Low Risk. A gi score below l0 with a one-year risk of GI ulceration below 0.3% is considered low risk. At this level, there is probably little, if any, role for the COX-l-spar-ing agents, because the annual risk of GI bleeding in this population is below 0.3%.

When patients are unable to tolerate two previously prescribed traditional nonselective NSAIDs, a more selective COX-2 agent should be considered. Although dyspepsia and NSAID-induced GI bleeding are marginally related at best, the COX-l-sparing medications, compared with their traditional counterparts, are associated with a rate of GI side effects that is approximately one-third lower. Apcalis Oral Jelly

Moderate Risk. A gi score of ll to l5 with a one-year GI ulceration risk of 0.3% to l% is considered moderate risk. The costs of all of the COX-l-selective agents are still prohibitive for patients at this risk level and should be considered only when therapy with at least one or two traditional nonselective NSAIDs has failed.

Table 3 Costs of COX-1-Sparing Agents per Hundred Patients According to Actual Prescribing Patterns

Drug Cost
(Boehringer Ingelheim)
Rofecoxib $37,275.62
(Vioxx®, Merck)
(Celebrex®, Pharmacia/Pfizer)
Valdecoxib $31,207.50
(Bextra®, Pharmacia/Pfizer)
Figures are based on “tablet splitting,” when possible, as com­monly practiced in the Veterans Affairs health system.

High Risk. A gi score of l6 to 20 with a one-year GI ulceration risk of l% to 2% is considered high risk. With the average cost of a significant GI bleeding episode estimated at $20,000 per patient, a cost analysis based on risk can determine the cost-effectiveness of these medications. According to current prices, the less expensive COX-l–sparing drugs, such as meloxicam, are cost-effective for preventing GI bleeding in this population.

The cost of meloxicam in the federal health system is $0.44 per day for the 7.5-mg dose (half that of the l5-mg tablet). Because approximately 75% of patients taking meloxicam tablet stay with the 7.5-mg dose and 25% of patients increase their dose to l5 mg, the estimated annual cost per l00 patients for meloxicam is $20,075. In this high-risk population, the ability to prevent an episode of GI bleeding, even in only one of l00 patients treated per year, would justify the additional expense. A reduced incidence of GI bleeding not only affects direct medical costs but also results in indirect cost advantages related to decreased morbidity and improved quality of life.

Very High Risk. A gi score above 20 with a one-year GI ulceration risk of 2% or higher is considered extremely high risk. For patients at this level, any of the COX-l-sparing agents at that dose would be considered cost-effective, according to the costs of these medications in the federal health system (Table 3). An algorithm based on these cost considerations for use in the federal health system is shown in Table 4.

Table 4 Proposed Algorithm for the Veterans Affairs Health System: Evaluation of Risk for a Serious NSAID-Induced Gastrointestinal Event within the Next Year

Risk Level Points Recommendations
1 No risk 0-10 Patients may use a nonselective formulary NSAID; if therapy fails, they may use meloxicam (Mobic canadian, Boehringer Ingelheim)
2 Moderate risk 11-15 Patients may use a nonselective formulary NSAID; if therapy fails, they may use meloxicam
3 Significant risk 16-20 For patients receiving therapy for less than 30 days or for those using therapy intermittently, a standard NSAID may be used. For patients receiving therapy for more than 30 days, salsalate (e.g., Disalcid®, 3M;Salflex®, Amarin) or meloxicam may be used. If therapy is unsuccessful or intolerance to the agent develops, a COX-2 inhibitor may be used.
4 Substantial risk >20 Meloxicam or a COX-2 inhibitor may be used.
COX = cyclooxygenase; NSAID = nonsteroidal anti-inflammatory drug.

Other Considerations: Note that the risk of GI bleeding with the COX-l-sparing agents is not reduced to zero; therefore, before prescribing any of these medications, physicians and pharmacists should consider the risks and benefits of these drugs in the patients with the highest potential for GI bleeding (e.g., patients with a history of bleeding, those of advanced age, and those with concomitant medical conditions).

Patients who have experienced significant GI ADEs within the last five years should probably start taking a COX-l-sparing drug regardless of their gi score. Cardiovascular prophylaxis with aspirin should not influence the decision as to whether or not to use a COX-l-sparing medication, because the risk of GI events in patients taking low-dose aspirin (up to 325 mg/day) does not approach that of a full-prescription strength NSAID.


The data clearly show that COX-l-sparing agents, such as canadian celecoxib (Pharmacia/Pfizer) meloxicam, rofe-coxib, and valdecoxib (Bextra®, Pharmacia/Pfizer) reduce the risk of significant GI adverse events. However, because these drugs are much more expensive than comparative NSAIDs, they should be used prudently.

Patients at the highest risk for GI bleeding are most likely to benefit from the COX-l-sparing drugs. Risk factors for GI bleeding include:l,6,7,l7

  • advanced age (older than 65 years).
  • a history of GI bleeding.
  • corticosteroid use.
  • other comorbidities.

Physicians and pharmacists are often reluctant to incorporate “clinical tools” into their practices because they often find them cumbersome and tedious and are concerned that such tools might be used as a substitute for clinical judgment. However, with the significant restrictions placed on the use of COX-l-sparing agents, simple methods that enable practitioners to avoid unnecessary paperwork, and the annoyance of going through the process of preauthorization, are usually welcomed. The first step in developing an algorithm that is widely accepted and easily used is to have a method of estimating risk for individual patients and for the population as a whole.

Both in vitro and in vivo data are necessary to determine whether an agent is COX-l-sparing. Many assays have been developed in an attempt to determine COX-2 selectivity and specificity (i.e., the COX-l-sparing effects) of a variety of new agents. This effort has led to much confusion regarding the selectivity of these drugs. By using various assays, almost any drug might be considered a “COX-2-selective inhibitor,” including drugs that are widely accepted as not being COX-2-selective (e.g., indomethocin).

At present, the most widely accepted test is the human whole blood assay, which uses an ex vivo model; this method specifically mimics in vivo activity. Warner et al. developed an examination that uses the human whole blood assay to determine the selectivity of various NSAIDs and other COX-l-sparing agents. They stated that the use of a so-called effective concentration at 80% (EC-80), rather than an EC-50, was more applicable for these patients (i.e., 80% inhibition of the COX-2 enzyme is necessary to provide the full therapeutic effect and is therefore more clinically relevant). After accessing all available NSAIDs, they determined that four existing agents—cele-coxib, meloxicam, rofecoxib, and etodolac (valdecoxib was not available when this study was performed)—were more selective for the COX-2 enzyme than the COX-l enzyme. In addition, models with these and other medications showing selectivity in therapeutic dosing ranges have been established.
buy antibiotics canada

The structure of the COX-2-binding receptor site and the theoretical mechanisms of how agents selectively bind to this receptor, versus the COX-l receptor site, have been elucidated. Molecular modeling has demonstrated that the COX-2 active site has a side pocket that allows drugs having an appropriate side chain to bind to this region without being able to bind to the COX-l active site.