Table 4.
Multivariable analysis of the sub-group of patients with PEF data available describing risk predictors for at least one asthma attack defined by the American Thoracic Society criteria
| Reference category | Comparison | Odds ratio (95% CI) | p Value | Overall p value |
|---|---|---|---|---|
| Blood eosinophil count | ||||
| ≤400/µL | >400/µL | 1.48 (1.15, 1.91) | <0.001 | 0.002 |
| GINA management step | ||||
| 0 | 1/2 | 1.04 (0.67, 1.61) | 0.867 | <0.001 |
| 3 | 1.65 (1.00, 2.73) | 0.050 | ||
| 4/5 | 2.53 (1.43, 4.46) | 0.001 | ||
| GP consults for LRTIs | ||||
| 0 | 1+ | 1.48 (1.02, 2.14) | 0.040 | 0.003 |
| Asthma attack | ||||
| 0 | 1 | 3.72 (2.73, 5.06) | <0.001 | <0.001 |
| 2+ | 7.81 (5.18, 11.76) | <0.001 | ||
| % predicted PEF | ||||
| Per 1% increase | 0.995 (0.990, 1.000) | 0.036 | 0.036 | |
| Age | ||||
| Per year of age | 0.92 (0.87, 0.97) | 0.004 | 0.004 | |
The results are from a multivariable analysis. The following variables were significantly associated with asthma attack in the univariate analyses but were not significant in the multivariate model: hayfever diagnosis ever, eczema diagnosis ever, overall asthma control, daily short-acting β-agonist dosage and average inhaled corticosteroid daily dose.
GINA Global INitiative for Asthma, LRTI Lower respiratory tract infection