An evaluation of community pharmacy records in the development of pharmaceutical care in The Netherlands.

An electronic patient database linking prescribes with a Dutch community pharmacy consortium was evaluated in a subset of the population of Noordwijk (11,760 patients out of a total of 25,600). The pharmacy database (a file of 41 disease contra‐indications representing a subset of the prescribers' medical diagnosis) was studied in order to assess its value for the accurate discrimination of target patient groups within the community and for support to the pharmaceutical care of individual patients. The aim was to examine the application of the pharmacy records to pharmaceutical care, tested by measuring the accuracy of the pharmacy database to predict the community public health profile; and, more specifically, the accuracy to identify three potential target groups for pharmaceutical care. The records of patients with angina, chronic respiratory disease and diabetes (n=1116), representing 65% of the total pharmacy morbidity records, were studied in detail and verified by the files and texts of the prescriber's individual patient records. From samples of patients (n=273) from the three patient groups, the extent and nature of co‐morbidity, polypharmacy (drug entities prescribed annually) and drug therapy instability (prescription changes to dose or dose form annually) were characterised. Angina patients showed the most co‐morbidity, 46% having three or more additional diseases; chronic respiratory disease patients showed most drug therapy instability; and insulin‐dependent diabetic patients received most polypharmacy per disease. The pharmacy database predicted the prevalence of 10 of 23 relevant disease categories (representing 51% of the total morbidity on the medical records). However, the prevalences of eight categories were underestimated and of five categories overestimated. Of the three patient groups, 73% of patients appeared on both the pharmacy and the medical database. Of the total co‐morbidity recorded for these patients, 68% of records were common to both databases. The database discrepancies (32%) were due to morbidity omitted (12%) and morbidity unverified (10%) on the pharmacy database, together with morbidity omitted from the medical database (10%). The current pharmacy database provides a limited view of morbidity. A strategic approach to pharmaceutical care requires pharmacists and prescribers to verify and share patient information if patient groups and individuals within a group are to be usefully targeted.