NPC Archive Item: Concomitant use of an SSRI and NSAID increases the risk of an upper GI bleed six times

NOTE – This is an archive post from the NPC and has not been updated since first publication. Therefore, some hyperlinks may no longer be working.
MeReC Rapid Review NPC Logo

Loke YK, Trivedi AN, Singh S. Meta-analysis: gastrointestinal bleeding due to interaction between selective serotonin reuptake inhibitors and non-steroidal anti-inflammatory drugs. Aliment Pharmacol Ther 2008;27:31–40

The relative risk of experiencing an upper gastrointestinal (GI) haemorrhage is doubled in people who take a selective serotonin reuptake inhibitor (SSRI) antidepressant. When SSRIs and non-steroidal anti-inflammatory drugs (NSAIDs) are taken at the same time the relative risk is increased six-fold. The absolute risk to the individual will depend on whether they have risk factors for GI bleeds present (see table for details of numbers needed to harm (NNH)).

Prescribers should consider alternative treatments to SSRIs for patients who have risk factors for upper GI bleeding. Patients taking both SSRIs and NSAIDs should have their treatment reviewed, particularly if other risk factors for GI bleed are present (such as being aged 65 or over). Options include stopping the NSAID or changing it to an alternative analgesic, using a different type of antidepressant, or co-prescribing a proton pump inhibitor or misoprostol.

What is the background to this?
The adverse effects of antidepressants were discussed in a MeReC Briefing on the management of depression in primary care. Observational studies have suggested that SSRIs increase the risk of developing GI bleeding. Patients treated with antidepressants which have the highest affinity for the serotonin transporter (fluoxetine, sertraline, clomipramine and paroxetine) appear to be at greatest risk. Compared to patients receiving no treatment, the absolute risk increase seems relatively small, resulting in around three extra cases of upper GI bleeds requiring hospitalisation per 1000-patient years of treatment. The risk increase is similar to that of NSAIDs, and is enhanced if SSRIs and aspirin or NSAIDs are taken concomitantly. People with a history of GI bleeding, or aged 80 years or more, are at greater risk.

The British National Formulary (BNF) advises that SSRIs should be used with caution in people with a history of bleeding disorders (especially GI bleeding) and if used with other drugs that increase the risk of GI bleeding. In particular, prescribers are warned of an increased risk of bleeding when SSRIs are given with NSAIDs.

What does this study claim?
This meta-analysis of four observational studies included 153,000 patients. It found that the chance of having an upper GI haemorrhage more than doubled in patients taking SSRIs (pooled odds ratio [OR] 2.36, 95% confidence interval [CI] 1.44 to 3.85; P=0.0006), and tripled in those taking NSAIDs (OR 3.16, 95% CI 2.40 to 4.18, P<0.00001), compared with controls who were not. When NSAIDs and SSRIs were used concomitantly the chance of suffering an upper GI bleed was increased six times (OR 6.33, 95% CI 3.40 to 11.8; P<0.00001), compared with controls who were not taking these drugs.

So what?
The meta-analysis has limitations, not least because of the differences in study populations, types of NSAID and SSRI used, and co-existing morbidity amongst participants. Also, observational studies are subject to bias and confounding, and can only indicate association, not prove causation. Nevertheless, the meta-analysis adds to the weight of evidence showing that SSRIs and NSAIDs, alone and in combination, increase the risk of upper GI haemorrhage. It also provides further confirmation of the magnitude of that risk.

The authors estimate that 411 patients (aged 50 years and over with no previous ulcer drug use, or hospitalisation for an upper GI event) would need to be treated with SSRIs per year for one to suffer an upper GI haemorrhage. This number needed to harm [NNH] is estimated to be 106 in patients treated with both SSRIs and NSAIDs. This NNH decreases (i.e. absolute risk increases) if other risk factors for GI bleed are present (see table).

As the authors point out, although the increase in absolute risk to the individual patient may be modest, the population impact of SSRI-associated upper GI bleeding may be substantial because a large number of people are prescribed these drugs.

Healthcare professionals should consider patients’ GI history and concomitant medication before prescribing an SSRI. Particular care should be taken with elderly patients because the risk of upper GI bleeding increases with age. Also, the elderly are more likely to have comorbid conditions such as osteoarthritis, which may be treated using prescribed or over-the-counter NSAIDs.

Because they are at particularly high risk of an upper GI haemorrhage, patients taking NSAIDs and SSRIs together should have their treatment reviewed, particularly if other risk factors for GI bleed are present (such as being aged 65 or over). Options include stopping the NSAID or changing it to an alternative analgesic; using a different type of antidepressant; or co-prescribing a proton pump inhibitor.

Further information on the management of depression is available in the MeReC Briefing and on the depression section of NPC. Information on pain management is also available on NPC.

Study details

Design: meta-analysis of 4 controlled observational studies (1 cohort study and 3 case-control studies) involving 153,000 patients, which evaluated the association between upper GI haemorrhage and SSRIs and NSAIDs

Outcomes and Results:

Odds ratios (ORs) of upper GI haemorrhage:

OR (95% CI)

P value

SSRIs alone

2.36 (1.44 to 3.85)


NSAIDs alone

3.16 (2.40 to 4.18)



6.33 (3.40 to 11.82)


Estimated number needed to harm (NNH) using the ORs above, based on varying risk factors for an upper GI haemorrhage:

Patient population

Baseline upper GI event/10,000/year

NNH/year with SSRI (95% CI)

(95% CI)

Unselected >50years, no NSAID exposure


318 (152 to 979)

82 (41 to 181)

Risk factor for upper GI haemorrhage:

No previous ulcer drug use or past history of admission for upper GI event


411 (196 to 1266)

106 (53 to 233)

Ulcer drug use (past or present)


177 (85 to 545)

46 (24 to 101)

Past history of admission for upper GI event


121 (58 to 370)

32 (17 to 69)

Ulcer drug use past or present) and past history of admission for upper GI event


70 (34 to 214)

19 (10 to 41)

Please comment on this blog in the NPC discussion rooms, or using our feedback form.