Background: For medication adherence, pill count is a reliable objective for evaluation. In Japan, at the time of hospitalization, all prescription medicines of patients are checked by medical reconciliation and a pill count is done. This study investigated factors affecting medical adherence by pill count using medical reconciliation.

Methods: This retrospective study included 76 patients (mean age 65.6±12.6 years, 60.5% of men) with type 2 diabetes mellitus who had been taking oral hypoglycemic agents (OHAs) for more than 48 weeks. Patients whose OHAs were managed by another person or those who were taking psychotropic drugs were excluded. To evaluate medication adherence, medication adherence value (MAV) was calculated: (total prescription days-prescription days of OHAs brought when admitted)/(days from start of OHAs to hospitalization). The relationship between MAV and patient characteristics was analyzed.

Results: Subjects were divided into MAV of ≤ 0.86 (those who forgot to take medicine on one or more days per week), > 0.86 ≤ 0.93 (who forgot to take on one or more days in two weeks and less than one day per week), and > 0.93 (who forgot less than one day in two weeks). HbA1c showed a significantly higher value in the group of ≤ 0.86 compared with the group of > 0.93 (P = 0.02). The patients were divided into good (MAV > 0.86) and poor (MAV ≤ 0.86) adherence groups. On multivariate analysis of patient characteristics with P < 0.2 in the univariate analysis as explanatory variables, lower age and no insulin therapy were independent factors for poor adherence (P = 0.004; P = 0.02). Dosing period, Dosing time, number of doses, and types of OHAs were non-independent factors. However, in the poor adherence group, taking 3 doses a day was an independent factor (P = 0.03).

Conclusion: Our findings suggest that poor adherence was independently associated with lower age and no insulin therapy in type 2 diabetes patients. In poor adherence patients, taking 3 doses a day was a factor for reduced adherence.


M. Shiomi: None. Y. Tanaka: None. M. Kurobuchi: None. T. Takada: None. K. Otori: None.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at