tested twice for hemagglutination. Influenza A and B were confirmed by immunofluorescence (Chemicon International). During the 2005–2006 and 2006–2007 seasons, a centrifugation-enhanced Madin-Darby canine kidney shell vial method was used for viral culture. A terminal immunofluorescent test was performed after 72 h of incubation.

Real-time RT-PCR was performed on nucleic acid extracts from samples by use of the LightCycler Real-Time PCR System (Roche Diagnostics). Sequence information for RT-PCR primers and probes was provided by the Center for Disease Control and Prevention (CDC) Influenza Division. The assay was a TaqMan-based, real-time detection of the matrix protein (M1)of influenza A and the nonstructural protein 1 (NS1) of influenza B; the sequences of both proteins are highly conserved. The human RNase P gene primer and probe set served as an internal positive control for human nucleic acids. Samples were tested at least twice per week during the enrollment period. A subset of influenza isolates (those from the 2004–2005 season) or all isolates (those from the 2005–2006 and 2006–2007 seasons) were antigenically characterized by the CDC by use of the hemagglutination inhibition test with a set of reference postinfection ferret antisera [2020].

Calculation of vaccine effectiveness. VE was defined as [100 × (1 − adjusted odds ratio)], where the odds ratio is for influenza vaccination among laboratory-confirmed influenza case subjects compared with control subjects. The odds ratio provided a valid estimate of the incidence rate ratio in the source population for analyses that used data from traditional control subjects, because incidence density sampling was used [21]. Traditional control subjects were matched to influenza case subjects with illness occurring in the same week. Exchangeable strata were pooled for the analysis that used data from traditional control subjects; conditional logistic regression was performed with strata defined by combinations of age group and week of case enrollment [22, 23]. The analysis that used data from traditional control subjects adjusted for age (as a continuous variable), presence of any high-risk condition, and frequency of health care use (sum of physician, laboratory, and radiology encounters) during the 12 months before the study population was established. The last covariate was included because vaccinated individuals were observed to use health-care services more often than unvaccinated individuals.

Test-negative control subjects were identified longitudinally throughout each influenza season in the same manner as case subjects, and independently of vaccination status. As a result, the corresponding odds ratio served as a valid estimate of the incidence rate ratio among participants even when the outcome (influenza) was not rare [21, 24]. Logistic regression models that used data from test-negative control subjects were adjusted for week of enrollment, age, interval from symptom onset to sample collection, and presence of any high-risk medical condition. Exact logistic regression was performed for subgroup analyses when indicated. All analyses were performed using SAS (version 9.1; SAS Institute); P <.05 was considered significant. This study was reviewed and approved by the Marshfield Clinic Institutional Review Board, and all participants or parents gave written informed consent for influenza testing.

Previous SectionNext Section

Results

The number of patients approached for enrollment ranged from 1097 in 2005–2006 to 2252 in 2006–2007 (table 1). The numberof enrollments was greatest in 2006–2007 and lowest in 2005– 2006. The latter season was atypical; onset of the influenza season was delayed in Wisconsin, and enrollment continued for only 6 weeks. Of those evaluated for eligibility, 25%–35% were excluded each season because of symptom duration ⩾10 days, and ∼20% declined to participate each season. The number of children excluded from the analyses because of partial immunization in 2004–2005 was 56; in 2005–2006, it was 10; and in 2006–2007, it was 61. After these exclusions, the number of enrollments in 2004–2005 was 762; in 2005–2006, it was 346; and in 2006–2007, it was 871. The proportion of enrolled patients with laboratory-confirmed influenza was highest in 2004–2005 and lowest in 2006–2007 (table 2).

Table 1.
View larger version:
Table 1.

Characteristics of study population and recruitment results, according to influenza season.

Table 2.
View larger version:
Table 2.

Viral culture and reverse-transcriptase polymerase chain reaction (RT-PCR) results for each influenza season.

Less than 10% of potential traditional control subjects declined to participate each season, and 54%–68% were matched to an influenza case. Exclusion due to prior respiratory illness was the primary reason for not being matched to a case. The demographic and clinical characteristics of case patients with laboratory-confirmed influenza and test-negative control subjects are shown in table 3.

Table 3.
View larger version:
Table 3.

Demographic and clinical characteristics of patients with influenza confirmed by culture or reverse-transcriptase polymerase chain reaction and test-negative (TN) control subjects, according to influenza season.

The VE estimate obtained by using data from test-negative control subjects was highest for the 2006–2007 season and lowest for the 2004–2005 season (table 4). The 95% confidence interval (CI) excluded 0 only during the 2006–2007 season. The VE estimate in the analysis based on data from traditional control subjects was also highest during the 2006–2007 season, although this method yielded lower estimates of VE during each season, compared with the analysis that used data from testnegative control subjects (table 5). VE was estimated separately in children and adults in analysis that used data from test negative control subjects, and the 95% confidence intervals included 0 in all age groups that were examined. However, the VE point estimates were highest for the 2006–2007 season (data not shown).

Table 4.
View larger version:
Table 4.

Crude and adjusted vaccine effectiveness (VE) estimates based on comparison of laboratory-confirmed influenza case subjects and test-negative control subjects.

Table 5.
View larger version:
Table 5.

Crude and adjusted vaccine effectiveness (VE) estimates based on comparison of laboratory-confirmed influenza case subjects and traditional control subjects.

VE was estimated for the subgroup of case subjects and testnegative control subjects who were tested 0–3 days after illness onset. In these subgroups, VE was 18% in 2004–2005, −40% in 2005–2006, and 29% in 2006–2007. The confidence interval included 0 for each season. VE was also estimated for prevention of hospitalization with laboratory-confirmed influenza. During 2006–2007, 0 of 4 hospitalized influenza case subjects (including 2 case subjects who were hospitalized after outpatient enrollenrollment) were vaccinated, compared with 56% of test-negative control subjects, yielding a VE estimate of 90% (95% CI, 10%– 100%) by use of exact logistic regression. The estimated VE was similar when hospitalized influenza case subjects were compared with hospitalized, influenza-negative control subjects: 10 (67%) of 15 hospitalized, influenza-negative control subjects had received influenza vaccination, yielding an unadjusted VE of 88% (95% CI, 13%–100%) by use of exact logistic regression. Three of four hospitalized case subjects and 12 of 15 hospitalized control subjects were >50 years old.

The proportion of influenza viruses isolated from patients that were well matched to vaccine strains was 3 of 62 (5%) in 2004–2005, 2 of 42 (5%) in 2005–2006, and 85 of 93 (91%) in 2006–2007 (table 6). During 2004–2005, the predominant virus recovered from study patients was A/California/7/2004-like(H3N2), a drift variant of the H3N2 vaccine strain A/Fujian/411/2002. During 2005–2006, all influenza B isolates belonged to the B/Victoria/2/87 lineage, which was antigenically distinct from the B/Yamagata/16/88 lineage vaccine virus B/Shanghai/361/2002. During 2006–2007, both A/H1N1 and A/H3N2 viruses recovered from study participants were closely matched to the vaccine strains.

Table 6.
View larger version:
Table 6.

Antigenic characteristics of influenza viruses isolated from study patients.

Previous SectionNext Section

discussion

Influenza VE varied substantially over 3 consecutive seasons for this central Wisconsin population. The seasonal variation in VE was generally consistent with the degree of antigenic match between viruses isolated from patients and vaccine strains. VE was lowest during the 2004–2005 season, when only 5% of viruses from study participants were well matched to vaccine strains. Although low illness rates limited the precision of the VE estimate during the 2005–2006 season, VE was less than 25% and the antigenic match was poor. In both 2004–2005 and 2005– 2006, the confidence intervals included 0. In contrast, VE was ∼50% during the 2006–2007 season when ∼90% of isolates were well matched to vaccine strains. We used data from 2 control groups to generate VE estimates. The traditional control subjects were randomly sampled each week, with age matching, from a defined population of individuals for whom influenza vaccination was recommended. Although control subjects of this type are commonly used for casecontrol studies, they may introduce a bias in a study of VE for medically attended influenza. The vaccinated individuals in our study population used health-care services more often than unvaccinated individuals, and they were more likely to seek care for acute respiratory illness during the influenza season. As a result, vaccinated individuals with influenza were enrolled and tested more often than unvaccinated individuals with influenza, resulting in underestimation of VE when traditional control subjects were used as the comparison group. This bias was avoided in the case-control analysis based on test-negative control subjects, who sought care during the same period as influenza case subjects.

Simulation models suggest that the use of test-negative control subjects for estimating VE yields a point estimate that is nearly identical to the “true” VE in a simulated population if the test specificity is nearly 100%, as is likely when RT-PCR is used for case detection [25]. However, the simulations did not evaluate the effects of bias or confounding on VE estimates. We found that VE was consistently higher in the analysis that used data from test-negative control subjects, and this control group was most similar to the case subjects in terms of health-care seeking behavior. These findings suggest that analysis of data from testnegative control subjects may yield more unbiased estimates of VE, compared with analyses based on traditional control subjects. Our results also suggest that differential health-care seeking behavior is a more important source of bias, compared with the potential misclassification of true case subjects as testnegative control subjects when using a diagnostic method that is <100% sensitive. Additional research is needed to assess the optimal comparison group in studies that use laboratory confirmed influenza as the outcome.

Several randomized, placebo-controlled studies demonstrated efficacy of 58%–88% for currently licensed trivalent inactivated influenza vaccines [26–30]. One of these studies found that VE was high (86%) during a season when the vaccine and circulating viruses were well matched, and it was not significantly greater than 0 during another season with a poor match between vaccine antigens and circulating viruses [29]. A trial involving healthy adults aged 30–60 years over 5 seasons (1983– 1984 through 1987–1988) found that vaccinated individuals had a significantly reduced risk of developing laboratory-confirmed influenza A illness during the 2 seasons with the best antigenic match [31]. A more recent randomized clinical trial that involved healthy adults 18–46 years old reported that inactivated influenza vaccine had 77% efficacy for preventing laboratory confirmed influenza illness during the 2004–2005 season, despite circulation of a drifted A(H3N2) virus in the study population [32]. The trial was performed in a nearby state (Michigan), and efficacy was based on testing of all participants with acute respiratory illness, including illnesses that were managed at home. In contrast, we measured effectiveness against medically attended, laboratory-confirmed influenza. Our study population also included a high proportion of older adults and individuals with a high-risk medical condition, and these groups were not included in the Michigan trial. Impaired humoral and cell-mediated vaccine response may contribute to reduced clinical effectiveness in older adults [26, 33], although a recent review concluded that persistence of vaccine-induced antibody is not reduced in elderly individuals [34].

Observational studies have suggested that inactivated influenza vaccine retains some effectiveness during seasons with a poor antigenic match [3– 8]. However, most studies used nonspecific outcome measures (i.e., influenza infections were not confirmed), and they did not assess the antigenic characteristics of viruses recovered from study participants. VE was estimated for the prevention of laboratory-confirmed influenza in British Columbia during the 2005–2006 season, when there was substantial mismatch between vaccine strains and influenza viruses recovered from study participants [35]. Sentinel physicians enrolled patients with influenzalike illness (ILI) and influenza was detected by culture or RT-PCR in 47% of subjects; VE was 61% for laboratory-confirmed influenza. The British Columbia study used an analytic approach that was similar to ours, but there are differences that limit direct comparison of the VE estimates in Wisconsin and British Columbia. First, physicians in British Columbia tested only half the patients presenting with ILI, and the criteria for testing were not clear. The high proportion of positive influenza tests suggests they may have preferentially selected patients with typical influenza symptoms for testing. Second, the age distribution was different in the 2 populations, with a higher proportion of elderly individuals in Wisconsin. The proportion of participants with a high-risk condition was also higher in Wisconsin (54%), compared with British Columbia (14%). Because the British Columbia study enrolled a younger and healthier population, it is plausible that VE may have been higher in that study.

Our study has several limitations. First, the estimation of VE using test-negative control subjects is a new approach with little precedent. Test-negative control subjects are similar to case subjects in terms of health-care seeking behavior, but they likely differ from the source population in other ways that affect generalizability. Second, the outcome measure used in this study is not directly comparable to outcome measures used in previous studies. We assessed VE for the prevention of medically attended, laboratory-confirmed influenza, and it is not known how this outcome measure compares with VE for preventing all influenza illness. Third, our study had limited power to estimate VE for the prevention of hospitalization or death. A larger, multisite study would be required to estimate VE for those outcomes. Finally, the sensitivity and specificity of real-time RTPCR for influenza virus detection have not been fully evaluated, although it is more sensitive than culture [36, 37 VE because of false-negative RT-PCR results. This is a particular concern for patients who were tested later in the course of illness, when viral shedding is reduced.

Assessments of VE against laboratory-confirmed influenza in populations for whom annual influenza vaccination is recommended may influence influenza control recommendations if results are available while the influenza season is underway. For example, increased use of influenza diagnostic tests and antiviral agents may be recommended if lower VE is demonstrated as a result of poor antigenic match, as was the case for influenza B during the first 3 weeks of the 2007–2008 season in Wisconsin [38]. In addition, the use of standard VE assessments during multiple seasons will help clarify the relationship between antigenic match and clinical VE. Similar assessments will be needed during the early stages of a pandemic vaccination program, when it is likely that pandemic or prepandemic vaccines will be administered after limited immunogenicity and safety trials have been completed.

Previous SectionNext Section

additionals

Other members of the Marshfield Influenza Study Group include Lorelle Benetti, Juanita Herr, Debra Kempf, Mary Vandermause with the Marshfield Clinic Research Foundation; and Nancy Cox, Paul Gargiullo, Alexander Klimov, Teresa Wallis, and Xiyan Xu with the Centers for Disease Control and Prevention.

Previous SectionNext Section

Acknowledgements

We thank the following individuals for their contribution to this project: Vicki Allison, Richard Berg, Craig Becker, Nicholas Berger, Carol Beyer, Marilyn Bruger, Laura Coleman, Autumn Deedon, Theresa Esser, Jayne Frahmann, Julie Freidhoff, Gregg Greenwald, Deborah Hilgemann, Stephanie Irving, Tamara Kronenwetter-Koeppel, Kate Konitzer, Adam Lobner, Paul Mitchell, Jordon Ott, Melanie Rayhorn, Jacklyn Salzwedel, Greg Simon, Patrick Stockwell, Sandra Strey, and Kari Weik.

Previous SectionNext Section

Footnotes

  • ↵a Additional members of the study group are listed after the text.

  • Potential conflicts of interest: none reported.

  • Financial support: Research Center for AIDS and HIV infection, New York Harbor Health Care System; National Institutes of Health (grant AI-36984); National Institute for Allergy and Infectious Diseases (contract NO1-AI-75320, entitled “Tuberculosis Research Materials and Vaccine Testing”).

  • Received May 19, 2008.
  • Accepted August 18, 2008.
Previous Section
 

References

  1. 1.↵
    1. Fiore AE,
    2. Shay DK,
    3. Haber P,
    4. et al
    . Prevention and control of influenza:recommendations of the Advisory Committee on Immunization Practices (ACIP), 2007. MMWR Recomm Rep 2007;56:1-54.
    Medline
  2. 2.↵
    World Health Organization. Recommended composition of influenza virus vaccines for use in the 2005–2006 influenza season. Wkly Epidemiol Rec 2005;80:71-5.
    Medline
  3. 3.↵
    1. Russell KL,
    2. Ryan MA,
    3. Hawksworth A,
    4. Freed NE,
    5. Irvine M,
    6. Daum LT,
    7. et al
    . Effectiveness of the 2003–2004 influenza vaccine among US military basic trainees: a year of suboptimal match between vaccine and circulating strain. Vaccine 2005;23:1981-5.
    CrossRefMedlineWeb of Science
  4. 4.
    1. Nordin J,
    2. Mullooly J,
    3. Poblete S,
    4. et al
    . Influenza vaccine effectiveness in preventing hospitalizations and deaths in persons 65 years or older in Minnesota, New York, and Oregon: data from 3 health plans. J Infect Dis 2001;184:665-70.
    Abstract/FREE Full Text
  5. 5.
    1. Hak E,
    2. Nordin J,
    3. Wei F,
    4. et al
    . Influence of high-risk medical conditions on the effectiveness of influenza vaccination among elderly members of 3 large managed-care organizations. Clin Infect Dis 2002;35:370-7.
    Abstract/FREE Full Text
  6. 6.
    Centers for Disease Control and Prevention. Assessment of the effectiveness of the 2003–04 influenza vaccine among children and adults—Colorado, 2003. MMWR 2004;53:707-10.
    Medline
  7. 7.
    1. Sugaya N,
    2. Nerome K,
    3. Ishida M,
    4. Matsumoto M,
    5. Mitamura K,
    6. Nirasawa M,
    7. et al
    . Efficacy of inactivated vaccine in preventing antigenically drifted influenza type A and well-matched type B. JAMA 1994;272:1122-6.
    Abstract/FREE Full Text
  8. 8.↵
    1. Pyhala R,
    2. Haanpaa M,
    3. Kleemola M,
    4. Tervahauta R,
    5. Visakorpi R,
    6. Kinnunen L
    . Acceptable protective efficacy of influenza vaccination in young military conscripts under circumstances of incomplete antigenic and genetic match. Vaccine 2001;19:3253-60.
    CrossRefMedlineWeb of Science
  9. 9.↵
    1. DeStefano F,
    2. Eaker E,
    3. Broste S,
    4. et al
    . Epidemiologic research in an integrated regional medical care system: the Marshfield Epidemiologic Study Area. J Clin Epidemiol 1996;49:643-52.
    CrossRefMedlineWeb of Science
  10. 10.
    1. Greenlee R
    . Measuring disease frequency in the Marshfield Epidemiologic Study Area (MESA). Clin Med Res 2003;1:273-80.
    Abstract/FREE Full Text
  11. 11.
    1. Greenlee R,
    2. Naleway AL,
    3. Vidaillet H
    . Incidence of myocardial infarction in a general population: the Marshfield Epidemiologic Study Area. WMJ 2002;101:46-52.
    Medline
  12. 12.↵
    1. Belongia EA,
    2. Chyou PH,
    3. Greenlee RT,
    4. Perez-Perez G,
    5. Bibb WF,
    6. DeVries EO
    . Diarrhea incidence and farm-related risk factors for Escherichia coli O157:H7 and Campylobacter jejuni antibodies among rural children. J Infect Dis 2003;187:1460-8.
    Abstract/FREE Full Text
  13. 13.↵
    Centers for Disease Control and Prevention. Interim influenza vaccination recommendations, 2004–05 influenza season. MMWR Dispatch 2004;53:1-2.
  14. 14.↵
    Centers for Disease Control and Prevention. Update: influenza vaccine supply and recommendations for prioritization during the 2005–06 influenza season. MMWR Morb Mortal Wkly Rep 2005;54:850.
    Medline
  15. 15.↵
    1. Smith NM,
    2. Bresee JS,
    3. Shay DK,
    4. Uyeki TM,
    5. Cox NJ,
    6. Strikas RA
    . Prevention and control of influenza: recommendations of the Advisory Committee on Immunization Practices (ACIP). MMWR Recomm Rep 2006;55:1-42.
    Medline
  16. 16.↵
    1. Sato M,
    2. Hosoya M,
    3. Kato K,
    4. Suzuki H
    . Viral shedding in children with influenza virus infections treated with neuraminidase inhibitors. Pediatr Infect Dis J 2005;24:931-2.
    CrossRefMedlineWeb of Science
  17. 17.↵
    1. Treanor JJ,
    2. Hayden FG,
    3. Vrooman PS,
    4. et al.,
    5. US Oral Neuraminidase Study Group
    . Efficacy and safety of the oral neuraminidase inhibitor oseltamivir in treating acute influenza: a randomized controlled trial. JAMA 2000;283:1016-24.
    Abstract/FREE Full Text
  18. 18.↵
    1. Heikkinen T,
    2. Salmi AA,
    3. Ruuskanen O
    . Comparative study of nasopharyngeal aspirate and nasal swab specimens for detection of influenza. BMJ 2001;322:138.
    FREE Full Text
  19. 19.↵
    1. Heikkinen T,
    2. Marttila J,
    3. Salmi AA,
    4. Ruuskanen O
    . Nasal swab versus nasopharyngeal aspirate for isolation of respiratory viruses. J Clin Microbiol 2002;40:4337-9.
    Abstract/FREE Full Text
  20. 20.↵
    1. Kendal AP,
    2. Periera MS,
    3. Skehel JJ
    1. Kendal AP,
    2. Periera MS,
    3. Skehel JJ
    . Concepts and procedures for laboratory-based influenza surveillance. In: Kendal AP, Periera MS, Skehel JJ, editors. Atlanta, GA: US Department of Health and Human Services and Pan-American Health Organization; 1982. p. B17-B35.9.
  21. 21.↵
    1. Pearce N
    . What does the odds-ratio estimate in a case-control study? Int J Epidemiol 1993;22:1189-92.
    Abstract/FREE Full Text
  22. 22.↵
    1. Brookmeyer R,
    2. Liang KY,
    3. Linet M
    . Matched case-control designs and overmatched analyses. Am J Epidemiol 1986;124:693-701.
    Abstract/FREE Full Text
  23. 23.↵
    1. Rothman K,
    2. Greenland S
    . 2nd ed. Philadelphia, PA: Lippincott-Raven Publishers; 1998. Modern epidemiology.
  24. 24.↵
    1. Miettinen O
    . Estimability and estimation in case-referent studies. Am J Epidemiol 1976;103:226-35.
    Abstract/FREE Full Text
  25. 25.↵
    1. Orenstein EW,
    2. De Serres G,
    3. Haber MJ,
    4. et al
    . Methodologic issues regarding the use of three observational study designs to assess influenza vaccine effectiveness. Int J Epidemiol 2007;36:623-31.
    Abstract/FREE Full Text
  26. 26.↵
    1. Govaert TM,
    2. Thijs CT,
    3. Masurel N,
    4. Sprenger MJ,
    5. Dinant GJ,
    6. Knottnerus JA
    . The efficacy of influenza vaccination in elderly individuals. A randomized double-blind placebo-controlled trial. JAMA 1994;272:1661-5.
    Abstract/FREE Full Text
  27. 27.
    1. Nichol KL,
    2. Lind A,
    3. Margolis KL,
    4. et al
    . The effectiveness of vaccination against influenza in healthy, working adults. N Engl J Med 1995;333:889-93.
    CrossRefMedlineWeb of Science
  28. 28.
    1. Wilde JA,
    2. McMillan JA,
    3. Serwint J,
    4. O'Riordan MA,
    5. Steinhoff MC
    . Effectiveness of influenza vaccine in health care professionals: a randomized trial. JAMA 1999;281:908-13.
    Abstract/FREE Full Text
  29. 29.↵
    1. Bridges CB,
    2. Thompson WW,
    3. Meltzer MI,
    4. et al
    . Effectiveness and costbenefit of influenza vaccination of healthy working adults: A randomized controlled trial. JAMA 2000;284:1655-63.
    Abstract/FREE Full Text
  30. 30.↵
    1. Edwards KM,
    2. Dupont WD,
    3. Westrich MK,
    4. Plummer WD Jr..,
    5. Palmer PS,
    6. Wright PF
    . A randomized controlled trial of cold-adapted and inactivated vaccines for the prevention of influenza A disease. J Infect Dis 1994;169:68-76.
    Abstract/FREE Full Text
  31. 31.↵
    1. Keitel WA,
    2. Cate TR,
    3. Couch RB,
    4. Huggins LL,
    5. Hess KR
    . Efficacy of repeated annual immunization with inactivated influenza virus vaccines over a five year period. Vaccine 1997;15:1114-22.
    CrossRefMedlineWeb of Science
  32. 32.↵
    1. Ohmit SE,
    2. Victor JC,
    3. Rotthoff JR,
    4. et al
    . Prevention of antigenically drifted influenza by inactivated and live attenuated vaccines. N Engl J Med 2006;355:2513-22.
    CrossRefMedline
  33. 33.↵
    1. Bernstein E,
    2. Kaye D,
    3. Abrutyn E,
    4. Gross P,
    5. Dorfman M,
    6. Murasko DM
    . Immune response to influenza vaccination in a large healthy elderly population. Vaccine 1999;17:82-94.
    CrossRefMedlineWeb of Science
  34. 34.↵
    1. Skowronski DM,
    2. Tweed SA,
    3. De Serres G
    . Rapid decline of influenza vaccine-induced antibody in the elderly: is it real, or is it relevant? J Infect Dis 2008;197:490-502.
    Abstract/FREE Full Text
  35. 35.↵
    1. Skowronski DM,
    2. Masaro C,
    3. Kwindt TL,
    4. et al
    . Estimating vaccine effectiveness against laboratory-confirmed influenza using a sentinel physician network: results from the 2005–2006 season of dual A and B vaccine mismatch in Canada. Vaccine 2007;25:2842-51.
    CrossRefMedlineWeb of Science
  36. 36.↵
    1. Smith AB,
    2. Mock V,
    3. Melear R,
    4. Colarusso P,
    5. Willis DE
    . Rapid detection of influenza A and B viruses in clinical specimens by Light Cycler real time RT-PCR. J Clin Virol 2003;28:51-8.
    CrossRefMedlineWeb of Science
  37. 37.↵
    1. Zitterkopf NL,
    2. Leekha S,
    3. Espy MJ,
    4. Wood CM,
    5. Sampathkumar P,
    6. Smith TF
    . Relevance of influenza a virus detection by PCR, shell vial assay, and tube cell culture to rapid reporting procedures. J Clin Microbiol 2006;44:3366-7.
    Abstract/FREE Full Text
  38. 38.↵
    Centers for Disease Control and Prevention. Interim within-season estimate of the effectiveness of trivalent inactivated influenza vaccine— Marshfield, Wisconsin, 2007–08 influenza season. MMWR 2008;57:393-8.
    Medline
| Table of Contents

This Article

  1. J Infect Dis. (2009) 199 (2): 159-167. doi: 10.1086/595861
  1. AbstractFree
  2. » Full Text (HTML)Free
  3. Full Text (PDF)Free

Classifications

Share

  1. Email this article

Navigate This Article

Current Issue

  1. March 1, 2012 205 (5)
  1. Journal of Infectious Diseases
  1. Alert me to new issues

Most