Elsevier

The Lancet

Volume 363, Issue 9423, 29 May 2004, Pages 1751-1756
The Lancet

Articles
Cyclo-oxygenase-2 inhibitors versus non-selective non-steroidal anti-inflammatory drugs and congestive heart failure outcomes in elderly patients: a population-based cohort study

https://doi.org/10.1016/S0140-6736(04)16299-5Get rights and content

Summary

Background

Non-selective, non-steroidal anti-inflammatory drugs (NSAIDs) are associated with an increased risk of congestive heart failure, but little is known about the cardiovascular effects of a newer group of NSAIDS called selective cyclo-oxygenase (COX)-2 inhibitors. We aimed to compare rates of admission for congestive heart failure in elderly patients who were newly dispensed COX-2 inhibitors or non-selective NSAIDs.

Methods

In this population-based retrospective cohort study we identified NSAID-naive individuals aged 66 years or older, who were started on rofecoxib (n=14,583), celecoxib (n=18,908), and non-selective NSAIDs (n=5,391), and randomly selected non-NSAID users as controls (n=100,000).

Findings

Relative to non-NSAID users, patients on rofecoxib and non-selective NSAIDS had an increased risk of admission for congestive heart failure (adjusted rate ratio 1·8, 95% CI 1·5–2·2, and 1·4, 1·0–1·9, respectively), but not celecoxib (1·0, 0·8–1·3). Compared with celecoxib users, admission was significantly more likely in users of non-selective NSAIDs (1·4, 1·0–1·9) and rofecoxib (1·8, 1·4–2·4). Risk of admission for rofecoxib users was higher than that for non-selective NSAID users (1·5, 1·1–2·1). Of patients with no admission in the past 3 years, only rofecoxib users were at increased risk of subsequent admission relative to controls (1·8, 1·4–2·3).

Interpretation

These findings suggest a higher risk of admission for congestive heart failure in users of rofecoxib and non-selective NSAIDs, but not celecoxib, relative to non-NSAID controls.

Introduction

Rapid adoption of a selective group of non-steroidal anti-inflammatory drugs (NSAIDs), known as cyclo-oxygenase (COX)-2 inhibitors, into clinical practice1 has been met with both enthusiasm and caution. Although recent evidence suggests that the COX-2 inhibitors celecoxib and rofecoxib are associated with a lower risk of gastrointestinal events than non-selective NSAIDs,2, 3, 4 the cardiovascular safety of these agents has been challenged.5, 6 In addition to the debate surrounding their possible association with acute myocardial infarction, selective COX-2 inhibitors might be associated with cardiovascular and renal adverse effects that are similar to those of non-selective NSAIDs, raising systemic vascular resistance and reducing renal perfusion in susceptible individuals.7

In non-selective NSAID users, increases in blood pressure and development of peripheral oedema have been consistently associated with the development of congestive heart failure.8, 9, 10 However, little is known about the association between the use of selective COX-2 inhibitors and this condition. Two small studies separately comparing high doses of celecoxib11 and rofecoxib12 to non-selective NSAIDs and placebo reported slight decreases in water, sodium, and potassium excretion, and slight increases in systolic blood pressure in the selective COX-2 inhibitor and non-selective NSAID groups relative to placebo. Published and unpublished secondary analyses from two large randomised trials separately examining celecoxib2, 13 and rofecoxib3, 14 suggest differences between various NSAIDs with respect to onset of hypertension and oedema. In the celecoxib trial, the incidence of hypertension and peripheral oedema associated with the drug was significantly lower than that of the non-selective NSAID comparator group.2 The incidence of congestive heart failure was not reported. In the rofecoxib trial, frequency of hypertension and oedema associated with rofecoxib was significantly greater than that of naproxen.14 Incidence of congestive heart failure in rofecoxib users was higher than that in naproxen users, although this difference was not statistically significant (0·5% vs 0·2%, respectively, p=0·07). Two small trials directly comparing celecoxib and rofecoxib failed to show significant differences between the two drugs with respect to the effect on blood pressure.15, 16 However, two large randomised trials in elderly osteoarthritis patients with longstanding hypertension reported significantly greater increases in systolic blood pressure and a higher likelihood of onset, or worsening of, oedema in those receiving rofecoxib than those receiving celecoxib.17, 18

In the absence of a large randomised controlled trial comparing the effects of selective COX-2 inhibitors with non-selective NSAIDs and non-NSAID users with respect to admission for congestive heart failure, we examined this association in more than 140 000 NSAID-naive elderly patients.

Section snippets

Patients and procedures

We undertook a population-based retrospective cohort study by linking administrative health-care databases, which covered over 1·3 million individuals aged 65 years and older in Ontario, Canada. These patients had universal access to prescription drug coverage, hospital care, and physician services. This study was approved by the ethics review board of Sunnybrook and Women's College Health Sciences Centre.

The administrative health-care databases in Ontario allowed for cohort identification,

Results

Of about 1·3 million potential patients aged 65 years and older, 364 686 (28%) were dispensed an NSAID during the study period. From these individuals, we identified 11 606 new users of non-selective NSAIDs, 18 908 users of celecoxib, 14 583 users of rofecoxib, and 100 000 non-NSAID users (table 1) meeting our inclusion criteria. In non-selective NSAID users, most patients were started on the combination of diclofenac plus misoprostol (5720 individuals, 49%), naproxen (1920, 17%), ibuprofen

Discussion

Compared with non-NSAID users, we have recorded higher rates of admission for congestive heart failure in elderly patients who were initiated on treatment with rofecoxib and non-selective NSAIDs, but not celecoxib. These differences are clinically important, in view of the large numbers of patients given NSAIDs of any type. In particular, there was a greater risk of readmission in individuals previously admitted for heart failure and who received rofecoxib or non-selective NSAIDs. Further, all

References (27)

  • ER Heerdink et al.

    NSAIDs associated with increased risk of congestive heart failure in elderly patients taking diuretics

    Arch Intern Med

    (1998)
  • J Feenstra et al.

    Association of nonsteroidal anti-inflammatory drugs with first occurrence of heart failure and with relapsing heart failure: the Rotterdam study

    Arch Intern Med

    (2002)
  • J Rossat et al.

    Renal effects of selective cyclooxygenase-2 inhibition in normotensive salt-depleted subjects

    Clin Pharmacol Ther

    (1999)
  • Cited by (417)

    View all citing articles on Scopus
    View full text