[Analysis of the difference between indirect calorimetry and predicted energy estimation in patients with chronic...

To analyze the difference between indirect calorimetry (IC) and predicted energy estimation in patients with chronic obstructive pulmonary disease (COPD), and its possible factors affecting the difference, to provide reasonable energy supply basis for COPD patients.

A prospective cohort study was conducted. Twenty-six patients with COPD undergoing mechanical ventilation (MV) admitted to intensive care unit (ICU) of Hangzhou City Fuyang District First People's Hospital in Zhejiang Province from January to December in 2016 were enrolled. The energy values of patients were calculated by IC and predicted energy estimation, respectively. According to the degree of IC values deviating from the predicted energy estimation, the patients were divided into energy approaching group (IC values deviating from the empirical energy estimation ≤15%) and energy deviation group (IC values deviating from the empirical energy estimation > 15%). Bland-Altman diagram was drawn, and the consistency of the energy target values assessing by two methods was analyzed. The factors influencing the energy value deviation of the two measuring methods were screened by the multivariate Logistic regression and linear regression analysis.

Twenty-six patients were enrolled in the final analysis. The energy target value of IC was significantly higher than that of predicted energy estimation (kJ: 7 079.3±1 213.4 vs. 6 527.0±949.8), and the difference between two values was statistically significant (P < 0.01). Bland-Altman heterogeneity analysis showed that the overall consistency of the energy values between the predicted energy estimation and IC was quite good. There were 14 patients in energy approaching group, and 12 in energy deviation group. There was no significant difference in gender, age, body mass index (BMI), type of COPD, or acute physiology and chronic health evaluation II (APACHE II) score between the two groups. In energy deviation group, the IC value of patients was significantly higher than predicted energy estimation (kJ: 7 711.1±1 125.5 vs. 6 556.3±907.9, P < 0.01). However, in energy approaching group, there was no significant difference between the energy values of IC and predicted energy estimation (kJ: 6 539.6±1 037.6 vs. 6 501.9±1 016.7, P > 0.05). Multivariate Logistic regression analysis showed that APACHE II score was an independent risk factor inducing the deviation of energy evaluation between IC and predicted energy estimation [odds ratio (OR) = 1.403, 95% confidence interval (95%CI) = 1.019-1.932, P = 0.038]. Multivariate linear regression analysis showed that the APACHE II score increased by 1, and the energy deviation increased by 2.0 kJ (β = 0.476, 95%CI = 0.004-0.956, P = 0.047).

For patients with COPD, there was a good correlation between predicted energy estimation and the resting energy expenditure measured by IC. APACHE II score was an independent risk factor inducing the deviation of energy evaluation between IC and predicted energy estimation. It is suggested that the target value of energy should be determined by IC for patients with high APACHE II score.