Study On Risk Factors And Outcome Of Refeeding Syndrome In Neurocritically Ill Patients

Background&aim:Neurocritically ill patients are prone to malnutrition and need nutritional support due to difficulties in swallowing,impaired consciousness,perceptual defects,cognitive dysfunction and increased metabolic demands.Malnutrition and inadequate nutritional support may lead to refeeding syndrome(RFS).In 2006,the National Institute for Health and Clinical Excellence(NICE)proposed risk factors for RFS for hierarchical management,but the literature reported low sensitivity.In 2018,the modified NICE(mNICE)risk stratification was proposed for predicting RFS,however,its sensitivity and specificity have not been validated.This study aimed to explore the incidence,risk factors and outcome of RFS in neurocritically ill patients,moreover,to verify the predictive efficacy of mNICE risk stratification in 2018 for RFS in neurocritically ill patients.Methods:We conducted a retrospective study among neurocritically ill patients from the Department of Neurology,Nanfang Hospital from January 2013 to April 2019.Inclusion criteria:patients from neurocritical care unit(NCU)who had enteral nutrition>72h and had serum phosphorus results both at admission and 72±12h.Exclusion criteria:(1)incomplete data;(2)age>85 years old or<18 years old;(3)blood phosphorus<0.65 mmol/L before NCU admission;(4)loss of follow-up;(5)terminal malignant tumor;(6)complications of diabetic ketoacidosis;(7)recent parathyroidectomy,renal replacement therapy,use of phosphorus binding agent,and mild hypothermia therapy.RFS was defined as the occurrence of new-onset hypophosphatemia(<0.65 mmol/L)within 72h of the commencement of nutritional support.The primary outcome was 6-month mortality.The secondary outcomes included 30-day mortality,NCU stay,and hospital length of stay.Categorative variables were represented by the Chi-square test or Fisher’s exact test with the number of cases(%),while continuous variables were represented by the independent sample t-test or Mann-Whitney U test based on the data distribution with the mean ±standard deviation or median[interquad interval].The samples were divided into two groups:RFS and non-RFS.Univariate analysis was used to identify candidate variables related to RFS,and then the candidate variables were included in multiple logistic regression to identify independent risk factors for RFS,and next we establish a regression model to predict RFS.The receiver operating characteristic(ROC)curve and the area under the curve(AUC)were used to evaluate the predictive efficacy of the regression model for RFS.mNICE risk stratification were collected based on clinical data and then divided into two categories:no risk,low risk,high risk and very high risk to evaluate the sensitivity and specificity of RFS prediction.The test level a was set at 0.05.Results:A total of 328 patients were enrolled,and 56(17.1%)of them developed RFS within 72 h of nutrition support.Significantly,we found that patients with high malnutrition universal screening tool(MUST)and sequential organ failure assessment(SOFA)scores were more likely to develop RFS.The occurrence of RFS was associated with a longer NCU stay,higher 30-day mortality and 6-month mortality,and poorer 6-month functional outcome.Moreover,RFS was identified as an independent risk factor for 6-month mortality(OR=1.940,95%CI 1.020-3.688,P=0.043).Of the 53 patients(16.2%)who were considered to be at high or very high risk of RFS,only 14(26.4%)actually developed RFS,with a sensitivity(95%CI)of 0.250(0.148-0.386)and specificity(95%CI)of 0.857(0.808-0.895).Conclusion:RFS is not rare in neurocritically ill patients and is more likely to occur in patients with nutritional risk and more severe conditions.RFS is an independent risk factor for 6-month mortality in neurocritically ill patients.The 2018 modified NICE score showed good specificity for identifying RFS,but poor sensitivity.