Home

Download PDF version of web site
Increase Font Size Decrease Font Size
STATISTICS RELATED TO ADHERENCE

So what happens after a prescription is written? Here are some startling insights into the depth of the medication adherence problem in the United States. Consider these statistics (American Heart Association):

  • 12% of Americans don't fill their prescription at all
  • 12% of Americans don't take medication at all after they fill the prescription
  • Almost 29% of Americans stop taking their medication before it runs out
  • 22% of Americans take less of the medication than is prescribed on the label

Persistence rates, especially among those with newly diagnosed disease, also decrease over time, and in persons with newly diagnosed high blood pressure have been reported to be as low as 78% after 12 months and 46% after 54 months (Caro et al., 1999).

TABLE 1.
MAIN REASON FOR NOT FILLING PRESCRIPTIONS, AMERICANS AGE 50 AND OLDER
Cost of the drug
40%
Side effects of drug
11%
Thought drug wouldn’t help much
11%
Didn’t think I needed it
8%
Drug did not help
6%
Don’t like taking prescription drugs
5%
Condition improved
4%
Already taking too many prescriptions
3%
Source: AARP, 2004

Surveys of older adults indicate that 55% do not follow, in some way, their medication regimens (Amaral, 1986). In an AARP survey of Americans aged 50 and older (AARP, 2004), 25% said they did not fill a prescription written by their doctor in the past two years; cost was cited as the main deterrent (Table 1). A 2002 study of 325 older persons (average age of 78 years) reported that 39% were unable to read the prescription labels, 67% did not fully understand the information given to them, and as a result 45% were nonadherent. These problems were especially prevalent in men and in people older than 85 years (Moisan et al., 2002). It is likely that the incidence of medication nonadherence is actually higher than published reports show, as methodological difficulties associated with conducting medication adherence studies may lead to an underestimation of the extent of the problem (Haynes et al., 1979).