This was a 150-day pre-post study (60-day run-in, 90-day intervention) at Lakeview Internal Medicine in West Des Moines, Iowa. Prescribers participate in pay-for-performance and receive quarterly prescribing reports. Prior to the main intervention, all prescribers (n = 5) attended an educational session presented by a clinical pharmacist reviewing current evidence on the study medication classes. Sample sign-out sheets were posted to demonstrate baseline sample use. All sampled statins, levothyroxine, and SSRIs were removed from the clinic for 90 days (December 1, 2007–February 28, 2008). No restriction was placed on the ability to prescribe these medicines. Data from a third-party payer were used to compare generic prescribing percentage during the sample-free period and a matched 90-day period prior to intervention (July 2, 2007–September 29, 2007). Data were analyzed using ² as a test of proportions. P < .05 was considered statistically significant.

Results
Six studied samples were removed from the clinic (3 statins, 2 levothyroxine products, and 1 SSRI). Generic statin prescribing increased by 16.4% (odds ratio [OR], 1.96; 95% confidence interval [CI], 1.51-2.56 [P < .001]) from baseline (Table). Generic SSRIs increased by 4.7% (OR, 1.47; 95% CI, 1.04-2.09 [P = .03]). Generic levothyroxine increased by 1.2% (OR, 1.05; 95% CI, 0.79-1.38 [P = .76]). Combining all 3 classes, generic prescribing increased by 8.6% (OR, 1.42; 95% CI, 1.2-1.67 [P < .001]). For any medication prescribed, generic prescriptions increased by 4.1% (OR, 1.19; 95% CI, 1.11-1.27 [P < .001]).

View this table: [in this window] [in a new window] [as a PowerPoint slide]   Table. Summary of Generic Prescribing per 90-Day Period

Prescribers elected not to replace samples at the study's conclusion. Thus, data from an additional 90 days were collected to determine sustainability of effect. Generic statin prescriptions increased by 22.8% from baseline (OR, 2.55; 95% CI, 1.95-3.34 [P < .001]), demonstrating significant sustainability. Similarly, generic SSRI prescriptions increased a total of 9.7% from baseline (OR, 1.65; 95% CI, 1.13-2.40 [P = .01]). Generic levothyroxine prescriptions increased by 3.1% from baseline (OR, 1.13; 95% CI, 0.84-1.52 [P = .43]). Combined, the 3 classes increased by 11.9% from baseline (OR, 1.63; 95% CI, 1.37-1.93 [P < .001]). All generic medications prescribed increased by 7.1% (OR, 1.35; 95% CI, 1.26-1.45 [P < .001]).

Comment
To our knowledge, this is the first study that has successfully demonstrated increases in generic prescribing by removing select sample medications from a private practice setting. Previous studies have been conducted in health maintenance organizations or where hospital policies and formularies affect clinic prescribing.^2-3 Such interventions may not reflect the realities of private practice. Counter sampling has proven ineffective in private clinics.^4-5

Our findings have implications for both patients and health care providers. Sample medication use has been shown to increase out-of-pocket costs for patients.⁶ Thus, sample removal and provider education may help mitigate increasing health care expenditures. In addition, clinics participating in pay-for-performance agreements with insurance companies may benefit financially. Overall, generic prescribing increased from 58% to 65.1%. According to one third-party payer ("Incent and Reward Best Practices" [internal document, part of the Pay-for-Performance agreement between Wellmark and the clinic physicians], January 2008), this clinic moved from not qualifying for a generic prescribing award to the highest level award offered in 180 days. Further study of the financial implications of sample removal from this setting should be considered.

AUTHOR INFORMATION
Correspondence: Dr Miesner, Drake University College of Pharmacy & Health Sciences, 2507 University Ave, Harvey-Ingham Hall 126, Des Moines, IA 50311 (andrew.miesner{at}drake.edu).

Author Contributions: Dr Miesner had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Miesner and Koenigsfield. Acquisition of data: Miesner and Allen. Analysis and interpretation of data: Miesner and Wall. Drafting of the manuscript: Miesner. Critical revision of the manuscript for important intellectual content: Allen, Koenigsfield, and Wall. Statistical analysis: Wall. Study supervision: Miesner and Koenigsfield.

Financial Disclosure: Dr Wall has received speaking honoraria from Merck, Sanofi-Aventis, and Ortho-McNeil as well as a grant (unrelated to this study) from TAP Pharmaceuticals. Dr Allen has received pay-for-performance participation awards from Wellmark during this study. Dr Koenigsfeld has received speaking honoraria from Merck, Sanofi-Aventis, Procter & Gamble, and the American Pharmacist Association, as well as a grant from the Community Pharmacy Foundation and TAP Pharmaceuticals.

Previous Presentations: Data in this article have been presented and published as an abstract at the 34th Annual Midwest Pharmacy Residents Conference; May 10, 2008; Omaha, Nebraska; and the 2008 Iowa Pharmacy Association Annual Meeting; June 21, 2008; Des Moines, Iowa.

Additional Contributions: Laura Arensdorf, PharmD, facilitated the data collection in this study.