Abstract
INTRODUCTION: Diabetic ketoacidosis (DKA) is a life-threatening complication characterized by hyperglycemia, metabolic acidosis, and ketonemia. Although the time to resolution of DKA has been estimated at 12 to 18 hours, the factors that could prolong it have not been fully studied. METHODS: Retrospective study of medical records of adult patients admitted to the general ward with a diagnosis of DKA. They were classified according to severity as mild, moderate and severe. The time to resolution of crisis (TRC) was defined as that necessary until normalization of metabolic parameters. A logistic regression analysis was performed to evaluate the association between TRC>12 hours and continuous variables. ROC analysis and survival analysis were performed using a Cox regression model. RESULTS: 85 patients were studied. 48.23% presented severe DKA. The TRC was 14 hours, being higher in severe DKA. Patients with TRC>12 hours had a lower pH and HCO3-, and a higher anion gap, white blood cells, and volume of crystalloids used. Logistic regression analysis showed that pH and crystalloid volume correlated with TRC>12 hours. ROC analysis determined a pH cutoff value of 7.13 for TRC>12 hours (sensitivity 77%, specificity 88%). The Cox regression showed that a pH<7.13 on admission is associated with a higher TRC (HR 3.30). CONCLUSION: A pH lower than 7.13 at the time of hospital admission identifies patients with DKA who will need a longer time to resolve their metabolic parameters.