500 Abstracts duration was analyzed using Kaplan Meier estimates. Statistical comparison of the survival curves was done with the Log-rank test. We pooled IMS Mediplus data (patient medical records completed by general practitio- ners) from Germany, France and the United Kingdom for analysis. Data was obtained for all irbesartan patients with at least one year of follow-up and from randomly selected samples of patients treated with either diuretics, beta blockers, calcium channel blockers, ACE inhibitors, or other AIIRA. Patients were newly diagnosed with hy- pertension between October 1, 1997 and November 30, 1998 and initiated antihypertensive treatment as mono- therapy. RESULTS: Patients included (2,416) were initiated on di- uretics (422), beta blockers (441), calcium channel block- ers (466), ACE inhibitors (333), losartan (188), irbe- sartan (380), other AIIRA (186). When irbesartan was prescribed, more patients remained on that therapy (60.8%) than with all other antihypertensives (44.2%, p = 0.001), or other AIIRAs (51.3%, p = 0.009). Compared to patients taking all other antihypertensives, fewer pa- tients receiving irbesartan required adjunctive therapy (16.1% vs 25.3%, p = 0.001) or switched therapy (9.0% vs 13.6%, p = 0.013). CONCLUSIONS: Irbesartan was superior compared to all antihypertensives in achieving persistence of initial therapy. By evaluating persistence rather than compli- ance, we captured both the patient’s and the physician’s role in determining the course of treatment. PCV31 A DECISION-ANALYSIS MODEL FOR ENHANCING MEDICATION ADHERENCE IN PATIENTS TAKING STATINS Peterson AM 1 , McGhan WF 2 1 Philadelphia College of Pharmacy, Philadelphia, PA, USA; 2 University of the Sciences, Philadelphia, PA, USA OBJECTIVE: Controlled clinical trials have demon- strated the positive impact of statin therapy on health outcomes in hyperlipidemic patients. The positive impact of adherence is seemingly intuitive and many programs have been designed to improve adherence. However, there are few studies analyzing the cost-effectiveness of medication-adherence interventions. This decision analy- sis model will examine the cost-effectiveness of programs designed to enhance medication adherence in patients taking statins. The model varies the medication-adher- ence rate and program costs to determine the differences in the expected outcome of three different types of inter- vention. METHODS: Data from published clinical and pharma- coeconomic studies were entered into a decision analysis model. A Monte Carlo simulation using 10,000 trials was used with beta distributions for the assumptive vari- ables. The baseline adherence rate was set at 67% and the cardiovascular event rate at 1.5%. Behavioral (B) type interventions were assumed to increase adherence by 8.61%, Educational (E) interventions by 11.22% and Combined (C) B and E by 17.04%. Program costs were estimated as follows: B = $200, E = $100, and C = $225 per patient. Cardiovascular and serious medication- related events were also used as outcomes. The cost-per- patient-per-event avoided was calculated. ANOVA was used to test for differences among intervention types. RESULTS: The results showed that the interventions in- creased the number of events avoided in the first year by 0.04 (95% CI = -0.04,0.12) for B, by 0.06 (95%CI = -0.02, 0.14) for E and by 0.08 (95%CI = 0.02, 0.14) for C. The cost-per-event-per-patient avoided in the first year of treatment was B = $6,038, E = $2,568 and C = $3,839. There was a statistical difference among all inter- vention types with respect to cost of events avoided. CONCLUSIONS: There was no difference in the number of events avoided in the first year of statin treatment with respect to intervention type. While C interventions yielded more events avoided, E interventions appear to be the most cost-effective. More study on the cost-effectiveness of medication adherence programs is required. PCV32 DEVELOPMENT OF TWO INSTRUMENTS: ONE TO MEASURE EDUCATIONAL MATERIAL ACCEPTABILITY AND ONE TO MEASURE KNOWLEDGE GAINED IN PATIENTS WITH HEART FAILURE Gwadry-Sridhar FH 1 , Guyatt GH 2 , Nadeau L 3 , Calhoun M 3 , Arnold JMO 3 1 McMaster University and University of Western Ontario, London, ON, Canada; 2 McMaster University, Hamilton, ON, Canada; 3 London Health Sciences Centre, UWO, London, ON, Canada Patients with heart failure suffer from poor health out- comes due to multiple co-morbidity requiring multiple medications. Patients who misunderstand their diagnosis and treatment plan may use their medication subopti- mally. Providing knowledge through education is one mechanism that can help patients improve medication compliance. OBJECTIVES: To develop and establish measurement properties of two distinct instruments that can be used in patients with heart failure (HF) to measure acceptability of educational materials and subsequent knowledge gained following an educational intervention. METHODS: Using an expert panel, we developed mini- mal criteria required to educate patients with HF. Subse- quently, we developed and tested two questionnaires. The first measure, which we refer to as a measure of edu- cational material acceptability (EMA) was developed to help us differentiate between booklets designed to edu- cate patients with HF so as to select which booklets were most acceptable to the patient. The second measure, which we refer to as the measure of knowledge acquisi- tion (MKA), was developed to help us evaluate the im- pact of our educational intervention on knowledge. MKA