People who do not take their medicines as directed by their doctor risk harming their health and also cost the NHS in the UK more than £500 million every year, a new study has reported.

The Aston Medication Adherence Study (AMAS), conducted by Aston Pharmacy School and believed to be the first large-scale research project on medication adherence in the UK, examined the extent of non-adherence and factors associated with lower levels of adherence with three specific patient groups - people diagnosed with dyslipidaemia, type 2 diabetes and or hypothyroidism - living in the Heart of Birmingham.

The researchers found that, overall, around a quarter to a third of the patients were non-adherent to their medication. Profiling of those patients who were most likely to benefit from targeted support to help them take their medicines as prescribed found that these included patients: - who are younger than 60 years of age; - of Islamic faith; - of Asian, Caribbean, African and "Other Black" origin; - whose primary language is Urdu or Bengali; and - who are living in the most socioeconomically deprived areas.

Results from exploratory focus groups revealed an array of issues and barriers faced by people on long-term medication, the researchers note. For example, patients expressed fears about side effects with their treatment and the need for better communication and information about medicines. The presence or absence of symptoms played a role in patients' medication-taking behaviour, and constant changes in generic forms of a medication decreased the levels of trust that patients had in pharmacists, the focus groups also showed.

"What is important about the AMAS is that it identifies adherence patterns within an ethnically-diverse inner-city area with high levels of deprivation - this is currently uncharted territory," said the study's principal investigator, Professor Chris Langley. He also noted that the development of an innovative software tool was pivotal to the study and facilitated analysis of "a vast volume of prescribing data."

The software programme calculated individual patient Medication Possession Ratios (MPRs) for all medication "runs" of interest - a "run" being the time between the first and last prescriptions for a medicine. Although MPRs are not a definitive indicator of adherence to a medication regimen, as the data does not indicate whether the prescribed medication was dispensed and/or subsequently taken by the patient, a low MPR value is a reliable indicator of non-adherence, the researchers point out.

"We identified numerous groups in which adherence levels were lower than in the general population, but it is not possible at this time to identify why adherence is lower in these groups," said AMAS investigator Dr Joe Bush. "While the focus groups suggested possible reasons for non-adherence, we hope to explore these issues further and identify the primary reasons for non-adherence in these patient groups in future research," he added.

- For the study, more than one million anonymised individual prescription issues were analysed, and just over 7,000 patient questionnaires were administered, in two mailings. Seven focus groups were conducted and a full review of the literature regarding adherence was undertaken.