TABLE 4.
TREATMENT FAILURES IN AFRICAN AMERICAN AND WHITE SUBJECTS ACCORDING TO ASTHMA THERAPIES RECEIVED IN ACRN TRIALS BY LOGISTIC REGRESSION
| African American |
White |
African American vs. White |
||||
| Therapy | Failures/Total | % Failure | Failures/Total | % Failure | Odds Ratio (95% confidence interval) | P Value |
| All treatments | 46/233 | 19.7 | 101/795 | 12.7 | 1.69 (1.15–2.48) | 0.0074* |
| All LABA | 36/81 | 44.4 | 55/217 | 25.4 | 2.36 (1.38–4.02) | 0.0017* |
| LABA + ICS | 10/43 | 23.3 | 13/114 | 11.4 | 2.35 (0.95–5.87) | 0.0611 |
| LABA + leukotriene | 20/29 | 69 | 13/47 | 27.7 | 5.81 (2.11–16.02) | 0.0007* |
| LABA only | 6/9 | 66.7 | 29/56 | 51.8 | 1.86 (0.42–8.19) | 0.4109 |
| No LABA | 10/152 | 6.6 | 46/578 | 8 | 0.81 (0.40–1.65) | 0.5702 |
| All vs. no LABA | Comparison of odds ratios | 0.0190* | ||||
| All ICS | 13/95 | 13.7 | 22/327 | 6.7 | 2.20 (1.06–4.55) | 0.0339* |
| LABA + ICS | See above LABA section | |||||
| ICS only | 3/52 | 5.8 | 9/213 | 4.2 | 1.39 (0.36–5.32) | 0.6326 |
| No ICS | 33/138 | 23.9 | 79/468 | 16.9 | 1.55 (0.98–2.45) | 0.0627 |
| All vs. no ICS | Comparison of odds ratios | 0.4244 | ||||
| All leukotriene | 20/46 | 43.5 | 14/97 | 14.4 | 4.56 (2.02–10.28) | 0.0003* |
| LABA + leukotriene | See above LABA section | |||||
| Leukotriene only | 0/17 | 0 | 1/50 | 2 | n/a | n/a |
| No leukotriene | 26/187 | 13.9 | 87/698 | 12.5 | 1.13 (0.71–1.82) | 0.6006 |
| All vs. no leukotriene | Comparison of odds ratios | 0.0037* | ||||
| All short-acting β-agonist (only) | 2/29 | 6.9 | 5/98 | 5.1 | 1.38 (0.25–7.50) | 0.7109 |
| No short-acting β-agonist | 44/204 | 21.6 | 96/697 | 13.8 | 1.72 (1.16–2.56) | 0.0073* |
| All vs. no short-acting β-agonist | Comparison of odds ratios | 0.8019 | ||||
| All placebo (only) | 5/54 | 9.3 | 31/217 | 14.3 | 0.61 (0.23–1.66) | 0.3341 |
| No placebo | 41/179 | 22.9 | 70/578 | 12.1 | 2.16 (1.40–3.31) | 0.0004* |
| All vs. no placebo | Comparison of odds ratios | 0.0228* | ||||
Definition of abbreviations: ACRN = Asthma Clinical Research Network; ICS = inhaled corticosteroids; LABA = long-acting β-agonists; n/a = not enough events in each category for analysis.
Significant exacerbations for DICE, IMPACT, SMOG, and PRICE (which had no treatment failures) were counted as treatment failures; P values differ from Figures 1–5 because logistic regression is used in Table 4 and survival analysis is used for the figures. The subset of subjects used for each treatment in each analysis is the same.
*
P value < 0.05.