Optimal recall period in assessing the adherence to antihypertensive therapy: a pilot study.

OBJECTIVE: To evaluate the validity of patient self-reported adherence, and to find the optimal length of recall period which best reflects the long-term adherence pattern of the patient.

SETTING: Patients were recruited from a general practitioner's practice in a Hungarian town.

METHOD: In this prospective study 30 patients, who had already been on antihypertensive treatment, were involved. The study was designed to monitor one antihypertensive medication per patient for 3 months. Patients received a 3-month supply of one antihypertensive medication in an electronic Medication Event Monitoring System (MEMS). At the end of the study period patients completed a structured questionnaire regarding their medication taking behavior during the last 7, 14 and 30 days. The results measured with MEMS were considered as the reference value, and other measures were compared using the Bland-Altman method.

MAIN OUTCOME MEASURES: Self-reported adherence, length of recall period, taking adherence and timing adherence measured by MEMS.

RESULTS: Of the 30 patients included, 29 patients (13 males and 16 females) completed the study. The mean age of the patients was 60.6 years, ranging between 36 and 86 years. Patients were monitored for an average of 89 days (ranging between 49 and 106 days). Fifteen patients were on once daily, 9 patients were on twice daily, and 5 patients were on 3 times daily dosing schedule. The total expected number of medication taking events was 4,281. The MEMS caps recorded a total of 4,071 openings, which showed only a 3.56% deviation from the pill counts of the remaining tablets. The overall taking adherence was 95.1%, timing adherence was 75.2%. Patients' adherence report using a visual analog scale and reporting the number of missed doses became more accurate as the length of the recall period increased. Increased number of chronically taken medications was associated with better adherence. Increased dosing frequency of the observed antihypertensive medication resulted in decreased adherence.

CONCLUSION: The results showed that the length of the recall period influences the accuracy of self-reported adherence. Patients seem to be able to report more precisely their medication taking behavior regarding a 30 day period than a 7 day period.