Short-term Intensive Glucose Control in Patients with Severe Acute Pancreatitis
Yan-yan Zuo,Yan Kang,Bo Wang,Wan-hong Yin
DOI: https://doi.org/10.3760/cma.j.issn.1003-0603.2012.01.006
2012-01-01
Abstract:OBJECTIVE:To investigate effects of intensive glucose control (IGC) on glucose variability (GV) and clinical outcomes in patients with severe acute pancreatitis (SAP), and to assess predictive values of different measures of GV on intensive care unit (ICU) death.METHODS:A prospective, randomized, non-blinded clinical trial was conducted. All adult patients with diagnosis of SAP, who were admitted to ICU of West China Hospital between July 1st 2010 and July 1st 2011, were enrolled. Eligible patients were randomly assigned to receive either IGC to maintain glucose level between 6.1 mmol/L and 8.3 mmol/L or control group (none intervention was given). GV, ICU mortality, the incidence of infection, length of mechanical ventilation and ICU stay of the two groups were compared. The standard deviation of blood glucose level [GLU(SD)], mean amplitude of glycemic excursion [GLU(MAGE)] and glycemic lability index [GLU(GLI)] were chosen as measures of GV to analyze the multi variable correlation between them and ICU mortality, and the ability of these three parameters was assessed in predicting ICU death using area under a receiver operating characteristic curve (AUC).RESULTS:One hundred and twenty-two patients were eligible for the study, and 30 patients with SAP were enrolled in the study, with 15 cases in each group. The basic data of the two groups were comparable. Compared with control group, IGC showed an effect to reduce GV, including GLU(SD) [mmol/L: 1.81 (0.97, 2.65) vs. 2.48 (1.29, 2.87)], GLU(MAGE) [mmol/L: 3.76 (3.67, 5.85) vs. 5.30 (4.35, 6.80)], GLU(GLI) [306.8 (220.6, 613.3) vs. 339.5 (218.4, 423.1)], and lower ICU mortality (13.3% vs. 40.0%), but the difference showed no significant difference (all P > 0.05). It could also shorten the length of ICU stay (days: 11.3 ± 9.9 vs. 15.8 ± 7.6, P < 0.01), decrease the incidence of infection of blood stream and the lung (6.7% vs. 40.0%; 33.3% vs. 73.3%, both P < 0.05). A positive correlation between GLU(GLU) and ICU mortality was found (r = 0.371, P = 0.044), but no correlation was found between GLU(SD) or GLU(MAGE) and ICU mortality (r value was -0.144 and -0.065, P value was 0.448 and 0.731). AUC for GLU(SD), GLU(MAGE) and GLU(GLI) was 0.594 [95% confidence interval (95%CI) 0.382-0.805], 0.543 (95%CI 0.287-0.798) and 0.751 (95%CI 0.548-0.954) respectively, and GLU(GLI) was the best predictor of ICU death.CONCLUSIONS:IGC had an effect to reduce GV, decrease ICU mortality, shorten length of ICU stay, and lower the incidence of infection. Compared with GLU(SD) and GLU(MAGE), GLU(GLI) was the best predictor of ICU death.