National estimates and predictors of prescription medication sample use in the United States, 1999-2005

Objectives: To determine the prevalence of free medication sample use in the United States and analyze the effects of socioeconomic status and drug safety actions. Design: Cross-sectional study. Setting: United States from 1999 to 2005. Participants: Survey respondents representative of the civilian noninstitutionalized population. Intervention: Analysis of data from the Medical Expenditure Panel Survey, a nationally representative longitudinal household survey. Main outcome measures: Identification of a medication as being provided as a sample at least once during a study year. Results: An annual average of 5.1% (range 4.4% in 2005 to 5.8% in 2002) of all prescription medications were provided as a sample at least once during a year, with 18.3% of all Americans who received at least one prescription drug receiving at least one drug as a sample. On multivariate analysis, sample use was greater among young (18-30 years) non-Hispanic whites and the uninsured but had minimal independent association with income. The proportion of sample use among users of hormone replacement therapy and cyclooxygenase-2 inhibitors remained relatively constant even as total use of these drugs declined after Food and Drug Administration regulatory action. Conclusion: Use of medication samples is common in the U. S. population. After adjusting for health insurance, sample use was not associated with income and samples were less frequently provided to racial/ethnic minorities and to the elderly. The putative economic benefits of free samples do not appear to go to patients with the greatest financial need. Drug regulatory actions did not have a disproportionate effect on provision of drugs as samples.