The epidemiology of influenza in children hospitalized in Canada, 2004-2005, in Immunization Monitoring Program Active (IMPACT) Centres

1 April 2006

Volume 32

Number 07

Introduction

Infection with influenza is common in children and associated with significant illness and with death. Although children with influenza infection often present with respiratory symptoms or fever, more serious complications, such as encephalopathy and death, can occur¹. Hospitalization rates among healthy young children and especially those < 6 months old approach hospitalization rates of high-risk adults^2-4.

The recommendations for influenza vaccination of children were recently expanded. During the 2003-2004 influenza season and before, children aged ≥ 6 months with high-risk conditions were recommended by the National Advisory Committee on Immunization (NACI) to receive influenza vaccination⁵. For the 2004-2005 season, all children aged 6 to 23 months were recommended to receive influenza vaccination whether or not high-risk conditions were present⁶.

In response to the severity of the 2003-2004 influenza season for children and the expanding vaccine recommendations, nine centres of the Immunization Monitoring Program B Active (IMPACT), in collaboration with the Public Health Agency of Canada and other sponsors, initiated a study to enhance surveillance for children hospitalized with influenza in Canada. IMPACT is a hospital-based surveillance system that captures information about vaccine-preventable diseases and adverse events following immunization among hospitalized children throughout Canada. Composed of 12 tertiary care pediatric centres drawing referrals from every province and territory, IMPACT represents over 90% of pediatric tertiary care beds in Canada. IMPACT centres are located in St. John's, Halifax, Quebec City, Montreal (two), Toronto, Ottawa,Winnipeg, Saskatoon, Calgary, Edmonton, and Vancouver.

This report summarizes the second year of this influenza surveillance study and includes all 12 IMPACT centres. Its goal is to:

1. describe the demographic and clinical characteristics of children hospitalized with influenza in Canada during the 2004-2005 season,

2. highlight differences between the 2003-2004 and 2004-2005 seasons, and

3. describe vaccine coverage among children in light of current NACI recommendations.

Methods

Case patients were defined as children aged 0 to 16 years who were admitted to an IMPACT care centre with laboratory-confirmed influenza between 24 August, 2004, and 27 August, 2005. Children hospitalized with influenza were identified through virology laboratory reports and/or admission records. Each hospital has a policy of routinely testing hospitalized children with respiratory symptoms for viral infection. Acceptable laboratory evidence of influenza infection included positive culture and/or rapid test. Once the cases had been identified, trained nurse monitors or infectious disease specialists reviewed each patient's hospital record to determine the reason for admission. Only children who were admitted because of influenza or a complication of influenza were included as case patients.

Case patient demographic information, health status, vaccination history, method of diagnosis, treatment, clinical manifestation, complications, and outcome data were collected using a standard case report form. To ensure that vaccination histories were complete, nurse monitors worked closely with unit nurses to request vaccination records from parents at the time of the admission, attached reminders to the medical chart alerting health care providers of needed records, or contacted vaccine providers directly with permission. Case report forms were reviewed and data entered by the IMPACT data management centre in Vancouver, BC. Data were analyzed using SAS v8.1 (SAS Institute, Cary, NC).

Results

From the 12 IMPACT centres, 391 children were reported with laboratory-confirmed influenza during 2004-2005. This represented at least a 22% decrease from 2003-2004, when 505 children were admitted with influenza at nine centres ⁷. The earliest reported case was on 11 September, 2004, in Toronto, however, paediatric hospitalizations as an indicator of influenza activity in Canada did not begin to increase until the end of November (Figure 1). Canada-wide activity included two sub-peaks in early January and mid-February. The last reported case was on 24 May, 2005, in Quebec City. Spanning 27 weeks, the 2004-2005 season occurred later than the 2003-2004 season, which started in late September.

The pattern of admissions varied among centres, and the largest number of children were reported from Toronto, Quebec City, and Ottawa (Table 1). In addition, peak admission activity within each centre was different from that in 2003-2004, when it followed a west to east progression, starting in Edmonton and followed byWinnipeg, Ottawa, Toronto, Halifax, and Quebec. In 2004-2005, peak activity was mirrored between eastern and western centres (Figure 2).

Figure 1. Number of children admitted with influenza by week of admission and type of influenza, 2004-2005

Figure 2. Number of children admittedwith influenza by week of admission and centre, 2004-2005

Table 1. Location of children admitted with influenza by number, diagnostic method, virus type, and antiviral therapy, 2004-2005

Examination of the age distribution (Table 2) revealed that 214 (54.7%) children were ≤ 23 months old. The sex ratio favored males in every age group (58.6% overall) except patients > 12 years, of whom 12 (60%) were female. In total, 180 (46.1%) were previously healthy, and 211 (53.9%) had underlying conditions. Of the latter, 140 (66%) had conditions that indicated they should be vaccinated against influenza (Table 3). Interestingly, hemoglobinopathies were the second most common underlying condition and mainly involved children with sickle cell disease (n = 19). Of children with hemoglobinopathies, the most common symptoms were respiratory distress (n = 11), fever (n = 7), or wheezing (n = 4). Children ≥ 2 years were significantly more likely to have an underlying condition for which vaccine was indicated (odds ratio [OR] = 5.0, 95% confidence intervals [CI] 3.2 to 8.2, p < 0.001). In contrast, healthy children represented over half (58%) of patients admitted during the 2003-2004 season, a season that was particularly severe for children.

Children presented with a variety of symptoms, the majority with respiratory symptoms including respiratory distress (n = 180), wheezing (n = 94), pneumonia (n = 68), apnea (n = 10), and croup (n = 14). Other symptoms included otitis media (n = 36) and seizures (n = 33), especially in children aged 6 to 23 months. Seizures were the only manifestation in nine children (Table 4), of whom five had had previous seizures. The mean symptom duration before hospitalization was 3.1 days (range 0 to 30), and this varied among age groups, the longest occurring in children aged 6 to 23 months (Table 2). Severe manifestations included encephalitis (n = 4) and hepatitis (n = 1).

Table 2. Features of 391 children admitted with influenza by age, including hospital stay, case parameters, and virus type, 2004-2005

* Intensive care unit

Table 3. Underlying conditions in children admitted with influenza according to official recommendation for influenza vaccination, 2004-2005

Table 4. Children with influenza who presented with seizures by additional symptoms and need for intubation, 2004-2005

The hospital course of children admitted with influenza is summarized in Table 5. Secondary bacterial infections were documented in 15 children (3.8%) and included pneumonia (n = 2), bacteremia (n = 2), cellulitis (n = 1), urinary tract infection (n = 7), and other (n = 3). The most common organism isolated from children with secondary bacterial infections was Escherichia coli (n = 8). Only 26 children (6.6%) received antiviral therapy, including oseltamivir (n = 26), amantadine (n = 3) and zanamivir (n = 1). In contrast, 271 (69.3%) received antibiotic therapy. Children with underlying conditions were significantly more likely to receive antiviral (OR = 11.4, 95% CI 2.8 to 100.7, p < 0.001) or antibiotic therapy (OR = 2.0, 95% CI 1.3 to 3.2, p < 0.001). Thus, the reasons for antibiotic use may have been related more to suspicion of bacterial infection than to microbiologic confirmation.

The mean hospital stay was 4.2 days (range 1 to 34), the longest occurring in children > 12 years (see Table 2). Of all admissions, 48 (12.3%) were to the intensive care unit (ICU). Children with underlying conditions were significantly more likely to be admitted to the ICU (OR = 2.0, 95%CI 1.0 to 4.2, p = 0.02). Of children admitted to ICUs, 25 (52%) required intubation for assisted ventilation. The majority of intubations occurred in children whowere 2 to 5 years old (n = 8), were healthy (n = 11), or did not have a secondary bacterial infection (n = 23). No children required extracorporeal membrane oxygenation. In total, 369 children (94.3%) recovered from their infection and were discharged home. Twenty children (5.1%) remained in hospital, 17 for reasons unrelated to influenza admission and three for further management or treatment completion at another facility.

There were two deaths: case A was a 7-year-old male with preexisting encephalopathy who was admitted with a 6-day history of respiratory distress. He was treated with antibiotics and died 3 days later. Influenza A was isolated by culture, and his vaccinations were reported to be "up to date"; however, more specific immunization details were not available. Case B was a 4-year-old institutionalized male with anoxic brain injury and permanent tracheostomy. He was admitted with a 2-day history of respiratory distress, wheezing, and pneumonia. He was admitted to the ICU, intubated, and treated with antibiotics on suspicion of having bacterial infection. Influenza type B was identified by rapid antigen test and culture. He had received influenza vaccination before admission, although the date of administration was not documented.

At the time of admission, 48 children (12.3%) had received influenza vaccination, and their complete immunization history was known; 13 (3.3%) had received influenza vaccination, but only a partial history was known; 294 (75.2%) had not been vaccinated; 17 (4.4%) had immunizations that were reported to be "up to date", but no further details were available; and 19 (4.9%) had unknown influenza vaccination histories. Reasons for not vaccinating were as follows: unknown (n = 121), age exclusion (n = 89), unaware of indication (n = 31), parental refusal (n = 16), contraindication (n = 3), counseled against by health provider (n = 2), or other (n = 32). Other reasons for 15 children were "too ill" or "sick" to receive vaccination. Using available data obtained from vaccine records, the vaccine status for each child was determined on the basis of age and vaccine doses. In total, 41 (10.5%) of 391 children were vaccinated appropriately, and 350 (89.5%) were not vaccinated or not appropriately vaccinated (Table 5). Of the children who were vaccinated appropriately, 30 (73.2%) had indications for vaccination, five of whom (12.2% of the 41) were immunocompromised. Further more, only 11 (8.9%) of 124 children age 6 to 23 months were vaccinated appropriately.

Table 5. Vaccination status and hospital course of children admitted with influenza, by health status, 2004-2005

* Comprised the following: children > 9 years old who had received one influenza vaccination, children < 9 years who had received one influenza vaccination if this was not the first time they had received the vaccine, or children < 9 years old who had received two doses if this was the first time they had received the vaccination.
† Comprised those children with unknown vaccination histories (n = 19).

In total, 271 children (69.3%) were admitted with influenza type A and 120 (30.7%) with type B. All sites reported more children with influenza type A except Vancouver, where 17 (54.8%) were type B. In contrast, only 1% of children had been infected with type B during the 2003-2004 season. Children with influenza type A presented earlier than those with type B, as peak activity was in December versus March (Figure 1). Overall, children < 2 years old were significantly more likely to have type A infection than children ≥ 2 years old (OR = 2.2, 95% CI 1.4 to 3.3), p < 0.001). The children's condition was diagnosed by rapid antigen test (n = 179), culture (n = 97), or both (n = 115), but the methods used varied by centre (Table 1).

Discussion

Influenza infection is the most common vaccine-preventable disease among children in Canada. Monitoring influenza admissions over two seasons highlighted the variability of influenza epidemiology from season to season. Examples of this variability included the decrease in number of admissions, the non west-east progression, and the decrease in proportion due to type A virus (69% vs. 99%). Thus, a surveillance system with adequate geographic representation and conducted over a number of influenza seasons is needed to demonstrate trends and disease impact given the variability of influenza seasons.

The burden of influenza infection in children continues to be large. As in 2003-2004, children aged 6 to 23 months made up the majority of admissions, followed by children < 6 months. This occurred despite NACI guidelines to vaccinate all children age 6 to 23 months and the care providers of infants and children. Vaccination campaigns should continue to emphasize this vulnerable subset of the pediatric population. Prevention of disease among children < 6 months is challenging, as neither influenza vaccine nor chemoprophylaxis is recommended for this age group. Immunization and prophylaxis of older children and of adults in close contact with infants are important means of protection ⁸.

Over half of the children had underlying conditions, and many had conditions not targeted for vaccination, specifically neurological or developmental disorders. In contrast, over half of the children hospitalized during the particularly severe 2003-2004 season were healthy. Thus, the profile of influenza admissions may be linked to the severity of the season. Further evaluation of children with neurological or developmental disorders revealed that six were admitted because of seizure activity alone. Neurological or developmental conditions that enhance seizure risk as opposed to just compromising respiratory function may be important targets for future NACI guidelines.

The rates of antiviral and antibiotic therapy were similar between the two seasons. Amantadine and oseltamivir are approved for treatment (vs. prophylaxis) in children ≥ 1 year and should be started as soon as possible after the onset of symptoms, however recommendations for use are dependant on the susceptibility of circulating influenza strains ⁸. In January 2006, the US CDC recommended that amantadine and rimantadine not be used for either treatment or prophylaxis for the remainder of the 2005-2006 influenza season due to high levels of resistance ⁹. Data on the benefit of delayed antiviral administration are lacking, especially for hospitalized pediatric patients. Thus, under-utilization of antivirals may have been related to the presence of prolonged symptoms. Antibiotic use was common despite lack of confirmed bacterial infection. Increased rates of multidrug-resistant bacteria throughout Canada underscore the importance of appropriate antibiotic use.

The overwhelming majority of children had not been vaccinated, including many with conditions indicated for vaccination. In particular, the majority of children most recently recommended to be vaccinated (healthy children aged 6 to 23 months) were not. The reasons for not vaccinating suggest that parents were not aware of indications or misunderstood contraindications. Only children with severe anaphylactic reactions to chickens or egg protein should not receive inactivated influenza vaccine. Mild acute illness with low-grade fever, receipt of antibiotic therapy, and recent exposure to infectious disease are not contraindications ⁸. This study has several limitations. Although most children presented with respiratory symptoms, there were several who did not. Practitioners caring for children with more unusual symptoms may not have suspected influenza. Those presenting with late complications may not have had detectable virus any longer. Thus, our surveillance system, which is laboratory-based, may have underestimated the true burden of influenza in children. Furthermore, inter-centre variability of diagnostic methods may have affected the sensitivity of the surveillance system. Standardizing methods throughout Canada would improve the ability to compare data from each centre.

This report summarizes the epidemiology of influenza in children hospitalized at 12 pediatric centres in Canada during 2004-2005. Although the total number of admitted children declined, the 2004-2005 influenza season represented a significant health burden in children. Despite expanded NACI recommendations, ~90% of hospitalized children had not been vaccinated or not appropriately vaccinated. Improved vaccine coverage, especially among high-risk children, may reduce the burden of influenza infection. Expansion of current guidelines to include children with a history of seizures may be of benefit. Furthermore, documentation of complete vaccine histories versus summary statements (e.g. up to date or "UTD") should be encouraged as standard of care by all levels of health care providers, so that accurate estimates of vaccine uptake can be determined and vaccine failures readily identified.

Acknowledgements

We thank the IMPACT nurse monitors, nurse liaison, and the data centre staff. Special thanks go to Dr. Y. Yau, Department of Laboratory Medicine, Hospital for Sick Children, Toronto, Ontario; Dr. B. Lee, Provincial Laboratory for Public Health, Edmonton, Alberta; and Dr. E. Thomas, Microbiology Program, BC Children's Hospital, Vancouver, British Columbia. Funding for this project was provided by the Public Health Agency of Canada.

IMPACT Participants

IMPACT investigators and participating centres include the following: Dr. S. Halperin (IWK Health Centre, Halifax, Nova Scotia); Dr. R. Morris (Dr. Charles A. Janeway Child Health Centre, St. John's, Newfoundland); Dr. P. Déry (Centre mère-enfant de Québec, Ste-Foy, Quebec); Dr. M. Lebel (Hôpital Sainte-Justine, Montréal, Quebec); Dr. D. Moore (Montreal Children's Hospital, Quebec): Dr. N. Le Saux (Children's Hospital of Eastern Ontario, Ottawa, Ontario);Dr. L. Ford-Jones (Hospital for Sick Children, Toronto, Ontario); Dr. J. Embree (Winnipeg Children's Hospital, Manitoba); Dr. B. Tan (Royal University Hospital, Saskatoon, Saskatchewan); Dr. T. Jadavji (Alberta Children's Hospital, Calgary, Alberta); Dr. W. Vaudry (Stollery Children's Hospital, Edmonton, Alberta); Dr. D. Scheifele (British Columbia Children's Hospital, Vancouver, British Columbia).

References

1. Bhat N, Wright JG, Broder KR et al. Influenza-associated deaths among children in the United States, 2003-4.N Engl J Med 2005;353:2559-67.

2. Neuzil KM, Mellen BG, Wright PF et al. The effect of influenza on hospitalizations, outpatient visits, and courses of antibiotics in children. N Engl J Med 2000;342:225-31.

3. Izurieta HS, Thompson WW, Kramarz PK et al. Influenza and the rates of hospitalization for respiratory disease among infants and young children. N Engl J Med 2000;342:232-9.

4. Griffin MR, Walker FJ, Iwane MK et al. Epidemiology of respiratory infections in young children. Insights from the new vaccine surveillance network. Pediatr Infect Dis J 2004;23(11):S188-92.

5. National Advisory Committee on Immunization (NACI). Statement on influenza vaccination for the 2003-2004 season. CCDR 2003;29(4):1-20.

6. National Advisory Committee on Immunization (NACI). Statement on influenza vaccination for the 2004-2005 season. CCDR 2004;30(3):1-32.

7. Moore D, VaudryW, Scheifele D et al. Canadian children hospitalized for influenza at Immunization Monitoring Program Active (IMPACT) centers, 2003-4 [Abstract]. In: 6th Canadian Immunization Conference, Montreal, 5-8 Dec, 2004.

8. American Academy of Pediatrics. Influenza. In: Pickering LK, Baker CJ, Overturf GD et al. (eds). The red book, 2003 Report of the Committee on Infectious Diseases. 26^th ed. Elk Grove Village: American Academy of Pediatrics, 2003:382-91.

9. CDC. High levels of adamantane resistance among influenza A (H3N2) viruses and interim guidelines for use of antiviral agents - United States, 2005-2006 influenza season. Morb Mortal Wkly Rep 2006;55:44-6.

Source: S Wootton, MD, D Scheifele, MD, University of British Columbia, Division of Infectious and Immunological Diseases, Department of Pediatrics, Vancouver, British Columbia; M Mozel, MSc, British Columbia Children's Hospital, Vaccine Evaluation Center, Vancouver, British Columbia; D Moore, MD, Infectious Disease Division, Montreal Children's Hospital, McGill University Health Centre, Montréal, Quebec; W Vaudry, MD, Department of Pediatrics, Stollery Children's Hospital, Edmonton, Alberta; S Halperin, MD, IWK Health Centre, Department of Pediatrics, Halifax, Nova Scotia; T Tam, MD, Immunization and Respiratory Infections Division, Centre for Infectious Disease Prevention and Control, Public Health Agency of Canada, Ottawa, Ontario.