Efficacy of a training programme for the management of diabetes mellitus in the hospital: A randomized study (stage 2 of GOVEPAZ healthcare).

AIMS: To assess the efficacy of a structured educational intervention for health professionals on the appropriateness of inpatient diabetes care and on some clinical outcomes in hospitalised subjects. METHODS: A multicentre (6 regional hospitals) cluster-randomized (2:1) two parallel-group pragmatic intervention trials, as a part of the GOVEPAZ study, was conducted in three clinical settings, that is, Internal Medicine, Surgery and Intensive Care. Intervention consisted of a 2-month structured education of clinical staff to inpatient diabetes care. Twelve wards - 2 for each hospital - and 6 wards - 1 for each hospital - were randomized to usual care and to the intervention arm, respectively. Consecutively hospitalised diabetic subjects (n = 524, age 74 ± 14 years, 57% males, median HbA1C 57 mmol/mol) were included. The clinical appropriateness of inpatient diabetes management was assessed by a previously validated multi-domain performance score (PS). Clinical outcomes included hypoglycemia, glucose control biomarkers, clinical conditions at discharge and inpatient mortality rate. RESULTS: A numerically, but not statistically significant, higher PS (+0.94; 95% C.I.: -0.53 - +2.4) was achieved in the intervention than in the usual care wards. Hypoglycemias (p = 0.32), glucose control (p = 0.89) and survival rates (p = 0.71) were similar in the two experimental arms. Plasma glucose on admission (OR = 1.52 per 1 SD; C.I. 1.07-2.17; p = 0.021) and the number of hypoglycemic events per patient (OR = 1.55 per 1 SD; C.I.:1.11-2.16; p = 0.011) were independently associated with the inpatient mortality rate. CONCLUSIONS: Structured education of the clinical staff failed to improve the inpatient appropriateness of diabetes care or clinical outcomes. In-hospital hypoglycemia was confirmed to be an independent indicator of death risk.