Medication Adherence Determinants

Drug therapies are now the front-line treatment for almost all chronic diseases and many other health conditions.  Compliance with prescribed medications, however, is often spotty.  Failure to comply is estimated (often in a very exaggerated way) to lower quality of care and to increase spending.  Therefore, increasing compliance is a focus of many health organizations.  Understanding why patients often don’t comply is a focus of a new paper at the National Bureau of Economics Research.   (NBER Paper)   The researchers used a comprehensive dataset from Denmark covering seven years to analyze behavior of patients and their physicians that might contribute to adherence or non-adherence.  Some reasons which have been identified for non-adherence include cost, side effects, regimen complexity, confusion and lack of cognitive ability, inadequate provider communication, and lack of social support.  One finding which may be due to the universal coverage available in Denmark, which pays for a large part of drug costs, is that financial considerations appear to account for little non-adherence, a finding which is unlikely to be replicated in the United States.  Drugs for hypertension, diabetes and cholesterol control were studied.  Adherence varies by drug, with beta blockers having the most problems, which is almost certainly due to side effects from that class.  The study also uses the medication possession ratio to determine adherence, which relies on medication availability and does not utilize any direct evidence regarding the taking of the drug.  While adherence rates were generally quite high, there was substantial variation.  Men were less compliant, as were younger people, although these effects are small.  Education-level had little effect but surprisingly, higher income was associated with poorer adherence.  Taking more drugs was associated with better compliance.  Having a longer relationship with the prescribing doctor improves adherence, as does a longer course of therapy with the medication.  There was a clear pattern of certain patients with certain doctors having more compliance issues.  The researchers do an interesting exploration of the possible interactions between patients inclined to be compliant or non-compliant and physicians who are more or less aggressive about emphasizing the need for compliance.  These effects also seem to vary by drug, so that, for example, beta blockers, with their more frequent side effects, need more pushing by doctors to ensure compliance.  More of the non-adherence appears to be accounted for by patient-specific factors than physician-specific ones, however, much of the variation is unexplained by either.  But there clearly is room within the patient-physician relationship to improve adherence by emphasizing its importance.