RESUMO
Objective: To investigate the effect of sedation on resting energy expenditure (REE) in patients with extremely severe burns and the choice of REE estimation formula during the treatment. Methods: A retrospective non-randomized controlled clinical study was conducted. From April 2020 to April 2022, 21 patients with extremely severe burns who met the inclusion criteria and underwent mechanical ventilation treatment were admitted to the Department of Burn and Wound Repair of Second Affiliated Hospital of Zhejiang University School of Medicine, including 16 males and 5 females, aged 60 (50, 69) years. Early anti-shock therapy, debridement, skin transplantation, nutritional support, and other conventional treatments were applied to all patients. Patients were sedated when they had obvious agitation or a tendency to extubate, which might lead to aggravation of the disease. REE measurement was performed on patients using indirect calorimetry on post-injury day 3, 5, 7, 9, 11, 14 and every 7 days thereafter until the patient died or being successfully weaned from ventilator. Totally 99 times of measurements were carried out, of which 58 times were measured in the sedated state of patients, and 41 times were measured in the non-sedated state of patients. The age, weight, body surface area, residual wound area, post-injury days of patients were recorded on the day when REE was measured (hereinafter briefly referred to as the measurement day). The REE on the measurement day was calculated with intensive care unit conventional REE estimation formula Thumb formula and special REE estimation formulas for burns including the Third Military Medical University formula, the Peng Xi team's linear formula, Hangang formula. The differences between the sedated state and the non-sedated state in the clinical materials, measured and formula calculated values of REE of patients on the measurement day were compared by Mann-Whitney U test and independent sample t test. The differences between the REE formula calculated values and the REE measured value (reflecting the overall consistency) in the sedated state were compared by Wilcoxon signed rank-sum test. The Bland-Altman method was used to assess the individual consistency between the REE formula calculated value and the REE measured value in the sedated state, and to calculate the proportion of the REE formula calculated value within the range of ±10% of the REE measured value (hereinafter referred to as the accuracy rate). Root mean square error (RMSE) was used to evaluate the accuracy of the REE formula calculated value relative to the REE measured value. Results: Compared with those in the non-sedated state, there was no statistically significant change in patient's age or post-injury days on the measurement day in the sedated state (P>0.05), but the weight was heavier (Z=-3.58, P<0.01), and both the body surface area and the residual wound area were larger (with Z values of -2.99 and -4.52, respectively, P<0.01). Between the sedated state and the non-sedated state, the REE measured values of patients were similar (P>0.05). Compared with those in the non-sedated state, the REE values of patients calculated by Thumb formula, the Third Military Medical University formula, the Peng Xi team's linear formula, and Hangang formula on the measurement day in the sedated state were significantly increased (with Z values of -3.58 and -5.70, t values of -3.58 and -2.74, respectively, P<0.01). In the sedated state, compared with the REE measured value, there were statistically significant changes in REE values of patients calculated by Thumb formula, the Third Military Medical University formula, and Hangang formula on the measurement day (with Z values of -2.13, -5.67, and -3.09, respectively, P<0.05 or P<0.01), while the REE value of patients calculated by the Peng Xi team's linear formula on the measurement day did not change significantly(P>0.05). The analysis of the Bland-Altman method showed that in the sedated state, compared with the REE measured value, the individual consistency of the calculated value of each formula was good; Thumb formula and Hangang formula significantly underestimated the patients' REE value (with the average value of the difference between the formula calculated value and the measured value of -1 463 and -1 717 kJ/d, the 95% confidence interval of -2 491 to -434 and -2 744 to -687 kJ/d, respectively), but the individual differences were small; the Third Military Medical University formula significantly overestimated the patients' REE value (with the average value of the difference between the formula calculated value and the measured value of 3 530 kJ/d, the 95% confidence interval of 2 521 to 4 539 kJ/d), but the individual difference was small; the Peng Xi team's linear formula did not significantly overestimate the patients' REE value (with the average value of the difference between the formula calculated value and the measured value of 294 kJ/d, the 95% confidence interval of -907 to 1 496 kJ/d), while the difference standard deviation was 4 568 kJ/d, which showed a large individual difference. In the sedated state, relative to the REE measured value, the accuracy rates of REE values calculated by Thumb formula, the Third Military Medical University formula, the Peng Xi team's linear formula, and Hangang formula were 25.9% (15/58), 15.5% (9/58), 10.3% (6/58), and 15.5% (9/58), respectively, and RMSE values were 4 143.6, 5 189.1, 4 538.6, and 4 239.8 kJ/d, respectively. Conclusions: Sedative therapy leads to a significant decrease in REE in patients with extremely severe burns undergoing mechanical ventilation treatment. When REE cannot be regularly monitored by indirect calorimetry to determine nutritional support regimens, patients with extremely severe burns undergoing sedation may be prioritized to estimate REE using Thumb formula.