Effect of enteral nutritional support on clinical outcomes of patients with sepsis / 中华危重病急救医学

ABSTRACT

Objective:To research whether clinical outcomes of patients with sepsis can be improved by higher enteral nutritional support.Methods:A retrospective cohort method was applied. 145 patients with sepsis who were hospitalized in intensive care unit (ICU) of Peking University Third Hospital from September, 2015 to August, 2021 and met inclusion criteria as well as exclusion criteria were selected, including 79 males and 66 females, the median age was 68 (61, 73). Researchers evaluated whether there was correlation between improved modified nutrition risk in critically ill score (mNUTRIC), daily energy intake and protein supplement of patients and their clinical outcomes through Poisson log-linear regression analysis and Cox regression analysis.Results:The median of mNUTRIC score of 145 hospitalized patients was 6 (3, 10), wherein 70.3% of patients (102 cases) were in high-score group (≥ 5 scores) and 29.7% of patients (43 cases) were in low-score group (< 5 scores); the average of daily protein intake in ICU was about 0.62 (0.43, 0.79) g·kg -1·d -1, and the average of daily energy intake was about 64.4 (48.1, 86.2) kJ·kg -1·d -1. As shown by Cox regression analysis, increase of mNUTRIC score, sequential organ failure assessment (SOFA), and acute physiology and chronic health evaluation Ⅱ (APACHE Ⅱ) were correlated to growth of in-hospital mortality [hazard ratio ( HR) = 1.12, 95% confidence interval (95% CI) was 1.08-1.16, P = 0.006; HR = 1.04, 95% CI was 1.01-1.08, P = 0.030; HR = 1.08, 95% CI was 1.03-1.13, P = 0.023]. Higher average daily intake of protein and energy as well as lower mNUTRIC, SOFA, and APACHE Ⅱ scores were also significantly correlated to lower 30-day mortality ( HR = 0.45, 95% CI was 0.25-0.65, P < 0.001; HR = 0.77, 95% CI was 0.61-0.93, P < 0.001; HR = 1.10, 95% CI was 1.07-1.13, P < 0.001; HR = 1.07, 95% CI was 1.02-1.13, P = 0.041; HR = 1.15, 95% CI was 1.05-1.23, P = 0.014); however, there was no significant correlation between gender as well as number of complications and in-hospital mortality. Within 30 days of attack of sepsis, the average daily intake of protein and energy were not correlated to days of non-ventilator ( HR = 0.66, 95% CI was 0.59-0.74, P = 0.066; HR = 0.78, 95% CI was 0.63-0.93, P = 0.073). Increase of patients' average daily intake of protein and energy were significantly correlated to a lower in-hospital mortality ( HR = 0.41, 95% CI was 0.32-0.50, P < 0.001; HR = 0.87, 95% CI was 0.84-0.92, P < 0.001), shorter ICU stay ( HR = 0.46, 95% CI was 0.39-0.53, P < 0.001; HR = 0.82, 95% CI was 0.78-0.86, P < 0.001), and hospital stay ( HR = 0.51, 95% CI was 0.44-0.58, P < 0.001; HR = 0.77, 95% CI was 0.68-0.88, P < 0.001). According to correlation analysis, among patients with mNUTRIC score ≥ 5, increasing daily intake of protein and energy can reduce in-hospital mortality ( HR = 0.44, 95% CI was 0.32-0.58, P < 0.001; HR = 0.73, 95% CI was 0.69-0.77, P < 0.001), and 30-day mortality ( HR = 0.51, 95% CI was 0.37-0.65, P < 0.001; HR = 0.90, 95% CI was 0.85-0.96, P < 0.001); the receiver operator characteristic curve (ROC curve) further confirmed that higher protein intake had good predictive value for inpatient mortality area under the curve (AUC) = 0.96 and 30-day mortality (AUC = 0.94); higher emergy intake had good predictive value for inpatient mortality (AUC = 0.87) and 30-day mortality (AUC = 0.83). By contrast, among patients with mNUTRIC score < 5, it is only discovered that increasing daily intake of protein and energy can reduce 30-day mortality of patients ( HR = 0.76, 95% CI was 0.69-0.83, P < 0.001). Conclusions:The increase of average daily intake of protein and energy for patients with sepsis is significantly correlated to reduction of in-hospital mortality and 30-day mortality, shorter ICU stay, and hospital stay. The correlation is more significant in patients with high mNUTRIC score, and higher intake of protein and energy can bring down in-hospital mortality and 30-day mortality. As for patients with low mNUTRIC score, nutritional support cannot improve prognosis of the patients significantly.