BACKGROUND: The extent to which pandemic H1N1 influenza (pH1N1) differed from seasonal influenza remains uncertain.
METHODS: By using active surveillance data collected by the Immunization Monitoring Program, Active at 12 Canadian pediatric hospitals, we compared characteristics of hospitalized children with pH1N1 with those with seasonal influenza A. We compared demographics, underlying health status, ICU admission, and mortality during both pandemic waves versus the 2004/2005 through the 2008/2009 seasons; influenza-related complications and hospitalization duration during pH1N1 wave 1 versus the 2004/2005 through the 2008/2009 seasons; and presenting signs and symptoms during both pH1N1 waves versus the 2006/2007 through the 2008/2009 seasons.
RESULTS: We identified 1265 pH1N1 cases (351 in wave 1, 914 in wave 2) and 1319 seasonal influenza A cases (816 from 2006/2007 through 2008/2009). Median ages were 4.8 (pH1N1) and 1.7 years (seasonal influenza A); P < .0001. Preexisting asthma was overrepresented in pH1N1 relative to seasonal influenza A (13.8% vs 5.5%; adjusted P < .0001). Symptoms more often associated with pH1N1 wave 1 versus seasonal influenza A were cough, headache, and gastrointestinal symptoms (adjusted P < .01 for each symptom). pH1N1 wave 1 cases were more likely to have radiologically confirmed pneumonia (adjusted odds ratio = 2.1; 95% confidence interval = 1.1–3.8) and longer median length of hospital stay (4 vs 3 days; adjusted P = .003) than seasonal influenza A. Proportions of children requiring intensive care and deaths in both pH1N1 waves (14.6% and 0.6%, respectively) were not significantly different from the seasonal influenza A group (12.7% and 0.5%, respectively).
CONCLUSIONS: pH1N1 in children differed from seasonal influenza A in risk factors, clinical presentation, and length of hospital stay, but not ICU admission or mortality.
- CI —
- confidence interval
- DFA —
- direct fluorescent antibody assay
- IMPACT —
- Immunization Monitoring Program, Active
- OR —
- odds ratio
- pH1N1 —
- pandemic H1N1 influenza
- PCR —
- polymerase chain reaction
What’s Known on This Subject:
Although several studies have demonstrated increased morbidity and mortality with pH1N1 in children, others have found its clinical course to be similar to seasonal influenza. Moreover, most studies were conducted at single centers, thus raising concerns about generalizability of findings.
What This Study Adds:
This analysis provides national-level active hospital-based surveillance data comparing pH1N1 with 5 previous years of seasonal influenza A and demonstrates differences in risk factors and clinical presentation but not in ICU admission or mortality.
While pandemic (H1N1) 2009 influenza (pH1N1) affected many more individuals,1,2 garnered more media attention, and elicited more anxiety than influenza in previous seasons, it is less clear to what extent pH1N1 differed from seasonal influenza in illness severity and risk groups. Several pediatric studies comparing pH1N1 with seasonal influenza have demonstrated higher rates of pneumonia, hospitalization, ICU admission, and mortality associated with pH1N1.3–5 Others have found the clinical course and outcome of pH1N1 to be similar to seasonal influenza.6–11 Moreover, many of these studies involved small sample sizes or were conducted at single centers, thus raising concerns about generalizability of findings.6,7,9–11 Similarly, asthma has been reported to be a more significant risk factor for adverse outcomes with pH1N1 infection than with seasonal influenza, but only in single-centered studies,6,9 or only when compared with seasonal influenza A and B combined.12
Because many now regard the pH1N1 virus as a seasonal influenza subtype, a better understanding of how it differs from previous seasonal influenza subtypes will inform changes required of current preventive and therapeutic guidelines. Insights into the extent to which at-risk groups for severe seasonal influenza can be extrapolated to a pandemic strain may also be useful for resource allocation planning by public health authorities in future pandemics. Given that genetic reassortment of the pH1N1 virus with cocirculating seasonal viruses has been shown, in an experimental model, to potentially lead to increased pathogenicity,13 a comparison of pH1N1 with previous seasonal influenza subtypes can serve as a baseline for ongoing monitoring of changes in disease severity.
The Canadian Immunization Monitoring Program, Active (IMPACT) has conducted surveillance of children hospitalized with influenza at 12 children’s hospitals since 2004,14 continuing through the 2009 pandemic.15 Thus, this study’s primary objective was to compare the characteristics of children admitted to pediatric hospitals across Canada with pH1N1 with those with influenza A in the 5 preceding seasons. Secondary objectives were to compare selected parameters in the second versus first wave of the pandemic, and identify factors associated with severe pH1N1 and seasonal influenza A infection.
Patients and Setting
Patients were ascertained through active surveillance for laboratory-confirmed influenza admissions at the 12 pediatric referral centers of IMPACT, a national surveillance initiative with centers in Newfoundland, Nova Scotia, Quebec, Ontario, Manitoba, Saskatchewan, Alberta, and British Columbia. These centers admit >75 000 children annually, account for ∼90% of pediatric tertiary care beds in the country, receive referrals from all provinces and territories, and serve a population of ∼50% of Canada’s children.16 All centers have institutional ethics approval for surveillance.
A trained nurse monitor at each center screens daily laboratory results for cases and reports aggregate weekly counts by age group and influenza type to the Public Health Agency of Canada via a web-based data reporting platform (Daciforms version 2.5.1). Case details are then abstracted from medical charts by using electronic standardized data collection forms. Data collected include demographics, preexisting medical conditions, vaccination history, clinical manifestations, treatment, complications, and level of care required.
Seasonal influenza A cases comprised children aged ≤16 years hospitalized for laboratory-confirmed influenza A from September 11, 2004 to March 31, 2009. Analysis was restricted to influenza A, because this was the type associated with the pandemic. Surveillance in seasons other than the pandemic typically lasted 7 months (from the start of each influenza season to June 30). For the 2008/2009 season, we included only cases reported up to March 31, 2009 to minimize the risk of case misclassification related to the potential circulation of pH1N1 in April before the availability of polymerase chain reaction (PCR) diagnosis for pH1N1. Consequently, 8 cases of influenza A reported in April 2009 were excluded. pH1N1 cases comprised admissions for laboratory-confirmed influenza A from May 1, 2009 to March 14, 2010, with those reported up to August 29, 2009 grouped as wave 1 cases.17 Whereas laboratory screening for pH1N1 cases used a combination of PCR, direct fluorescent antibody assay (DFA), and viral culture, all cases were confirmed by using PCR specific for pH1N1, which was in place at all IMPACT hospitals by June 2009. DFA- or culture-positive, but pH1N1 PCR-negative patients were excluded; 1 such patient was excluded in wave 2. Seasonal influenza A cases were defined as patients who tested positive by PCR, DFA, and/or viral culture. These case definitions reflect the varying laboratory-testing algorithms used at participating sites.
Variables of Interest
We focused on demographics, health status, seasonal influenza vaccination status, presenting signs and symptoms, treatments (antiviral and antibiotic), and measures of illness severity (influenza-related complications, mortality, ICU admission, mechanical ventilation, extracorporeal membrane oxygenation, and lengths of ICU and hospital stay). Some of these parameters were not captured in all seasons. Hence, comparison between pH1N1 and seasonal influenza A was confined to both pandemic waves vs 2008/2009 season for ethnicity; wave 1 vs 2006/2007 through 2008/2009 seasons for presenting signs and symptoms; and wave 1 versus all 5 preceding seasons for vaccination status, treatments, complications, and length of hospital stay. For comparison of pandemic wave 2 with wave 1, analysis of vaccination status, treatments, complications, and length of hospital stay was restricted to patients admitted to ICU as during wave 2, these details were captured only for patients needing intensive care because of resource constraints from the large volume of reported cases. Radiologically confirmed pneumonia and ICU admission were the measures of disease severity for which we constructed models of risk for pH1N1 and seasonal influenza A; the risk model for pH1N1-associated pneumonia was based on data from wave 1 of the pandemic.
All cases, grouped by health status, were classified as having no underlying health condition or at least 1 underlying condition. Children in the latter category were further classified as those with (1) asthma (without other underlying conditions), (2) neurologic condition (as the primary underlying condition associated with the admission), (3) any vaccine-indicated condition other than asthma or neurologic condition, or (4) any other underlying condition. Neurologic conditions included those that compromise the management of respiratory secretions and are associated with increased risk of aspiration, such as cerebral palsy, seizure disorders, and neuromuscular disorders. Vaccine-indicated conditions were those considered to be at high risk for influenza-related complications (excluding age <2 years)18,19 and for which seasonal influenza vaccination was recommended by Canadian guidelines up to the 2009/2010 season.18
To examine differences between pH1N1 and seasonal influenza A, we first compared each variable of interest across the seasonal influenza season(s) and pandemic wave(s) for which data were available. Significant results for variables other than ethnicity were then retested without the pandemic cases to determine whether changes over time existed in the seasonal data or were due to the pandemic; for ethnicity, retesting was conducted without the seasonal cases.
We analyzed differences in continuous variables by using Student's t test or analysis of variance (for normally distributed data) and Mann-Whitney or Kruskal-Wallis methods (for nonnormally distributed data). Comparisons of categorical variables were made by using the χ2 or Fisher's exact test as appropriate. Where potential confounding variables were present, we performed multivariate logistic or linear regression and/or stratified analysis as indicated in the main text and footnotes of tables.
To identify independent predictors of pneumonia and ICU admission, we conducted univariate and multivariate logistic regression. Given the significant missing data for ethnicity and vaccination status, we constructed multivariate models that included and excluded these variables. We considered for the multivariate models all variables with univariate P values <.1. Most analyses were determined a priori based on previous literature or clinical experience. All tests were 2-sided and a P value of <.05 was considered statistically significant. Data were analyzed by SPSS statistical software (version 16.0, SPSS Inc, Chicago, IL).
We identified 1265 hospitalized pH1N1 patients distributed in 2 distinct waves: 351 in wave 1 (May 1 to August 29, 2009) and 914 in wave 2 (August 30, 2009 to March 14, 2010) (Fig 1). There were 1319 children hospitalized with seasonal influenza A during the preceding 5 seasons, with a mean of 263.8 admissions per season (range, 192–313). Of these, 816 cases were from 2006/2007 to 2008/2009. Because there were no significant differences in the variables of interest across the seasonal influenza seasons, these data are presented in aggregate.
Age distributions of hospitalized and ICU-admitted pH1N1 patients differed from those of seasonal influenza A patients, with a shift toward older children among pH1N1 patients (Tables 1 and 2). Non-European descent was overrepresented among children hospitalized with pH1N1 relative to those with seasonal influenza A in 2008/2009, the only previous season for which information on ethnicity was available. Both hospitalized and ICU pH1N1 patients, compared with their seasonal influenza A counterparts, had a higher age-adjusted odds of having asthma but similar age-adjusted odds of having no underlying condition, any vaccine-indicated condition, a neurologic condition, or any vaccine-indicated condition other than asthma or a neurologic condition. To account for the challenge of diagnosing asthma at younger than age 2 and the potential for increased influenza testing among older patients with asthma during the pandemic, we also restricted the comparison of proportions of patients with asthma to the 2 to 9 year age stratum and found a similar odds ratios (odds ratio [OR] = 2.63 [95% confidence interval (CI), 1.83–3.78]; P < .001 for hospitalized patients and OR = 3.622 [95% CI, 1.260–10.415]; P = .017 for ICU patients). For patients admitted to ICU, previous seasonal influenza vaccination was more frequently documented in pH1N1 (wave 1) relative to seasonal influenza A, even after adjustment for age (<2 vs ≥2 years) and presence of a vaccine-indicated condition. Differences observed between the pandemic waves were a shift toward younger children in wave 2 and overrepresentation of non-European ethnic groups in wave 1 for hospitalizations, but not for ICU admissions.
Clinical Presentation, Treatment, and Illness Severity
After adjustment for potential confounders (including age), proportions of children hospitalized with pH1N1 who presented with headache, cough and gastrointestinal symptoms were significantly greater than those with seasonal influenza A from 2006/2007 through 2008/2009(Table 3). Antivirals were prescribed more often for pH1N1-infected children admitted to hospital during wave 1 and to ICU during both waves, whereas antibiotic administration was more likely in hospitalized pH1N1 patients in wave 1 but not in ICU admissions in both waves (Tables 2 and 4). Among ICU patients with pH1N1 infection, antiviral use was also more frequent during wave 2. These treatment differences remained statistically significant after adjustment for potential confounders.
Radiologically confirmed pneumonia was diagnosed more frequently among both hospitalized and ICU-admitted pH1N1 patients, after adjustment for age (continuous variable) and presence of asthma. In contrast, the age-adjusted odds of croup was lower in both hospitalized and ICU pH1N1 patients. Median duration of hospitalization for children with pH1N1 was 4 days compared with 3 for those with seasonal influenza A (P < .001). In a linear regression model of duration of hospital stay (after logarithmic transformation) that included influenza subtype (pH1N1 from wave 1 versus seasonal influenza A), age (continuous variable), and presence of asthma, influenza subtype remained an independent predictor of length of stay, with pH1N1 subtype associated with a longer length of stay (P = .003). Proportions of ICU admissions and deaths were not significantly different between pH1N1 and seasonal influenza A when comparisons included both pandemic waves. To explore the potential confounding effect of differential use of antiviral agents on the need for ICU admission and mortality, we repeated the analyses without wave 2 cases. Although univariate analysis revealed a significantly higher proportion of pH1N1 cases admitted to ICU (P = .04), this difference was no longer significant when adjusted for age (continuous variable), asthma, and antiviral use, or when adjusted for antiviral use alone (Table 4). Similar multivariate analyses also demonstrated no significant difference in either all-cause or influenza-attributable mortality. Among ICU cases, there was no significant difference in length of ICU stay, mechanical ventilation or occurrence of other serious influenza-associated complications between pH1N1 and seasonal influenza A. A comparison of disease severity between the pandemic waves revealed a greater proportion of children requiring mechanical ventilation during wave 2. All other measures of severity were similar between the 2 waves for ICU cases, whereas the need for ICU admission and mortality in hospitalized patients were similar between wave 1 and 2.
Risk Factors for Pneumonia and ICU Admission in pH1N1 and Seasonal Influenza A Infection
In multivariate models with ethnic origin included as an independent variable, aboriginal ethnicity and underlying neurologic condition were independent predictors of pH1N1-associated pneumonia (in wave 1) and ICU admission (in both waves), with asthma also independently associated with pH1N1-associated pneumonia (Supplemental Table 5). For seasonal influenza A (2004/2005–2008/2009), multivariate models without ethnic origin as an independent variable demonstrated underlying neurologic condition to be an independent predictor of seasonal influenza A–associated pneumonia and ICU admission, and seasonal influenza vaccination to be associated with decreased likelihood of ICU admission (Supplemental Table 6).
Our study found that children hospitalized with pH1N1, compared with those admitted with seasonal influenza A, were more likely to (1) present with cough, headache, and gastrointestinal symptoms; (2) be diagnosed with pneumonia and experience a longer hospital stay but not be at increased risk of ICU admission or death; and (3) be older, be from a non-European ethnic group, and have preexisting asthma. Our analysis of risk factors for severe disease revealed that independent predictors of influenza-associated pneumonia, and ICU admission, as well, were aboriginal ethnicity (for pH1N1) and underlying neurologic conditions (for pH1N1 and seasonal influenza A).
Although early reports highlighted increased frequency of gastrointestinal manifestations with pH1N1 infection,20,21 increased likelihood of cough (even in the absence of pneumonia) in pH1N1 relative to seasonal influenza A has not been previously reported. While this statistically significant difference may not be clinically significant, the pH1N1 virus has been shown to bind to α-2,3-linked receptors resulting in more extensive replication in the lower respiratory tract.22 It is not known, however, whether the virus replicates in the lower respiratory tract of patients without causing pneumonia. Similarly, the pathogenesis of gastrointestinal symptoms in pH1N1 patients remains unclear. Animal and human studies have demonstrated shedding of viable virions from the gastrointestinal tract23–25 but without viremia,24,26–28 suggesting that the pH1N1 virus likely does not disseminate hematogenously to the gastrointestinal tract after a primary respiratory tract infection. Regardless of the exact pathophysiology of fecal pH1N1 excretion, these observations highlight a difference between pandemic and seasonal influenza strains that could have important implications on prevention of transmission.
Our finding that children with pH1N1 had an increased risk of pneumonia compared with seasonal influenza A is consistent with 2 previous prospective community-based studies4,8 and 1 retrospective emergency department study.29 What has not been described is a decreased likelihood of upper respiratory tract disease such as croup with pH1N1 infection. This observation was unexpected given that the enhanced lower respiratory tract replication seen in animal models has not been associated with less efficient binding (via α-2,6-sialic acid) and replication in the upper respiratory tract as has been demonstrated for pathogenic H5N1 influenza virus infection.30 We also demonstrated a longer hospitalization among children admitted with pH1N1. However, proportions of ICU admissions and deaths between the 2 groups were similar. In contrast, an Argentinean study of 204 children hospitalized with pH1N1 documented a 10-fold increase in mortality compared with the 2007 season.3 This discrepant finding may reflect increased host genetic predisposition in Argentinean children and/or socioeconomic factors (the study was conducted at 6 public hospitals offering free care to children without private medical insurance).
Consistent with previous descriptions of pH1N1-infected children who required hospitalization31,32 or ICU admission,9 or died of severe pneumonia,33 our analysis demonstrated a shift in age distribution away from infants and young children with pH1N1 for both hospital and ICU admissions. Immune-complex–mediated pulmonary disease resulting from high levels of cross-reacting but nonneutralizing antibodies has been proposed as a novel mechanism for severe pH1N1 disease in healthy middle-aged adults.34 The investigators attributed these nonprotective, low-avidity antibodies in middle-aged adults to repeated previous seasonal influenza infections, and suggested that the older adults were protected by preexisting neutralizing cross-reactive antibodies elicited by an H1N1 virus that circulated before 1957. This mechanism may also explain the relative sparing of infants and young children from severe pandemic disease, because they are more likely than older children and adolescents to lack previous exposures to seasonal influenza viruses and, thus, lack pathogenic immunity. The increased risk of pH1N1-related morbidity in the aboriginal group in our study may reflect local management of mild disease in remote Aboriginal communities, whereas severe cases were referred to IMPACT centers, or represent a spurious association from missing ethnicity data. However, this observation agrees with data from other countries during the 2009 pandemic, previous pandemics, and seasonal influenza epidemics.35,36 Other non-European ethnic groups have also been noted to be disproportionately affected by severe pH1N1 infection.36 Moreover, our analysis indicated that this disparity in susceptibility to severe influenza among minority ethnic groups widened during the pandemic.
With >3 times the combined sample size of 2 previous single-center studies,6,9 we confirm asthma as a more significant risk factor for hospitalization and ICU admission in children with pH1N1 than with seasonal influenza A. In another large multicenter analysis, asthmatic children hospitalized with pH1N1 were more likely to have pneumonia (based on International Classification of Diseases, Ninth Revision coding) and require ICU admission in comparison with those hospitalized with seasonal influenza A and B combined, but not A alone.12 However, this study did not exclude asthmatic children with additional medical conditions who comprised 27% of the sample. The mechanism by which asthma confers an increased relative risk of hospitalization and ICU admission with pH1N1 compared with seasonal influenza A remains poorly understood and merits further investigation.
Although pH1N1 may have had a greater impact (in comparison with seasonal influenza A) on asthmatics than other medical conditions, we found that underlying neurologic conditions conferred increased risks (in comparison with healthy individuals) of pneumonia and ICU admission with both pH1N1 and seasonal influenza A, whereas asthma only conferred an increased risk of pneumonia with pH1N1. This may represent poorer lung reserve and/or additional occurrence of nonrespiratory complications (eg, status epilepticus) that require ICU admission in children with neurologic conditions.
Our study has several limitations. Confirmation of pH1N1 by PCR and seasonal influenza A by nonmolecular assays, insufficient seasonal influenza A subtype information for subtype-specific comparisons, late availability of the pH1N1 vaccine relative to readily available vaccines in previous seasons,17 and the influence of publicity surrounding the pandemic on hospital and ICU admission criteria could all have introduced bias into our analyses. Widely disseminated recommendations by public health agencies (including the US Centers for Disease Control and Prevention37) to provide early antiviral treatment to all patients hospitalized with influenza and all at-risk patients with influenza, regardless of illness severity, might have obscured any existing difference in disease severity between pH1N1 and seasonal influenza. Incomplete data collection during wave 2 of the pandemic limited our ability to compare all variables. Our adjustment for differences in antiviral use between the 2 groups was hampered by lack of data on timing of antiviral prescription. Lack of ethnicity data for the prepandemic years other than 2008/2009 and missing ethnicity data for 40% to 45% of cases could have led to biased estimates of risk. Although our ethnicity findings should be interpreted with caution, the increased risk for non-European ethnic groups has been documented in other studies.35,36
Taken together, the observations from our study are consistent with animal and human data, suggesting that pH1N1 differs from seasonal influenza A viruses in its virulence and mechanisms by which it causes severe disease, potentially resulting in certain risk groups being disproportionately affected during a pandemic. This study demonstrates the value of ongoing surveillance at a national level by using standardized case finding and uniform data collection. Because genetic reassortment of the pH1N1 virus with cocirculating seasonal viruses could lead to increased pathogenicity, it is important for surveillance systems such as ours to continue monitoring for any changes in clinical features and disease severity during the postpandemic period.
IMPACT is a national surveillance initiative conducted by the IMPACT network of pediatric investigators. We gratefully acknowledge the expert assistance provided by the Monitor Liaison (Heather Samson), the IMPACT nurse monitors, and staff of the data center (Kim Marty, Wenli Zhang, Shu Yu Fan, Engy Grove, and Debbe Heayn). Investigators and centers participating in this IMPACT project included: R. Morris MD, Janeway Children’s Health & Rehabilitation Centre (St. John’s, NL); S. Halperin MD, IWK Health Centre (Halifax, NS); P. Déry MD, Centre Mere-Enfant de Quebec, CHUL (Quebec City, PQ); D. Moore MD, The Montreal Children’s Hospital (Montreal, PQ); M. Lebel MD, CHU Ste-Justine (Montreal, PQ); N. Le Saux MD, Children’s Hospital of Eastern Ontario (Ottawa, ON); D. Tran MD, The Hospital for Sick Children (Toronto, ON); J. Embree MD, Winnipeg Children’s Hospital (Winnipeg, MB); B. Tan MD, Royal University Hospital (Saskatoon, SK); T. Jadavji MD, Alberta Children’s Hospital (Calgary, AB); W. Vaudry MD, Stollery Children’s Hospital (Edmonton, AB); L. Sauvé MD, BC Children’s Hospital (Vancouver, BC).
- Accepted April 23, 2012.
- Address correspondence to Wendy Vaudry, MD, Division of Infectious Diseases, Stollery Children’s Hospital, 3-588D Edmonton Clinic Health Academy, 11405-87 Ave, Edmonton, Alberta, Canada T6G 1C9. E-mail:
Drs Tran, Vaudry, Moore, Bettinger, Halperin, and Scheifele conceived and designed the study; Drs Vaudry, Moore, and Bettinger designed and revised the case report forms; Dr Tran conducted the statistical analysis, performed the final literature review and wrote the initial draft of the paper; Drs Bettinger and Scheifele oversaw data integrity at the Vaccine Evaluation Centre; and all authors oversaw data collection at the sites, contributed to data interpretation, critically revised the manuscript, and approved the final version submitted for publication.
The authors have no commercial or other association that might pose a conflict of interest.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: The Canadian Immunization Monitoring Program, Active (IMPACT) is managed by the Canadian Paediatric Society, which receives ongoing funding from the Public Health Agency of Canada's Centre for Immunization and Respiratory Infectious Diseases for IMPACT. Canadian Paediatric Society had no role in study design, data collection, and analysis or decision to publish. Public Health Agency of Canada provided input into the study design and was involved in the review and approval of the manuscript.
- Gordon A,
- Saborío S,
- Videa E,
- et al
- O’Riordan S,
- Barton M,
- Yau Y,
- Read SE,
- Allen U,
- Tran D
- Dawood FS,
- Kamimoto L,
- D’Mello TA,
- et al.,
- Emerging Infections Program Network
- Moore DL,
- Vaudry W,
- Scheifele DW,
- et al
- Helferty M,
- Vachon J,
- Tarasuk J,
- Rodin R,
- Spika J,
- Pelletier L
- Maines TR,
- Jayaraman A,
- Belser JA,
- et al
- Yoo SJ,
- Moon SJ,
- Kuak EY,
- et al
- Munster VJ,
- de Wit E,
- van den Brand JM,
- et al
- Ocaña-Macchi M,
- Bel M,
- Guzylack-Piriou L,
- et al
- ↵Centers for Disease Control and Prevention. Updated Interim Recommendations for the Use of Antiviral Medications in the Treatment and Prevention of Influenza for the 2009-2010 Season. 2009. Available at: www.cdc.gov/h1n1flu/recommendations.htm. Updated December 7, 2009. Accessed March 9, 2012
- Copyright © 2012 by the American Academy of Pediatrics