Therapeutic alliance and glycaemic control in type 1 diabetes: a pilot study.

Aim: Effective diabetes care requires integrating physicians' clinical expertise with patients' concerns and resources. This prospective study examined whether or not two measures of therapeutic alliance could predict glycaemic control after 1 year of follow-up in patients with type 1 diabetes.

Methods: Consecutive type 1 diabetic outpatients were recruited, and their age, gender, level of education, marital status and age at the time of diabetes diagnosis were self-reported. The presence of diabetes complications was ascertained by the patients' physicians. Both patients and physicians completed the revised Helping Alliance Questionnaire (HAQ-R) and the 12-item Working Alliance Inventory (WAI-12) to assess therapeutic alliance. Patients also completed the Center for Epidemiological Studies Depression scale to assess depressive mood. HbA(1c) was measured at baseline and 1 year later.

Results: Sixty-four type 1 diabetic outpatients (32 men, 32 women; mean age±standard deviation [S.D.]: 38.2±8.0 years) were included. HbA(1c) level at follow-up (mean±S.D.: 7.56±1.18%) was positively correlated with the HbA(1c) level at baseline (r=0.698, P<0.001), and associated with presence of retinopathy at baseline (8.18±1.24% versus 7.41±1.13%, P=0.036). In addition, the HbA(1c) level at follow-up was negatively correlated with therapeutic alliance, as assessed at baseline by the physicians using either the HAQ-R (r=-0.431, P<0.001) or the WAI-12 (r=-0.365, P=0.003), even after controlling for the HbA(1c) at baseline.

Conclusion: Although the observational nature of the present study prevents causal conclusions to be drawn, these preliminary results suggest that promoting therapeutic alliance can improve glycaemic control in type 1 diabetes.