TABLE 3.
Sensitivity Analyses
| Sensitivity Analyses | Vaccinated / Cases (%) |
Vaccinated / Controls (%) |
Adjusted Odds Ratio (95% CI) |
Estimated vaccine effectiveness (95% CI) |
|---|---|---|---|---|
| Overall estimate | 28 / 162 (17%) | 766 / 2605 (29%) | 0.43 (0.28, 0.68) | 56.7 (31.9, 72.5) |
| Including self-reported vaccination | 56 / 190 (29%) | 1241 / 3270 (38%) | 0.52 (0.37, 0.75) | 47.5 (25.3, 63.2) |
| Influenza season defined using >4% positive tests per week | 28 / 154 (18%) | 579 / 1563 (37%) | 0.41 (0.26, 0.65) | 59.1 (35.3, 74.1) |
| Influenza season defined using >5% positive tests per week | 28 / 153 (18%) | 523 / 1458 (36%) | 0.40 (0.25, 0.63) | 60.1 (36.8, 74.8) |
| Season 2010–2012 | 28 / 156 (18%) | 721 / 2300 (31%) | 0.44 (0.28, 0.69) | 56.4 (31.2, 72.3) |
| Restricted to 7 days of symptoms onset | 23 / 136 (17%) | 624 / 2071 (30%) | 0.41 (0.25, 0.68) | 58.9 (32.3, 75.0) |
| Restricted to 14 days of symptoms onset | 26 / 154 (17%) | 710 / 2379 (30%) | 0.43 (0.27, 0.68) | 57.3 (31.7, 73.2) |
| Restricted to radiographic consolidation or pleural effusion | 24 / 139 (17%) | 700 / 2389 (29%) | 0.43 (0.27, 0.71) | 56.6 (29.2, 73.4) |
| Use of influenza (−), other virus (+) as controls | 28 / 162 (17%) | 368 / 1196 (31%) | 0.37 (0.23, 0.61) | 62.8 (39.5, 77.1) |
| Use of influenza (−), other virus (−) as controls | 28 / 162 (17%) | 398 / 1409 (28%) | 0.46 (0.29, 0.74) | 53.8 (25.5, 71.4) |
| Exclude subjects with previous use of antivirals | 27 / 155 (17%) | 750 / 2546 (29%) | 0.44 (0.28, 0.70) | 56.2 (30.4, 72.4) |
| Propensity score-adjusted analysis | 28 / 162 (17%) | 766 / 2605 (29%) | 0.45 (0.29, 0.72) | 54.9 (28.5, 71.5) |
| Use of respiratory syncytial virus-pneumonia as cases | 125 / 396 (32%) | 641 / 2209 (29%) | 1.18 (0.88, 1.58)* | −18.0 (−57.6, 11.7) |
Footnote: Vaccine effectiveness was estimated as (1-adjusted odds ratios) × 100; where odds ratios compared the odds of vaccination between cases and controls while controlling for age, gender, race/ethnicity, presence of children at home, smoking status, insurance status, use of oxygen supplementation at home, timing of admission relative to disease onset, timing from beginning of the influenza season to admission, the specific influenza season, and the presence of immunosuppressive conditions and other chronic medical conditions associated with influenza-associated complications including cardiopulmonary disease, diabetes mellitus, chronic liver/kidney disease, and neurological disease. Sensitivity analyses included: inclusion of self-reported vaccination data; influenza seasons defined as having at least 4% or 5% positive influenza tests per week; restriction to complete seasons 2010–2011 & 2011–2012; restriction to hospitalizations within 7 & 14 days of disease onset; restriction to patients with independent radiologic assessment indicating consolidation, infiltrate or pleural effusion; defining control pneumonia patients as those who were influenza negative but tested positive for other respiratory viruses; defining control pneumonia patients as those who tested negative for all respiratory viruses; excluding patients with any antiviral use before admission; propensity score-adjusted analysis included 16 additional variables for individual chronic medical conditions; and, a separate analysis that evaluated the association between influenza vaccination and respiratory syncytial virus-pneumonia.
Although this odds ratio was not statistically different from 1, the point-estimate odds ratio greater than 1 (and the corresponding negative vaccine effectiveness point-estimate) indicates that the odds of vaccination among cases was numerically higher than the odds of vaccination among controls.