Influenza-attributable deaths, Canada 1990-1999

Area of Science:

• Epidemiology
• Public Health
• Biostatistics

Background:

• Influenza mortality is often underestimated by vital statistics.
• Accurate assessment of influenza's impact on mortality is crucial for public health planning.

Purpose of the Study:

• To estimate the true number of deaths attributable to influenza in Canada during the 1990s.
• To analyze the relationship between influenza-certified deaths and overall mortality patterns.

Main Methods:

• Analysis of weekly mortality data in Canada (1989/1990-1998/1999) by cause, age, and place of death.
• Utilized Poisson regression to model all-cause and cause-specific mortality based on influenza-certified deaths, controlling for seasonality and trend.
• Calculated influenza-attributable deaths by subtracting baseline-predicted deaths from predicted deaths.

Main Results:

• Estimated average of just under 4000 annual influenza-attributable deaths in Canada during the 1990s (13/100,000 persons).
• Observed significant variation, with 6000-8000 deaths in severe seasons (1997/1998, 1998/1999) and minimal excess mortality in others.
• Only 8% of influenza-attributable deaths were certified as influenza; pneumonia accounted for 15% and all respiratory causes for 40%.

Conclusions:

• Influenza significantly contributes to excess mortality beyond officially certified influenza deaths.
• The weekly pattern of influenza-certified deaths serves as a reliable indicator of excess all-cause mortality.
• Underdiagnosis and undercertification of influenza as a cause of death are prevalent, necessitating improved surveillance and reporting.