Study design and data collection
This study used a database from a prospective, multi-centre, observational study that investigated the epidemiology of acute kidney injury (AKI) in critically ill patients in 30 ICUs at 28 tertiary hospitals in Beijing, China, from March 1 to August 31, 2012 (the Beijing Acute Kidney Injury Trial (BAKIT) [19] (for a complete list of these hospitals and the persons responsible for the data acquisition, see Additional file 1). Study subjects included all adult patients (age ≥ 18 years) admitted consecutively to the ICU. Only the initial ICU admission was considered in this study. The following patients were excluded: patients with preexisting end-stage chronic kidney disease, patients already receiving renal replacement therapy (RRT) before admission to the ICU, and patients who had received kidney transplantation in the previous 3 months [20]. Pre-existing comorbidities were diagnosed based on the International Classification of Diseases (ICD-10) codes. Patients were followed up until death, until hospital discharge, or for 28 days. Among the 9079 patients who were admitted consecutively, 3107 patients were included in our study (Fig. 1).
Thorough follow-up of all patients included in the study was conducted in the first 10 days after ICU admission. The collected data included demographics, anthropometrics, admission diagnosis, comorbidities, daily vital signs and laboratory data, which were used to automatically calculate the APACHE II score, the Simplified Acute Physiology Score II (SAPS II) score [21] and the SOFA score, days from hospital to ICU admission, ICU length of stay (LOS), hospital LOS, use of vasoactive drugs, and length of mechanical ventilation. RRT data were also reported.
Mortality data were collected up to 28 days after ICU discharge from hospital records, including records from hospital admissions and visits to outpatient clinics.
Outcomes
The primary outcome was 28-day mortality, and the secondary outcome was the occurrence of the AKI.
Nutritional support
Nutritional support methods were based on the guidelines for enteral and parenteral nutrition issued by the European and American Society of Enteroprotective Nutrition [22], combined with our accumulated clinical experience, individualized nutritional support was given to all patients. The patients began enteral nutrition (EN) 20–25 kcal/(kg.d) and a protein requirement of 1.2–2.0 g/(kg.d) within 24–48 h of admission to the ICU (on average). If the patient was intolerant of EN or had contraindications to EN, parenteral nutrition (PN) support was given within 24—48 h. If EN could not fully meet the nutritional needs of patients, appropriate intravenous supplementation with glucose, amino acids, or fat emulsion was given, that is, the combination of EN and PN.
Definitions
We used the modified 9-point scale of the NUTRIC score, the mNUTRIC score [8]. We defined the scores from 0 to 4 as “low scores”, which indicated a low level of risk of malnutrition, and the scores from 5 to 9 as “high scores”, which were associated with worse clinical outcomes [8]. Because the mNUTRIC score includes APACHE II score, it was calculated only once at ICU admission.
AKI severity was classified according to the KDIGO guidelines [23]. AKI occurring within 10 days is defined as AKI, and more than 10 days is defined as non-AKI.
Statistical analysis
Non-normally distributed continuous variables were expressed as the medians with interquartile ranges (IQRs) and were compared using the Mann–Whitney U test or Kruskal–Wallis analysis of variance with Bonferroni correction. Categorical variables were expressed as the number of cases and proportions and were compared using the Mantel–Haenszel Chi-square test.
A multivariate Cox regression analysis was performed using a backward stepwise selection method, with P value < 0.05 as the entry criterion, and P value ≥ 0.10 as the removal criterion. The assumption of proportional hazards was checked graphically using log (-log (survival probability)) plots and was found to be appropriate. Because the mNUTRIC score includes APACHE II and SOFA score, to avoid the duplicates, we did a collinearity analysis on the mNUTRIC score, APACHEII and SOFA, and found that there was no collinearity between them. Variables considered for multivariable analysis included age, sex, body mass index (BMI), APACHE II score, SAPS II, SOFA score, mNUTRIC score, use of vasoactive drugs, mechanical ventilation, AKI, RRT and underlying diseases. We tested for collinearity among all variables using a Cox regression analysis to generate hazard ratios (HR) and 95% confidence intervals (CIs).
The receiver operating characteristic (ROC) curve was drawn according to the sensitivity and specificity of the mNUTRIC score in predicting the 28-day mortality risk of patients and the best cut-off value was determined by the maximum of the Youden index (i.e., sensitivity plus specificity minus one) calculated from the ROC analysis. Using Hosmer- lemeshow goodness of fit to test the calibration of the scoring system. The 28-day survival stratified by low and high mNUTRIC scores was additionally evaluated graphically using the Kaplan–Meier product limit survival plot, we used Log-rank (Mantel–Cox) test for the comparison of survival curves. To verify the predictive effect of the mNUTRIC score on 28-day mortality in different populations, subgroup analysis was performed, we divided the study population into mechanical ventilation, medical mechanical ventilation (Because surgical patients are intubated for surgery, unlike medical patients who are intubated for serious medical conditions, we separately list medical patients who require intubation), sepsis, AKI and RRT patients, respectively.
All statistical analyses were performed using SPSS software (IBM Corp., Statistics for Windows, version 22.0, Armonk, NY, USA), with a two-sided P value < 0.05 considered statistically significant.
