RESUMO
OBJECTIVE: To develop and validate a diagnostic-therapeutic guideline (DTG) for the management of hospitalized diabetic patients, and apply it to a sample of medical files. MATERIAL AND METHODS: The DTG was constructed and then validated by a consensus of internist physicians, afterwards it was applied to a sample of 97 randomly selected medical files of diabetic patients discharged from a internal medicine service. Hospitalization time, studies requested, average glycemia, glycemic prior to the discharge, time for to achieve glycemic control and mortality were compared at different percentages of the DTG's application. The results were analyzed by U of Mann-Whitney and ANOVA of Kruskal-Wallis. RESULTS: The intra-class correlation coefficient among the physicians for the guide validation was 0.94, with minimal modifications of content and format. The time, in hours, to achieve glycemic control was minor with a higher application percentage of the complete guide, p < 0.001, and also of the therapeutic domain, p = 0.05. The last glycemic prior to the discharge was lower with a higher application than 75% of the diagnostic domain, p < 0.001. A higher application percentage of the entire guide was found in living subjects more than in deceased subjects, p = 0.001, and also of the therapeutic domain, p = 0.008, and of the diagnostic domain p = 0.05. CONCLUSIONS: A higher level of application of the DTG is associated with a shorter time to achieve glycemic control, lower glycemic prior to the discharge and lower mortality.