There have been concerns regarding the use of selective cyclooxygenase-2 inhibitors (cox-2 inhibitors) in place of conventional non-steroidal anti-inflammatory drugs (NSAIDs) due to an increased risk of adverse cardiovascular events [1]. This has been highlighted by the recent withdrawal of rofecoxib [2]. We wished to establish current prescribing patterns and characteristics of use for non-selective NSAIDs (nsNSAIDs) and cox-2 inhibitors, particularly rofecoxib, in Scottish general practice.

Using 55 geographically representative general practices [3], which collect reliable demographic, prescribing and morbidity information, we established the prescribing patterns for selective and non-selective NSAIDs over a 10-year period (1^st April 1992 to 31^st March 2002) and the characteristics of use for cox-2 inhibitors over a 1-year period (1^st April 2001 to 31^st March 2002) for 362 155 registered patients. All registered patients who had ever been prescribed an NSAID, excluding aspirin, were identified and relevant demographic, morbidity and prescribing data obtained.

Of the 362 155 registered patients, nsNSAIDs had been prescribed at least once for 90511 (25%) patients (M : F, 39 384 : 51 127; mean age 50.1 years, 95% CI: 50.0 years, 50.3 years) with 9313 (10%) also being prescribed a cox-2 inhibitor (M : F, 3231 : 6082; mean age 59.0 years, 95% CI: 58.7 years, 59.3 years). Between April 1992 and March 2002, 6% of all prescriptions in general practice were attributed to NSAIDs. Over the 10-year period, a decrease in nsNSAID prescribing was observed with a corresponding increase in cox-2 inhibitor prescribing subsequent to their introduction. Since April 1999, 30.9% of all NSAID prescriptions were for ibuprofen, 33.4% for diclofenac sodium, 10.7% for naproxen, 4.8% for meloxicam, 4.4% for rofecoxib, 1.8% for celecoxib and 1.0% for etodolac. Rofecoxib prescribing had increased to 8.4% of all NSAID prescriptions during 2001/02 and celecoxib followed a similar pattern, accounting for 4.4% of a1l NSAID prescribing in the same year. Rofecoxib was prescribed to 3031 patients in 2001/02 (M : F, 1004 : 2027; mean age 59.6 years, 95% CI: 58.9 years, 60.0 years). Of these rofecoxib users, 1.6% had a history of stroke and 5.2% had a diagnosis of coronary heart disease (CHD) during the same period. Aspirin was prescribed to 63.8% and 71.7% of rofecoxib users with a history of stroke and CHD, respectively.

We noted a significant uptake in the use of cox-2 inhibitors in Scottish general practice since their introduction, with a significant number of patients with a history of stroke or CHD being prescribed these agents.

1. Mamdani M, et al. Lancet 2004; 363: 1751.
2. Anonymous. http://www.vioxx.com/rofecoxib/vioxxd/consumer/index.jsp
    [Accessed October 2004].
3. Milne RM, et al. J Epidemiol Commun Health 1998; 52 (Suppl 1): 20s.