FACTORS AFFECTING COMPLIANCE WITH MEDICATION REGIMES
With regard to medication-related factors, the review by the APhA (2003) generally found that decreases in compliance have been associated with increases in:
Dunbar-Jacob et al. (2003) found that medication adherence (measured as the percentage of prescribed doses taken, the percentage of days with correct dosing, and the percentage of expected doses with correct timing) in a sample of 169 community-dwelling people 62 and older with cardiovascular conditions significantly declined as the dosing frequency increased. The average adherence (percentage of prescribed doses taken) was 95 percent for once dosing daily, 84 percent for twice dosing daily, 84 percent for three doses daily, and 78 percent for four doses daily. As the number of doses per day increased, adherence to correct timing decreased. Subjects who were on a once-per-day regimen had an average compliance rate of 83 percent compared to 53 percent for those on a twice-per-day regimen and 27 percent for those on a four-times-per-day regimen.
Compliance rates for Type 2 diabetes patients taking antihyperglycemic drugs did not reach 50 percent in a study by Dailey et al. (2001), even with a simple monotherapy dosing regime; however, compliance with a monotherapy regime was 36 percent higher than compliance with a polytherapy regime. Daily et al. (2001) used pharmaceutical claims data from Medi-Cal to examine antihyperglycemic drug use patterns among Medicaid recipients with Type 2 diabetes and to compare compliance and persistence for simple one-drug antihyperglycemic regimes to more complex multiple-drug regimens. There were 37,431 patients in the 1-year follow-up period (mean age = 61.3, range 30 to 107.9 years) and 16,452 patients (mean age = 70, range 30 to 107.9) in the 2-year follow-up period. The treatments consisted of either metformin alone or sulfonylurea alone (monotherapy) or a combination of metformin plus sulfonylurea (polytherapy). Results showed that patients on the monotherapy regime had approximately 65 percent more days of continuous (or persistent) treatment (129 days) per patient per year than did patients taking polytherapy (78 days). Monotherapy subjects had a 36-percent higher compliance rate than polytherapy subjects (177 days versus 130 days.) At the 1-year follow up period, patients on monotherapy (either metformin or sulfonylurea alone) had identical compliance rates of 49.3 percent. This compares to a compliance rate of 36 percent for patients on a polytherapy regime (metformin plus sulfonylurea). At the 2-year follow-up, compliance dropped to 42.4 percent for patients on metformin alone, 42.1 percent for patients on sulfonylurea alone, and 29 percent for patients on both metformin and sulfonylurea.
Balkrishnan et al. (2003) reported lower compliance in patients 65 and older with Type 2 diabetes who were taking injectable antidiabetic medications as compared to older Type 2 diabetic patients taking oral antidiabetic medications. Oral antidiabetic medication use was associated with a 0.28-point increase in medication possession ratio (MPR, based on pharmacy data, calculated as the number of days of antidiabetic prescription supply dispensed divided by the number of days between prescription refills, ranging from 0-1), which translates to a 62-percent increase over the MPRs of patients not using oral antidiabetics. Other factors associated with lower adherence included a self-reported emergency room visit in the year before enrollment in the Medicare HMO (a 0.43-point decrease in the MPR, or a 6% decrease versus the mean) and an increase in the severity of comorbidity (the Charleston index).
Even with a simple dosing regime (e.g., once-daily dosing), a review of studies on hypertensive patients found that on average, only 75 percent of drugs prescribed for the hypertensive condition were taken (Cramer, 2004). Cramer describes the compliance scenario as an abrupt discontinuation of treatment, lasting several days or more, followed by abrupt resumption. She reports on the consequences of erratic compliance with antihypertensive therapy by citing Psaty et al. (1990), who found that patients who took less than 80 percent of their antihypertensive medication were at a four-fold elevated risk for an event than good compliers. As an example, when a hypertensive patient falls and breaks a hip, the cause is more likely due to missed doses that lead to very high blood pressure and lightheadedness than to an effect of the drug.
Murray et al. (2004) indicate that congestive heart failure (CHF) is an exemplary disease for illustrating difficulties associated with adherence. The incidence of CHF increases with aging, and in older adults, it is most often the result of a long history of poorly maintained hypertension, which requires multiple medications. With the onset of CHF, patients may be required to take 5 or more cardiovascular medications, in addition to any medications needed for other chronic diseases. With the greater number of medications taken by older people, and the increased likelihood that many of these individuals will have cognitive impairment, Murray et al. (2004) state that older adults with CHF would appear to be at risk for problems with medication adherence.
Although the association between higher numbers of concurrent medications and poorer adherence is the prevalent finding in the literature, the counterintuitive finding that higher numbers of medications predicted better adherence was found in a study using subjects taking cardiovascular medicines (Shalansky and Levy, 2002) and in a study using subjects taking rheumatoid arthritis medications (Treharne, Lyons, and Kitas, 2004). Subjects in both studies were patients recruited from outpatient hospital clinics (heart failure clinic, lipid clinic, and a rheumatology clinic). Attending an outpatient clinic in and of itself may be associated with greater adherence. Also, in the cardiovascular clinic study, all subjects demonstrated persistence to therapy before being surveyed (e.g., subjects who were not taking cardiovascular medications for at least 3 months were excluded). In addition, subjects’ perceptions of their disease severity and their need to take medicines has been shown to be an important predictor of adherence, and subjects who believe they are more ill (e.g., through being prescribed a large number of medications) may be more likely to take steps necessary to maintain their health. Taking higher numbers of medications requires more attention to the medication-taking regime, and in the cardiac study, use of compliance aids was higher in adherent subjects. In the RA study, patients taking more medicines perceived that their medications were more necessary. These patient-related factors (described below) may explain the counterintuitive association between higher numbers of medications and greater compliance.
In a study evaluating compliance with lipid-lowering medication in 510 patients with a mean age of 64, Kaplan et al. (2004) found that medication side effects reported as “frequent and very unpleasant” were associated with noncompliance (in analyses adjusted for age, sex, and race/ethnicity).