Introduction
Adherence to drug regimens is a challenge in patients with type 2 diabetes1 and might jeopardise therapeutic outcomes. We studied the potential impact of medication regimen complexity and demographic factors on achievement of therapeutic goals (glycaemic control, blood pressure, and lipid levels) in patients with type 2 diabetes.

Methods
A pre-piloted postal questionnaire was sent to 407 type 2 diabetics receiving oral hypoglycaemics identified from repeat prescribing records of three medical practices in Aberdeen. Patients were asked details of current medicines, most recent values of blood pressure (BP), total cholesterol (TC), glycosylated haemoglobin (HbA1c) and demographics. Access to medical notes was requested to permit data validation. Complexity of the patient stated medication regimens was assessed using the Medication Regimen Complexity Index (MRCI).2 The MRCI comprises three sections to account for dosage forms (section A), dosing frequencies (section B) and additional directions (section C). Each section is scored, with scores weighted depending on the complexity; the complexity index is the sum of the three scores. HbA1c, TC and BP were classified into 'satisfactory' and 'less satisfactory' based on recommended targets.3 Data were analysed using SPSS (version 13.0). Factors associated with poor outcomes were identified using Student t-test/Mann-Whitney U statistic.

Results
Responses were received from 194 (47.7%) patients, 11 of whom were excluded (not meeting inclusion criteria/incomplete questionnaires). The majority of respondents were male (n=99, 54.1%) and were over 60 years (n=121, 66.1%). Patients were using a median of 5 prescribed medicines (range1-15) requiring a median of eight tablets/ capsules to be taken daily (range 1-27). Upon scoring each regimen using the MRCI, mean scores (± SD) for various sections of the MRCI were: section A (dosage forms) 1.5 ± 1.25; section B (dosing frequency) 7.8 ± 3.86; section C (additional instructions) 6.1 ± 2.91; and total MRCI score 15.4 ± 6.58. The lower scores achieved in section A of the MRCI were due to most patients taking oral solid dosage forms; higher scores in sections B and C were due to frequent dosing and medicines requiring additional directions. The number of respondents with less satisfactory HbA1c, TC and BP were 98, 107, and 86, respectively. 'Less satisfactory' HbA1c values were associated with: higher MRCI total scores (p=0.001), section B (p=0.001) and section C (p=0.03); greater number of total medications (p=0.001); and age <60 years (0.004). Age <60 years was the only predictor of 'less satisfactory' TC. Fewer number of total medications (p=0.02) and age <70 years (p=0.002) were found to be associated with 'less satisfactory' BP.

Discussion
This study has highlighted the complex medication regimens and the high variability of the MRCI scores in type 2 diabetic patients. Although we have identified an association between complexity and disease control, it may be the disease control which influences the medication regimen and thus its complexity; these issues merit further investigation in larger numbers of patients. Simplifying the prescribed medication regimen by reducing dosing frequency will reduce the MRCI score and might enhance outcomes. Further evaluation of the MRCI is warranted.

Acknowledgements: We acknowledge the input of A Davie, L Juroszek and A Taylor