Background: It is generally believed that selective serotonin reuptake inhibitor (SSRI) drugs increase the risk of abnormal bleeding and decrease the risk of ischemic heart disease events by blocking the uptake of serotonin into platelets, leading to an impairment in the platelet hemostatic response.

Objective: To perform a detailed qualitative review of existing literature on the association of abnormal bleeding with the use of SSRIs.

Data sources: We conducted a PubMed search during June 2009 using the search terms antidepressants and SSRIs (including the names of individual SSRIs: fluoxetine, sertraline, paroxetine, fluvoxamine, citalopram, and escitalopram) in association with bleeding, platelets, hemostasis, nonsteroidal anti-inflammatory drugs (NSAIDs), aspirin, antiplatelet drugs, proton pump inhibitors, peptic ulcer, premenstrual dysphoric disorder, menstruation, pregnancy, postpartum hemorrhage, surgery, tooth extraction, dental bleeding, stroke, ischemic heart disease, and other terms related to the field. We then searched the reference lists of identified studies.

Study selection: We provide a qualitative discussion of all studies that would inform clinicians about the mechanisms of bleeding and bleeding risks associated with these drugs in different clinical contexts.

Results: Epidemiologic studies show that SSRI use is associated with roughly doubled odds of upper gastrointestinal (GI) bleeding; bleeding at other sites has been less commonly described, as has a possibly increased risk of bleeding associated with surgical procedures. The risk of SSRI-associated GI bleeding is increased with the concurrent use of NSAIDs, anticoagulants, and antiplatelet agents and is decreased by concurrent proton pump inhibitors. The risk of bleeding is increased in patients with cirrhosis of the liver or liver failure. There is, curiously, little literature on use of SSRIs and menstrual or postpartum blood loss. Selective serotonin reuptake inhibitors appear protective against ischemic heart disease events. The data are too limited to allow interpretations about influences on ischemic and hemorrhagic stroke.

Conclusions: On the basis of the findings of our literature search, we suggest that SSRI-induced increase in gastric acid secretion may explain the GI bleeding risk and that SSRI-related effects on platelet reactivity, endothelial reactivity, and inflammatory markers may explain the ischemic heart disease protective effect. Because the absolute risk of GI bleeds with SSRIs is low, precautions are probably necessary only in high-risk patients, such as those with acid-peptic disease and those with a history of bleeds. We discuss management issues and areas for future research.