Table 3. Sensitivity analyses of adjusted hazard ratios for risk of death.
| Hazard ratio (95% CI) | |||
|---|---|---|---|
| Antidepressant | Primary analysis | Excluding patients with cancer | Adjusted for HdPS |
| Citalopram | 1.00 | 1.00 | 1.00 |
| SSRI | |||
| Escitalopram | 0.95 (0.89–1.02), p = 0.137 |
0.94 (0.86–1.03), p = 0.192 |
0.96 (0.89–1.04), p = 0.356 |
| Sertraline | 0.96 (0.91–1.01), p = 0.088 |
0.98 (0.92–1.04), p = 0.506 |
1.02 (0.96–1.07), p = 0.575 |
| Fluoxetine | 0.86 (0.77–0.96), p = 0.006 |
0.85 (0.74–0.98), p = 0.028 |
0.95 (0.84–1.07), p = 0.380 |
| Paroxetine | 0.79 (0.71–0.89), p < .001 |
0.80 (0.69–0.93), p = 0.004 |
0.90 (0.79–1.03), p = 0.119 |
| SSNRI | |||
| Venlafaxine | 0.92 (0.85–0.99), p = 0.032 |
0.98 (0.89–1.09), p = 0.750 |
0.99 (0.91–1.08), p = 0.835 |
| Duloxetine | 0.63 (0.58–0.70), p < .001 |
0.72 (0.63–0.83), p < .001 |
0.83 (0.74–0.92), p < .001 |
| TCA | |||
| Amitriptyline | 1.15 (1.10–1.20), p < .001 |
0.88 (0.82–0.94), p < .001 |
1.04 (0.95–1.14), p = 0.353 |
| Doxepin | 0.79 (0.73–0.86), p < .001 |
0.68 (0.61–0.77), p < .001 |
0.87 (0.80–0.95), p = 0.001 |
| Trimipramine | 0.61 (0.55–0.69), p < .001 |
0.54 (0.46–0.64), p < .001 |
0.69 (0.61–0.77), p < .001 |
| Opipramol | 0.57 (0.53–0.61), p < .001 |
0.57 (0.51–0.63), p < .001 |
0.71 (0.65–0.76), p < .001 |
| NASSA | |||
| Mirtazapine | 0.94 (0.92–0.97), p < .001 |
0.92 (0.88–0.95), p < .001 |
0.98 (0.95–1.02), p = 0.335 |
| Herbal AD | |||
| St. John's wort | 0.42 (0.38–0.47), p < .001 |
0.42 (0.37–0.48), p < .001 |
NA |
Abbreviation: AD, antidepressant; SSRI, selective serotonin reuptake inhibitor; TCA, tricyclic ADs; SSNRI, selective noradrenalin reuptake inhibitor; NASSA, noradrenergic and specific serotonergic ADs; HdPS, high-dimensional propensity score; NA, not applicable due to non-overlap of propensity score distributions; CI, confidence interval.