15 March 2006 Volume 32 Number 06

Influenza in Canada - 2004-2005 Season

[Table of Contents] [Next]

Introduction

National surveillance of influenza in Canada is coordinated through the FluWatch program at the Immunization and Respiratory Infections Division (IRID), Centre for Infectious Disease Prevention and Control (CIDPC), Public Health Agency of Canada (PHAC). The primary objectives of FluWatch are the early detection of influenza outbreaks across the country; provision of timely and up-to-date information on influenza and influenza-like illness (ILI) activity in Canada and abroad to professionals as well as the public; monitoring of circulating strains of influenza virus, including new subtypes; and contribution of virologic surveillance information to the World Health Organization (WHO) global surveillance initiative, to assist with decision-making for vaccine composition for the following season. This report provides an epidemiologic summary of influenza activity in Canada during the FluWatch 2004-2005 season (the ninth FluWatch season).

Methods

The FluWatch program consists of a network of sentinel laboratories, sentinel primary-care clinical practices, provincial and territorial ministries of health, and pediatric hospitals. On a weekly basis throughout the season, the network reports on five main indicators of influenza activity:

1. sentinel laboratorybased respiratory virus detections and virus strain identifications in Canada;
2. sentinel surveillance of ILI consultation rates across Canada;
3. regional influenza activity levels as assigned by provincial and territorial FluWatch representatives;
4. pediatric influenza-related hospital admissions and mortality data;
5. a synopsis of international influenza activity. A brief summary of the sources of data for influenza surveillance in Canada is provided in the next sections.

Respiratory Virus Detections

Laboratory-based respiratory virus detections are reported through the sentinel laboratory Respiratory Virus Detections Surveillance System (RVDSS), which operates year-round (September-August). Laboratories participating in the FluWatch program report weekly to IRID the total number of influenza tests performed (by viral culture and direct antigen detection) as well as the total number of positive influenza tests (aggregate data).

On a less timely basis (bi-monthly to monthly), a proportion of the laboratories report case-by-case data on influenza virus detections, providing additional detailed epidemiologic and laboratory information, including date of onset, age, virus type, and subtype. These data represent a subset of the cumulative weekly detections, supplemented with a small proportion of positive detections by seroconversion (i.e. four-fold rise in antibody titre by any method).

Influenza Virus Strain Identification

The National Microbiology Laboratory (NML), PHAC, conducts national surveillance on human influenza virus strains in collaboration with provincial laboratories and other Canadian hospitaland university-based laboratories, and provides this information to the FluWatch program. A proportion of the weekly influenza detections across Canada are referred to the NML for strain characterization. NML virologic surveillance detects and describes antigenic changes in the circulating strains of influenza virus. Surveillance information on Canadian influenza viruses and representative strains are shared with the WHO collaborating centres for influenza. This information contributes to global influenza monitoring and to decisions about vaccine recommendations for the coming season, and provides a comparison of the antigenic match between the circulating and vaccine strains.

ILI Consultations Reported by Sentinel Clinical Practices

The College of Family Physicians of Canada, National Research System (NaReS), is responsible for recruiting sentinel physicians and nurses and managing their participation in seven provinces and three territories across Canada. In the other three provinces (British Columbia, Alberta, and Saskatchewan), sentinel recruitment and reporting is managed by independent provincial programs, and the surveillance information is forwarded directly to FluWatch. The FluWatch objective is to have at least one sentinel recruited from each of the census divisions across Canada; in the case of Quebec, where there are 99 census divisions, representative recruitment is accomplished by coverage of health regions (n = 18) rather than by census division. In addition, for the more densely populated census divisions/health regions, the objective is to have at least one sentinel recruited per 250,000 population. For one clinic day each week, sentinels are asked to report the total number of patients seen for any reason (denominator) and the total number of patients meeting a standard case definition for ILI (numerator). Age group information is also collected. In Alberta, however, age group information is collected only on the numerator, and for the denominator the age distribution of the Canadian population is applied.

Data from sentinels are weighted by the estimated population for each of the census divisions being represented each week. This is done in order to produce a summary ILI rate for the Canadian population. Each week the weights are recalculated on the basis of the actual census divisions with data to report. Weighted rates are summed to create a national ILI rate each week.

Regional Influenza Activity Levels Assessed by Provincial and Territorial Epidemiologists

Provinces and territories are subdivided into influenza surveillance regions as defined by provincial or territorial FluWatch representatives, who assess the weekly influenza activity level in their respective jurisdictions according to laboratory reports of influenza detections, ILI rates, and reports of outbreaks occurring in long-term care facilities (LTCF), hospitals, schools, and/or worksites. Influenza activity levels are reported to IRID as one of four standard categories: no activity reported, sporadic activity, localized activity, or widespread activity ^*.

Influenza-related Hospitalizations in Children

Since the 2003-2004 influenza season, hospital-based surveillance of influenza in children has been reported to FluWatch through the Immunization Monitoring Program Active (IMPACT) network of pediatric hospitals. This national network involves 12 centres across Canada, which account for approximately 90,000 admissions every year and represent over 90% of all tertiary care pediatric beds. Laboratory-confirmed cases of influenza that require admission to hospital are identified by a trained nurse monitor at each participating centre. Upon identification, detailed case report forms are completed on the basis of information from the hospital chart. Case reports are sent to a data centre in Vancouver, British Columbia, for review, data entry, and analysis. Starting from the 2004-2005 season, aggregate data of age group, influenza type, and death due to influenza in children are reported to IRID on a weekly basis.

International

FluWatch also reports on international influenza activity, as assessed from surveillance reports published by other countries (e.g. US Centers for Disease Prevention and Control [CDC] in Atlanta) and international surveillance systems (e.g. European Influenza Surveillance Scheme and theWHO's FluNet).

Dissemination

FluWatch disseminates weekly reports during the influenza season and biweekly reports during the low season (May-June to September) to health professionals and the public through a variety of media and mechanisms, including the CIDPC FAXlink, fax, e-mail, and PHAC's Viral Respiratory Diseases Web site at <http://www.phac-aspc.gc.ca/fluwatch/index.html>.

In addition, summaries of laboratory detection data for influenza, respiratory syncytial virus, parainfluenza virus, and adenovirus are made available weekly throughout the year on the Respiratory Virus Detections Web site, <http://www.phac-aspc.gc.ca/bid-bmi/dsd-dsm/ rvdi-divr/index.html>. Summaries of worldwide influenza activity are included periodically in the CIDPC Infectious Diseases News Brief <http://www.phac-aspc.gc.ca/bid-bmi/dsd-dsm/nb-ab/index.html>, and annual summaries and periodic updates on influenza surveillance in Canada are published in the Canada Communicable Disease Report, <http://www.phac-aspc.gc.ca/publicat/ ccdr-rmtc/index.html>.

Results

Influenza Virus Detections from the RVDSS (aggregate data)

Between 22 August, 2004, and 27 August, 2005, a total of 101,258 influenza tests were performed by 33 laboratories across Canada (samples from the Yukon, Northwest Territories, and Nunavut were tested by laboratories in nearby provinces and were included in the aggregated provincial results from the laboratory conducting the test). Approximately 12,879 (12.7%) of the results were positive for influenza, and of these 10,766 (83.6%) were influenza A, and 2,113 (16.4%) were influenza B.

Influenza Virus Detections from the RVDSS (case-by-case data)

Nineteen laboratories in nine provinces reported a total of 10,006 case-by-case records, which are presented in Table 1, broken down by the province or territory from which the specimen originated. The largest number and proportion of cases were reported by Quebec at 4,157 (41.6%), Ontario at 2,745 (27.4%), and Alberta at 962 (9.6%).

The majority of influenza cases (8,592 or 85.9%) were laboratory confirmed by virus isolation. Less commonly reported methods of laboratory confirmation included direct antigen detection (1,395 or 13.9% of cases) and serologic testing (19 or 0.2% of cases). The same three methods of laboratory confirmation were used in previous seasons; however, use of virus isolation has been generally increasing over the years.

Table 1. Influenza virus detections and case-by-case influenza data reported to the Centre for Infectious Disease Prevention and Control, by province/territory or region, Canada, 2004-2005

Of the total number of isolations, 8,567 (85.6%) were confirmed as influenza type A and 1,439 (14.4%) as influenza type B. Of the 8,567 influenza A identifications, 681 (8.0%) were subtyped, and all of them were H3N2. The percentage of influenza type A in each province/territory varied from 47% in Nunavut to 100% in Newfoundland and Labrador and New Brunswick; for nine of the 13 provinces and territories more than 80% of isolates were typed as influenza A. However, the total numbers of cases reported in some provinces/territories are small. Of all the influenza B cases identified during the 2004-2005 season, 675 (46.9%) occurred in Ontario and 413 (28.8%) in Quebec (Table 2).

Table 2. Case-by-case influenza data, by province/territory, influenza type, and sub-type, Canada, 2004-2005

Nationally, 75% of all influenza cases had an onset date in the 10-week period between week 53 and week 9 (late December of 2004 to early March of 2005), and 50% of the reports had an onset date during the peak period from week 1 to week 6 (early January to mid February of 2005), with onsets peaking during week 5 (913 cases, see Figure 1). Regional peaks in case-by-case data were also evident, except in the Territories, where relatively very few cases were reported. The first peak in the case-by-case data occurred during week 1 in Quebec and British Columbia, followed by the Prairies during week 2 and the Atlantic provinces during week 5. Ontario was the last to reach peak activity, and this occurred during week 9, when 27% of cases were identified as influenza B. Because influenza A accounted for 85% of all influenza cases, its distribution followed a similar pattern to that of the combined (influenza A and influenza B) case distribution. Influenza B, however, circulated later in the season, with the peak period from week 9 to week 13 (late February to early April) and onsets peaking during week 11 (179 cases).

Figure 1. Case-by-case influenza data by type, region andweek of onset, Canada, 2004-2005

During the 2004-2005 season, the ≥ 65-year age group represented 45.9% of the cases in the case-by-case dataset, and the age groups 45 to 64, 25 to 44, and 0 to 4 years represented 13.3 %, 13.1%, and 12.4% of the total cases, respectively. Note that the 0 to 4-year age group is also the narrowest age group. Of the influenza A cases, 51% were identified in the ≥ 65-year age group (Figure 2). Influenza B cases were more evenly distributed across age groups, the highest proportion (22%) occurring in the ≥ 65-year age group and the lowest proportion (5%) in the 10 to 14-year age group (Figure 2).

Figure 2. Proportionate distributions of case-by-case influenza data, by influenza type and by age group, Canada, 2004-2005

Influenza Virus Strain Identification, NML

During the period 21 September, 2004, to 31 August, 2005, the NML antigenically characterized 8.7% (1,125/12,879) of all respiratory virus detections of influenza A and B samples identified by sentinel provincial and hospital laboratories (Table 3): 911 influenza A (H3N2) (81%), one influenza A (H1N1) (< 1%), and 214 influenza B viruses (19%). The NML identified no influenza A (H1N2) viruses this season. Of the influenza A (H3N2) isolates, 515 (57%) had haemagglutinin proteins that were antigenically similar to those of the A/Fujian/411/02-like strain, which was the recommended influenza A (H3N2) component for the 2004-2005 season, and 395 (43%) had haemagglutinin proteins that were similar to those of the A/California/07/04-like strain, which is the new H3N2 component of the 2005-2006 vaccine. Of the 214 influenza B viruses tested, 170 (79%) belonged to the B Yamagata lineage and were antigenically similar to the 2004-2005 vaccine strain B/Shanghai/361/02, and 44 (21%) belonged to the B/Victoria/02/87 lineage and were similar to the B/HongKong/ 330/01-like virus, which was included in the 2002-2003 and 2003-2004 vaccines.

It should be pointed out that of the 911 influenza A (H3N2) viruses tested, 420 had been characterized as A/Fujian/411/ 02-like before the new variant, A/California/07/04-like, was reported. Of the 420 H3N2 viruses that were received from 21 September, 2004, to 15 February, 2005, 387 isolates were retested after the NML had received the A/California/07/04-like antiserum and antigen from CDC on 17 February, 2005. Revised virus characterization results from the re-testing indicated that 260 isolates (67.2%) were similar to A/California/07/04-like, and 127 (32.8%) were similar to A/Fujian/411/02-like. Nevertheless, since A/Fujian/411/02-like viruses and A/California/07/04-like viruses are still antigenically related to each other, the H3N2 component of the 2004-2005 vaccine was expected to have provided a certain level of protection against A/California-like viruses.

Table 3. Distribution of influenza strains characterized by the Respiratory Virus Section of the NationalMicrobiology Laboratory, by province/territory, Canada, 2004-2005

Since 1996-1997, the first year of the FluWatch program, there have been six influenza A predominant seasons (a range of 55% to 99.5% of influenza detections from the RVDSS were due to influenza A), two mixed seasons (both with influenza A and B detected in an approximately 60% to 40% ratio, respectively), and only one influenza B predominant season, in 2000-2001 (when 68% of influenza detections were influenza B). Using the NML influenza characterization data, of the six influenza A predominant seasons, all were predominated by H3N2 subtypes (ranging from 82% to 100% of influenza A characterizations); in the two mixed seasons, the predominant influenza A subtypes were H3N2 (> 99% of influenza A characterizations in the 1996-1997 season and 63% of influenza A characterizations in the 2002-2003 season); and in the predominantly B season, the predominant influenza A subtype was H1N1 (> 99% of NML influenza A characterizations in the 2000-2001 season). Predominant strains for the nine seasons from 1996 to 2005 in Canada are presented in Figure 3.

Figure 3. Seasonal distribution of case-by-case data by influenza type and week of onset, Canada, 1996-2005.

[click on the image to enlarge]

ILI Consultations Reported by Sentinel Clinical Practices

Overall, recruitment for the 2004-2005 season represented most of the well-populated urban and rural regions across Canada, except Quebec. A total of 21 sentinels (1/350,000 population) were recruited in seven (39%) of the 18 health regions in Quebec, amounting to considerable under-representation. In all other provinces and territories, 218 sentinels (1 per 80,000 population) were recruited in 125 (66%) of the 189 census divisions outside of Quebec. During the 2004-2005 season, CIDPC received, on average, ILI data from 66% (158/239) of FluWatch sentinels weekly across Canada. Of the total population of sentinels, 92% (219/239) provided ILI data for at least 50% of the reporting weeks, and 36% (86/239) provided ILI data for at least 90% of the reporting weeks.

During the 2004-2005 influenza season, there was no clearly defined ILI peak; however, the reported ILI rates reached a plateau during the period between week 53 and 10 (late December of 2004 to mid March of 2005), with a rate of 49 per 1,000 patients seen during week 53 and 48 per 1,000 patients seen during week 10, dropping off thereafter. Between week 41 and week 52 (late September to late December 2004), ILI rates remained below the 1996-2004 mean rates. Between weeks 1 and 20, ILI rates remained within or below baseline values, with the exception of weeks 6, 10, 12, 13, and 18, when ILI rates were higher than expected values.

The rate and timing of the peak in ILI rates varied by age group, ranging from 49 ILI consultations per 1,000 patients seen in the 20 to 64-year age group (peak in week 10), 58 per 1,000 in those aged ≥ 65 years (week 7), 103 per 1,000 in those 5 to 19 (week 12), and 149 per 1,000 in those aged 0 to 4 years (week 9).

Assessment of Influenza Activity Level

Alberta was the first province to report localized influenza activity in one of its surveillance regions, during week 42, followed by British Columbia in week 43 (see footnote for definition of “localized”). The number of surveillance regions reporting localized activity increased gradually over the following weeks, first becoming widespread during week 52 in British Columbia and peaking for the country during week 6 of 2005. Most reporting of influenza activity level (64% of the widespread activities and 43% of the localized activities) occurred over an 8-week period, spanning week 2 to week 9 (mid-January to early March of 2005), with 20 or more regions reporting localized or widespread activity during each of these weeks (Figure 5).

Provincial and territorial weekly reporting of activity levels was high; activity level assessments were reported for 100% of regions in 49 of the 53 weeks. Widespread activity was reported 111 times, whereas localized activity was reported 323 times.

During the 2004-2005 influenza season, 10 provinces reported a total of 762 laboratory-confirmed influenza outbreaks in LTCF. Most of these (701 or 92%) were reported between week 51 and week 14 (mid December to early March of 2005). The peak occurred in week 6, when 72 LTCF outbreaks were reported.

Figure 4. Census division weighted, age-standardized influenza-like illness rates, by week, Canada, 2004-2005, compared with seasons 1996 to 2004 (average with 95% confidence intervals)

[click on the image to enlarge]

Figure 5. Number of surveillance regions reporting widespread or localized influenza activity, by week,
Canada, 2004-2005

Figure 6. Number of laboratory-confirmed influenza outbreaks reported in long-term care facilities, by week, Canada, 2004-2005

[click on the image to enlarge]

Influenza-related Hospitalizations in Children

During the 2004-2005 influenza season, 12 IMPACT centres in eight provinces reported 391 influenza-associated hospitalizations. Two deaths were reported (case fatality of 0.5%), one in week 2, due to influenza A, in a child aged ≤ 5 years and the other in week 12, due to influenza B, in a child aged 2 to 4 years. Both had pre-existing conditions that put them at high risk of influenza complications. Influenza A represented 69.3% (271) of the total hospitalized pediatric cases, and influenza B represented 30.7% of cases. As often happens, influenza A predominated early in the season, peaking during week 51, and influenza B predominated later in the season, peaking during week 11 (see Figure 7).

Figure 7. Weekly hospital admissions due to influenza in children, by influenza type, Canada, 2004-2005

Reporting of influenza hospitalizations in children began at different times among the 12 IMPACT hospitals: the first hospitals to report cases were in Ontario (week 37) and Alberta (weeks 41 and 46); all cases were associated with influenza A. Six of the 12 centres had reported at least one case by week 50. The number of cases peaked during weeks 53 and 1 (both with 34 cases of influenza A) and again during week 7 (37 cases, 28 influenza A and nine influenza B). The last cases of the season were reported by hospitals in Saskatchewan (week 20), and British Columbia and Quebec (week 21); two cases were influenza A (Saskatchewan and Quebec), and one was influenza B (British Columbia).

The distribution of cases by age was as follows: 0 to 5 months, 90 cases (23.0%); 6 to 23 months, 124 cases (31.7%); 2 to 4 years, 79 cases (20.2%); and ≥ 5 years, 98 cases (25.1%). Reported influenza immunization rates among cases were as follows: 0 to 5 months, 0 cases (0%); 6 to 23 months, 20 cases (16.1%); 2 to 4 years, 20 cases (25.3%); and ≥ 5 years, 21 cases (21.4%) (see Table 4). An association between age and hospitalization for infection due to influenza A was noted: children 0 to 23 months were more likely to be hospitalized with influenza A than influenza B as compared with children ≥ 2 years (odds ratio 2.15, 95% confidence intervals 1.35-3.04, p < 0.001) (see Figure 8).

Table 4: Distribution by age and immunization status of paediatric hospital admissions due to influenza, Canada, 2004-2005

United States: CDC

Between week 41 and week 20 (3 October, 2004, to 21 May, 2005),WHO and the National Respiratory and Enteric Virus Surveillance System laboratories in the United States tested a total of 149,131 specimens for influenza viruses, 23,023 (15.4%) of which were positive. Of the 23,023 influenza viruses, 17,432 (75.7%) were influenza A viruses and 5,591 (24.3%) were influenza B viruses. Influenza activity in the United States peaked during week 7 (week ending 19 February, 2005). Both viruses exhibited a similar distribution across the season, influenza A peaking in week 6 and influenza B peaking in week 7.

Figure 8. Proportionate distribution of pediatric hospital admissions due to influenza, by type and age group, Canada, 2004-2005

International: WHO¹

Influenza activity was generally mild to moderate between September 2004 and August 2005. In the northern hemisphere, activity started in December 2004 in North America and increased rapidly in January 2005. In Europe, activity began in December 2004 and increased in January and February 2005. In general, activity started late and was low compared with the 2003-2004 influenza season in the northern hemisphere. In the southern hemisphere, activity started in April 2005 and increased in May in both Oceania and South America. The overall levels of activity in the southern hemisphere were similar to those in the 2004 season.

Influenza A (H1N1), A (H3N2), and B viruses co-circulated and caused outbreaks. While most of the outbreaks (regional or widespread) were associated with influenza A (H3N2) viruses, influenza B viruses circulated widely and caused outbreaks in some countries in Africa, Asia, eastern Europe, Oceania, and South America. Influenza A(H1) viruses circulated to a lesser extent and caused outbreaks in a few countries in Africa, central Asia, and eastern Europe.

The majority of influenza A (H1N1) viruses were antigenically closely related to A/New Caledonia/20/99. Very few detections of A (H1N2) were reported during this period. The majority of A (H3N2) viruses characterized were antigenically closely related to A/California/7/2004, although some earlier viruses were more closely related to A/Fujian/411/2002. Two lineages of influenza B viruses co-circulated. In the final months of the 2004-2005 season the proportion of B/Victoria/2/87 lineage viruses, which caused outbreaks in some countries in the southern hemisphere, had been increasing.

Avian Influenza

The first recorded cases of human infection with the avian influenza A (H5N1) virus occurred in Hong Kong in 1997; 18 persons were infected, of whom six died (33% fatality)².

Since 1997, the H5N1 virus has crossed the species barrier on at least two other occasions: in Hong Kong (2003, two cases with one death) and in the current outbreak that began in December 2003 ³. By the end of August 2005, human cases were reported in four countries: Viet Nam, Thailand, Cambodia, and Indonesia. The first case occurred in Viet Nam (onset December 2003), and as of August 2005 a total of 66 persons had been infected and 23 had died (35% fatality). Thailand reported its first case on 23 January, 2004, and has since reported 17 persons infected, of whom 12 have died (71% fatality). Cambodia reported its first case on 2 February, 2005, and has since reported four cases (100% fatality). Indonesia reported its only fatal case on 21 July, 2005. All human cases have coincided with outbreaks of highly pathogenic H5N1 avian influenza in poultry, and avian influenza H5N1 in humans remains a rare disease³. To date, there has been no evidence of sustained human-to-human transmission, and the WHO influenza pandemic preparedness level remains at Phase 3¹.

Discussion

The 2004-2005 influenza season appeared to be a mild-tomoderate influenza season in terms of severity, which is comparable to the 2003-2004 season. However, influenza activity in the 2004-2005 season started 2 months later than in the previous season and had a higher proportion of influenza B identifications. The 2004-2005 season had more impact on the elderly as opposed to children, as reflected by the large number of outbreaks in LTCF compared with last season.

The number of influenza tests performed for the 2004-2005 season was greater than for the 2003-2004 season. Overall, the number of tests performed has increased considerably since 1996-1997, the first year of the FluWatch program, and earlier^4-13. This is likely related to a change in testing practices due to increased awareness of the importance of influenza and pandemic preparedness. The overall percentage of positive tests in the 2004-2005 season (12.7%) was higher than for the previous four seasons (range 6% to 12.3%). As there is no other indication that this season was more severe than recent seasons, the increase in the percentage of influenza tests that are positive may be related to physicians' and nurses' improved swabbing technique resulting from more influenza specimens being collected (as noted by the increase in influenza tests conducted) and their better assessment of patients who meet the ILI case definition and require swabbing.

Comparison of case-by-case data for different seasons may not be reasonable, as the reporting practices for respiratory virus data and case-by-case data change over the years. The proportion of available case-by-case data from respiratory virus detection data increased from approximately 67% in the 2003-2004 season to 78% (10,006/12,879) this season but was lower compared with 2000 to 2003 seasons (ranging approximately from 84% to 94%). The variation in the number of case-by-case records and their distribution by province/territory should also be interpreted with caution, as these numbers are likely to reflect differences in population size and distribution, testing and reporting practices and criteria, and availability of diagnostic services, which vary across the regions.

Influenza viruses began circulating late in Canada this season with a peak of case-by-case data during week 5, compared with week 50 in the 2003-2004 season. Quebec and British Columbia were the first two provinces to reach the peak of case-by-case reports, during week 1 of 2005. The greatest number of influenza surveillance regions reporting localized or widespread activity occurred during week 6 of 2005.

The distribution of influenza virus type this season was different from that of the 2003-2004 season, (83.6% this season versus 98.7% of influenza detections during the 2003-2004 season being influenza type A). The proportion of influenza A for the case-bycase data (85.6%) was slightly higher than for respiratory virus detection data (83.6%), which is, at least partly, due to respiratory virus detection data having a larger proportion of cases reported by Ontario (40%) compared with the case-by-case data (27%). As well, Ontario had a higher proportion of influenza B cases than the other provinces and territories. Influenza A predominated across the country and reached its peak during week 5. Starting in early 2005, the onsets of influenza B reported by provinces/territories increased steadily and overtook influenza A in the weekly proportion of positive tests for week 11.

Laboratory surveillance data from CDC showed a similar influenza distribution across the season to Canada's, although influenza B circulated earlier and accounted for a higher proportion of laboratory-confirmed cases in the United States. Similarly, the WHO influenza surveillance system indicated a mild to moderate season, with a late start and a predominance of influenza A (H3N2), which was also the Canadian experience.

According to the strain identification data from NML, the predominant strains circulating this season were A/Fujian/411/02-like virus (46% of the total identifications), A/California/07/04-like virus (35%) and B/Shanghai/361/02 virus (15%). A different subtype distribution was seen in the previous season, when 92% of the viruses were identified as A/Fujian/411/02-like virus. Among the influenza A isolates subtyped after NML had received the A/California/07/04-like antiserum and antigen (February 2005), 81% were characterized as A/California/07/04-like and 19% were characterized as A/Fujian/411/02-like.

CDC antigenically characterized 920 influenza viruses collected by US laboratories between 1 October, 2004, and 21 May, 2005, reporting nine influenza A (H1) viruses (1%), 646 influenza A (H3N2) viruses (70%), and 265 influenza B viruses (29%). The haemagglutinin proteins of the influenza A (H1) viruses were similar antigenically to the haemagglutinin of the vaccine strain A/New Caledonia/20/99. Of the 646 influenza A (H3N2) isolates, 156 (24%) were characterized as antigenically similar to A/Wyoming/3/2003 (A/Fujian/411/2002-like, the (H3N2) component of the 2004-05 influenza vaccine), and 490 (76%) were characterized as A/California/7/2004-like. Of the influenza B viruses isolated in the United States during 2004-2005, 174 (66%) were characterized as B/Shanghai/361/2002-like, which was the influenza B component recommended for the 2004-2005 influenza vaccine, and 33 (12%) showed somewhat reduced titres to ferret antisera produced against B/Shanghai/361/2002. Fifty-eight influenza B viruses (22%) belonged to the B/Victoria lineage. Strain identification data from CDC indicated that the predominant circulating strain was A/California/7/2004-like virus and that more influenza B case were circulating in the United States than in Canada in the 2004-2005 season.

The ILI rates reported by sentinels nationwide peaked during the period of week 53 to week 10, and the highest rate, of 49 consultations per 1,000 patients, occurred during week 53. With similar reporting practices of sentinel physicians but fewer representative participating census divisions than last season, the ILI rate in 2004-2005 was lower and later than in the 2003-2004 season, with a peak of 80 visits per 1,000 patient visits during week 52.

Unlike ILI rates, increased influenza activity levels were reported by more surveillance regions across Canada in 2004-2005, 323 regions reporting localized and 111 regions reporting widespread activity compared with 219 and 70 regions, respectively, in the 2003-2004 season. The discrepancy between ILI rates and activity level can be partly explained by the significant increase in the reporting of outbreaks occurring in LTCF in 2004-2005 (762 outbreaks in 2004-2005 compared with 368 outbreaks in 2003-2004), as ILI surveillance does not capture influenza activity occurring in the elderly in LTCF, children who visit pediatricians, or the majority of consultations that occur in emergency departments and after-hours clinics; as well, there was more complete reporting of activity levels by provinces and territories in 2004-2005 than in 2003-2004 (in 2004-2005 there were 16 instances of a province or territory not reporting an activity level for one or more of its surveillance regions versus 213 instances in 2003-2004; in both years there were 52 surveillance regions).

Of the case-by-case influenza data reported to PHAC this season, the age group accounting for the greatest proportion of cases was the ≥ 65-year age group (46%), as compared with the 2003-2004 season, when 43% of cases were represented by the 0 to 14-year age group. A similar age distribution to that of the 2004-2005 season occurred in the three seasons spanning the years from 1997 to 2000, 41% to 42% of laboratory-confirmed cases being reported in the ≥ 65-year age group. In the seasons spanning 2000 to 2004, 42% to 62% of the laboratory-confirmed cases were reported in the 0 to 14-year age group, the highest proportion occurring in the 0 to 4-year age group. The distribution of influenza type A varied across age groups in 2004-2005, ranging from 93% among those aged ≥ 65 years to 59% among those 5 to 9 years, as the effectiveness of the H3N2 component of the season's vaccine in preventing A/California/07/04 (H3N2)-like virus infections among residents of LTCF may have been relatively low.

However, in terms of ILI rates, the age group with the highest ILI peak rate was children aged 0 to 18 years, and the lowest was in the adult group aged 20 to 64 years. This age distribution is similar to that seen in the previous season, when the peak ILI rates were, overall, higher and earlier. However, caution should be used when interpreting age-specific data because of possible age-related biases in health care utilization and physician testing behavior.

As the 2004-2005 season appears to have had less impact on youth than the elderly, the reported number of influenza hospitalizations for children consequently decreased from 526 cases in the 2003-2004 season to 391 cases in 2004-2005, even though the number of reporting centres increased from nine hospitals in six provinces to 12 hospitals in eight provinces over the same period. Compared with only 1% of the pediatric hospitalizations attributed to influenza B in the 2003-2004 season, 29% were attributed to influenza B in 2004-2005. The age distribution of pediatric hospitalizations was similar for the two seasons, in that the majority were reported in children < 2 years old (54% to 58%).

References

* No activity reported: based on sporadic or no ILI and no laboratory-confirmed influenza or outbreaks within the surveillance region.
Sporadic activity: sporadically occurring ILI and laboratory-confirmed influenza (within previous 4 weeks) with no outbreaks detected within the surveillance region.
Localized activity: sporadically occurring ILI and laboratory-confirmed influenza (within previous 4 weeks) together with outbreaks of ILI in schools and worksites or laboratory-confirmed influenza in residential insitutions occurring in < 50% of the influenza surveillance region(s).
Widespread activity: sporadically occurring ILI and laboratory-confirmed influenza (within previous 4 weeks) together with outbreaks of ILI in schools and worksites or laboratory-confirmed influenza in residential institutions occurring in ≥ 50% of the influenza surveillance region(s).

1. World Health Organization. Summary of influenza activity, September 2004-August 2005.WHO Weekly Epidemiological Record 2005;80:353-6.
2. World Health Organization. Avian influenza frequently asked questions. www.who.int/csr/disease/avian_influenza/avian_faqs/ en/index.html#whyso. Accessed 5 December, 2005.
3. World Health Organization. Avian influenza fact sheet. URL: www.who.int/csr/disease/avian_influenza/avianinfluenza_factsheetJan2006/en/index.html. Accessed 20 January, 2006.
4. Public Health Agency of Canada. Influenza in Canada 2003-2004 season. CCDR 2005;31:1-20.
5. Health Canada. Influenza in Canada - 2001-2002 season. CCDR 2003;29:45-60.
6. Health Canada. Influenza in Canada - 2000-2001 season. CCDR 2002;28:17-28.
7. Health Canada. Influenza in Canada - 1999-2000 season. CCDR 2001;27:1-9.
8. Health Canada. Influenza in Canada - 1998-1999 season. CCDR 1999;25:185-92.
9. Health Canada. Influenza in Canada - 1997-1998 season. CCDR 1998;24:169-76.
10. Health Canada. Influenza in Canada - 1996-1997 season. CCDR 1997;23:185-92.
11. Health Canada. Influenza in Canada - 1995-1996 Season. CCDR 1996;22:193-99.
12. Health Canada. Influenza in Canada - 1994-1995 Season. CCDR 1995;21:205-12.
13. Health Canada. Influenza in Canada - 1993-1994 Season. CCDR 1994;20-21:185-91.

Source: L Xie (MHSc candidate), SG Squires, MSc, JF Macey, MSc, S Aziz, MSc, B Winchester, MSc, H Zheng, TWS Tam, MD, FRCPC, Immunization and Respiratory Infections Division, CIDPC, Public Health Agency of Canada, Ottawa, Ontario.

[Table of Contents] [Next]