OBJECTIVE: To describe the process of creating a cohort from refill data from a pharmaceutical assistance program and evaluate the feasibility of using such data in the construction of measures that assess patients' use and adherence to this program. METHODS: Retrospective cohort of patients admitted to a pharmaceutical care program in Alzheimer's disease between 2002 and 2005 in the state of Rio de Janeiro. The refill data records were taken from the Authorization of High Complexity Procedures of the Outpatient Information System of the Unified Health System. The probabilistic record linkage methodology was used to identify the multiple entries for each individual and the statistics software Statistical Package for the Social Sciences (SPSS) to build the cohort RESULTS: After completion of all processing steps, a database was obtained where the unit of analysis was the individual, containing demographics, dispensing unit, the multiple supply with their respective dates of submission, primary and secondary ICD codes, the performed procedures (in this case medications and dose per unit) and the number of units dispensed. CONCLUSIONS: The formed measures allow the assessment of medicines utilization and exposure using secondary data as unique information source, plus open new possibilities in the pharmacoepidemiology and healthcare program evaluation field.
medication adherence; pharmacoepidemiology; program evaluation; health information systems; medical record linkage