Abstract Background Does an intervention designed to foster patient-centered communication and shared decision making among GPs and their patients with poorly controlled type 2 diabetes mellitus reduce the level of HbA1c. Methods The DEBATE trial is a cluster-randomized controlled trial conducted in German primary care and including patients with type 2 diabetes mellitus having an HbA1c level of 8.0% (64 mmol/mol) or above at the time of recruitment. Data was measured before intervention (baseline, T0), 6–8 months (T1), 12–14 months (T2), 18–20 months (T3), and 24–26 months (T4) after baseline. Main outcome measure is the level of HbA1c. Results In both, the intervention and the control group the decline of the HbA1c level from T0 to T4 was statistically significant (− 0.67% (95% CI: − 0.80,-0.54%; p < 0.0001) and − 0.64% (95% CI: − 0.78, − 0.51%; p < 0.0001), respectively). However, there was no statistically significant difference between both groups. Conclusions Although the DEBATE trial was not able to confirm effectiveness of the intervention tested compared to care as usual, the results suggest that patients with poorly controlled type 2 diabetes are able to improve their blood glucose levels. This finding may encourage physicians to stay on task to regularly approach this cohort of patients. Trial registration The trial was registered at ISRCTN registry under the reference ISRCTN70713571.