275 Abstracts OBJECTIVES: To identify hypertension prescription patterns in a Los Angeles County Latino population, evaluate prevalence of use of foreign antihypertensive drugs, and its association with blood pressure control. METHODS: Data was obtained from the Los Angeles Latino Eye Study (LALES), a study designed to assess the prevalence of eye disease and self-reported visual func- tion in Latinos age 40 and older. In total, 1076 patients with self- reported history of hypertension, clinic blood pressure values and available medication profile of drugs taken within the past month were identified for analysis. Antihypertensive drugs were identified and assessed for origin as US or foreign (Mexican, Central or South American), drug class, and total number of antihypertensives used. Chi-square tests were used to test for dif- ferences between US and foreign drug groups in terms of uti- lization of different drug classes, demographic differences, and blood pressure control. RESULTS: Only 50.5% (533/1076) of subjects with history of hypertension were on at least one anti- hypertensive medication. 34.9% of subjects had taken two or more antihypertensives in the past month. A total of 30.2% (161/533) had taken a beta-blocker within the previous month, 32.8% (175/533) an ACE-inhibitor, and 25.0% (133/533) a calcium channel blocker. The prevalence of foreign and US sup- plied drug use were 1.5% (16/1076) and 48.0% (517/1076) respectively. The difference in blood pressure control between the US and foreign drug use groups was not statistically significant (p = 0.1160). CONCLUSIONS: Although foreign medications may be easily accessible, there does not appear to be significant use of foreign antihypertensive medications within this Los Angeles Latino population. The relationship between foreign drug use and blood pressure control was not found to be statis- tically significant but the result may be due to the limited number of foreign drug use observations. PCV47 USE OF ANTIHYPERTENSIVE MEDICATIONS AMONG VETERANS NEWLY DIAGNOSED WITH HYPERTENSION Vincze G 1 , Barner JC 2 1 Novartis Pharma AG, Basel, Switzerland; 2 University of Texas, Austin, TX, USA OBJECTIVES: To assess type, duration, and medication avail- ability of antihypertensive therapy among veterans. METHODS: Electronic medical records from the Central Texas Veterans Health Care System were extracted for adults newly diagnosed with hypertension between 1995 and 2003. Information on demographic (age, gender, and race/ethnicity) and clinical (diag- noses, height, weight, labs, and blood pressure values) charac- teristics were collected. Drug utilization was measured as a time-dependent variable; thus, the use of combination therapies was captured for any given day for each patient in the sample. Medication availability was calculated as the ratio of “total days supply” and “number of days between first and last refill.” RESULTS: The average age of the participants (N = 11,187) was 60.7 (SD: 12.7). Half (50.1%) of the patients could be catego- rized as having controlled blood pressure. Veterans were fol- lowed for an average of 3.6 years (total of 51,549 person-years). Overall, 29,561 treatment episodes were identified; an average of about 2.6 per patient. Over forty percent (41.6%) of these episodes involved one medication only, but patients frequently used dual (26.9%) and three or more (15.9%) combination ther- apies. Patients took “single therapy alone” for a median of 1.3 years, “dual therapies alone” for a median of 1.0 years, whereas “three or more therapies” for a median of 0.9 years. Medication availability ratios were relatively high, ranging from 0.88 to 0.93. Overall, calcium channel blockers (22.4%), angiotensin converting enzyme inhibitors (21.4%), and beta-blockers (19.9%) were used most frequently. CONCLUSIONS: Our results indicate high medication availability ratios; i.e., veterans filled their antihypertensive medications as it was prescribed by their physicians. Nevertheless, the average blood pressure control rate was only 50 percent, which requires improvements in patients’ antihypertensive therapy. The necessity of better selection of antihypertensive therapies is also supported by the finding of frequent therapy changes. PCV48 ASSESSMENT OF THE VALIDITY OF TWO CLAIMS-BASED ALGORITHMS IN IDENTIFYING HYPERTENSIVE PATIENTS Bullano MF 1 , Kamat SA 1 , Barlas S 2 , Willey VJ 1 , Makin C 1 ,Watson DJ 2 , Brenneman SK 3 1 HealthCore, Inc, Wilmington, DE, USA; 2 Merck Research Laboratories, Blue Bell, PA, USA; 3 Merck & Co., Inc,West Point, PA, USA OBJECTIVES: Despite inherent limitations, administrative claims databases have been widely used to assess quality of care and evaluate medical care appropriateness in several therapeutic areas including hypertension (HTN). The objective of this study was to assess the validity of two commonly used claims-based algorithms in identifying the presence of HTN, using patient medical records (MR) as the gold standard. METHODS: Managed care claims data spanning January 1, 2000 through March 31, 2003 were used. HTN patients (test-positives) were identified using two algorithms—algorithm A used medical claims only and algorithm B used both pharmacy and medical claims. Test-positives and test-negatives (health plan members without HTN medical claims in algorithm A and without a HTN medical AND pharmacy claim in algorithm B) were 1 : 1 matched on their propensity for having a HTN claim. Propensity scores were computed using a logistic regression model including demo- graphic, clinical and physician characteristics to allow for mul- tivariate matching. MR were reviewed for the presence of HTN, and the validity of claims-based criterion was evaluated by com- puting its sensitivity and specificity, and comparing with MR data. RESULTS: A total of 258 and 138 matched pairs of test- positives and test-negatives for algorithms A and B, respectively, were identified. Baseline characteristics were similar for test-pos- itives and test-negatives. The sensitivity (95% CI) and specificity (95% CI) of algorithm A were calculated as 71% (65.1–76.1) and 75% (68.8–80.3) respectively, and those of algorithm B were 76% (69.1–82.3) and 93.3 (86.6–97.3) respectively. The kappa score for assessing strength of agreement between claims data and MR was 0.45 for algorithm A and 0.65 for algorithm B. CONCLUSION: The results demonstrate the validity of using claims based algorithms in identifying hypertensive patients. However, the higher sensitivity and specificity of algorithm B demonstrate that using a combination of medical and pharmacy claims will help reduce the risk of patient misclassification, when compared with using medical claims only. PCV49 ADVANTAGES AND LIMITATIONS OF USING A WEB-BASED SURVEY TO OBTAIN PATIENT-REPORTED OUTCOMES Chen KS 1 , Plauschinat CA 2 , Frech F 2 , Dubois RW 1 , Doan Q 1 1 Cerner Health Insights, Beverly Hills, CA, USA; 2 Novartis Pharmaceutical Corporation, East Hanover, NJ, USA OBJECTIVE: To assess the advantages and limitations of using a web-based survey to obtain patient-reported outcomes (PRO). METHODS: A web-based patient survey was conducted in a population diagnosed with hypertension to assess the correlation between blood pressure control and patients’ satisfaction with antihypertensive treatment. Patients were selected from the