Abstract:

Objective To investigate the relationship between fasting blood glucose at admission and outcomes in hospitalized patients. Methods The clinical data of 186 192 hospitalized patients between 2006 and 2009 were screened, and patients confirmed with diabetes were excluded. Patients who received fasting blood glucose measurement on the second day of admission and died in hospital were selected as death case group (n=1 731), and 1 731 matched controls were recruited according to International Classification of Diseases (ICD-10). The associations of fasting blood glucose at admission with hospital mortality and hospital cost were statistically analysed. Results The fasting blood glucose at admission and hospital cost in death case group were significantly higher than those in control group (P<0.05). Patients with fasting blood glucose at admission of 5.0 mmol/L to 7.9 mmol/L had the lowest hospital mortality and hospital cost, while the hospital mortalities of those with fasting blood glucose at admission lower than 5.0 mmol/L or 7.9 mmol/L significantly increased (P<0.01), and the hospital cost also markedly increased. The risk of hospital mortality was the highest when fasting blood glucose at admission was lower than 3.0 mmol/L, with odds ratio (OR) of 1.67 and 95% confidence interval (CI) of 1.42 to 1.98 (P<0.01). Fasting blood glucose at admission was independent risk factor for hospital mortality after adjustment by age, sex and ICD (OR=1.486，95%CI =1.423 to 1.552)（P<0.001). Conclusion Fasting blood glucose at admission is independent risk factor for hospital mortality in hospitalized patients, and hypoglycemia may be more risky for hospital mortality.

Key words: hospitalized patients, fasting glucose, hospital mortality, hospital cost