Back

Clinician Article

---

Risk of acute myocardial infarction with NSAIDs in real world use: bayesian meta-analysis of individual patient data.

---

---

• Bally M
• Dendukuri N
• Rich B
• Nadeau L
• Helin-Salmivaara A
• Garbe E, et al.

Read abstract Read evidence summary Read full text

---

---

---

Clinical Comments

Cardiology

1. Observational (vs recent randomized trial suggesting no increased risk). 2. Similar to previous findings. 3. Even if increased risk, may be small in absolute terms for the average patient.

Cardiology

Most physicians are aware of an increased risk for cardiovascular events with NSAIDs. This sophisticated meta-analyses had three findings that I found surprising: 1. the risk seems to have a class effect and the much debated differences between NSAIDs were not evident in this analysis; 2. the magnitude of the risk is high; 3. while there is a dose-response relationship with risk, even short-term NSAID use is associated with substantial risk.

Family Medicine (FM)/General Practice (GP)

More evidence to be cautious with NSAIDs.

General Internal Medicine-Primary Care(US)

I think most of us knew we were kidding ourselves thinking naproxen would be safer. This seems to imply that if a patient wants to take NSAIDs and makes it through the first month, the risk is less after that - like passing a stress test?

General Internal Medicine-Primary Care(US)

This study`s Bayesian meta-analysis methodology is extremely complex, and likely beyond the full understanding of most clinicians. The results demonstrate a consistent increased risk for MI in NSAID and Cox-2 inhibitors. No particular drug was better than the others. The risk was particularly high in the first month of use. The risk was dose-related and higher doses (celecoxib >200mg; diclofenac >100 mg; ibuprofen >1200 mg; naproxen >750 mg;) clearly increased the overall risk.

General Internal Medicine-Primary Care(US)

The article highlights the association between NSAIDs and the risk for myocardial infarction, even at low doses and for less than seven days. Many clinicians are aware of the risk, but this analysis helps quantify the risks in ways that were not done before. After reading this article, I would strongly consider advising patients at an elevated risk for a coronary event to consider another medication for pain relief besides NSAIDs.

Internal Medicine

I think of an acute gout scenario, where this studies' findings of increased myocardial infarction likelihood with all NSAIDS suggests other MI risk factors should temper a decision to prescribe adequate dose NSAID rather than steroid or colchicine. Applying Bayesian method to retrospective data rather than new data collection is probably appropriate. Statistical description and defense of individual patient data use in meta-analyses is elegant, with appraisal of its accuracy deferred to experts in that field. Comment on diminished statistical power from exclusion of redacted IPD by (presumed HIPAA) policy is thought provoking. The huge patient group and granular analysis confirms a class increase of acute myocardial infarction with initial and maintenance NSAID prescription, more important than small differences in likelihood for individual medications.

Internal Medicine

This is useful confirmation that this is, in essence, a class effect.

Public Health

It's important to note that what this adds to the use of NSAIDS and risk of AMI. Practitioners should be more aware of the risks associated with higher dose problems and length of time, as noted, short term use and high dose spell increased risk.

Public Health

Cohort study using prescription/clinical database evaluating the association of NSAIDs use including selective COX inhibitors. The use of NSAIDs, any type, was associated with increased AMI risk especially within the first 1 month of initiation as compared to non-users. Increased doses also associated with increased risk.

Public Health

This IPD-level analysis provides interesting information relevant to practice that recent NSAID use is associated with increased risk for AMI. Are we seeing two different signals? Is the underlying acute exacerbation of an inflammatory condition the reason for plaque rupture or erosion rather than NSAID use at lower doses? Is that why chronic use is not so strongly associated as recent use? Is the second wave and the association with higher dose more related to interference with the balance between the different products of the cyclooxygenase pathway? We know that inflammation is associated with plaque rupture and erosion. The outcome of the trial of MTX in chronic ischemic heart disease may add insight. For now, I would warn patients against high-dose NSAIDs and look carefully at the CAD risk factor optimization when patients have a reason to start NSAIDs. This also raises public health concerns as NSAIDs are available OTC.