[Evaluation of the effectiveness of a protocol designed to improve glycemic control in hospitalized internal medicine patients with hyperglycemia].

BACKGROUND: Our aims were to assess the effectiveness of a diabetes (DM) management protocol to increase scheduled insulin therapy and to improve glycemic inpatient control.

PATIENTS AND METHODS: We designed an analytical retrospective cohort study comparing 2 groups of medical services hospitalized patients with a primary of secondary discharge diagnosis of DM, before (group PRE) and after (group POS) the delivery of a DM management protocol. We analyzed the quality of attention by process indicators (cumulative probability of receive scheduled insulin therapy, evaluated with Kaplan-Meier analysis) and result indicators (adjusted glucose differences (group POS - group PRE), evaluated with multivariate regression models).

RESULTS: A number of patients (355) were included (228 group PRE and 127 group POS). The median time to scheduled insulin regimen beginning was 1 (CI 95%: 0-2.5) day in group POS and 4 (CI 95%: 2-6) days in group PRE (p=0.056). First 48 hours mean glucose in patients without scheduled insulin therapy was lower in group POS than in group PRE (163.9 versus 186.7 mg/dl; p=0.025). The first 24 hours mean glucose was significantly lower in patients of group POS, with a difference between both groups of -24.8 mg/dl (CI 95%: -40.5-(-9); p=0.002). Stratified analysis showed statistically significant mean in-hospital glucose difference only in the nothing by mouth situation (-29.8 mg/dl; CI 95%: -58.9-(-0.6); p=0.045).

CONCLUSION: The delivery of an institutional protocol can improve hospitalized DM patients management quality.