PFIZER's FELDENE v

ASPIRIN COST COMPARISON STUDY concludes that the nonsteroidal anti-inflammatory agent has a 31.1% lower total cost than aspirin because of the decreased incidence of gastrointestinal side effects requiring treatment. According to a Pfizer-sponsored retrospective study of 199 patients, using New Jersey Medicaid claims data, investigators Bernard Bloom, et al, (from the University of Pennsylvania and Pracon, Inc.) concluded that "the difference in expected G-I related costs more than offsets the price difference between these two anti-arthritis medications." The study consisted of 142 piroxicam treatment periods (114 patients) matched with 142 aspirin treatment periods (85 patients). Drug treatment periods were defined as three months in duration. According to the investigators, the cost of drug therapy for each piroxicam patient averaged $67.38 while the cost for each aspirin patient averaged $35.80. On top of that cost, "mean expenditures related to G-I adverse reactions among piroxicam-treated patients were $45.37, compared with $213.43 among aspirin-treated patients," the study found. Bloom, et al, noted that G-I therapy was received by 30.3% of the aspirin patients and 20.4% of the piroxicam patients. Bloom, et al, calculated that the expected cost per person of treating G-I adverse events in aspirin patients was $64.64, while the expected cost per person in piroxicam patients was $9.26. Thus, the base cost of Feldene ($67.38) plus treatment for G-I toxicity ($9.26) totaled $76.64 per person. Similarly, the base cost for aspirin ($35.80) plus treatment for G-I toxicity ($64.64) totaled $100.44 per person. The price difference thus favors aspirin by 31.1%. The study was presented at a Pfizer-sponsored symposium on cost containment & arthritis held in San Francisco April 4-6. Those attending the symposium included state formulary cmte. members and hospital administrators. Pfizer said the study will be published in Hospital Therapy in the fall. Last year, Pfizer sponsored a cost comparison study in support of its cephalosporin antibiotic Cefobid. One of the factors influencing results of the study was the hospitalization of three aspirin patients. "Mean G-I related drug expenditures for aspirin patients were more than twice that of piroxicam patients ($59.32 versus $25.59, p < 0.01)," the investigators stated. "The large difference in total G-I costs between the two groups ($213.32 for aspirin patients versus $45.37 for piroxicam patients) is largely attributable to the three hospitalized aspirin patients." There were no G-I-related hospitalizations among piroxicam patients. Bloom, et al, said the difference was tested for statistical significance by use of a Wilcoxon Rank-Sum test which found that hospitalization "had no undue effect in the determination of statistical significance." Another factor influencing study results was inclusion criteria. Only treatment periods in which the patient received at least a second prescription were retained for analysis. Therefore, Bloom, et al, said that by definition patients were excluded "who suffered an adverse reaction and for whom the original therapy was discontinued prior to a refill." The investigators commented that the results of the study are also "influenced by the special characteristics of Medicaid reimbursement policies." They explained that "the Medicaid data base used in this analysis provides conservative estimates of the total costs of care since providers are generally reimbursed at a much lower rate than by other third party payers." They added that "it is also likely, particularly in other insurance systems, that some treatment related costs will be born by the patient; hence, the cost estimates presented here may be significantly lower than those for non-Medicaid patients."