Table 3.
Multivariable Cox analysis of the entire population.
| Variable | Risk of primary outcome HR (CI, p-value) |
|---|---|
| Age | 1.01(1.00–1.03) p = 0.50 |
| Gender male vs. female |
1.04(1.00–1.09) p = 0.03 |
| Collection of promethazine | 1.89(1.76–2.04) p < 0.0001 |
| Collection of melatonin | 1.31(1.19–1.45) p < 0.0001 |
| Comorbid asthma | 1.05(1.01–1.10) p = 0.03 |
| Collection of inhaled long-acting muscarinic receptor antagonist (LAMA) | 1.03(.97–1.10) p = 0.30 |
| Collection of inhaled corticosteroids (ICS) | 1.06(1.01–1.12) p = 0.03 |
| Charlson comorbidity index | 1.01(1.00–1.02) p = 0.18 |
| Tobacco exposure ‘never smoking’, ‘previous smoking’, ‘active smoking’ |
1.02(.99–1.05) p = 0.20 |
| Medical Research Council (MRC) Dyspnea Scale | 1.04(1.02–1.06) p = 0.0003 |
| Decreasing body mass index (BMI) | 1.04(1.01–1.06) p = 0.01 |
| FEV1% GOLD stages | 1.07(1.04–1.10) p < 0.0001 |
Note: Hazard ratios were calculated by adjusted Cox analysis adjusting for all variables included in the propensity score match and collection of prescriptoins of promethazine and melatonin (N = 56,523).
Gender (male vs. female) and collection of promethazine, melatonin, long-acting muscarinic receptor antagonist (LAMA) and inhaled corticosteroid (ICS) were analyzed as binary variables.
Age (≤75 years, >75 years and ≤80 years, >80 years and ≤85 years, and >85 years), Charlson comorbidity index (CCI, range 0–13), tobacco exposure (‘never smoking’, ‘previous smoking’, and ‘active smoking’, MRC (range 1–5), BMI (≥25 kg/m2, <25 kg/m2 and ≥20 kg/m2, <20 kg/m2 and ≥15 kg/m2, and <15 kg/m2) and FEV1% GOLD stages (range 1–4) were analyzed as semi quantified variables.