Volume 28-03
1 February 2002

[Table of Contents]

INFLUENZA IN CANADA: 2000 -2001 SEASON

Introduction

The Centre for Infectious Disease Prevention and Control (CIDPC), Public Health Agency of Canada, Health Canada, maintains a national influenza surveillance network through the FluWatch program. The objective of FluWatch is to monitor and report on influenza activity across Canada during the influenza season, which runs approximately from October until May. The FluWatch program has four main components, which are available on a weekly basis as aggregate data: 1) laboratory-based influenza virus detection and identification in Canada; 2) sentinel surveillance of influenza like illness (ILI) consultation rates in Canada; 3) regional influenza activity levels, as assigned by provincial and territorial epidemio logists (FluWatch representatives), and 4) international influenza activity summarized from surveillance reports published by other countries (e.g., United States [U.S.] Centers for Disease Control and Prevention [CDC] in Atlanta) and international surveillance systems (e.g., European Influenza Surveillance Scheme and the World Health Organization). Additional epidemiologic information on laboratory-confirmed cases of influenza is available less frequently, on a case-by-case basis. These data provide additional epidemiologic information (i.e., age and gender distribution) on influenza cases by the type and sub-type of influenza virus.

FluWatch disseminates information on influenza activity to health professionals and the public through a variety of mechanisms including the CIDPC FAXlink, fax, e-mail and Health Canada's Division of Respiratory Diseases' website <http://www.phac-aspc.gc.ca/fluwatch/index.html>. FluWatch reports are made available on a weekly basis during the influenza season and summaries of laboratory surveillance data are made available weekly throughout the year. Summaries of worldwide influenza activity are included periodically in the weekly Infectious Diseases News Brief which is available on the Division of Disease Surveillance website <http://www.phac-aspc.gc.ca/bid-bmi/dsd-dsm/nb-ab/index.html>. Periodic updates on influenza surveillance in Canada are published during the influenza season in the Canada Communicable Disease Report.

This report provides a summary of influenza activity in Canada during the 2000-2001 influenza season, including: 1) laboratory detections of influenza from sentinel laboratories; 2) case-by-case epidemiologic and laboratory data (supplemental information on laboratory detections from individual case reports); 3) ILI consultations from sentinel physician reporting and; 4) provincial and territorial activity level reporting. Comparisons are made with previous seasons throughout^(1-4).

Methods

Laboratory detections of influenza: Thirty laboratories across Canada participated in the FluWatch surveillance program during the 2000-2001 season. While the influenza season commonly runs from October through April, laboratory detections are performed year round and analyzed and reported on from August to August. All participating laboratories are asked to report the total number of influenza tests performed as well as the total number of tests positive for influenza infection to the Division of Disease Surveillance, CIDPC, on a weekly basis. Methods used in the detection of influenza included: viral culture, direct antigen detection and seroconversion (i.e., >= 4-fold rise in antibody titre by any method). Laboratory detections data were presented as aggregate data by testing laboratories and analyzed by the province performing the testing (territorial samples were tested by laboratories in nearby provinces) and type of influenza.

Case-by-case epidemiologic and laboratory data: On a less timely basis (bi-monthly to monthly), additional epidemiologic and laboratory information were reported to CIDPC by 21 of the 30 laboratories that perform influenza testing across Canada. Approximately 10% of influenza isolations were referred to the National Microbiology Laboratory for strain identification. Laboratory-confirmed case-by-case data were presented by the province/ territory from which the specimen originated (some laboratories received out-of-province samples), and were analyzed by week of onset of illness, age and gender of the case and influenza type and strain.

ILI consultations reported by sentinel physicians: The College of Family Physicians of Canada (CFPC), National Research System (NaReS), was responsible for recruiting sentinel physicians in nine out of 13 provinces and territories across Canada. In the other four provinces (Quebec, British Columbia, Alberta, and Saskatchewan), sentinel recruitment and reporting was managed by independent provincial programs. FluWatch was able to form a network which included all of these provincial surveillance systems. The FluWatch objective was to have at least one physician recruited from each of the census divisions across Canada or, for census divisions with large populations, to have one sentinel physician recruited per 250,000 population. The case definition for ILI was &"acute onset of respiratory illness with fever and cough and with one or more of the following -sore throat, arthralgia, myalgia, or prostration which could be due to influenza virus. In children < 5 years of age, gastrointestinal symptoms may also be present. In patients < 5 or >= 65 years of age, fever may not be prominent&". For 1 clinic day each week, between 1 October, 2000 and 21 April, 2001, sentinel sites were asked to complete a report form, including 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 was collected for all patients (for both the numerator and denominator) seen through sentinel physicians recruited by NaReS, and for patients seen through provincial surveillance systems in British Columbia and Saskatchewan. In Alberta, age group information was collected only on numerator data; age group for denominator data was generated by applying the Canadian population distribution. In Quebec, the provincial surveillance system was not set up to collect age group information and therefore ILI rates could only be compared to the crude rates for the rest of Canada. Sentinel report forms were either returned by fax, or the information was conveyed via e-mail or telephone to CIDPC on a weekly basis for data collation, analysis and dissemination.

Regional influenza activity levels assessed by provincial and territorial epidemiologists: Most provinces and territories are subdivided into influenza surveillance regions as defined by the provincial or territorial epidemiologist. For the 2000-2001 influenza season, there were 53 surveillance regions: British Columbia (four), Alberta (three), Saskatchewan (three), Manitoba (12), Ontario (five), Quebec (one), New Brunswick (seven), Nova Scotia (four), Prince Edward Island (one), Newfoundland (10), Yukon (one), Northwest Territories (one), and Nunavut (one). Provincial and territorial FluWatch representatives assessed the influenza activity level in their respective jurisdictions, weekly, using a variety of sources of information which included: laboratory reports of influenza detection, sentinel physician reports of ILI surveillance and reports of outbreaks. In addition, school and work-site absenteeism and emergency department and hospital admission data may also have been be used in assessing the level of influenza activity. Influenza activity levels* were reported as: 1) no activity reported, 2) sporadic activity, 3) localized activity, and 4) widespread activity.

Results

Laboratory detections of influenza (aggregate data): Between 27 August, 2000 and 25 August, 2001, a total of 55,085 influenza tests were performed by 30 laboratories receiving samples from all provinces and territories across Canada. Approximately 8% (4,204) of tests were positive, including 1,349 (32%) influenza A and 2,855 (68%) influenza B. While the number of tests performed during 2000-2001 is comparable to the past 2 years and considerably increased from 1996-1997 and 1997-1998, the percentage of positive tests in 2000-2001 was lower than in the 4 previous years of laboratory surveillance (range 9% to 13%).

Case-by-case epidemiologic and laboratory data: Of the 4,204 positive influenza tests, 3,935 (94%) had laboratory-confirmed case records with epidemiologic and laboratory details reported to CIDPC by 21 laboratories in 10 provinces (Table 1). This compared with 5,907 laboratory-confirmed case-by-case records reported by 16 laboratories in 10 provinces for the previous season (1999-2000). The variation in numbers of confirmed cases and the distribution of virus type and subtype among provinces/territories should be interpreted with caution; these numbers are likely to reflect differences in population size and distribution, testing and reporting practices and criteria, and availability of diagnostic services.

Table 2 shows the case-by-case laboratory-confirmed data, by province/territory and influenza type and subtype for cases reported during the 2000-2001 season. The largest number and proportion of cases were recorded in Quebec (949 cases, 24%); Alberta (744 cases,19%), Ontario (654 cases, 17%) and Saskatchewan (611 cases, 16%). The distribution of influenza types for laboratory-confirmed case-by-case data was the same as for the more timely laboratory detections data; 68% (2,668/3,935) of case-by-case records were confirmed as influenza type B and 32% (1,267/3,935) were confirmed as influenza type A. Compared to previous years, this distribution represents a drop in the proportion of laboratory confirmed influenza A virus infections and a marked increase in the proportion of influenza B virus infections. In 1999-2000, 95% of influenza isolates were type A and only 5% of isolates were type B⁽¹⁾. Similarly, in the previous three seasons (1997-1998 through 1999 2000) only a minority (range 1.5% to 36%) of influenza infections were due influenza B.

Further subtyping of the influenza A viruses isolated during the 2000-2001 season also showed variation in the proportions of viral subtype compared to the previous season. Of the 1,267 influenza A identifications, 28% (357/1,267) were subtyped: 99.4% (355/357) were of the H1N1 subtype and 0.6% (2/357) were of the H3N2 subtype. In the previous season, only 6.5% of influenza A subtypes were H1N1 and 93.5% were H3N2.

Although confirmed cases were consistently reported earlier in the Prairies, 62% of all cases in Canada were reported in January and February, with 37% of all cases reported in the 3-week period from 28 January to 17 February 2001. However, seasonal peaks in laboratory-confirmed cases were evident nationally as well as regionally, except for the Territories. In most regions these peaks represented an initial peak in influenza B virus infections followed >= 2 weeks later by a peak in influenza A virus infections (Figure 1).

Figure 1. Laboratory-confirmed cases of influenza by region, type and week of onset, Canada, 2000-2001

Figure 2. Proportionate distribution of laboratoryconfirmed influenza cases, by age group, Canada, 2000-2001

During the 2000-2001 season, the greatest proportion of cases, 24% (931/3,935), occurred in the youngest age group, <= 5 years, which is also the narrowest age grouping. Over 60% of cases occurred in those <=  24 years. This represents a much younger distribution of cases compared to the previous four seasons. For example in 1999-2000, only 26% of laboratory-confirmed cases were reported in those aged <= 24 years, whereas 42% of cases were reported in the >= 65 years age group. In 2000-2001, only 8% of laboratory-confirmed cases were reported for those in the >= 65 years of age group (Figure 2).

Laboratory confirmations: The majority of influenza cases (84%) were laboratory-confirmed by virus isolation. Less commonly reported methods of laboratory confirmation included direct antigen detection (15.5% of cases) and serology (0.5% of cases). The same three methods of laboratory confirmation were used in previous seasons; however, use of virus isolation has been increasingly reported (compared to 78% in 1999-2000 and 54% in 1997-1998).

Types of influenza virus circulating during the 2000-2001 season: Two peaks in activity were observed during the 2000 2001 season. The first peak, due to influenza B, occurred during week 1 of 2001, followed by a peak in influenza A during week 5. Strain characterization was completed on 497 isolates (12% of all isolates), between October 2000 and April 2001: 243 influenza A isolates and 254 influenza B isolates. Of the 243 influenza A influenza isolates, the following strains were identified: 236 (97%) A/New Caledonia/20-99(H1N1)-like, five (1%) A/Johannesburg/ 82/96(H1N1)-like; and two (0.4%) A/Panama/2007/99 (H3N2)-like. While < 1% of influenza A isolates characterized during the 2000 2001 season were H3N2, between 83% (1999-2000) and 100% (1996-1997) were H3N2 during the previous four seasons (1996 1997 through 1999-2000). Of the 254 influenza B isolates, 253 (99.6%) were characterized as B/Yamanashi/ 166/98-like and one (0.4%) was characterized as B/Beijing/243/97 like. Table 3 shows the provincial and territorial distribution of characterized strains for the 2000-2001 season. B/Yamanashi/166/98-like isolates were identified in all provinces and the one B/Beijing/243/97 like virus was identified in Quebec. A/New Caledonia/20/99-like (H1N1) isolates were identified in all provinces and territories except for Prince Edward Island and Nunavut (Dr. Yan Li, National Microbiology Laboratory, Winnipeg: personal communication, 2001). The seasonal distribution of laboratory-confirmed influenza infections in Canada, by type, is shown in Figure 3.

ILI reported by sentinel physicians: In Quebec, where there are 99 census divisions, representative recruitment was accomplished by coverage of health regions (n=18) instead of by census division. A total of 19 sentinel clinics, Centres locaux de services communautaires (CLSC), participated in 12 (67%) of the 18 health regions in Quebec, with more CLSCs recruited in regions with higher population density. In all other provinces and territories, representative recruitment was accomplished by coverage of census divisions; 231 sentinel physicians and sentinel clinics (one per 150,000 population) were recruited in 141 (75%) of the 189 census divisions outside of Quebec. Overall, recruitment represented most of the well-populated urban and rural regions across Canada. Each week between late October and mid April, CIDPC received ILI data from an average of 50% of FluWatch sentinels (not including CLSC sentinel sites in Quebec). This response rate was down from an average of 68% reporting each week in 1999-2000. During this same period of time, 91 sentinels reported ILI data for at least 50% of the reporting weeks and 10 sentinels reported ILI data for at least 90% of the reporting weeks. Participation rates in 2000-2001 were also down from last season; in 1999-2000, 239 sentinels reported at least 50% of the reporting weeks and 39 sentinels reported 90% of the reporting weeks.

There was no prominent peak in ILI reporting rates during the 2000-2001 season and these rates remained well below the 1996 2000 mean rate throughout most of the season (Figure 4). These data do not include Quebec, where age-specific ILI rates are not collected. Nevertheless, the crude rate (non age-standardized) for Quebec is similar to the age-standardized rate for Canada, throughout 2000-2001 and there is close correspondence in the seasonal trend of ILI (data not shown). The highest weekly rate was 31 cases of ILI per 1,000 patient visits which occurred during week 7, the week ending 17 February 2001, (Figure 4). Peak activity was considerably milder and later than last season's peak of 149 cases per 1,000 during week 52. During week 7, 2001 the age groups with the highest ILI rates were those 5 to 19 years of age (27/1,000) and <= 5 years of age (25/1,000); those > 65 years had the lowest ILI rate (3/1,000). Over the 2000-2001 ILI surveillance period, 2% of (4,611/202,898) patients seen were were diagnosed with ILI, with an overall ILI rate of 23 per 1,000 patients seen (compared to a rate of 41 per 1,000 patients seen during the 1999 2000 season). The highest rates of ILI were in children, with 43 cases of ILI per 1,000 patients seen in those 0 to 4 years of age and 38 per 1,000 patients seen in those 5 to 19 years of age.

Figure 3. Seasonal distribution of laboratory-confirmed influenza infections by type, Canada, 1996-2001

Figure 4. Census division weighted age-standardised influenza-like illness (ILI) rates*, Canada, by report week for the 2000-2001 Influenza season compared to 1996-1997 through 1999-2000 seasons (mean ratewith 95% confidence intervals)

Figure 5. Number of surveillance regions reporting widespread or localized influenza activity, Canada, by week and year, 7 October 2000 through 21 April 2001

Influenza activity level assessment: Saskatchewan and Alberta were the first provinces to report localized influenza activity in the week ending 16 December, 2000 (week 50). Widespread activity was first reported in Saskatchewan and the Yukon during the week ending 30 December, 2000 (week 52). The number of regions reporting localized or widespread activity increased gradually over the next 6 weeks and peaked at 28 (53% of the 53 influenza surveillance regions) during the week ending 3 February, 2001 (week 5) (Figure 5).

Discussion

During the 2000-2001 influenza season, the Prairie provinces were the first to report laboratory-confirmed influenza and increased regional influenza activity levels. After a peak in the reporting of laboratory-confirmed cases in early January, resulting from influenza activity in the Prairies, there was a nationwide increase in cases reported between late January and early March. Similarly, the greatest number of influenza surveillance regions reporting localized or widespread activity began in late January (week 5). Increased ILI activity began slightly earlier and continued into early March, with the highest ILI rates reported in mid-February. Overall, the increases in influenza activity indicators occurred over a longer time period, due to overlapping influenza B and influenza A activity occurring over several weeks. As such, there was no prominent peak in influenza activity indicators compared to the well defined, single peak experienced during 1999-2000. Compared to previous years, lower than usual activity was reported for all three indicators of influenza activity. In particular, ILI rates were considerably lower during the 2000-2001 season, compared to the previous four seasons.

In contrast to the predominance of influenza A, and in particular the A/Sydney/5/97(H3N2)-like virus, seen over the past 3 years, influenza B predominated overall this season and influenza A (H3N2) accounted for < 1% of all characterized isolates. A/New Caledonia/20/99(H1N1)-like virus, which was first isolated in Canada during 1999-2000, was more prominent in 2000-2001, accounting for 47% of characterized isolates (up from 16% last year). Most strains identified in Canada during the 2000-2001 season matched the virus components of the 2000-2001 trivalent influenza vaccine: A/New Caledonia/20/99(H1N1)-like, A/Panama/2007/99 (H3N2)-like and B/Yamanashi/166/98-like. In addition, five viruses were identified as A/Johannesburg/82/96(H1N1)-like. Although these H1N1 viruses are antigenically distinct from A/New Caledonia/20/99(H1N1)-like virus, the A/New Caledonia/20/99 vaccine strain produces high titres of cross-reactive antibodies. Finally, a single B/Beijing/243/97-like virus was identified from a patient who first experienced symptoms while traveling in Asia during March 2001. This virus belongs to the B/Victoria/2/87 lineage whereas the other influenza B viruses identified, like the vaccine strain, belong to the B/Yamagata/16/88 lineage. No other viruses of the B/Victoria/2/87 lineage have been identified in North America in recent years⁽⁶⁾.

The trends observed in Canada were similar to those observed in the U.S.; however, influenza A (H1N1) predominated in the U.S. overall, despite the predominance of influenza B in some regions and during some weeks. As well, the CDC identified an additional influenza A (H1N1) strain, A/Bayern/07/95 (H1N1)-like, and an antigenic drift variant of the B/Yamanashi/166/98 vaccine strain in the isolates characterized from the U.S. laboratories. The influenza B variant, B/Sichuan/379/99-like virus, accounted for 89% of influenza B strains identified by CDC, yet produced only low titres of cross reacting antibodies to the B/Yamanashi/166/98 vaccine strain. As a result of this finding, it was recommended that the influenza B component of the 2001-2002 vaccine be updated to include the B/Sichuan/379/99-like virus⁽⁷⁾.

In 2000-2001, children had the highest ILI rates, which is comparable to previous years. However, in terms of laboratory-confirmed influenza, children < 14 years of age accounted for a higher proportion of cases (46%) compared to the previous year (21%). Only 8% of laboratory-confirmed influenza cases were from seniors >= 65 years of ager as compared to 42% during the previous season. The higher ILI rates in children < 14 years of age during 2000-2001 likely reflects a higher proportion of influenza infections in relation to other respiratory viruses, than in the previous season. Nevertheless, caution should be used when interpreting age-specific data due to possible age-related biases in health care utilization and physician testing behavior.

The FluWatch program provides an overall picture of influenza activity in Canada. While each component of the program has its limitations, as a whole they appeared to complement each other. The main limitations were: 1) specimen collection and submission to the National Microbiology Laboratory were subject to the individual practices of the attending physicians and the availability of the test within and between provinces/territories; 2) weekly age-specific ILI data were not available from Quebec and therefore could not be included in many of the analyses; however, a retrospective analysis of crude rates was possible; 3) the background data used as a baseline for comparison of 2000-2001 ILI rates is based only on four previous seasons and is thus somewhat unstable due to wide confidence intervals (the baseline will become more stable over time); 4) the activity level provided by the provincial/territorial epidemiologists, although based on several standardized indicators, is somewhat subjective; and, 5) participation rates for activity level reporting and ILI reporting were lower during 2000-2001 than in recent years and may have underestimated influenza activity or skewed the timing of peak activity.

Acknowledgements

We would like to thank the staff of the laboratories who participated in the Respiratory Virus Detection program during the 2000-2001 season, and Dr. Yan Li and Carol Stansfield of the Respiratory Viruses Section, National Microbiology Laboratory, for information regarding influenza virus strain characterization. We also wish to thank all the physicians and nurse practitioners who contributed to the ILI surveillance program in association with the College of Family Physicians of Canada, NaReS, and the sentinel influenza surveillance programs in British Columbia, Alberta, Saskatchewan and Quebec. Finally, we wish to express our thanks to the provincial and territorial epidemiologists and FluWatch representatives for providing information about the influenza activity level in their jurisdictions.

Laboratories wishing to participate in the FluWatch surveillance program should contact Mr. Peter Zabchuk, Division of Disease Surveillance, Bureau of Infectious Diseases, CIDPC, at 613-952-9729.

References

1. Health Canada. Influenza in Canada, 1999-2000 season. CCDR 2001;27:1-9.

2. Health Canada. Influenza in Canada, 1998-1999 season. CCDR 1999;25:185-92.

3. Health Canada. Influenza in Canada, 1997-1998 season. CCDR 1998;24:169-76.

4. Health Canada. Influenza in Canada, 1996-1997 season. CCDR 1997;23:185-92.

5. CDC. Update: Influenza activity -United States and worldwide, May-September 2001. MMWR 2001;50:822-25.

6. CDC. Update: Influenza activity -United States and worldwide, 2000-01 season, and composition of the 2001-02 influenza vaccine. MMWR 2001;50:466-70.

Source: JF Macey, MSc, P Zabchuk, B Winchester, MSc, TWS Tam, MD, FRCPC, Division of Respiratory Diseases and Division of Disease Surveillance, Centre for Infectious Disease Prevention and Control, Health Canada, Ottawa, Ontario.

_____
* For the 2000-2001 influenza surveillance season, FluWatch program activity levels were defined as follows:

1 = No activity reported.

2 = Sporadic: sporadically occurring ILI and confirmed influenza^† with no outbreaks detected within the surveillance region.

3 = Localized: sporadically occurring ILI and confirmed influenza^† and outbreaks of ILI in < 50% of the surveillance region(s).

4 = Widespread: sporadically occurring ILI and confirmed influenza^† and outbreaks of ILI in >= 50% of the surveillance region(s).

† confirmation of influenza within the surveillance region at any time within the prior 4 weeks.

[Previous] [Table of Contents]