Table 1.
GRADE evaluation of interventions for depression in adults: drug and physical treatments
| Important outcomes | Symptom severity, relapse rates, adverse effects | ||||||||
| Number of studies (participants) | Outcome | Comparison | Type of evidence | Quality | Consistency | Directness | Effect size | GRADE | Comment |
| What are the effects of drug and physical treatments in mild to moderate or severe depression? | |||||||||
| At least 150 (at least 16,000)[14] [15] [16] [17] [18] [19] [20] | Symptom severity | Prescription antidepressant drugs v placebo | 4 | 0 | 0 | –1 | 0 | Moderate | Directness point deducted for issues affecting generalisability of trial results to clinical practice (placebo issue, sponsorship, publication bias) |
| At least 24 (at least 2115)[23] [24] [25] [26] | Symptom severity | Prescription antidepressant drugs v placebo in older adults | 4 | –1 | 0 | –2 | 0 | Very low | Quality point deducted for incomplete reporting of results. Directness point deducted for diverse populations included and short-term RCTs |
| 1 (unclear)[27] | Symptom severity | Prescription antidepressant drugs v placebo in psychotic depression | 4 | –1 | 0 | –2 | 0 | Very low | Quality point deducted for incomplete reporting of results. Directness points deducted for imprecision of result (CI 0.5 to 147) and small number of comparators (amitriptyline only) |
| 4 (250)[28] | Symptom severity | Prescription antidepressant drugs v placebo in atypical depression | 4 | 0 | 0 | –1 | 0 | Moderate | Directness point deducted for 1 RCT with extreme result compared with other RCTs, which may affect robustness of overall result |
| 84 (5376)[38] | Symptom severity | Tricyclic antidepressants v each other | 4 | 0 | –1 | –1 | 0 | Low | Consistency point deducted for significant heterogeneity among RCTs. Directness point deducted for inclusion of treatments other than tricyclic antidepressants in analysis |
| 6 (551)[15] | Symptom severity | Low-dose tricyclic antidepressants v standard-dose tricyclic antidepressants | 4 | 0 | 0 | –2 | 0 | Low | Directness points deducted for short-term results and unclear clinical relevance of results |
| Unclear (unclear)[39] | Symptom severity | Tricyclic antidepressants plus benzodiazepines v tricyclic antidepressants alone | 4 | –1 | 0 | –2 | 0 | Very low | Quality point deducted for incomplete reporting of results. Directness points deducted for inclusion of treatments other than tricyclic antidepressants in analysis and unclear clinical relevance (improvement at 1 week, not apparent at 6 weeks) |
| 1 (unclear)[27] | Symptom severity | Tricyclic antidepressants in people with psychotic depression | 4 | –1 | 0 | –2 | +1 | Low | Quality point deducted for incomplete reporting of results. Directness points deducted for small number of comparators (imipramine and fluvoxamine only) and unclear outcome. Effect-size point added for RR >2 |
| 117 (25,928)[43] | Symptom severity | SSRIs v each other | 4 | 0 | 0 | –1 | 0 | Moderate | Directness point deducted for generalisability issues (applies to acute phase treatment only, differences between direct and multiple treatments meta-analysis analysis) |
| At least 49 (at least 4073)[20] [38] [44] | Symptom severity | SSRIs v tricyclic antidepressants | 4 | 0 | 0 | 0 | 0 | High | |
| At least 8 (at least 597)[20] [28] | Symptom severity | SSRIs v monoamine oxidase inhibitors | 4 | 0 | 0 | 0 | 0 | High | |
| 1 (50)[45] | Symptom severity | SSRIs plus benzodiazepines v SSRIs alone | 4 | –2 | 0 | –1 | 0 | Very low | Quality points deducted for sparse data and incomplete reporting of results. Directness point deducted for small number of comparators |
| 3 (102)[27] | Symptom severity | SSRIs v each other or other antidepressants in people with psychotic depression | 4 | –1 | 0 | –1 | 0 | Low | Quality point deducted for sparse data. Directness point deducted for unclear generalisability (small trials, widely different response rates between RCTs [response to fluvoxamine 30% in 1 RCT, 82% in another]) |
| At least 13 (at least 2153)[20] [22] [28] | Symptom severity | Monoamine oxidase inhibitors v tricyclic antidepressants | 4 | 0 | –1 | 0 | 0 | Moderate | Consistency point deducted for inconsistent results between RCTs depending on analysis |
| At least 15 (at least 773)[20] [22] [42] | Symptom severity | Venlafaxine v tricyclic antidepressants | 4 | –1 | 0 | 0 | 0 | Moderate | Quality point deducted for incomplete reporting of results in 1 review |
| At least 29 (at least 3692 people)[20] [22] [42] [43] | Symptom severity | Venlafaxine v SSRIs | 4 | –1 | 0 | 0 | 0 | Moderate | Quality point deducted for incomplete reporting of results in 1 review |
| 6 (256)[64] | Symptom severity | Electroconvulsive therapy v simulated electroconvulsive therapy | 4 | 0 | 0 | –1 | 0 | Moderate | Directness point deducted for clinical heterogeneity |
| 18 (1144)[64] | Symptom severity | Electroconvulsive therapy v prescription antidepressant drugs | 4 | 0 | 0 | –1 | 0 | Moderate | Directness point deducted for clinical heterogeneity |
| 25 (1635)[64] [66] [67] [68] | Symptom severity | Bilateral v unilateral electroconvulsive therapy | 4 | 0 | 0 | –1 | 0 | Moderate | Directness point deducted for clinical heterogeneity |
| 1 (64)[69] | Symptom severity | Different types of electroconvulsive therapy v each other in older adults | 4 | –3 | 0 | 0 | 0 | Very low | Quality points deducted for sparse data, post-hoc analysis, and no intention-to-treat analysis |
| 18 (3064)[41] | Symptom severity | St John’s wort v placebo | 4 | –1 | –1 | –2 | 0 | Very low | Quality point deducted for weak methods. Consistency point deducted for statistical heterogeneity. Directness points deducted for variation in preparations used and variation in results depending on location of trial |
| 5 (1016)[41] | Symptom severity | St John’s wort v tricyclic antidepressants | 4 | –1 | 0 | –2 | 0 | Very low | Quality point deducted for weak methods. Directness points deducted for variation in preparations used and variation in results depending on location of trial |
| 12 (1794)[41] | Symptom severity | St John’s wort v SSRIs | 4 | –1 | 0 | –2 | 0 | Very low | Quality point deducted for weak methods. Directness points deducted for variation in preparations used and variation in results depending on location of trial |
| At least 28 (at least 1267)[74] [75] [76] [77] [78] | Symptom severity | Exercise v control or different forms of exercise or other treatments | 4 | –2 | –1 | –1 | 0 | Very low | Quality points deducted for weak methods and incomplete reporting of results. Consistency point deducted for statistical heterogeneity. Directness point deducted for inclusion of non-clinical populations |
| 4 (298)[79] | Symptom severity | Exercise v control or different forms of exercise or other treatments in older adults | 4 | –2 | 0 | –1 | 0 | Very low | Quality points deducted for weak methods and incomplete reporting of results. Directness point deducted for heterogeneity among RCTs |
| What are the effects of interventions in treatment-resistant depression? | |||||||||
| At least 10 (at least 269)[80] [82] [81] | Symptom severity | Lithium augmentation v placebo augmentation | 4 | –2 | 0 | –1 | 0 | Very low | Quality points deducted for incomplete reporting of results and weak methods. Directness point deducted for clinical heterogeneity between RCTs |
| 4 (148)[80] [83] | Symptom severity | Augmentation with pindolol v placebo augmentation | 4 | –2 | 0 | 0 | 0 | Low | Quality points deducted for sparse data and for incomplete reporting of results |
| Which interventions reduce relapse rates? | |||||||||
| At least 31 (at least 4410)[85] [86] [87] [42] [84] | Relapse rates | Continuing prescription antidepressant drug v placebo | 4 | –1 | 0 | 0 | 0 | Moderate | Quality point deducted for incomplete reporting of results |
| 4 (419)[88] [89] [90] [91] | Relapse rates | Continuing prescription antidepressant drug v placebo in older adults | 4 | –1 | –1 | 0 | 0 | Low | Quality point deducted for incomplete reporting of results. Consistency point deducted for conflicting results |
Type of evidence: 4 = RCT. Consistency: similarity of results across studies.Directness: generalisability of population or outcomes.Effect size: based on relative risk or odds ratio.