Factors influencing diabetes treatment satisfaction in the INtegrating DEPrEssioN and Diabetes treatmENT randomized clinical trial: A multilevel model analysis

Aims: Patient satisfaction is associated with positive diabetes outcomes. However, there are no identified studies that evaluate both patient- and clinic-level predictors influencing diabetes care satisfaction longitudinally.

Methods: Data from the INtegrating DEPrEssioN and Diabetes treatmENT trial was used to perform the analysis. We used fixed and random effects models to assess whether and how changes in patient-level predictors (treatment assignment, depression symptom severity, systolic blood pressure, body mass index, LDL cholesterol, and haemoglobin A1C) from 0 to 24 months and clinic-level predictors (visit frequency, visit cost, number of specialists, wait time, time spent with healthcare provider, and receiving verbal reminders) measured at 24 months influence diabetes care satisfaction from 0 to 24 months.

Results: Model 1 (patient-level predictors) accounted for 7% of the change in diabetes satisfaction and there was a significant negative relationship between change in depressive symptoms and care satisfaction (β = -0.23, SE = 0.12, p < 0.05). Within Model 1, 2% of the variance was explained by clinic-level predictors. Model 2 included both patient- and clinic-level predictors and accounted for 18% of the change in diabetes care satisfaction. Within Model 2, 9% of the variance was attributed to clinic-level predictors. There was also a cross-level interaction where the change in depression had less of an impact on the change in satisfaction for those who received a verbal reminder (β = -0.11, SE = 0.21, p = 0.34) compared with those who did not receive a reminder (β = -0.62, SE = 0.08, p < 0.01).

Conclusions: Increased burden of depressive symptoms influences diabetes care satisfaction. Clinic-level predictors also significantly influence diabetes care satisfaction and can reduce dissatisfaction in primary care, specifically, reminder calls from clinic staff.