Abstract: Objective To apply the nutritional risk screening 2002 (NRS2002) and the modified version of NRS2002 (the modified NRS2002) in the nutritional assessment of patients with traumatic fractures, and to evaluate the efficacy of the two screening expressions in order to rationally select screening tools, detect nutritional deficiencies as early as possible and provide nutritional support for patients with traumatic fractures. 
Methods A total of 271 fracture patients admitted for acute trauma were selected for nutritional risk screening by using NRS2002 and the modified NRS2002. The sensitivity and specificity of the two screening tools were analyzed, the consistency of the two screening tools was judged by calculating Kappa values, and the correlation of physical examination indicators and laboratory indicators with nutritional risk was verified. 
Results Among the 271 patients, 68 (25.1%) were screened for nutritional risk by NRS2002, and 103 (38.0%) were screened for nutritional risk with the modified NRS2002. In patients with nutritional risk (≥3 points) screened by NRS2002 and the modified NRS2002, their upper arm muscle circumference (AMC), triceps skinfold (TSF), grip strength (GS), total protein (TP), albumin (ALB), prealbumin (pALB), calcium, phosphorus, magnesium, red blood cells (RBCs), and hemoglobin (HGB) were lower than those without nutritional risk, while calf circumference (CC), chloride, white blood cells (WBC), and C-reactive protein (CRP) were higher than those of those without nutritional risk (P＜0.05). There was no difference in the physical examination indexes and laboratory indexes of patients with nutritional risk (≥3 points) screened by the modified NRS2002 and NRS2002. The consistency test of the modified NRS2002 and NRS2002 results was good (Kappa=0.673, P＜0.05), and the area under the receiver operating characteristic (ROC) curve of the modified NRS2002 was 0.894. At one week afterwards, the positive rate of NRS2002 re-screening of patients without nutritional risk (＜ 3 points) was significantly higher than that of the modified NRS2002 (P＜0.05). AMC, TSF, GS, ALB, pALB, and HGB were negatively correlated with NRS2002 and the modified NRS2002 (P＜0.05), while CC was positively correlated with NRS2002 and the modified NRS2002 (P＜0.05). 
Conclusion The screening results of the modified NRS2002 are generally consistent with those of NRS2002. The sensitivity of the modified NRS2002 is higher in nutritional risk screening, and the positive rate of re-screening after one week is low, indicating that its ability to detect the nutritional risk of patients is superior to that of NRS2002. BMI, AMC, TSF, GS, ALB, pALB, HGB and other factors related to the occurrence of nutritional risk. Therefore, it can be used in combination with the modified NRS2002 to assess the nutritional risk of patients. 

Key words: fractures, nutritional status, nutritional risk screening