 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
a British Columbia Centre for Disease Control, Vancouver, British Columbia, Canada
b Westcoast Clinical Research, Coquitlam, British Columbia, Canada
c Direction de Santé Publique, Chicoutimi, Quebec, Canada
d Vancouver Coastal Health Authority, Vancouver, British Columbia, Canada
e Fraser Health Authority, Surrey, British Columbia, Canada
f Direction Générale de la Santé Publique, Ministère de la Santé et des Services Sociaux du Québec, Québec, Québec, Canada
g Institut National de Santé Publique du Québec, Québec, Québec, Canada
h Public Health Agency of Canada, Ottawa, Ontario, Canada
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONREFERENCES |
|---|
METHODS. Parents bringing infants and toddlers for influenza immunization to clinics in Quebec or British Columbia consented to structured telephone interview 5 to 10 days later. One adult provided information for all household members. Symptom experience commencing before and after immunization was assessed. Non-immunized persons also served as a comparison group for immunized household members.
RESULTS. Sample included 690 immunized infants and toddlers and 1801 household members, 1374 immunized. Only fussiness, fever, decreased appetite, drowsiness, and nasal congestion/coryza were reported for >5% of infants/ toddlers within 72 hours of immunization, but only arm discomfort was reported among >5% of immunized household contacts. In multivariate analysis, muscle ache was the only systemic symptom reported more often by immunized household members compared to non-immunized persons. Oculorespiratory symptoms were infrequent and there was no difference between immunized and non-immunized household members in their report. Less than 1% of adults required time off work because of adverse events following influenza immunization in the household. Less than 2% of subjects experiencing an adverse event following influenza immunization were considered unlikely to be vaccinated again.
CONCLUSION. Influenza vaccine is well-tolerated by infants, toddlers and their household members. Post-marketing observational designs are an expedient way to assess adverse events following influenza immunization. These methods should be established and rehearsed annually in preparation for a pandemic.
Key Words: influenza • trivalent inactivated vaccine • immunization • adverse events • infants • toddlers
Abbreviations: ORS—oculorespiratory syndrome • AEFII—adverse event following influenza immunization • TIV—trivalent inactivated vaccine • OR—odds ratio • CI—confidence interval
In 2004, the Advisory Committee on Immunization Practice in the United States and the National Advisory Committee on Immunization in Canada first recommended annual influenza vaccination for infants and toddlers 6 to 23 months of age and their household contacts.1,2 Influenza vaccine is generally considered safe for children, although the number of infants and toddlers included in published safety studies to date has been small (<1000).1,3 Systemic symptoms after receipt of influenza vaccine have been noted most often among persons (eg, young children) with no previous experience with influenza virus antigens.1
During the 2000 immunization campaign in Canada, oculorespiratory syndrome (ORS) was first identified as an adverse event following influenza immunization (AEFII).4–9 The syndrome consisted of bilateral red eyes, respiratory symptoms, and/or facial swelling beginning within 24 hours after immunization.5–9 The cause was not determined, although the implicated vaccine (Fluviral S/F; Biochem Pharma, Laval, Canada) was found to contain a proportion of aggregated unsplit virions greater than expected.4,8 In descriptive evaluations, rates of ORS were highest among middle-aged women (16%) and lowest among elderly persons (<3%).5 A similar pattern was observed for immunized children, with rates of ORS close to the highest observed for adults (13%), higher among first-time recipients (24%) than among repeat recipients (10%).10 An association between first-time vaccination and reports of fever, aching muscles, nausea, and vomiting was observed previously in association with influenza vaccine and was also noted among immunized children in Canada in 2000.1,10 In placebo-controlled trials to assess reformulated vaccine for 2001, the vaccine-attributable risk of ORS was 2.9% among adults unaffected previously and 33% among adults affected previously in 2000.11,12 Through a phase IV (postmarketing) telephone survey, we assessed AEFIIs, including ORS, among immunized infants, toddlers and their household contacts during the first year of implementation of this publicly funded program in Canada in 2004.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
2 test for univariate analyses and logistic regression for multivariate analyses.
Recruitment and Interview
Participants were recruited from among parents bringing their infant or toddler to immunization clinics conducted by public health nurses in 2 Canadian provinces, British Columbia and Quebec. British Columbia is the westernmost province, whereas Quebec is located nearly 5000 km away, in central Canada. In accordance with published guidelines of the National Advisory Committee on Immunization, influenza vaccine was also offered and provided free at the same time to siblings, parents, and other household members.2
In Quebec, the immunizing public health nurse also obtained study consent. In British Columbia, separate research staff members obtained consent before vaccination by a public health nurse. Details of vaccine lot, concurrent immunizations, and numbers of persons in the household were recorded at the time of vaccination.
Separate study staff members (British Columbia) or public health nurses (Quebec) conducted telephone interviews with a single household member 5 to 10 days after infant or toddler immunization, with the same structured questionnaire in English or French. One respondent was interviewed about his or her own experience and also acted as a surrogate respondent, providing information on symptom experiences for all other household members. The questionnaire elicited information about immunization history and specific symptoms after the current immunization for the vaccinated infant or toddler. Experiences with previous doses (repeat recipient) or subsequent doses (first-time recipient) were not elicited. Symptom experiences were elicited for all members of the household, whether vaccinated or not. Except for the infant or toddler, the immunization status of household members was not confirmed beyond surrogate report. In addition to solicited symptoms, respondents were invited to volunteer any other unsolicited symptoms experienced within the household during the same time period. For vaccinated household members, the reference date was the date of their own immunization. For nonvaccinated household members, the reference date was the date of the infant's or toddler's immunization. The questionnaire also elicited information on the impact of AEFIIs on health care use, activities of daily living, and likelihood of being vaccinated again if the vaccine was provided free (very likely, likely, undecided, unlikely, or very unlikely).
Case Definitions
ORS was defined as bilateral red eyes, respiratory symptoms (cough, wheeze, chest tightness, difficulty breathing, sore throat, hoarseness, or throat tightness/difficulty swallowing), or facial swelling beginning within 24 hours after vaccination.8 Otherwise, to be considered an AEFII, solicited or unsolicited symptoms had to begin within 72 hours after vaccination.
Vaccine Components
Vaxigrip (Aventis Pasteur, Val De Reuil, France), a lower thimerosol-containing, split trivalent inactivated vaccine (TIV) for intramuscular injection was provided for infants and toddlers, at a dose of 0.25 mL. Fluviral S/F was provided as split TIV for older children and adults, at a dose of 0.25 mL (administered intramuscularly) for children <35 months of age and a dose of 0.5 mL (administered intramuscularly) for persons 
3 years of age. Two doses of vaccine, administered 1 month apart, were provided to first-time recipients <9 years of age.2
Vaccine antigens for both Vaxigrip and Fluviral S/F included 15 µg of hemagglutinin for each of A/New Caledonia/20/99 (H1N1), A/Wyoming/3/2003 (H3N2), and B/Jiangsu/10/2003.2 In addition, in a 0.5-mL dose, Fluviral S/F contained 50 µg of thimerosol (0.01%, w/v) and trace amounts of sodium deoxycholate and formaldehyde. Vaxigrip contained 2 µg of thimerosol (0.0004%, w/v), 
30 µg of formaldehyde, and trace amounts of Triton X-100, sucrose, and neomycin.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Baseline Characteristics
Infants and Toddlers
In Quebec, 316 immunized infants and toddlers received the same lot of Vaxigrip and 4 received Fluviral S/F. In British Columbia, 282 received the same Vaxigrip lot as in Quebec, 70 received another Vaxigrip lot, and 18 received Fluviral S/F.
Baseline characteristics of the infants and toddlers are shown in Table 1. In Quebec, compared with British Columbia, the mother more often was the respondent and the child more often was white, more often had an underlying respiratory condition, more often received concurrent vaccines, and more often was a first-time influenza vaccine recipient.
|
Baseline characteristics of the household contacts are shown in Table 2. Household contacts in Quebec more often were first-time recipients and more often had an underlying condition (notably respiratory). In both British Columbia and Quebec, fathers were significantly less likely to have been immunized (British Columbia: 70%; Quebec: 60%), compared with mothers (British Columbia: 85%, P < .001; Quebec: 86%, P < .001) or siblings (British Columbia: 78%, P = .002; Quebec: 86%, P < .001).
|
|
2 respiratory symptoms, and 6% (1 child) experienced a combination of red eyes and respiratory symptoms. Fever was present with ORS symptoms for 38% (8 infants and toddlers), and other systemic symptoms, such as fussiness, drowsiness, loss of appetite, nausea, vomiting, or diarrhea, occurred among another 28%. ORS symptoms lasted 6 days in 28% of the cases. None of the 21 infants and toddlers with ORS symptoms had an underlying respiratory condition or other condition. All ORS symptoms were observed in association with the same commonly administered Vaxigrip lot number. Twenty reports were in association with first-ever influenza immunization. Three children had received other vaccines concurrent with influenza vaccine.
Only fussiness (21%), fever (11%), decreased eating/drinking (8%), and drowsiness (7%) were reported among >5% of immunized infants and toddlers (Table 3). No seizures or other neurologic symptoms were reported. In addition to the specifically solicited symptoms shown in Table 3, the symptoms volunteered most commonly as other symptoms were coryza/nasal congestion (6% in Quebec and 10% in British Columbia) and limb sensitivity (5% in Quebec and 3% in British Columbia).
Multivariate Analysis
In logistic regression analysis combining British Columbia and Quebec (n = 681) and adjusting for province of residence, first versus repeat immunization, gender, and underlying condition, the only variable associated independently with ORS symptoms among immunized infants and toddlers was residence in Quebec, with an adjusted odds ratio (OR) of 7.7 (95% CI: 2.1–27.8). Infants and toddlers in Quebec more often experienced other symptoms, such as fussiness (OR: 1.5; 95% CI: 1.03–2.2), crying (OR: 4.1; 95% CI: 1.2–14.7), and cough (OR: 6.1; 95% CI: 1.4–27.8). No other associations were found.
Social and Health Care Effects
Eight adults required time off work (range: 1–3 days) because of an AEFII involving an immunized infant or toddler (adult work loss was associated with 1% of immunized infants and toddlers). Medical consultation was sought for 38 (6%) of 690 immunized infants and toddlers in relation to an AEFII (24 at a doctor's office, 5 at the emergency department, and 9 through telephone consultation). Among 296 infants and toddlers with AEFIIs, 274 (92%) were considered very likely or likely to be immunized again if vaccine was provided free; 7% of respondents were undecided, and 1% were unlikely to have their infant or toddler immunized again.
Household Contacts
Frequencies and Univariate Associations
Among immunized household contacts, the rates of cough and sore throat were significantly higher in Quebec than in British Columbia; in neither province, however, nor overall, were these rates significantly different between immunized and nonimmunized persons (Table 4). Unlike immunized infants and toddlers, other immunized household members in Quebec did not experience more cough (or sore throat) with onset before immunization, compared to British Columbia. Only arm discomfort (26%) was reported by >5% of immunized household contacts. Persons receiving influenza vaccine for the first time, compared with repeat recipients, experienced higher rates of fever (4% vs 2%; OR: 2.2; 95% CI: 1.1–4.7), vomiting (2% vs 0.2%; OR: 10; 95% CI: 1.4–209), diarrhea (2% vs 0.6%; OR: 3.2; 95% CI: 0.98–11.5), headache (5% vs 2%; OR: 2.7; 95% CI: 1.4–5.6), ORS (2.5% vs 1%; OR: 2.6; 95% CI: 0.95–7.4), and muscle aches (5% vs 2%; OR: 2.0; 95% CI: 1.05–4.0).
|
Immunized household contacts in Quebec reported more diarrhea (2%) than did immunized contacts in British Columbia (0.7%; P < .001). There were no significant differences between British Columbia and Quebec in the report of any symptom among nonimmunized persons. In combined analysis, higher rates were found among immunized persons, compared with nonimmunized persons, for headache, muscle ache, and limitation of limb movement (Table 4). Nasal congestion/coryza were reported by 1% of both immunized and nonimmunized household contacts.
Overall, the rates of AEFIIs were highest among immunized mothers, compared with immunized fathers or siblings, for swelling at the injection site (3% vs 1% vs 1%; P = .03), limitation of limb movement (7% vs 4% vs 2%; P = .004), headache (5% vs 4% vs 1%; P = .004), and muscle aches (5% vs 4% vs 1%; P = .01). Persons acting as respondents for the entire household reported more swelling at the injection site and limitation of limb movement for themselves, compared with other household members (swelling at the injection site: 3% vs 1%, P = .002; limitation of limb movement: 7% vs 3%, P = .003), but did not otherwise differ in their symptom experiences.
Multivariate Analysis
In logistic regression analysis among household contacts (n = 1675), with adjustment for province of residence, immunization status, gender, underlying condition, relationship to infant (mother, father, sibling, or other), and whether the participant was the household respondent, injection site reactions and limited limb movement were reported more often by immunized persons. Muscle aches were the only systemic symptom associated significantly with influenza immunization (OR: 2.9; 95% CI: 1.0–8.2). ORS (OR: 3.7; 95% CI: 1.5–8.9), sore throat (OR: 4.9; 95% CI: 1.4–17.8), and diarrhea (OR: 2.7; 95% CI: 1.01–7.2) were associated with residence in Quebec. When only immunized household contacts were included (n = 1278) and "whether immunized" was replaced by first-time immunized or repeat immunized, fever (OR: 2.7; 95% CI: 1.3–5.6), headache (OR: 2.7; 95% CI: 1.4–5.3), muscle aches (OR: 2.2; 95% CI: 1.2–4.1), and vomiting (OR: 9.2; 95% CI: 1.2–71.4) were associated independently with first-time immunization.
Social and Health Care Effects
Among immunized household contacts, 16 (<1%) required medical consultation in relation to AEFIIs; none was hospitalized. Twenty-five immunized household contacts (1%) experienced sleep disturbances after immunization. Four immunized siblings missed school. Four adults required time off work because of their symptoms or those of an immunized child. Among immunized household contacts who reported any AEFII, 1% of mothers, 3% of fathers, 2% of siblings, and 0% of other household contacts were considered unlikely or very unlikely to be reimmunized if vaccine was provided free.
Among nonimmunized household contacts, 3 (<1%) required medical consultation for symptoms they experienced; none was hospitalized. Six nonimmunized household contacts (1%) experienced sleep disturbances because of their own symptoms. One nonimmunized sibling missed school. Five adults missed work because of symptoms of a nonimmunized household member.
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Fussiness (21%) and fever (11%) were reported most often, but symptoms were otherwise infrequent among immunized infants and toddlers in our survey. Previous studies with TIV among children reported similar rates.1 In a study of 791 healthy children 1 to 15 years of age, 11.5% of children 1 to 5 years of age experienced fever.13 In another study of 52 children 6 months to 4 years of age, 27% experienced fever and 25% experienced irritability and insomnia.14 Among very young children, unrelated causes of fussiness or irritability, such as teething, should also be considered. Runny nose or nasal congestion was reported previously for 20% to 75% of children and 28% to 78% of adults who received live attenuated inactivated vaccine.1 These symptoms were not highlighted previously in safety reviews of TIV. In our survey, nasal symptoms were reported spontaneously for 5% to 10% of immunized infants and toddlers but were not reported at a significantly higher rate for other immunized household members, compared with nonimmunized household members (1%).
With the exception of arm discomfort (26%), no single AEFII was reported by >5% of immunized household contacts, and there were few differences between immunized and nonimmunized persons. Even so, we likely overestimated the vaccine-attributable risk, because our control group did not receive placebo as a baseline for comparison. Headache, muscle ache, and limited limb movement were reported more often by immunized versus nonimmunized household contacts in univariate analyses, but muscle ache was the only systemic symptom associated with immunization in multivariate analysis. In previous placebo-controlled studies among adults, the most frequent AEFII has been soreness at the injection site (10–64%), although another study of nearly 2000 adults and children also identified body aches as the only systemic symptom reported more frequently by TIV recipients than control subjects.1,15–18 As in previous surveys, we found first-time influenza vaccine recipients experienced higher rates of systemic symptoms such as fever, headache, muscle ache, and vomiting compared to repeat vaccine recipients. Overall, AEFIIs resulted in minimal societal or health care effects, with no significant difference between immunized and nonimmunized household contacts. Less than 1% of adults required time off work because of an AEFII involving themselves or their children.
We did not include an age-specific control group for infants and toddlers in this survey, but provincial differences in symptom experiences are illuminating. Quebec reported more infants and toddlers with ORS symptoms within 24 hours after influenza immunization, compared with British Columbia. Because the vast majority of infants and toddlers received vaccine from the same lot, variation in vaccine product is unlikely to explain provincial differences. Although the difference may be explained by random variation, we think it is most likely attributable to background illness, unrelated to vaccine, as evidenced by the greater proportion of infants and toddlers in Quebec compared to British Columbia, with cough or other ORS symptoms preceding immunization. The current case definition for ORS, which is a compilation of several nonspecific symptoms, is especially prone to the cumulative effects of other causes. In 2000, ORS was first brought to the attention of health authorities in Canada through passive surveillance, without special prompting. It was characterized as a syndrome consisting of multiple ocular, respiratory, and facial symptoms vivid enough to be reported spontaneously and clustered dramatically around receipt of one manufacturer's vaccine.5,6,10,19 Since then, ORS reports to the passive surveillance system in Canada have declined each year.19,20 The experience with ORS in 2000 should thus be distinguished clearly from isolated reports of cough or sore throat identified through active survey and individual symptom prompting in 2004. Among infants and toddlers in our survey, the high proportion of cases with only 1 ORS symptom, associated with fever and other systemic symptoms and with prolonged duration, suggests a prominent role for other causes. Of the 21 infants with ORS symptoms in this survey, only 1 experienced multiple symptom categories, which suggests either a very low incidence of vaccine-attributable ORS or mild presentation if it does occur.
Rates of ORS symptoms in Quebec were not substantially higher than expected background rates of acute respiratory illness in the community in late fall and early winter, especially among young children.21–23 Provincial differences in reported rates of cough, sore throat, and diarrhea may reflect variability in virus circulation. National summaries of virus identification in November and December 2004 showed significantly higher rates of detection of respiratory syncytial virus and adenovirus among specimens tested in Quebec (7% and 7%, respectively), compared with British Columbia (2% and 0.5%, respectively; P < .001). Quebec also reported more norovirus outbreaks during the study period. We did not assess smoking prevalence among households in our survey, but this could also explain differences in respiratory symptoms. Smoking rates are consistently highest in Quebec (25%) and lowest in British Columbia (16%), compared with the rest of Canada (22%).24 In summary, we think that the proportion of participants with ORS symptoms in British Columbia (<1%) may be more representative of experiences after influenza vaccination, compared with the rate found in Quebec. Even the British Columbia rate likely represents an overestimate and includes background causes unrelated to vaccine. Although rates are low overall, the provincial differences in AEFII rates illustrate the importance of including nonimmunized, age-matched and geographically matched, control subjects in vaccine safety assessments, or comparing experience in the same subject before and after immunization.
Reports of AEFIIs were also higher among immunized household contacts in Quebec, compared with British Columbia, but this difference was not evident among nonimmunized persons. We did not record the lot numbers of vaccine received by household contacts; therefore, it is not known whether this could explain the differences between provinces among immunized household members. Response, recall, or interview biases or those related to losses to follow-up monitoring might have played a role. Mothers more often were immunized, more often reported symptoms, and more often were the respondents in Quebec. Surrogate responders could amplify the individual effects of biases at the household level, when all members are included but their experiences are relayed entirely through 1 person. As immunized responders, mothers might have been more attuned to their own symptom experiences or those of other immunized household members, rather than those of nonimmunized household members. A smaller proportion of eligible parents initially consented to participate in Quebec, and fewer consenting Quebec participants ultimately completed a telephone interview. British Columbia hired specific research staff members dedicated to study-related tasks, whereas Quebec relied on existing public health nurses to conduct the study in addition to their busy service duties. Bias might have been introduced into results from Quebec if the 15% of initially consenting parents who subsequently declined or dropped out had no symptoms to report. It is reassuring to note that, despite these factors, each of which could have skewed results toward higher estimates, our survey still found only low rates of AEFIIs overall.
Given the speed with which influenza campaigns are conducted each fall, timeliness is essential for identifying and responding to potential concerns. The methods and infrastructure for rapid investigation should be established before mass campaigns begin and should be used with each season's new influenza vaccine. Such approaches may become critical during a pandemic when vaccine is administered to large segments of the population, with many being vaccinated for the first time. Repeating safety surveys annually will enhance public confidence, test protocols, and establish expected variations in actively solicited AEFII rates. Postmarketing surveys are an efficient way to pace vaccine development and safety monitoring with public expectations and the goal of administering influenza vaccine rapidly to as many eligible persons as possible. Our survey, which was planned in the summer and completed within 4 weeks in the fall, rapidly confirmed low rates of AEFIIs and ruled out serious events in the week after immunization for several new cohorts of previously unvaccinated persons. This population-based approach is more efficient than clinical trials and could be repeated annually and during a pandemic. Public health investment in the infrastructure to support ongoing postmarketing research and evaluation is required. Future modifications could include verification of immunization status, longer follow-up monitoring, inclusion of age-matched and geographically-matched control (non-immunized) subjects, or comparison of symptom experience over the same time period before and after immunization.
|
ACKNOWLEDGMENTS |
|---|
We thank the parents who participated in this survey, public health nurses in British Columbia and Quebec for their assistance, and Lisan Kwindt for database management. We also thank Aleina Tweed and Dr Eleni Galanis for providing thoughtful feedback on earlier versions of the manuscript.
|
FOOTNOTES |
|---|
Address correspondence to Danuta M. Skowronski, MD, British Columbia Centre for Disease Control, 655 West 12th Ave, Vancouver, British Columbia, Canada V5Z 4R4. E-mail: danuta.skowronski{at}bccdc.ca
Financial Disclosure: Drs Skowronski and De Serres previously received funding for unrelated research from Sanofi Pasteur (formerly Aventis Pasteur, manufacturer of Vaxigrip) and GlaxoSmithKline.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. J. Hambidge, J. M. Glanz, E. K. France, D. McClure, S. Xu, K. Yamasaki, L. Jackson, J. P. Mullooly, K. M. Zangwill, S. M. Marcy, et al. Safety of Trivalent Inactivated Influenza Vaccine in Children 6 to 23 Months Old JAMA, October 25, 2006; 296(16): 1990 - 1997. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
||
|
||
Home | My Pediatrics | Journal Information | Current Issue | Past Issues | Subscriptions & Services | Contact Us Click here for faster international access © 2006 American Academy of Pediatrics. All rights reserved. |
||
|
||
