Table 2.
Results of a Meta-Regression Analysis Conducted to Identify Variables That Influence Estimates of Influenza-Associated Mortality
| Variable | Change in Annual No. of Excess Deaths (per 100,000 Persons) | 95% Confidence Interval | P Value |
|---|---|---|---|
| Age group, yearsa | |||
| Children (ages ≤14 years) | 0 | Referent | |
| Adults (ages 15–64 years) | 2 | −10, 15 | 0.694 |
| Older adults (ages ≥65 years) | 57 | 46, 67 | <0.001 |
| Whether the estimate was for seasonal periods or pandemic periods | |||
| Seasonal | 0 | Referent | |
| 2009 pandemic | −23 | −33, −13 | <0.001 |
| Cause of death | |||
| Pneumonia and influenza | 0 | Referent | |
| Respiratory diseases | 11 | −2, 25 | 0.105 |
| Respiratory and cardiovascular diseases | 29 | 16, 42 | <0.001 |
| All causes | 43 | 32, 54 | <0.001 |
| Statistical modeling technique | |||
| Multiplier method | 0 | Referent | |
| Regression model with a proxy for influenza activity | 30 | −5, 65 | 0.096 |
| Serfling-type model | 40 | 3, 77 | 0.033 |
| Other | 30 | −6, 67 | 0.101 |
| Income level of country/regionb | |||
| High | 0 | Referent | |
| Middle | 3 | −14, 19 | 0.766 |
| Climatic zone | |||
| Temperate | 0 | Referent | |
| Tropical/subtropical | 5 | −7, 18 | 0.408 |
a Only studies reporting age-specific estimates of the influenza-associated mortality rate were included in the analysis.
b Income level in the middle year of the study period as reported by the World Bank (121) was used to classify the income level of a country or region. If no national income data were available for a specific year, the year closest to the study period with national income reported by the World Bank was used for the analysis.