Table 3.
Logistic regression analysis of hydroxychloroquine AEs.
| Coefficients | Estimate | Std. Error | Adjusted OR | 95% CI | p value |
|---|---|---|---|---|---|
| Model 3a: Hydroxychloroquine usage and disease state as predictors of cardiac adverse reactions | |||||
| n = 704,994 AIC = 320,950 | |||||
| Hydroxychloroquine | 0.196 | 0.0161 | 1.22 | (1.18–1.25) | < 2 × 10–16* |
| SLE | 0.522 | 0.0242 | 1.68 | (1.61–1.77) | < 2 × 10–16* |
| Model 3b: Hydroxychloroquine, NSAID, and aspirin usage and disease state as predictors of cardiac adverse reactions | |||||
| n = 704,994 AIC = 314,075 | |||||
| Hydroxychloroquine | 0.011 | 0.0166 | 1.01 | (0.98–1.04) | 0.505 |
| SLE | 0.566 | 0.0245 | 1.76 | (1.68–1.85) | < 2 × 10–16* |
| Number of Concurrent NSAIDs | 0.533 | 0.0096 | 1.70 | (1.67–1.74) | < 2 × 10–16* |
| Aspirin | 1.096 | 0.0175 | 2.99 | (2.89–3.10) | < 2 × 10–16* |
| Model 4a: Hydroxychloroquine usage, age, sex, and disease state as predictors of cardiac adverse reactions | |||||
| n = 509,229 AIC = 234,519 | |||||
| Hydroxychloroquine | 0.316 | 0.0185 | 1.37 | (1.32–1.42) | < 2 × 10–16* |
| SLE | 0.945 | 0.0293 | 2.57 | (2.43–2.72) | < 2 × 10–16* |
| Male | 0.450 | 0.0132 | 1.57 | (1.53–1.61) | < 2 × 10–16* |
| Age (odds per year) | 0.026 | 0.0005 | 1.03 | (1.03–1.03) | < 2 × 10–16* |
| Model 4b: Hydroxychloroquine, NSAID, and aspirin usage, age, sex, and disease state as predictors of cardiac adverse reactions | |||||
| n = 509,229 AIC = 229,653 | |||||
| Hydroxychloroquine | 0.136 | 0.0190 | 1.15 | (1.10–1.19) | 8.2 × 10–13* |
| SLE | 0.969 | 0.0297 | 2.64 | (2.49–2.79) | < 2 × 10–16* |
| Male | 0.420 | 0.0134 | 1.52 | (1.48–1.56) | < 2 × 10–16* |
| Age (odds per year) | 0.024 | 0.0005 | 1.02 | (1.02–1.03) | < 2 × 10–16* |
| Number of Concurrent NSAIDs | 0.574 | 0.0109 | 1.77 | (1.74–1.81) | < 2 × 10–16* |
| Aspirin | 0.919 | 0.0200 | 2.51 | (2.41–2.61) | < 2 × 10–16* |
* = coefficients with significant p values.