Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Noncompliance with antihypertensives


Bandolier is interested in whether patients actually take the medicines they are prescribed. This used to be called compliance, though concordance is now the favoured term. Anecdotally, not taking medicines is thought to be common, and one experienced pharmacist recently told Bandolier that in the UK up to 40% of medicines prescribed are not actually taken when the prescriptions are filled.

There are several steps, though:

  1. Having a prescription
  2. Having the prescription filled
  3. Taking the medicine
  4. Reporting how often the medicine prescribed is actually taken

A new US study from Harvard [1] suggests that, for antihypertensive medicines, there is little concordance between step 1 and steps 2 and 4.


Patients with diagnosed hypertension treated in a health maintenance or veterans organisation in New England formed the population. Medicines were either free, or had a minimal co-payment to be made by the patient.

Computer systems captured diagnoses, prescriptions, physician and hospital visits. Records for patients with a diagnosis of hypertension during 1996 and who were in the system for a full year after the first antihypertensive medication prescription were collected, and a random sample asked to participate. Of the 500 of about 1,000 patients who were willing to participate, 200 patients taking a single antihypertensive agent were selected.

A telephone survey instrument was developed after literature review and discussion with physicians, and tested for feasibility. It assessed self-reported compliance and other factors. Data on prescriptions filled for days of supply over the 365 days following the first prescription was collected.


The 200 patients were mostly (80%) over 55 years, about 60% were men, and had been hypertensive for an average of 12 years. Almost all of them reported a high level of compliance, with only five of the 200 reporting using their medicines on fewer than 300 of 365 days.

In fact, over half of them filled prescriptions for fewer than 300 days (Figure 1), with 11% collecting less than half the medicine they should have taken, and six (3%) filling no prescriptions at all while claiming a very high degree of compliance.

Figure 1: Actual filled prescriptions over one year (percentage of total patients)

Those who filled prescriptions for more than 80% of days were no different from those filling prescriptions for fewer than 80% of days, in terms of age, sex, education, or any other of a range of variables, though the number of provider contacts in the previous six months may have been related.


No surprise there, then. But there is a quantification of unfilled prescriptions, and excellent evidence that, at least for hypertension, that self-report of compliance by patients is likely to be unhelpful. In other areas, like anticoagulation with warfarin where INR can be used as a check, things may be different. What this study does not address is prescriptions filled but medicines not taken.

Compliance (concordance) is an important issue to which Bandolier would like to return. Readers who are aware of good studies can help by reporting them, as searching is not straightforward.

Is this at all important? It may have major economic implications. Bandolier hears of the amount of returned medicines being weighed in tons in just one area. Yet we have constraints on effective medicines. Waste in healthcare should not be allowed.


  1. PS Wang et al. How well do patients report noncompliance with antihypertensive medications?: a comparison of self-report versus filled prescriptions. Pharmacoepidemiology and Drug Safety. 2003. Published Online: 28 Feb 2003 DOI 10.1002/pds.819

previous or next story