INTRODUCTION — The use of nonsteroidal antiinflammatory drugs (NSAIDs) is associated with an increased risk of adverse cardiovascular events. Both nonselective NSAIDs and cyclooxygenase (COX)-2 selective NSAIDs (coxibs) increase the risk of such events. Although the risk of adverse cardiovascular events is heterogeneous across NSAIDs, the precise absolute and relative risk data for many of these agents are limited.
The most important adverse cardiovascular events include cardiovascular death, myocardial infarction (MI), and stroke. Other patient-important events include heart failure (HF), blood pressure elevation, atrial fibrillation, and venous thromboembolism. The effect of NSAIDs on blood pressure is reviewed separately. (See "NSAIDs and acetaminophen: Effects on blood pressure and hypertension".)
The adverse cardiovascular effects of NSAIDs will be reviewed here. Other adverse effects of NSAID therapy and the adverse effects of aspirin are presented separately. (See "Nonselective NSAIDs: Overview of adverse effects" and "Overview of COX-2 selective NSAIDs", section on 'Toxicities and possible toxicities' and "Aspirin: Mechanism of action, major toxicities, and use in rheumatic diseases" and "NSAIDs (including aspirin): Primary prevention of gastroduodenal toxicity".)
The pharmacology, mechanism of action, approach to therapeutic use, and the variability of responses to NSAIDs are also discussed in detail elsewhere. (See "NSAIDs: Pharmacology and mechanism of action" and "NSAIDs: Therapeutic use and variability of response in adults".)
PATHOPHYSIOLOGY — The mechanism by which nonsteroidal antiinflammatory drugs (NSAIDs) lead to an increase in cardiovascular events, such as myocardial ischemia and stroke, is likely related to their impact on inhibition of cyclooxygenase (COX)-2, which is associated with reduced prostaglandin I2 (PGI2 or prostacyclin) production by vascular endothelium with little or no inhibition of potentially prothrombotic platelet thromboxane A2 production . The relatively selective reduction in prostacyclin activity could predispose to endothelial injury . (See "The endothelium: A primer" and "Coronary artery endothelial dysfunction: Basic concepts".)
However, some analyses have failed to show a clear relationship of increased cardiovascular risk with COX-2 selectivity, and other factors are also important, especially as they impact blood pressure; renal function, including effects on fluid and electrolytes; endothelial cells; and nitric oxide production . (See "NSAIDs and acetaminophen: Effects on blood pressure and hypertension" and "NSAIDs: Acute kidney injury" and "NSAIDs: Electrolyte complications".)
NSAID TYPES — Over 20 nonsteroidal antiinflammatory drugs (NSAIDs) are grouped into six major classes based on chemical structure (table 1). The nonselective NSAIDs include ibuprofen, naproxen, ketoprofen, flurbiprofen, oxaprozin, etodolac, sulindac, indomethacin, ketorolac, meloxicam, piroxicam, meclofenamate, mefenamic acid, and nabumetone. The cyclooxygenase (COX)-2 selective NSAIDs (coxibs) include celecoxib and etoricoxib, although the latter agent is not available in the United States (see "NSAIDs: Pharmacology and mechanism of action" and "NSAIDs: Therapeutic use and variability of response in adults"). Most coxibs, other than celecoxib and etoricoxib, have been withdrawn from the market in many countries. Diclofenac and meloxicam are not easily categorized as a nonselective NSAID or coxib [4,5]. (See "Overview of COX-2 selective NSAIDs".)
DETERMINANTS OF RISK — The use of nonsteroidal antiinflammatory drugs (NSAIDs) is associated with a small increase in the absolute risk of adverse cardiovascular events [3,6-15]. The risk of adverse cardiovascular events varies with both patient and drug characteristics. For example, the risk of myocardial infarction (MI) increases with higher doses, frequency of use, and established cardiovascular disease [3,8,12,13,16,17].
Patient characteristics — Patient factors that impact the risk of adverse cardiovascular events include the presence or absence of prior cardiovascular disease, history of systemic inflammatory disorder, older age, and male sex, as well as hypertension, hyperlipidemia, diabetes, and smoking. (See 'NSAIDs and cardiovascular events' below.)
In addition, the degree of adherence by patients to prescribed treatment regimens in routine clinical practice is likely to affect their level of risk, especially when it affects frequency of use.
While in most patient groups NSAIDs increase cardiovascular disease risk, this is less clear in patients with systemic inflammatory disorders such as rheumatoid arthritis. Baseline cardiovascular risk is increased in patients with rheumatoid arthritis and systemic lupus erythematosus. In a study of cardiovascular risk from diclofenac, patients with spondyloarthritis exhibited more than twofold greater cardiovascular risk than those with osteoarthritis . Some disease-modifying antirheumatic drugs (DMARDs) appear to have a protective cardiovascular effect based on observational studies, but it is not clear if NSAIDs provide a relative benefit or increase in risk. This issue is discussed separately. (See "Coronary artery disease in rheumatoid arthritis: Implications for prevention and management", section on 'DMARDs for control of inflammation due to RA' and "Coronary artery disease in rheumatoid arthritis: Implications for prevention and management", section on 'Nonsteroidal antiinflammatory drugs'.)
NSAID characteristics — NSAID characteristics that impact the risk of adverse cardiovascular events include the duration, frequency, and dose of therapy, and possibly the degree of selectivity for inhibition of cyclooxygenase (COX)-2 relative to COX-1 and other drug characteristics  (see 'NSAID types' above). The evidence regarding the potential impact of COX selectivity is discussed below. (See 'NSAIDs and cardiovascular events' below.)
With regard to duration, the risk of adverse cardiovascular events such as MI, stroke, or cardiovascular death is extremely small due to the limited exposure experienced over a short course of therapy, as might be used for patients with an acute but limited musculoskeletal injury.
●The use of most nonsteroidal antiinflammatory drugs (NSAIDs) is associated with an increased risk of adverse cardiovascular events. To minimize this risk, NSAIDs should be prescribed at the lowest effective dose for the shortest duration possible for the given indication. (See 'Determinants of risk' above and 'NSAIDs and cardiovascular events' below.)
●For most patients without known cardiovascular disease and without risk factors for gastrointestinal (GI) bleeding, we usually prescribe naproxen or ibuprofen, rather than other NSAIDs, because of their low cost, availability without a prescription, and flexible dosing. Other NSAIDs are reasonable alternatives. (See 'Patients without cardiovascular disease' below.)
For patients without known cardiovascular disease who are at increased risk of GI bleeding, we prefer a COX-2 selective NSAID (a coxib; eg, celecoxib or, where available, etoricoxib) to other NSAIDs, particularly if long-term use is anticipated. This is related to greater long-term GI safety of celecoxib compared with naproxen and ibuprofen and comparable cardiovascular safety. We assume GI safety of etoricoxib would be similar to celecoxib.
The use of gastro-prophylaxis in these patients is discussed elsewhere. (See "NSAIDs (including aspirin): Primary prevention of gastroduodenal toxicity" and "NSAIDs (including aspirin): Secondary prevention of gastroduodenal toxicity".)
●For most patients with established cardiovascular disease, we suggest using alternative analgesic agents (eg, topical NSAIDs, acetaminophen, or topical capsaicin, where appropriate depending upon the condition being treated) whenever possible. (See 'NSAIDs and cardiovascular events' below.)
For these patients with cardiovascular disease but no recent events who will be taking NSAIDs, particularly for long-term (greater than a month) use, we use either celecoxib at doses of up to 200 mg per day (due to the lower risk of bleeding) or naproxen at doses up to 500 mg twice daily. In these patients, we usually prescribe a proton pump inhibitor as well to reduce the risk of GI bleeding. We also prefer these agents for short-term (no more than a month or intermittent) use. (See "NSAIDs (including aspirin): Primary prevention of gastroduodenal toxicity" and "NSAIDs (including aspirin): Secondary prevention of gastroduodenal toxicity".)
●Other factors may also influence drug choice and the use of adjunctive therapies. In addition to baseline cardiovascular risk, other potential NSAID toxicities, such as risk of GI bleeding and renal disease, must be considered when making the choices of using an NSAID or between the various NSAIDs, as should the specific indication, the likely required dose of the drug, the other categories of medications that could be effective, and the duration of therapy. In patients using an NSAID plus either aspirin or another antiplatelet agent and those with a history of GI bleeding, we would generally also treat with a proton pump inhibitor to reduce the risk of a GI bleed. (See "NSAIDs (including aspirin): Primary prevention of gastroduodenal toxicity" and "NSAIDs (including aspirin): Secondary prevention of gastroduodenal toxicity" and "NSAIDs: Acute kidney injury" and "NSAIDs: Electrolyte complications" and "COX-2 inhibitors and gastroduodenal toxicity: Major clinical trials", section on 'Combined use of COX-2 inhibitors and PPIs'.)
NSAIDS AND CARDIOVASCULAR EVENTS — The use of nonsteroidal antiinflammatory drugs (NSAIDs) is associated with an increased risk of cardiovascular disease events in patients with and without cardiovascular disease. The increase in absolute risk is higher in those with established cardiovascular disease due to the higher baseline risk.
Patients without cardiovascular disease — In patients without cardiovascular disease, NSAIDs are associated with an increased relative risk for adverse cardiovascular disease outcomes such as myocardial infarction (MI), stroke, and cardiovascular death. This translates into a very small increase in absolute risk because the baseline risk is low in this patient group [3,6-15,17,20]. The increase in absolute risk appears to be greater and roughly similar with most NSAIDs when they are used at high doses [8,12,13,17,21].
The magnitude of the increase in relative and absolute risks is best illustrated by a 2013 meta-analysis of predominantly individual participant data from randomized trials that compared nonselective NSAIDs or selective cyclooxygenase (COX)-2 inhibitors (coxibs) with either placebo or another nonselective NSAID or coxib . The analysis (Coxib and traditional NSAID Trialists' [CNT] Collaboration) utilized data from over 300,000 individuals in over 600 trials. In these trials, only 9 percent had a history of cardiovascular disease. After follow-up of approximately one year, the meta-analysis found:
●Major cardiovascular events (a composite of nonfatal MI, nonfatal stroke, or vascular death) were increased compared with placebo for high-dose diclofenac (adjusted rate ratio [RR] 1.41, 95% CI 1.12-1.78) and for the coxibs (RR 1.37, 95% CI 1.14-1.66), largely due to increases in major coronary heart disease events. A similar, but statistically nonsignificant, increase in risk was observed with high-dose ibuprofen compared with placebo (RR 1.44, 95% CI 0.89-2.33). The use of high-dose naproxen did not result in an increase in major cardiovascular events (RR 0.93, 95% CI 0.69-1.27).
There was a trend toward an increase in vascular death, as an individual endpoint, with diclofenac, the coxibs, and ibuprofen compared with placebo (RR 1.65, 99% CI 0.95-2.85; 1.58, 99% CI 1.00-2.49; and 1.90, 99% CI 0.56-6.41, respectively). Risk of vascular death was not significantly increased by use of naproxen compared with placebo (RR 1.08, 99% CI 0.48-2.47).
●The study data permitted estimates of the excess absolute risk of a major vascular event or death associated with use of these agents. For patients at low cardiovascular risk (baseline risk of 0.5 percent per year), the use of diclofenac, ibuprofen, or a coxib resulted in an excess of two events per 1000 persons per year; naproxen was not associated with any excess risk. This small increase in risk will rarely influence the decision whether to use one of these drugs for short-term analgesia.
In the Prospective Randomized Evaluation of Celecoxib Integrated Safety versus Ibuprofen or Naproxen (PRECISION) trial, which was published after the meta-analysis, three-quarters of the patients did not have established cardiovascular disease but were at moderate to high risk . No differences were seen in cardiovascular outcomes between celecoxib, naproxen, and ibuprofen in the doses used in the trial. PRECISION randomly assigned 24,081 patients with osteoarthritis (90 percent) or rheumatoid arthritis to celecoxib (100 to 200 mg twice daily), naproxen (375 to 500 mg twice daily), or ibuprofen (600 to 800 mg three times daily). The mean (±standard deviation [SD]) daily doses for the drugs were: celecoxib, 209±37 mg; naproxen, 852±103 mg; and ibuprofen, 2045±246 mg. The primary composite outcome of cardiovascular death (including hemorrhagic death), nonfatal MI, or nonfatal stroke was similar across treatment arms: 2.3 percent in celecoxib, 2.5 percent in naproxen, and 2.7 percent in ibuprofen; hazard ratio (HR) for celecoxib versus naproxen, 0.93, 95% CI 0.76-1.13; HR for celecoxib versus ibuprofen, 0.85, 95% CI 0.70-1.04. The trial had some limitations . The numbers of patients who stopped the drugs were very high in the study as a whole over the duration of follow-up (68.8 percent of the patients stopped taking the study drug, and 27.4 percent of the patients discontinued follow-up), although it was similar across the three groups, as were the proportion who were lost to follow-up. The mean follow-up period was approximately 34 months. One other question raised about this trial is whether results may have been affected by systematic differences in the dosing of the agents. The trial also had important strengths, including the extremely large number of patients, their clinical relevance, and the applicability of the study protocol to typical practice.
Drug comparisons — For any given patient, the baseline risks for cardiovascular events and gastrointestinal (GI) bleeding need to be considered prior to recommending a specific agent. Celecoxib has a good profile with regard to GI risk, and the data available for the Coxib and Traditional NSAID Trialists' (CNT) Collaboration analysis suggested that high-dose naproxen might have the best cardiovascular risk profile among the NSAIDs analyzed [24,25].
In the Coxib and Traditional NSAID Trialists' (CNT) Collaboration analysis, which had a mean duration of follow-up of approximately one year (see 'Patients without cardiovascular disease' above), the risk of cardiovascular disease events was similar for all comparisons with the exception of high-dose naproxen. However, in PRECISION, where the duration of follow-up was nearly three years, no significant difference was found between naproxen and the other agents (ibuprofen and celecoxib). Other smaller studies have come to differing conclusions as to whether longer-term, high-dose naproxen might be safer than other nonselective NSAIDs [3,6,7,9-11,15,17,26-33]. CNT and PRECISION found a similar rate of adverse cardiovascular outcomes between some coxibs and some nonselective NSAIDs. The cardiovascular risk with celecoxib at doses of 200 mg twice daily is similar to nonselective NSAIDs [3,6,8-10,22].
However, it is possible that there are within-group differences. As an example, multiple observational studies have demonstrated consistently higher risk of a cardiovascular event for diclofenac compared with other NSAIDs (diclofenac was not studied in PRECISION) (see 'NSAID types' above). In a large, propensity-matched, nationwide cohort study of patients without established cardiovascular disease, the initiation of diclofenac was associated with increased adverse cardiovascular event rates compared with noninitiators (incidence rate ratio [IRR] 1.5, 95% CI 1.4-1.7) and with initiators of ibuprofen (IRR 1.2, 95% CI 1.1-1.3) and naproxen (IRR 1.3, 95% CI 1.1-1.5) . Rates were increased for MI, cardiac death, heart failure (HF), ischemic stroke, and atrial fibrillation and flutter. Two network meta-analyses of data from randomized trials and some observational studies also support this contention [3,9,15]. (See 'NSAID characteristics' above and 'NSAIDs and atrial fibrillation' below.)
Etoricoxib, a coxib available in many countries but not the United States, may be associated with a higher cardiovascular risk, including cardiovascular death, than some other NSAIDs. In the Multinational Etoricoxib and Diclofenac Arthritis Long-term (MEDAL) trials, there was no difference between diclofenac and etoricoxib in the risk of thrombotic cardiovascular events . Similar but imprecise results of comparable cardiovascular risk for etoricoxib and diclofenac were seen in a network meta-analysis of randomized trials and in a meta-analysis of observational studies [3,8].
Patients with known coronary heart disease — Due to the relatively high baseline risk for cardiovascular events in patients with established disease, we try to avoid using NSAIDs, particularly long term. In addition, as most of these patients are taking at least one antithrombotic agent (aspirin), the potential for GI bleeding needs to be considered.
In particular, we avoid NSAIDs in patients with recent acute MI, unstable angina, and during the perioperative period in patients undergoing coronary artery bypass surgery [19,36-39]. (See 'Recommendations of others' below and 'NSAIDs and heart failure' below.)
The increased risk of cardiovascular events described for NSAIDs in patients without cardiovascular disease (see 'Patients without cardiovascular disease' above) has also been reported in large randomized trials and observational studies including patients with established coronary artery, peripheral artery, or cerebrovascular disease, or those at high cardiovascular disease risk [22,40-43]. (See "Overview of established risk factors for cardiovascular disease".)
The following three studies included patients with established cardiovascular disease:
●In the PRECISION trial, which compared naproxen, ibuprofen, and celecoxib, about 25 percent of patients had established disease. In subgroup analyses, comparisons between drugs led to similar findings to the entire population . Thus, no significant difference in cardiovascular safety among the three drugs evaluated (naproxen, celecoxib, and ibuprofen) was demonstrated. (See 'Patients without cardiovascular disease' above.)
●The absolute excess risk in patients at high risk of cardiovascular disease was estimated in the 2013 CNT meta-analysis of randomized trial data  (see 'Patients without cardiovascular disease' above). In patients at high cardiovascular risk (baseline risk of 2 percent per year), the increased number of major vascular events for patients taking diclofenac or ibuprofen was estimated at seven to eight excess events per 1000 persons per year, including two fatal events. Coxibs, diclofenac, and ibuprofen increased the absolute annual excess risk of a major vascular event to a similar degree in these patients. However, in this meta-analysis, high-dose naproxen did not increase such risk. This issue is discussed above. (See 'Patients without cardiovascular disease' above.)
●A 2020 observational study from Korea evaluated cardiovascular and bleeding events in 108,232 patients with MI . During a mean follow-up of 2.3 years after the diagnosis of MI, comparing users of NSAID to non-users, the following were noted:
•NSAID use increased the risk for thrombotic cardiovascular events including recurrent MI, ischemic stroke or transient ischemic attack, or systemic arterial embolism (HR 6.96, 95% CI 6.24-7.77).
•NSAID use increased the risk for bleeding events (HR 4.08, 95% CI 3.51-4.73).
•Among the various NSAIDs tested, the point estimates for the level of increased risk for cardiovascular and bleeding events, compared with no NSAID use, were lowest with celecoxib (HR 4.65, 95% CI 3.17-6.82 and 3.44, 95% CI 2.20-5.39, respectively) and meloxicam (HR 3.03, 95% CI 1.68-5.47 and 2.80, 95% CI 1.40-5.60, respectively). Limitations of this study include enrollment of patients treated with older antithrombotic therapy, lack of data regarding over-the-counter use of NSAIDs and specific drug dosing, and unmeasured confounding.
Aspirin and other antithrombotic agents — Most cardiovascular disease patients and some without cardiovascular disease receive long-term antithrombotic therapy including aspirin, aspirin plus a P2Y12 receptor blocker, or an oral anticoagulant. The addition of NSAID therapy leads to an increased risk of bleeding, particularly from the GI tract.
All patients who are prescribed an NSAID and one or more antithrombotic agent should be counseled on the potential negative interaction and advised of potential prophylactic measures, including the use of proton pump inhibitors. (See "NSAIDs (including aspirin): Pathogenesis and risk factors for gastroduodenal toxicity" and "NSAIDs (including aspirin): Primary prevention of gastroduodenal toxicity" and "Nonselective NSAIDs: Overview of adverse effects", section on 'Gastrointestinal effects' and "Coronary artery disease patients requiring combined anticoagulant and antiplatelet therapy", section on 'Bleeding'.)
Despite the potential for an increased risk of GI bleeding with aspirin-plus-NSAID therapy, we believe the use of NSAIDs is reasonable in patients for whom they are clearly needed. In these patients, we prefer celecoxib at doses of up to 200 mg per day to other NSAIDs based upon its lesser GI risk and lack of known interaction with low-dose aspirin . Celecoxib demonstrated superior GI safety compared with naproxen and ibuprofen among aspirin users and nonusers [24,25]. Naproxen up to 500 mg twice daily with appropriate gastro-protection is an alternative, particularly for patients needing a high-dose NSAID, given some evidence for lower cardiovascular risk at high doses compared with other NSAIDs. (See "COX-2 inhibitors and gastroduodenal toxicity: Major clinical trials", section on 'Combined use of COX-2 inhibitors and PPIs' and "NSAIDs (including aspirin): Pathogenesis and risk factors for gastroduodenal toxicity" and "NSAIDs (including aspirin): Primary prevention of gastroduodenal toxicity" and 'Patients without cardiovascular disease' above.)
The beneficial antiplatelet effects of aspirin for prevention of cardiovascular events result from irreversible acetylation of the active site of COX in platelets. There is some evidence that suggests that some nonselective NSAIDs, such as ibuprofen and naproxen, attenuate this effect of aspirin [45-48] (see "Aspirin in the primary prevention of cardiovascular disease and cancer"). A warning in the ibuprofen package insert suggests using low-dose aspirin at least 30 minutes before ibuprofen or 8 hours after because of concerns regarding this interaction.
When NSAID therapy is necessary, some experts advise the patient that aspirin be taken at least two hours before taking a nonselective NSAID [46,49,50], although the clinical relevance of the capacity for the nonselective NSAIDs to interfere with the effects of aspirin on COX is uncertain. One clinical trial  has suggested an adverse effect on clinical outcomes from this interaction, while in a subgroup analysis of the PRECISION trial, there was no significant difference in cardiovascular risk between those receiving or not receiving aspirin  (see 'Patients without cardiovascular disease' above). The results found in several observational studies have been inconsistent [15,26,49-54].
The presumed mechanism of interference with the antiplatelet activity of aspirin is that nonselective NSAIDs compete with aspirin for a common binding site on COX-1 and prevent aspirin from binding [55,56]. As a result, aspirin is unable to acetylate a serine residue on COX-1, an irreversible reaction that inhibits COX-1 for the remaining life of the platelet . This pharmacodynamic interaction is not seen with celecoxib or diclofenac [46,57,58]. The blockade of the acetylation site caused by ibuprofen is less complete than that of aspirin at the doses used and declines rapidly between doses . The inhibitory effect of naproxen on platelet function is greater than that of ibuprofen, and its half-life is substantially longer . (See "Aspirin for the secondary prevention of atherosclerotic cardiovascular disease".)
Comparison with alternative agents — Acetaminophen or opioids have been suggested as alternative analgesics to avoid the cardiovascular risks associated with NSAID use when the antiinflammatory effects of NSAIDs are not required ; however, data from some observational studies suggest that use of these alternative agents may also be associated with various risks, including adverse cardiovascular effects [12,34,60].
There are no randomized trials that directly compare the cardiovascular risks of NSAIDs with acetaminophen or opioids, but retrospective observations in large cohorts or databases have examined this question in selected populations. As examples:
●In one large analysis of multiple population-based cohorts, the risk of major adverse cardiovascular events was very similar for acetaminophen and ibuprofen when each was compared with diclofenac using propensity matching and emulated trial design (IRRs for both diclofenac versus acetaminophen and for diclofenac versus ibuprofen, 1.2, 95% CI 1.1-1.3) .
●In another study, in a large cohort of women followed longitudinally, the relative risk of major cardiovascular events with acetaminophen was elevated compared with no therapy (nonusers) among frequent (≥22 days/month) users of acetaminophen (RR 1.35, 95% CI 1.14-1.59) and increased to greater degree among women who used ≥15 tablets per week (RR 1.68, 95% CI 1.10-2.58) .
●The relative risk of adverse cardiovascular events in older adults using opioids was increased compared with nonselective NSAIDs (HR 1.77, 95% CI 1.39-2.24) .
Even if the actual degree of risk with opioids is not as high as might be suggested from the limited data available, the potential for abuse limits the use of opioids; this must be factored into decision making. (See "Opioid use disorder: Epidemiology, pharmacology, clinical manifestations, course, screening, assessment, and diagnosis" and "Pharmacologic management of chronic non-cancer pain in adults", section on 'Opioids'.)
PATIENTS WITH HYPERTENSION — Nonsteroidal antiinflammatory drugs (NSAIDs), including both nonselective NSAIDs and coxibs, can raise the blood pressure and worsen control of hypertension in patients already being treated. The effects differ between agents . NSAIDs should be used with caution in such patients, and if required, they should be used at the lowest effective dose and for the shortest duration necessary for the given indication. If blood pressure becomes elevated while using NSAIDs, discontinuing or reducing the dose of the NSAID is generally preferred over adding antihypertensives, when possible. The impact of these drugs on blood pressure is reviewed in detail separately. (See "NSAIDs and acetaminophen: Effects on blood pressure and hypertension".)
NSAIDS AND HEART FAILURE — The use of most nonsteroidal antiinflammatory drugs (NSAIDs) is associated with a small increased absolute risk of the development of new heart failure (HF) in most  but not all studies [63,64]. However, among patients with prior HF, the risk appears significantly increased [63-65]. (See "NSAIDs: Acute kidney injury" and "Drugs that should be avoided or used with caution in patients with heart failure", section on 'Nonsteroidal anti-inflammatory drugs'.)
All nonselective and cyclooxygenase (COX)-2 selective NSAIDs (coxibs) can induce fluid retention (see "NSAIDs: Electrolyte complications", section on 'Edema'). The most common situation in which HF might develop would be the patient with hypertension and diastolic dysfunction. (See "Heart failure with preserved ejection fraction: Clinical manifestations and diagnosis".)
The studies presented below, in aggregate, suggest a small increase in absolute risk of HF among patients at low baseline risk:
●In a 2013 collaborative meta-analysis of data from randomized trials involving over 300,000 patients with a low baseline prevalence of atherosclerosis (9 percent), the risk of the development of HF in the placebo group was less than 1 percent . This risk of HF doubled with all coxib and nonselective NSAID therapy, and all nonselective NSAIDs had a lower risk when compared with the coxibs.
●In one analysis, the absolute risks from celecoxib of new-onset HF in patients without known heart disease were also very low; as an example, in an analysis of six randomized trials in which 88 percent of subjects had no history of prior cardiovascular events, the absolute risk from use of celecoxib (400 mg total daily dose) was 0.9 per 1000 patient-years . The incidence of edema ranged from approximately 1 to 10 percent of patients. In clinical trials, coxibs (other than rofecoxib) have also been associated with the development of lower-extremity edema at rates similar to those for nonselective NSAIDs [1,67-70], although several studies have shown higher rates of edema with rofecoxib [67,71].
●In the Multinational Etoricoxib and Diclofenac Arthritis Long-term (MEDAL) trial program, the incidence of HF was similar with etoricoxib and diclofenac (0.1 to 0.7 percent in the several trials), as was the incidence of drug discontinuation due to edema (0.8 to 1.9 and 0.4 to 0.8 percent in the trials, respectively) .
The use of NSAIDs also increases the risk of HF and death in patients with preexisting HF. The best data in patients with known prevalent HF come from an observational study of over 40,000 patients in Denmark who had survived a first hospitalization for HF and had subsequently used NSAIDs as determined by a nationwide registry of drug dispensing by pharmacies . The outcomes in these patients were compared with over 70,000 controls with HF who had not received NSAIDs. The following findings were reported:
●The adjusted risk of rehospitalization for HF was increased in patients on diclofenac (adjusted hazard ratio [HR] 1.35) or ibuprofen (adjusted HR 1.16), and to a numerically higher, but not statistically significant, level for users of naproxen (adjusted HR 1.18).
●There was a dose-dependent increase in risk of death, which was highest with diclofenac (adjusted HR 2.08). Higher doses of ibuprofen (>1200 mg/day) and naproxen (>500 mg/day), but not lower doses, were also associated with an increased risk of death (adjusted HR 1.31 and 1.22, respectively, for the higher doses).
NSAIDS AND ATRIAL FIBRILLATION — The possible association of nonsteroidal antiinflammatory drug (NSAID) use with the development of atrial fibrillation is discussed separately. (See "Epidemiology of and risk factors for atrial fibrillation", section on 'Medications'.)
RECOMMENDATIONS OF OTHERS — Numerous regulatory agencies and professional groups have made recommendations regarding the management of cardiovascular risk associated with the use of NSAIDs [4,19,36-38,72-74]. Our recommendations are broadly consistent with those of others.
The US Food and Drug Administration (FDA) has periodically updated their warnings for all NSAIDs [75,76].
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Side effects of anti-inflammatory and anti-rheumatic drugs".)
INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5th to 6th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10th to 12th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.
Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)
●Basics topic (see "Patient education: Nonsteroidal antiinflammatory drugs (NSAIDs) (The Basics)")
●Beyond the Basics topic (see "Patient education: Nonsteroidal antiinflammatory drugs (NSAIDs) (Beyond the Basics)")
SUMMARY AND RECOMMENDATIONS
●In patients with and without cardiovascular disease, the use of most nonsteroidal antiinflammatory drugs (NSAIDs), including nonselective NSAIDs and cyclooxygenase (COX)-2 selective NSAIDs (coxibs), is associated with an increased risk of adverse cardiovascular events, including death, myocardial infarction (MI), heart failure (HF), and stroke. The increase in absolute risk varies depending upon the baseline cardiovascular risk of the patient, the NSAID chosen, and its dose. (See 'Determinants of risk' above and 'NSAIDs and cardiovascular events' above.)
To minimize the risk for an adverse cardiovascular event in patients treated with an NSAID, it should be prescribed at the lowest effective dose for the shortest duration possible.
●For most patients without known cardiovascular disease in whom an NSAID is chosen for short-term (no more than a month) or intermittent use, we consider all available NSAIDs as reasonable choices. We often choose naproxen or ibuprofen for reasons of cost or convenience. (See 'Patients without cardiovascular disease' above and 'Our approach' above.)
In patients without known cardiovascular disease and without gastrointestinal (GI) risk factors who require an NSAID for long-term (over a month) use, we suggest naproxen or ibuprofen, rather than other NSAIDs (Grade 2C).
In patients without known cardiovascular disease but with GI risk factors (eg, history of GI bleeding or use of an antiplatelet agent) who require an NSAID, particularly for long-term (over a month) use, we suggest celecoxib rather than other NSAIDs (Grade 2C). Etoricoxib, where available, is a reasonable alternative to celecoxib in these patients. (See 'Patients with known coronary heart disease' above.)
●For most patients with established cardiovascular disease who require an NSAID, we suggest using alternative analgesic agents (eg, acetaminophen) whenever possible (Grade 2C). (See 'Comparison with alternative agents' above.)
For the patient with cardiovascular disease without a recent event who will be taking NSAIDs, particularly for long-term use but for short-term use as well, we prefer celecoxib at doses of up to 200 mg per day due to the lower risk of bleeding or naproxen, up to 500 mg twice daily. (See 'Patients with known coronary heart disease' above.)
In patients using prescription or over-the-counter NSAIDs on an occasional short-term basis, some experts advise their patients to take aspirin at least two hours before the NSAID. The beneficial effect of aspirin may be attenuated by prior or ongoing administration of some nonselective NSAIDs, including ibuprofen and naproxen, and patients taking aspirin for cardiovascular prevention should be cautioned regarding the use of NSAIDs as over-the-counter medications. (See 'Aspirin and other antithrombotic agents' above.)
●In patients taking an NSAID long term with or without aspirin, gastro-prophylaxis should be considered. (See "NSAIDs (including aspirin): Primary prevention of gastroduodenal toxicity" and "NSAIDs (including aspirin): Secondary prevention of gastroduodenal toxicity" and "COX-2 inhibitors and gastroduodenal toxicity: Major clinical trials", section on 'Combined use of COX-2 inhibitors and PPIs'.)