Introduction and Objectives Patients with cystic fibrosis are often prescribed complex time-consuming medication regimens. We know from studies in chronic diseases that most patients will be at best only partially adherent with their medication. We studied adherence with nebulised therapy in our adult CF clinic, by measuring the actual quantity of medication prescribed as recorded in General Practitioner and hospital computer prescribing records. The results were compared with the CF centre staff's predictions of adherence on an individual patient basis.
Methods We requested prescription issue data for 26 patients on nebulised dornase α, colistin and tobramycin both from the patient's general practitioners and from the hospital pharmacy. We compared the actual quantity of medication prescribed to each patient to the expected number of doses that they should have received during a 12-month period and calculated a percentage adherence based on these figures. Based on previous work in our difficult asthma clinic we assigned obtaining =80% of doses as being adherent; =50% but <80% as partially adherent and <50% as non-adherent with the medication.
Results Full prescription data were received for 22 of the 26 patients included in the study. Of these 15% (2/13) were adherent; 23% (3/13) were partially adherent and 62% (8/13) were non-adherent with nebulised colistin. 26% (5/19) were adherent; 26% (5/19) were partially adherent and 48% (9/19) were non-adherent with dornase α 100% (8/8) patients were adherent with nebulised tobramycin. The cystic fibrosis team correctly predicted only 40% (111/280) of the respective adherence rates found.
Conclusions The use of prescription data provides useful information regarding adult CF patients' adherence with high-cost nebulised therapy. In our clinic adherence rates were low, and CF clinicians were poor at predicting adherence rates in individual patients. Used in isolation prescription data are a useful indicator of non-adherence, but does not accurately measure positive adherence without consideration of other methods of measurement. We are currently extending this study to compare pharmacy issue data with other methods of measuring adherence in adult CF patients.