Table 1.
Baseline characteristics of the included studies in the meta-analysis.
| Author, year | Study design | Country | Sample size | Population with antithrombotic therapy | Study period | Age in years, mean ± SD | Male sex, n (%) | Exposure: SRI antidepressants | Outcomes |
|---|---|---|---|---|---|---|---|---|---|
| Kurdyak et al. [43], 2005 | Nested case-control | Canada | 16734 | Elderly patients (>65 years) treated with warfarin for ≥ 1 years | January 1994– December 2002 | 80.8 ± 6.6 | NR | SSRIs: citalopram, fluoxetine, fluvoxamine, paroxetine, sertraline | Gastrointestinal bleeding (UGIB) |
| Kharofa et al. [39], 2007† | Case–control | United States | 2692 | Treated with aspirin 2 weeks before index date | May 1997– October 2005 | Range: 50.9–70.2 | NR | SSRIs: citalopram, escitalopram, fluoxetine, paroxetine, sertraline | Brain haemorrhage (ICH and SAH) |
| de Abajo et al. [44], 2008† | Nested case–control | United Kingdom | 11321 | Current use of antiplatelet agents (primary low-dose aspirin) or oral anticoagulants (within 0–30 days of index date) | January 2000– December 2005 | 40–59 (27.2%) 60–69 (21.6%) 70–79 (34.7%) 80–84 (16.5%) |
6446 (56.9%) | SSRIs: citalopram, escitalopram, fluoxetine, fluvoxamine, paroxetine, sertraline SNRIs: duloxetine, venlafaxine |
Gastrointestinal bleeding (UGIB) |
| Schalekamp et al. [24], 2008 | Nested case–control | The Netherlands | 7666 | New users of coumarin: acenocoumarol (90.4%) and phenprocoumon (9.6%) | January 1991– December 2004 | 72.8 ± 9.8 | 4166 (54.3%) | SSRIs: citalopram, escitalopram, fluvoxamine, fluoxetine, paroxetine, sertraline | Major bleeding, Brain haemorrhage (ICH), Gastrointestinal bleeding |
| Dall et al. [49], 2009† | Case–control | Denmark | 40154 | Current use of aspirin (within the past 90 days) | August 1995– July 2006 | 72.1 ± 14.1 | 20541 (51.2%) | SSRIs: citalopram, escitalopram, fluoxetine, fluvoxamine, paroxetine, sertraline | Gastrointestinal bleeding (UGIB) |
| Wallerstedt et al. [25], 2009 | Retrospective cohort | Sweden | 234 | Treated with warfarin due to AF | January 1999– September 2005 | 72.0 ± 7.0 | 122 (52.1%) | SSRIs: citalopram, escitalopram, fluoxetine, fluvoxamine, paroxetine, sertraline | Major bleeding |
| Cochran et al. [26], 2011† | Retrospective cohort | United States | 100 | Treated with warfarin in an outpatients for ≥ 6 months | January 2007– November 2009 | 58.5 ± 16.0 | 25 (25.0%) | SSRIs: citalopram, escitalopram, fluoxetine, paroxetine, sertraline | Major bleeding, any bleeding |
| Labos et al. [37], 2011 | Retrospective cohort | Canada | 27058 | ACS with antiplatelet therapy: aspirin, clopidogrel, DAPT (aspirin and clopidogrel) | January 1998– March 2007 | 72.7 ± 10.6 | 19087 (70.5%) | SSRIs: citalopram, escitalopram, fluoxetine, fluvoxamine, paroxetine, sertraline | Major bleeding, Gastrointestinal bleeding |
| Schelleman et al. [45], 2011 | Nested case–control | United States | 666235 | Treated with warfarin | January 1999 – December 2005 | 18–50 (12.9%) 51–60 (12.2%) 60–70 (18.9%) 70–80 (27.5%) ≥81 (28.5%) |
242984 (36.5%) | SSRIs: citalopram, escitalopram, fluoxetine, paroxetine, sertraline SNRIs: venlafaxine |
Gastrointestinal bleeding |
| Vitry et al. [27], 2011 | Retrospective cohort | Australia | 17661 | Veterans who were new users of warfarin | July 2002– June 2006 | 81.8 ± 4.4 | 11277 (63.8%) | SSRIs: not specified | Major bleeding |
| Baillargeon et al. [28], 2012 | Nested case–control | United States | 3192 | Treated with warfarin for at least 180 days | January 2007 – December 2008 | 66–70 (10.4%) 71–75 (19.3%) 76–80 (22.2%) 81–85 (22.6%) ≥86 (25.5%) |
1116 (35.0%) | SSRIs: citalopram, escitalopram, fluoxetine, fluvoxamine, paroxetine, sertraline SNRIs: desvenlafaxine, duloxetine, milnacipran, venlafaxine |
Major bleeding |
| Lin et al. [50], 2013 | Retrospective cohort | Taiwan | 3238 | Treated with clopidogrel with an average dose of > 150 DDD per one-half year | January 2001 – December 2010 | 68.6 ± 11.6 | 1899 (58.6%) | SSRIs: not specified | Gastrointestinal bleeding (UGIB, LGIB) |
| Mosholder et al. [29], 2013 | Retrospective cohort | United States | 324356 | Treated with warfarin for at least 1 months | June 2006– October 2010 | <65 (13.0%) 65–74 (28.2%) 75–84 (37.4%) >84 (21.4%) |
213803 (65.9%) | SSRIs: not specified | Major bleeding, Brain haemorrhage (ICH), Gastrointestinal bleeding |
| Seitz et al. [54], 2013‡ | Retrospective cohort | Canada | 8568 | Treated with antiplatelet agents or warfarin in the 120 days preceding index | April 2003 – December 2009 | 82.4 ± 7.0 | 1927 (22.5%) | Current users of high-affinity SRIs: citalopram, escitalopram, clomipramine, duloxetine, fluoxetine, fluvoxamine, paroxetine, sertraline, venlafaxine | Perioperative blood transfusion |
| Giang et al. [52], 2014 | Retrospective cohort | United States | 162 | Treated with DAPT (aspirin and P2Y12 inhibitors) following coronary stent placement | October 2010 – January 2012 | NR | NR | SSRIs: citalopram, fluoxetine, sertraline | Any bleeding |
| Nguyen et al. [48], 2014 | Retrospective cohort | United States | 3153 | Veterans who were prescribed warfarin | October 2009 – September 2011 | NR | NR | SSRIs: not specified | Any bleeding |
| Quinn et al. [30], 2014† | Retrospective cohort | United States | 9186 | Treated AF with warfarin among the ATRIA study | Diagnosed AF from Jul 1996 – Dec 1997, and followed up to 6 years | ≥75 (53.3%) | 5337 (58.1%) | SSRIs: citalopram, escitalopram, fluoxetine, fluvoxamine, paroxetine, sertraline SNRIs: venlafaxine |
Major bleeding, Brain haemorrhage |
| Rashid et al. [38], 2016 | Retrospective cohort | Australia | 839 | ACS underwent angioplasty and received DAPT | January 2014 – December 2015 | 61.8 ± 12.5 | 663 (79.0%) | SSRIs: not specified | Major bleeding, any bleeding |
| Lai et al. [46], 2017 | Retrospective cohort | United States | 21503 | Treated with DOACs: apixaban (25.3%), dabigatran (25.9%), rivaroxaban (48.5%) | November 2010 – December 2015 | 18–64 (22.5%) 65–74 (30.7%) ≥75 (46.8%) |
11597 (53.9%) | SSRIs: not specified | Gastrointestinal bleeding |
| Laursen et al. [55], 2017† | Prospective registry | Denmark | 14343 | Treated with low-dose aspirin (≤150 mg/d) | August 2006 – August 2014 | 75.0 ± 27.6 | 7727 (53.9%) | SSRIs: not specified | Endoscopy-refractory bleeding, re-bleeding in peptic ulcer bleeding |
| Renoux et al. [40], 2017† | Nested case–control | United Kingdom | 92738 | Current use of antiplatelet agents or oral anticoagulants ( within 1 month before index date) | January 1995– June 2014 | 66.6 ± 16.6 | 36305 (39.1%) | SSRIs: citalopram, escitalopram, fluoxetine, fluvoxamine, paroxetine, sertraline | Brain haemorrhage (ICH) |
| Samuel et al. [31], 2017 | Retrospective cohort | United States | 575 | Primary or secondary diagnosis of an VTE and treated with full dose enoxaparin | October 2009 – October 2014 | 59.0 ± 38.3 | 310 (53.9%) | SSRIs: citalopram, escitalopram, fluoxetine, fluvoxamine, paroxetine, sertraline | Major bleeding |
| Scheitz et al. [41], 2017 | Prospective registries | Finland, France, Germany, Netherlands, Switzerland | 6242 | Preadmission with anticoagulants among acute ischaemic stroke patients treated by thrombolysis | June 1998–August 2016 | 70.1 ± 14.0 | 3501 (56.1%) | SSRIs: citalopram, escitalopram, fluoxetine, fluvoxamine, paroxetine, sertraline | Brain haemorrhage (post-thrombolysis symptomatic ICH) |
| Quinn et al. [32], 2018§ | Prospective cohort: using data from the ROCKET AF trial | International collaboration | 1474 | AF patients treated with rivaroxaban or warfarin for the prevention of stroke/systematic embolism | December 2006– June 2009 | 73.8 ± 8.5 | 703 (47.7%) | SSRIs: citalopram, escitalopram, fluoxetine, fluvoxamine, paroxetine, sertraline SNRIs: desvenlafaxine, duloxetine, venlafaxine |
Major bleeding, any bleeding |
| Iasella et al. [53], 2019 | Retrospective cohort | United States | 6819 | Treated with DAPT (clopidogrel-based) after PCI | July 2010– December 2014 | 66.8 ± NR | 4516 (66.2%) | SSRIs: citalopram, escitalopram, fluoxetine, fluvoxamine, paroxetine, sertraline | Any bleeding |
| Luo et al. [51], 2019 | Retrospective cohort | Taiwan | 11105 | Treated with aspirin with an average dose of > 14 DDD per month | January 2001 – December 2010 | 64.0 ± 12.7 | 5864 (52.8%) | SSRIs: not specified | Gastrointestinal bleeding (UGIB) |
| Gaist et al. [42], 2020† | Case–control | Denmark | 446264 | Current use (supply with grace period extended [60 days] up to cover index date) of antiplatelet agents (low-dose aspirin, clopidogrel) or oral anticoagulants (phenprocoumon, warfarin, apixaban, dabigatran, rivaroxaban) | January 2000– December 2016 | 71.3 ± 14.8 | 290280 (65.0%) | SSRIs: citalopram, escitalopram, fluoxetine, fluvoxamine, paroxetine, sertraline | Brain haemorrhage (SDH) |
| Komen et al. [33], 2020 | Retrospective cohort | Sweden | 30595 | AF patients with a new prescription of warfarin or DOACs | July 2011– December 2017 | 74.4 ± 11.0 | 17139 (56.0%) | Antidepressant: SSRIs (61.0%), TCA (11.3%), other (27.7%) | Major bleeding, Gastrointestinal bleeding, brain haemorrhage |
| Lee et al. [34], 2020 | Nested case–control | Korea | 25893 | AF patients with a new prescription of DOACs (apixaban, rivaroxaban, edoxaban, dabigatran) | January 2013 – December 2017 | 76.3 ± 9.1 | 11949 (46.1%) | SSRIs: escitalopram, fluoxetine, fluvoxamine, paroxetine, sertraline | Major bleeding, Gastrointestinal bleeding (UGIB, LGIB), brain haemorrhage (ICH) |
| Marchena et al. [35], 2020 | Prospective registry | Spain | 47050 | Adult patients receiving anticoagulant therapy for VTE (VKAs, LMWH, DOACs) | February 2009– September 2019 | 66.5 ± 17.8 | 23999 (51.0%) | SSRIs: citalopram, escitalopram, paroxetine, sertraline, SNRIs: duloxetine, venlafaxine Mixed: mirtazapine, trazodone |
Major bleeding, brain haemorrhage (ICH) |
| Mawardi et al. [47], 2020 | Retrospective cohort | United States | 248 | LVAD patients treated with warfarin and aspirin (81 mg or 325 mg) | January 2009 – January 2016 | 52.4 ± 8.4 | 142 (57.2%) | SSRIs: citalopram, escitalopram, fluoxetine, paroxetine, sertraline SNRIs: duloxetine, venlafaxine Mixed: bupropion, mirtazapine, trazodone |
Gastrointestinal bleeding |
| Zhang et al. [36], 2020 | Nested case–control | United Kingdom | 1887 | Adult patients with new users of DOACs (dabigatran, apixaban, rivaroxaban) | January 2008– December 2015 | 78.7 ± 10.2 | 1175 (62.3%) | SSRIs: citalopram, escitalopram, fluoxetine, nefazodone, paroxetine, sertraline, SNRIs: venlafaxine, duloxetin | Major bleeding, Gastrointestinal bleeding |
†On the basis of the entire study population.
‡On the basis of the current and former serotonergic users.
§On the basis of the propensity-score matching.
ACS: acute coronary syndrome; AF: atrial fibrillation; ATRIA: AnTicoagulation and Risk factors In Atrial fibrillation; DAPT: dual antiplatelet therapy; DDD: defined daily dose; DOACs: direct oral anticoagulants; ICH: intracerebral haemorrhage; LMWH: low-molecular weight heparin); LVAD: left ventricular assist device; NR: not reported; PCI: percutaneous coronary intervention; ROCKET AF: Rivaroxaban once daily Oral direct factor xa inhibition Compared with vitamin K antagonism for prevention of Embolism and stroke Trial in Atrial Fibrillation; SAH: subarachnoid haemorrhage; SD: standard deviation; SDH: subdural haematoma; SRIs: serotonin reuptake inhibitors; LGIB: lower gastrointestinal tract bleeding; UGIB: upper gastrointestinal tract bleeding; VKAs: vitamin K antagonists; VTE: venous thromboembolism.