Over and under-utilization of cyclooxygenase-2 selective inhibitors by primary care physicians and specialists: the tortoise and the hare revisited.

OBJECTIVES: To compare prescribing trends and appropriateness of use of traditional and cyclooxygenase-2 selective (COX-2) nonsteroidal anti-inflammatory drugs (NSAIDs) by primary care physicians (PCPs) and specialists. DESIGN: Retrospective cohort study. PATIENTS: One thousand five hundred and seventy-six adult patients continuously enrolled for at least 1 year with an independent practice association of a University-associated managed care plan who were started on a traditional NSAID or a COX-2 inhibitor from 1999 to 2002 and received at least 3 separate medication fills. MEASUREMENTS: Physician specialty was identified from office visits. Appropriateness of utilization was based on gastrointestinal risk characteristics. RESULTS: Primary care patients were younger and less likely to have comorbid conditions. Despite similar GI risk, COX-2 use among patients seen by PCPs was half that of patients seen by specialists (21% vs 44%, P<.001). While PCPs overused cyclooxygenase-2-specific inhibitors (COX-2s) less often than specialists (19% vs 41%, P<.001), they also tended to underuse COX-2s in patients who were at increased GI risk (46% vs 32%, P=.063). This represents a 3-fold and 8-fold difference in overuse versus underuse for PCPs and specialists, respectively. CONCLUSIONS: Using COX-2s as a model for physician adoption of new therapeutic agents, specialists were more likely to use these new medications for patients likely to benefit but were also significantly more likely to use them for patients without a clear indication. This study demonstrates the tension between appropriate adoption of innovative therapies for those individuals who would benefit from their use and those individuals who would receive no added clinical benefit but would incur added cost and be placed at increased risk.

Self-reported adherence in patients with asthma.

BACKGROUND: Not all patients with asthma adhere to their prescribed drug regimens. The Behavioral Model of Health Services Utilization, organized as predisposing, enabling, and need variables, may be useful in identifying variables related to adherent medication-taking behavior. OBJECTIVE: To examine the relationship between predisposing, enabling, and need variables and self-reported adherence with asthma controller drugs. METHODS: A secondary analysis was conducted from a cross-sectional study using survey and claims data obtained from adults with asthma enrolled in a managed care organization. Independent variables included predisposing (age, gender, race, years with asthma, number of comorbidities, health beliefs), enabling (income, number of metered-dose inhaler [MDI] instructors, inhaler technique, perceived physician access), and need (patient-perceived severity, symptom-derived severity, health-related quality of life). Adherence was measured using a 4 item questionnaire scored as the mean of the responses, with 5 denoting highest adherence. Multivariate regression analysis was used to identify variables with statistically significant relationships to self-reported adherence. Stepwise backward elimination was used, with the final model consisting of variables considered significant at p less than 0.05. RESULTS: The 573 respondents (1270 packets sent; 45% response rate) were primarily white (89.5%) and female (71.0%), with an average age of 40.5 +/- 12.4 years (mean +/- SD) and average asthma duration of 18.3 +/- 14.2 years. The mean adherence scale score was 3.7 +/- 1.1, with 84.6% indicating some level of nonadherence (score <5). The final model had an adjusted R(2) of 0.26 and included 6 independent variables. Better adherence was associated with stronger beliefs in the benefits of treatment and trigger avoidance, greater perceived asthma severity, longer asthma duration, more MDI instructors, and higher scores on the Short-Form 36 mental component summary. CONCLUSIONS: Complex beliefs, perceptions, and experiences constitute the variables associated with adherent medication-taking behavior. Future longitudinal studies should include these variables to determine the predictive strength of the model.