Abstract：
Both of neurological emergencies and hyperglycemia are independently associated risk factors of mortality in the ICU patients. In critically ills, hyperglycemia is secondary to already existing DM or stress-induced hyperglycemia (SIH). Admission glycemic gap (AGG) is considered as a reliable indicator of SIH. This study aimed to explore the association of AGG on diabetic neuro-critical patients\' short-term mortality, and understand the potential of AGG as the predictor of outcome. Sixty adult diabetic neuro-critical patients admitted in ICU and stayed at least for 24 hours, were prospectively observed for 30 days, or until discharge or death, whichever came first. The patients\' initial clinical assessment and HbA1c, CBC, ABG, and blood glucose level were done within 24 hours of admission. A1c derived admission glucose (ADAG) was calculated as, ADAG = (1.59 × HbA1c) - 2.59 (mmol/L). The AGG was calculated by subtracting ADAG from admission blood glucose level (ABGL). Death or survival of 30 days was our primary outcome and participants were divided between survivor or non-survivor groups according to primary outcome. Statistical comparisons of the study variables between the groups were performed and the relationship between parameters derived from blood glucose and mortality was prospected. Among the 60 patients enrolled, 35(58.3%) were non-survivors and 25(41.7%) were survivors. Age, sex, residence, primary diagnosis, co-morbidity, or drug history had no association with survival/non-survival. Among the initial clinical assessment parameters, lower GCS had significant association with non-survival. AGG, HbA1c, ADAG and ABGL were significantly different between the groups, with higher values in the non-survivors. Lower GCS, and higher AGG, HbA1c, ADAG and ABGL showed significant odds of non-survival. The highest odds of non- survival was for AGG (OR 2.95, 95% CI: 1.83-4.75; p<0.001). For ABGL and HbA1c the OR were 2.03 (95% CI: 1.44-2.86; p<0.001) and 1.93 (95% CI: 1.04-3.58; p<0.04) respectively. The final adjusted odds (aOR) of non-survival for higher AGG was 3.25 (95% CI: 1.71-6.16; p<0.001), signifying that AGG is independently associated with non-survival. AGG, GCS level, ABGL, HbA1c level, and ADAG can predict short-term outcome (mortality). However, AGG has the greatest potential to predict short-term outcome in diabetic neuro-critical patients.
摘要：
神经系统急症和高血糖是ICU患者死亡的独立相关危险因素。在危重病中，高血糖是继发于已经存在的DM或应激诱导的高血糖(SIH).入院血糖差距（AGG）被认为是SIH的可靠指标。本研究旨在探讨AGG与糖尿病神经危重患者短期死亡率的关系。并了解AGG作为结果预测因子的潜力。60名成年糖尿病神经危重患者入住ICU并至少停留24小时，前瞻性观察了30天，或者直到出院或死亡，以先到者为准。患者的初始临床评估和HbA1c，CBC,ABG,和血糖水平在入院后24小时内进行。A1c得出的入院血糖（ADAG）计算为，ADAG=(1.59×HbA1c)-2.59(mmol/L)。通过从入院血糖水平(ABGL)减去ADAG来计算AGG。我们的主要结果是死亡或存活30天，根据主要结果将参与者分为幸存者或非幸存者组。对各组间的研究变量进行统计比较，并对血糖与死亡率的关系进行了展望。在60名患者中，35人（58.3%）为非幸存者，25人（41.7%）为幸存者。年龄,性别,residence,初步诊断,合并症，或药物史与生存/非生存无关.在初始临床评估参数中，较低的GCS与非生存显著相关.AGG,HbA1c,ADAG和ABGL在组间有显著差异，非幸存者的价值更高。降低GCS,和更高的AGG，HbA1c,ADAG和ABGL显示出显著的非生存几率。AGG的非生存几率最高(OR2.95,95%CI:1.83-4.75;p<0.001)。对于ABGL和HbA1c，OR分别为2.03（95％CI：1.44-2.86；p<0.001）和1.93（95％CI：1.04-3.58；p<0.04）。较高AGG的非生存的最终校正赔率（aOR）为3.25（95％CI：1.71-6.16；p<0.001），这表明AGG与非生存独立相关。AGG,GCS等级,ABGL,HbA1c水平,ADAG可以预测短期结果（死亡率）。然而,AGG具有预测糖尿病神经危重患者短期预后的最大潜力。
