Consistency of indirect calorimetry and HB coefficient method for measuring energy expenditure of pa-tients with sepsis / 中华临床营养杂志

ABSTRACT

Objective To evaluate the difference and correlation of 24 h energy expenditure in patients with sepsis by indirect calorimetry (IC) and HB coefficient equation. Methods A prospective comparative study including 60 patients with sepsis who was suitable for nutritional support and respiratory indirect calorimetry in the intensive care unit (ICU) from January to October 2015 was conducted. Resting energy expenditure (REE) was measured simultaneously by respiratory indirect calorimetry and HB coefficient (Harris-Benedict formula×stress coefficient) in 60 patients with sepsis at 0 day, 3 day, 7 day, and 14 day after nutritional support, and the differences in dynamic REE were compared between the two methods. The consistency of REE by indirect calorimetry (IC) and HB coefficient equation was evaluated by Bland-Altman. The correlation of IC and HB methods was determined by Pearson analysis. The linear regression equation was determined by linear regression analysis. Results Within 2 weeks after nutritional support, 188 times of measures by IC method and HB method respectively were finished in all patients with sepsis. The dynamic REE in respiratory indirect calorimetry group was significantly higher than that in HB coefficient method group (P<0. 05). The average bias of REE between the IC method and the HB method was (1930. 9±597. 7) kJ/24 h. For the consistency boundaries was over large and beyond the scope of clinical acceptance, there was a bias between the two methods and they could not directly substituted. There was a linear correlation of REE between the IC method and the HB coefficient (r =0. 757, P =0. 000). The equation associated with the HB coefficient method is fitted using a oneway regression: Y=1. 17X+ 834. 11 (kJ/24 h), and X was the 24 h energy expenditure measured by the HB coefficient method. Conclusion The energy metabolism of patients with sepsis is not obvious in the first 2 weeks. The HB coefficient method significantly underestimates the 24 h energy expenditure of patients with sepsis. The prediction equation can be used to correct the HB coefficient method and improve the HB coefficient method to predict the energy expenditure of patients with sepsis.