[The influence of serum magnesium level on the prognosis of critically ill patients].

To investigate the incidence of hypomagnesemia and the effect of serum magnesium levels on the prognosis of critically ill patients in intensive care unit (ICU).A single-center prospective observation was conducted. The adult patients admitted to ICU of Zhejiang Provincial People's Hospital from January 2012 to January 2014 were enrolled, and they were expected to stay in hospital for more than 48 hours. All the patients who had been diagnosed with hypomagnesemia before ICU admission, or those who had received magnesium supplement therapy were excluded. All patients were monitored for serum magnesium levels within 24 hours after ICU admission, and they were divided into three groups: normomagnesemic group (serum magnesium levels 0.7-1.2 mmol/L), hypomagnesemic group (serum magnesium levels < 0.7 mmol/L), and hypermagnesemic group (serum magnesium levels > 1.2 mmol/L). Various parameters were recorded for every patient, including general information, disease composition, laboratory indexes, duration of mechanical ventilation, ICU stay days and final outcome. The acute physiology and chronic health evaluation II (APACHEII) score and sequential organ failure assessment (SOFA) score during the first 24 hours after ICU admission were calculated. The risk factors for the death in critically ill patients were postulated by logistic regression analysis.A total of 374 critically ill patients were enrolled, with 242 patients (64.71%) in normomagnesemic group, 102 (27.27%) in hypomagnesemic group, and 30 (8.02%) in hypermagnesemic group. As to the disease composition, although the patients in normomagnesemic group and hypomagnesemic group were mainly consisted of patients with nervous system diseases (33.06%, 31.37%) or pneumonia (25.62%, 25.49%), the proportion of patients with major abdominal and thoracic surgery (15.69% vs. 5.78%, χ (2) = 8.837, P = 0.003) or severe sepsis (7.84% vs. 1.65%, χ (2) = 9.935, P = 0.007) was significantly greater in the hypomagnesemic group compared with that of normomagnesemic group, and most hypermagnesemic patients were complicated by renal dysfunction in different degrees. Compared with the normomagnesemic group, the hypomagnesemic group was found to have higher SOFA scores (6.86±3.12 vs. 5.46±2.75, t = -2.930, P = 0.004), longer stay in ICU (days: 15.98±13.29 vs. 12.43±7.14, t = -2.318, P = 0.034) and higher mortality [54.90% (56/102) vs. 33.88% (82/242), χ (2) = 6.587, P = 0.010], but no statistically significant differences were found in gender composition, age, levels of other electrolytes (natrium, potassium, calcium, phosphorus), and APACHEII score. As shown by the result of the logistic regression analysis, APACHEII score [odds ratio (OR) = 1.129, 95% confidence interval (95%CI) = 1.064-1.197, P = 0.000] and serum magnesium level (OR = 2.163, 95%CI = 1.015-4.610, P = 0.046) were independent risk factors for death in critically ill patients.Serum magnesium levels are closely related to mortality rate in patients in ICU, so more attention should be paid to the occurrence of hypomagnesemia in critically ill patients.