Public Health Agency of Canada
Symbol of the Government of Canada

Share this page

Surveillance Protocol for Severe Respiratory Illness

Surveillance for Influenza in Hospitalized Adults 2009- 2010

Working Group: B. Amihod, C. Frenette, D. Gravel, A. McGeer, D. Moore, K.Suh, G. Taylor (chair), J. Vayalumkal, A. Wong

Table of Contents

  1. Introduction
  2. Methodology
  3. Ethics

Appendices

  1. Laboratory-confirmed influenza surveillance, patient questionnaire (PDF Document - 230 KB - 4 pages)
  2. Weekly record of cases (PDF Document - 59 KB - 1 page)
  3. Laboratory-confirmed influenza denominator information form (PDF Document - 44 KB - 1 page)
  4. Case definitions for laboratory-confirmed influenza surveillance (PDF Document - 30 KB - 1 page)
  5. Instructions for completing patient questionnaire (PDF Document - 35 KB - 3 pages)

Introduction

Influenza remains a substantial cause of morbidity and mortality, and a significant contributor to hospital costs during influenza season. CNISP has conducted laboratory-confirmed influenza surveillance in hospitalized patients 16 years of age and older for the past 3 seasonal flu seasons. Eligible patients were those with episodes of influenza that were laboratory confirmed by a positive rapid antigen test, positive polymerase chain reaction (PCR) result, or positive culture for influenza A or B.

A combined total of 532 patients were identified with laboratory-confirmed influenza during the first two seasonal influenza seasons; where 182 were identified in 2006-2007 and 350 in 2007-2008. Of these, 371 (70%) patients tested positive for influenza type A and 161 (30%) patients tested positive for influenza type B. An additional 13 patients in 2006-2007 and 8 in 2007-2008 patients were identified with influenza-like illness (ILI) during an outbreak. The mean age of the patients with laboratory-confirmed influenza was 67 ± 21 years (median, 74; range, 16-99). The source of the illness was found to be community-acquired in 412 (76%) patients and healthcare-associated in 133 (24%) patients. The overall mean incidence of CA-influenza in hospitalized patients during the first seasonal period was 10.9 per 10,000 patient admissions (range of 2.0 to 23.0) and 26.9 per 10,000 patient admissions (range 0.8 to 47.2) for the second seasonal period. The rate of HA-influenza was 4.3 per 100,000 patient-days (range 0 to 15.5) and 5.4 per 100,000 patient-days (range 0 to 26.9) for the second seasonal period. The influenza illness was the cause or contributed to the death of 9 (5%) patients; all of whom are 62 years of age and older.  Data from 2008-2009 is not yet available.

Towards the end of the 2008-2009 seasonal flu period, a  novel strain of swine-origin influenza A (H1N1) virus was identified as the cause of outbreaks of febrile respiratory infection in Mexico.Footnote 1 Footnote 2 Now formally called H1N1 flu virus, this novel strain was found to have genes from human, swine and avian influenza A viruses. It was soon identified elsewhere in the world, including Canada and the United States, prompting WHO to declare a Phase 5 Pandemic signifying widespread human infection on April 25, 2009.Footnote 3  By May 29, 2009, 53 countries have officially reported 15,510 cases of influenza A(H1N1) infection, including 99 deaths.Footnote 4 In Canada, a total of 921 laboratory-confirmed cases of H1N1 flu virus have been reported in 9 provinces and 1 territory in Canada. (Up to date reports of H1N1 flu virus is available on the PHAC website). The Public Health Agency of Canada (PHAC) responded quickly to the emergence of H1N1 flu virus by activating the Health Portfolio Emergency Operations Centre (HP-EOC) Level 4 (Full Activation) on April 24, 2009 in order to support the provinces and territories and provide assistance to Mexico.

H1N1 Flu Virus, is transmitted in ways similar to other influenza viruses.Footnote 5 The incubation period ranges from 1-7 days and the spectrum of illness ranges from self-limited mild respiratory illness to severe outcomes including death.1,5 Clinical manifestations of H1N1 flu virus are typical of seasonal influenza with fever (94% of cases), cough (92%) and sore throat (66%). However, 25% of patients had vomiting and another 25% had diarrhea, neither of which is typical of seasonal influenza.Footnote 1 Footnote 6 Chills, rhinorrhea, headaches, shortness of breath, myalgias, arthralgias and fatigue have also been reported.Footnote 6

By mid-May 2009, a decision was made to de-escalate the HP EOC to activation Level 2, Increased Vigilance, effective May 29, 2009 at 1700 EDT. An alternate model for ongoing surveillance coordination of H1N1 flu virus response was proposed which would transition from case-by-case reporting to community monitoring. The specific goals of this approach are to continually assess the geographic spread of H1N1 flu virus in Canada and monitor the severity and extent of transmission in order that effective public health measures can be implemented when needed to mitigate the spread of disease and reduce the associated burden of illness. Although most surveillance requirements are currently covered through the FluWatch Surveillance System, it was suggested that a number of enhancements were needed which included, amongst others, include a comprehensive, hospital-based surveillance system for patients admitted with influenza or acquiring influenza during hospitalization. All provincial and territorial authorities supported the enhancement of hospital-based surveillance through CNISP.

In this context, CNISP has been asked to extend their seasonal laboratory-confirmed influenza surveillance to a yearly reporting of both confirmed cases of flu and hospitalizations with possible influenza-associated admitting diagnosis. A real-time hospital based surveillance system that will capture influenza-associated hospitalizations and deaths is useful for monitoring trends and characterizing severe influenza related disease.

1.1 Rationale:

An effective and time sensitive surveillance infrastructure for the detection and monitoring of emerging respiratory infections in Canada will help policy-makers, hospitals and clinicians prepare for the emergence of a new novel influenza virus or the re-emergence of H1N1 flu virus. It is also recognized that a hospital's ability to appropriately manage pandemic influenza will be dependent on their understanding of the burden of both community and healthcare associated severe respiratory infections.  This surveillance will assist in understanding the burden of influenza illness in adults in Canadian hospitals, and in assessing how hospitals currently manage seasonal influenza in adults.

In addition, a hospital based surveillance system that will capture possible influenza-associated hospitalizations and deaths is useful for monitoring trends and characterizing severe influenza related disease. Tracking respiratory illness in hospitalized patients is a more sensitive method for identifying a virulent strain of influenza that is circulating in the community although it is recognized that not all patients who are admitted with respiratory illness are swabbed for influenza.  A real-time hospital based system will be based on two main strategies:

  1. Monitoring influenza associated hospitalizations by tracking the number of patients admitted with a possible influenza-related diagnosis. A substantial proportion of influenza cases are admitted for reasons such as pneumonia, exacerbation of COPD and asthma, and even congestive heart failure. Therefore the purpose of tracking the number of hospitalizations who have a diagnosis from the prescribed list below (see 2.3.2) will serve as an alert system to detect increasing admissions that are potentially related to influenza and to monitor influenza related admissions during the epidemic/pandemic period (increasing, peak and decreasing trend). This will be used as a monitoring tool only.
  2. Surveillance of laboratory-confirmed influenza including H1N1 swine flu hospitalized cases in order to identify high-risk groups, outcomes, and effectiveness of treatment or interventions. This will inform decision-makers on a numbers of issues such as who should be the priority groups for vaccine and antiviral treatment, hospital bed management, etc

1.2 Objectives

  1. To describe the incidence, outcome, epidemiology, and infection control measures of laboratory-confirmed influenza for hospitalized adults in CNISP hospitals,
  2. To monitor trend of potential influenza associated hospitalizations according to a list of select admitting diagnosis.

2. Methodology

  • 2.1 Surveillance period:
    Surveillance will start June 1, 2009 to May 31, 2010 for all laboratory-confirmed influenza and influenza associated admitting diagnosis in CNISP hospitals that are currently participating in the 2008-2009 laboratory-confirmed influenza surveillance. Other hospitals participating in the CNISP network will be asked to join the surveillance on July 1, 2009 and continue until May 31, 2010, however, if desired, hospitals may choose to collect retrospective data starting June 1, 2009. Hospitals will be given the choice of doing surveillance for 1) laboratory-confirmed influenza; 2) influenza associated admitting diagnosis, or 3) both.
  • 2.2 Eligible facilities:
    All CNISP hospitals who care for adult (age ≥16 yrs) patients.
  • 2.3 Eligible cases:
    2.3.1 Surveillance for laboratory-confirmed influenza
    Inpatients age ≥ 16 years with episodes of influenza that meet the following criteria:
    1. they have laboratory confirmed influenza:
      • that is, there is a positive rapid antigen test, positive PCR (polymerase chain reaction) result, or positive culture for influenza (Note: For false positive rapid antigen tests - if the CHEC member or their laboratory judge a positive rapid antigen or PCR test to be a false positive, the case will be excluded)

        or
      • the case is part of an outbreak of hospital-acquired influenza where testing was not performed.
    2. The specimen with the positive test result was obtained between June 1, 2009 and May 31, 2010, inclusive.
    3. The patient either required admission to the participating CNISP hospital in association with the illness for which the influenza test was done, or the illness was acquired within the participating hospital during the surveillance period. All hospital wards are included in the surveillance, e.g. Long-term care, alternative level of care (awaiting placement) units, psychiatric wards and maternity wards.

    2.3.2 Surveillance for influenza associated admitting diagnosis
    Inpatients age ≥ 16 years admitted to a participating CNISP hospital with one or more of the following admission diagnosis:

    • Influenza
    • Pneumonia (any)
    • Bronchitis
    • Bronchiectasis
    • Acute exacerbation of chronic obstructive pulmonary disease
    • Acute exacerbation of asthma and/or shortness of breath
    • Congestive heart failure
    • Myocardial infarction or chest pain (angina)
    • Other lower respiratory tract infection, not specified

    No other admitting diagnosis will be included in the surveillance at this time, although the list may be revised as information regarding H1N1 flu virus or other seasonal influenza virus is made available.

  • 2.4 Exclusion:

    Any case not admitted to a CNISP hospital. For example, patients seen in out-patient clinics, ambulatory care, emergency department, etc., for visits which did not result in hospitalization.

  • 2.5 Data collection:

    2.5.1 Laboratory-confirmed influenza
    Case data collection may be concurrent or, as required, by retrospective chart review. The patient questionnaire data form will be completed for each case of laboratory-confirmed influenza (Appendix A). Forms will be completed on paper, and submitted to PHAC; photocopies should be saved at the CHEC site. Completed data forms should be submitted monthly. Outcome data can be submitted when available.  Aggregate numbers of laboratory-confirmed cases will be forwarded to PHAC by fax every Monday at the close of business day which will include all cases identified in the previous week, i.e. Sunday to Saturday (Appendix B).

    2.5.2. Influenza associated admitting diagnosis
    Data collection will be performed weekly from Sunday to Saturday and will consist of completing a line listing for each day (Appendix B). Information will include date of admission, admitting diagnosis, date of birth or age of the patient and sex. Whenever possible, total weekly admissions will also be collected. The weekly line listing will be forwarded to PHAC by fax every Monday at the close of business day (as per 2.5.1)

  • 2.6 Denominator data:
    Each participating facility will submit at the end of each quarter (every 3 months) a completed, laboratory-confirmed influenza surveillance denominator information form (Appendix C). Each participating facility will submit the following denominators:
    1. number of adult admissions (if not submitted with the weekly reporting),
    2. number of adult patient days,
    3. number of influenza tests performed in the reporting hospital.
  • 2.7 Analysis:
    The primary analysis will be descriptive. Rate of admitted influenza cases per 1000 admissions, and rate of healthcare associated cases per 10,000 patient days will be calculated. Variability between geographic region (Western, Central and Eastern Canada) and/or provinces and territories, in rates of testing and in rates of disease will be described. In the event of a re-activation of HP-EOC to Level 3 or 4, rates may be calculated by Canadian city. Burden of illness will be estimated as mortality (influenza as primary or contributing cause) and acute length of stay (truncated at day 30).  Descriptive analysis of antibiotic and antiviral management may assist pandemic planners in permitting a comparison of current practice to draft pandemic recommendations, and data on bacteremia may assist specifically in understanding the contribution of staphylococcal and streptococcal infections to serious morbidity.

  • 2.8 Limitations:

    This surveillance will systematically underestimate the burden of serious illness due to influenza in adults. Viral diagnostics are used to a limited and variable extent in adult medicine, so that the number of influenza cases diagnosed as such will represent only a small fraction of the actual number of cases. However, in the current context of increased awareness of H1N1 flu virus, it is expected that more viral testing is being done on patients with symptoms suggestive of influenza-like illness. The overall rate of laboratory-confirmed influenza may be higher than during the previous seasonal surveillance however the increase may be related to increased testing.

    In addition, we will not detect hospitalizations that may be due to influenza, but whose direct cause is a complication occurring after viral shedding had decreased to an undetectable level. The surveillance will also have a very limited ability to detect transmission of influenza; respiratory illness is common in both staff and patients during the winter season, and it is likely that this surveillance will miss the majority of transmission.

    There are significant limitations to using admitting diagnosis of hospitalized patients for early detection of a novel flu virus or the re-emergence of H1N1 flu virus. However, this method will be useful during the epidemic/pandemic period to detect increasing, peak and decreasing trend. The early detection of increasing admission of influenza associated diagnosis will enable clinicians and federal/provincial/territorial authorities to prepare for the re-emergence or emergence of a novel flu virus. Despite these limitations, this information will add substantially to our knowledge of influenza and its management in hospitals.

3. Ethics

This proposal may or may not require approval for the institutional review board (IRB) at each participating site, depending on whether the data collected is part of the hospital's routine infection control program, and on whether ethics approval is needed for participation in CNISP data sharing. Each facility will be responsible for obtaining IRB approval if needed.


References

Footnote 1
Novel Swine-Origin Influenza A (H1N1) Virus Investigation Team. Emergence of a novel swine-origin Influenza A (H1N1) virus in humans. N Engl J Med 2009;361. http://content.nejm.org/cgi/reprint/NEJMoa0903810.pdf
Footnote 2
CDC. Update: Novel influenza A (H1N1) virus infections - worldwide, May 6, 2009. MMWR 2009;58(17):453-458. http://www.cdc.gov/mmwr/preview/mmwrhtml/mm5817a1.htm
Footnote 3
Coker R. Swine flu: Fragile health systems will make surveillance and mitigation a challenge. BMJ 2009;338:b1791 http://www.bmj.com/cgi/section_pdf/338/apr30_3/b1791.pdf
Footnote 4
>WHO. Influenza A (H1N1) – update 33. http://www.who.int/csr/don/2009_05_19/en/index.html
Footnote 5
CDC. H1N1 Flu: Interim guidance for clinicians on identifying and caring for patients with swine-origin influenza A (H1N1) virus infection. http://www.cdc.gov/h1n1flu/identifyingpatients.htm
Footnote 6
Weintrub PS. The CDC's report on the new swine-origin influenza A (H1N1 virus). Journal Watch Pediatrics and Adolescent medicine. May 8, 2009. http://pediatrics.jwatch.org/cgi/content/full/2009/508/1