The use of selective cyclooxygenase (COX)-2 inhibitors has increased substantially since this new class of medication was approved by the U.S. Food and Drug Administration. COX-2 inhibitors are significantly less likely to cause gastrointestinal ulcers than the nonselective nonsteroidal anti-inflammatory drugs (NSAIDs). The efficacy of these two classes of medications is the same. This decrease in gastrointestinal toxicity comes at a substantial cost, with the average wholesale price of a COX-2 inhibitor being 10 to 20 times greater than the price of generic ibuprofen. Because little is known about what factors influence the use of COX-2 inhibitors, Solomon and colleagues examined the effects of patient risk factors for gastrointestinal toxicity, other patient characteristics, and physician prescribing preferences on the decision to prescribe a COX-2 inhibitor.

Several characteristics may be used to identify patients who are at increased risk of NSAID-induced gastrointestinal toxicity. These characteristics include older age, a history of peptic ulcer disease, previous gastrointestinal hemorrhage, and use of oral glucocorticoids or warfarin. Two professional medical organizations have recommended that physicians use these risk factors to determine which patients are at risk for NSAID-induced gastrointestinal toxicity and should receive co-therapy with a gastro-protective agent (i.e., proton pump inhibitor, misoprostol) or a COX-2 inhibitor.

The study was a retrospective evaluation of Medicare beneficiaries enrolled in a state-wide pharmacy benefits program. Patients were included in the study if they filled a prescription for an NSAID or a COX-2 inhibitor during the index study year and if they continued to use the health care system. The data collected on NSAID users were limited to one randomly selected person for each COX-2 inhibitor user. Data were collected for the participants using diagnosis and procedure codes for inpatient and outpatient services. Known risk factors for NSAID-induced gastrointestinal toxicity were identified and recorded. In addition, other demographic patient data were collected.

Physician preference for prescribing COX-2 inhibitors was identified by determining which physicians had prescribed a COX-2 inhibitor for at least 10 patients and including them in the study. The physician preference was calculated using these data. Predictors of COX-2 inhibitor use were identified using multivariable logistic regression models.

The study included 28,190 patients, of which one half filled a prescription for a COX-2 inhibitor. Of patients with no risk factors for NSAID-induced toxicity, 17 percent received a prescription for a COX-2 inhibitor, while 23 percent received a prescription for an NSAID. Of patients who had one risk factor, 48 percent received NSAIDs, but only 7 percent were given a gastroprotective medication. In addition, 14 percent of the patients with two or more risk factors received NSAID prescriptions. Among the patients with two or more risk factors, the use of COX-2 inhibitors was significantly higher than the use of NSAIDs. When calculating the risk model for COX-2 inhibitor use, the risk factors for NSAID-induced gastrointestinal toxicity were poor predictors. When adding patient characteristics and physician prescribing preferences, the model had an excellent ability to discriminate between the two groups.

The authors conclude that known risk factors for NSAID-induced gastrointestinal toxicity are poor predictors of which patient will be given a prescription for a COX-2 inhibitor. In this study, the COX-2 inhibitors were overused and underused. Other idiosyncratic physician factors in the prescribing practices of the COX-2 inhibitor medications may be present that shape the use of this class of medication.

editor’s note: The COX-2 inhibitors are marketed as being safer and equally effective alternatives to NSAIDs. The emphasis should be on the term safer, because even COX-2 inhibitors have potential adverse effects. Newer medications can be overused for a variety of reasons. In this study, Solomon and colleagues demonstrated that the use of the COX-2 inhibitors was not necessarily based on the risk of NSAID-induced gastrointestinal toxicity. When using a new class of medications, physicians should be aware of the agent’s indications and of which patients would best benefit from a new drug.—K.E.M.