Admission blood glucose as a prognostic indicator in patients with acute myocardial infarction

Objective: To investigate the utility of admission blood glucose for predicting major adverse cardiac events (MACE) during hospitalisation and 6 months’ postdischarge in acute myocardial infarction (AMI) patients. Methods and results: This study recruited 2878 AMI patients admitted to the Coronary Care Unit at R. K. Khan Hospital, Durban, South Africa, from 2002 - 2014. Demographic and clinical data stored in an electronic database were obtained from all patients. Admission blood glucose levels were sub-divided into 3 groups; low (<7.8), medium (7.8-10.9) and high (≥11) mmol/l. The mean age of the study population was 57.18 ± 7 years of whom 65% were males. Self-reported diabetes was found in 59%, while 377 patients were diagnosed with diabetes based on HbA1c levels ≥6.5%, increasing the overall prevalence to 72% (n=2070). More patients were in the low admission blood glucose group (49%), medium group (16%), and high group (35%). The highest prevalence of MACE was seen in the high group (42%) compared to either the medium (39%) or low groups (26%; p<0.001), particularly for cardiogenic shock (p<0.001), cardiac failure (p<0.001) and death (p<0.001). Following multivariable logistic regression analyses of clinical and laboratory parameters associated with mortality, high admission blood glucose conferred a significantly increased odds of mortality (p=0.001). The optimal cut-off admission blood glucose value as determined via the receiver operating characteristic curve for predicting in hospital and 6 months’ mortality was 8.5mmol/l (AUC of 0.63) and 8.1mmol/l (AUC of 0.61) for MACE. Conclusions: This study shows that patients have multiple risk factors for AMI with diabetes playing a central role. Although elevated admission blood glucose is an important predictor for in hospital and shortterm MACE, the cut-off value for predicting MACE and mortality has only modest predictability and further research is required to improve the performance of these measures for routine clinical use.