Table 1 Systematic reviews reporting a quantitative synthesis of the evidence for the outcomes selected by the working group
Review | Type of study | Population | Intervention | Comparison | Outcome | Summary | AMSTAR-2 |
|---|---|---|---|---|---|---|---|
Clegg et al. 2011 [23] | Systematic review of observational studies and RCTs (no meta-analysis) | Patients from medicine and surgery settings, mostly elderly | Antidepressants Antipsychotics Benzodiazepines | No exposure or placebo | Risk of delirium | Increased risk for benzodiazepines as a class (1 case–control study; n = 1341; OR 3.0, 95% CI 1.3 to 6.8), with higher risk for longer-acting agents and higher doses; antipsychotics as a class (1 prospective cohort study; n = 325; OR 4.5; 95% CI 1.8 to 10.5), but not for haloperidol (1 RCT; n = 430; OR 0.9, 95% CI 0.6 to 1.3); and tricyclic antidepressants (1 RCT; n = 111; RR 1.7, 95% CI 1.4 to 2.1). | Critically low |
Dragioti et al. 2019 [30] | Umbrella review (meta-analyses of observational studies) | Mixed (general population; people with depression) | Antidepressants | No exposure | Cardiovascular risk | Increased risk of coronary heart disease for TCAs (N = 14; n = 347,750; OR 1.51, 95% CI 1.07 to 2.12; CE IV). | Critically low |
Risk of acute heart disease not increased for SSRIs (N = 14; n = 818,337; RR 1; 95% CI 0.83 to 1.22). | Critically low | ||||||
Risk of myocardial infarction not increased for antidepressants as a class (N = 21; n = 1,793,877; RR 1.03, 95% CI 0.88 to 1.22). | Critically low | ||||||
Increased risk of cerebrovascular disease for SSRIs (N = 6; n = 280,784; RR 1.26, 95% CI 1.14 to 1.39; CE III). | Critically low | ||||||
Risk of cerebrovascular disease not increased for TCAs (N = 4; n = 278,749; RR 1.06, 95% CI 0.96 to 1.17). | Critically low | ||||||
Coagulation risk | Increase risk of severe bleeding at any site for SSRIs and SNRIs taken together (N = 44; n = 1,443,029; OR 1.41, 95% CI 1.27–1.57; CE II). | Critically low | |||||
Dzahini et al. 2018 [32] | Systematic review and meta-analysis of observational studies | Mixed (general population; people with schizophrenia, bipolar disorder, depression), mostly elderly | Antipsychotics | No exposure | Risk of infections | Risk of pneumonia is increased by FGAs (N = 5; n = 29,510, RR 1.69, 95% CI 1.34 to 2.15), SGAs (N = 6; n = 30,656; RR 1.93, 95% CI 1.55 to 2.41) and antipsychotics as a class (N = 7; n = 30,760; RR 1.83, 95% CI 1.60 to 2.10). No differences emerged between FGAs and SGAs. | Critically low |
Huhn et al. 2019 [48] | Systematic review and network meta-analysis of RCTs | Adults with multi-episode schizophrenia | Antipsychotics | Placebo | Cardiovascular risk | Significantly increased risk of QTc prolongation (N = 51; n = 15,467): quetiapine (OR 3.43, 95% CI 0.94 to 6.0); olanzapine (OR 4.29, 95% CI 1.91 to 6.68); risperidone (OR 4.77, 95% CI 2.68 to 6.87); iloperidone (OR 6.93, 95% CI 4.49 to 9.36); ziprasidone (OR 9.7, 95% CI 7.43 to 12.04); amisulpride (OR 14.1, 95% CI 7.71 to 20.45); serenditole (OR 23.9, 95% CI 20.56 to 27.33). | Low |
Kunutsor et al. 2018 [55] | Systematic review and meta-analysis of observational studies | General population | Antidepressants | No exposure | Coagulation risk | Increased risk of venous thromboembolism for antidepressants as a class (N = 6; n = 828,327; OR 1.27; 95% CI 1.06 to 1.51), TCAs (N = 4; n = 59,161; OR 1.16; 95% CI 1.06 to 1.27), SSRIs (N = 4; n = 58,088; OR 1.12; 95% CI 1.02 to 1.23), and other antidepressants (N = 3; n = 3198; OR 1.59, 95% CI 1.21 to 2.09). | Critically low |
Lu et al. 2016 [59] | Systematic review and meta-analysis of RCTs | Adults with insomnia and COPD | Benzodiazepines | Placebo | Respiratory risk | No differences between benzodiazepines (i.e., triazolam and temazepam) and placebo in terms of percentage of time below 90% arterial oxygen saturation during sleep (N = 3; n = 94; weighted MD 1.32; 95% CI − 7.33 to 9.97) and other respiratory outcomes during sleep (sleep apnea, Apnea–Hypopnea Index, arterial oxygen saturation). | Critically low |
Ostuzzi et al. 2019 [105] | Systematic review and meta-analysis of RCTs | Adults with depression and ischemic heart disease | Antidepressants | Placebo | Cardiovascular risk | No differences emerged for antidepressants as a class (SSRIs the most represented) in terms of mortality because of cardiovascular events (N = 14; n = 2674; RD 0.0, 95% CI − 0.01 to 0.01) and nonfatal cardiac events (N = 9; n = 1869; RD − 0.01, 95% CI − 0.04 to 0.02). | High |
Papola et al. 2019 [70] | Umbrella review (meta-analysis of observational studies) | Mixed (general population; people with dementia, schizophrenia, or other psychiatric conditions), mostly elderly | Antipsychotics | No exposure | Cardiovascular risk | For antipsychotics as a class, there was an increased risk of sudden cardiac death (N = 6; n = 677,488; OR 2.24; 95% CI 1.45 to 3.46; CE III), myocardial infarction (N = 9; n = 399,868; OR 2.21; 95% CI 1.41 to 3.46; CE III), and stroke (N = 9; n = 65,700; OR 1.45, 95% CI 1.24 to 1.7; CE III). Meta-regression showed that the risk of myocardial infarction and stroke was higher in the elderly. | Moderate |
Coagulation risk | For antipsychotics as a class, there was an increased risk of venous thromboembolism (N = 14; n = 31,417,175; OR 1.55, 95% CI 1.31 to 1.83; CE II). | Low | |||||
Pollok et al. 2018 [72] | Systematic review and meta-analysis of RCTs | Adults with depression and COPD | Antidepressants | Placebo | Respiratory risk | No increased risk of respiratory impairment for TCAs (change in dyspnea during walk according to the Pulmonary Functional Status Instrument: N = 1; n = 30; MD 0.50; 95% CI − 1.34 to 2.34) and SSRIs (change in FEV1: N = 2; n = 148; MD 0.01; 95% CI − 0.03 to 0.05). | High |
Schneider-Thoma et al. 2019 [78] | Systematic review and meta-analysis of RCTs | Mixed (83% adults; 8% elderly; 42% schizophrenia; 30% bipolar disorder; 11% depression; 6% dementia) | Antipsychotics (mostly second-generation) | Placebo | Respiratory risk | Increased risk of respiratory, thoracic, and mediastinal serious adverse events in studies with at least one serious adverse event according to a maximum estimate (worst-case scenario) (N = 38; n = 13,007; OR 1.72; 95% CI 1.02 to 2.89). | Low |
Cardiovascular risk | Risk of cardiac (N = 54; n = 19,642; OR 1.22; 95% CI 0.85 to 1.75) and vascular serious adverse events (N = 33; n = 12,842; OR 1.82; 95% CI 0.97 to 3.41) were not increased in studies with at least one serious adverse event according to a maximum estimate (worst-case scenario). | Low | |||||
Risk of infections | Increased risk of infections in studies with at least one serious adverse event according to a maximum estimate (worst-case scenario) (N = 88; n = 28,479; OR 1.43; 95% CI 1.06 to 1.92). | Low | |||||
Sun et al. 2018 [83] | Systematic review and meta-analysis of observational studies | General population, mostly elderly | Benzodiazepines | No exposure | Risk of infections | Increased risk of pneumonia for benzodiazepines and related medications (e.g., zolpidem) (N = 10; n = 1,520,285; OR 1.25; 95% CI 1.09 to 1.44). The risk was confirmed for current and recent exposure (not for past exposure); for long-acting, intermediate-acting and short-acting agents; and for younger and older patients. | Critically low |
Wu et al. 2019 [95] | Systematic review and network meta-analysis of RCTs | Medical and surgical patients at risk of delirium | Antipsychotics Benzodiazepines Mood stabilizers | Placebo/treatment as usual | Risk of delirium | Decreased incidence of delirium as compared to placebo or treatment as usual (N = 38; n = 8168) emerged for olanzapine (OR 0.25; 95% CI 0.09 to 0.69) and risperidone (OR 0.27; 95% 0.07 to 0.99), while no differences emerged for lorazepam, haloperidol, and gabapentin. A higher incidence emerged for midazolam hydrochloride (OR 2.98; 95% CI 1.30 to 6.80). | Low |