[Stratified outcomes of "Kidney Disease: Improving Global Outcomes" serum creatinine criteria in critical ill patients: a secondary analysis of a multicenter prospective study].

Objective: To investigate the different outcomes of two types of acute kidney injury (AKI) according to standard of Kidney Disease: Improving Global Outcomes-AKI (KDIGO-AKI), and to analyze the risk factors that affect the prognosis of intensive care unit (ICU) patients in China.

Methods: A secondary analysis was performed on the database of a previous study conducted by China Critical Care Clinical Trial Group (CCCCTG), which was a multicenter prospective study involving 3 063 patients in 22 tertiary ICUs in 19 provinces and autonomous regions of China. The demographic data, scores reflecting severity of illness, laboratory findings, intervention during ICU stay were extracted. All patients were divided into pure AKI (PAKI) and acute on chronic kidney disease (AoCKD). PAKI was defined as meeting the serum creatinine (SCr) standard of KDIGO-AKI (KDIGO-AKI) and the estimated glomerular filtration rate (eGFR) at baseline was ≥ 60 mL×min×1.73 m, and AoCKD was defined as meeting the KDIGO-AKI standard and baseline eGFR was 15-59 mL×min×1.73 m. All-cause mortality in ICU within 28 days was the primary outcome, while the length of ICU stay and renal replacement therapy (RRT) were the secondary outcome. The differences in baseline data and outcomes between the two groups were compared. The cumulative survival rate of ICU within 28 days was analyzed by Kaplan-Meier survival curve, and the risk factors of ICU death within 28 days were screened by Cox multivariate analysis.

Results: Of the 3 063 patients, 1 042 were enrolled, 345 with AKI, 697 without AKI. The AKI incidence was 33.11%, while ICU mortality within 28 days of AKI patients was 13.91% (48/345). Compared with PAKI patients (n = 322), AoCKD patients (n = 23) were older [years old: 74 (59, 77) vs. 58 (41, 72)] and more critical when entering ICU [acute physiology and chronic health evaluation II (APACHE II) score: 23 (19, 27) vs. 15 (11, 22)], had worse basic renal function [eGFR (mL×min×1.73 m): 49 (38, 54) vs. 115 (94, 136)], more basic complications [Charlson comorbidity index (CCI): 3 (2, 4) vs. 0 (0, 1)] and higher SCr during ICU stay [peak SCr for diagnosis of AKI (μmol/L): 412 (280, 515) vs. 176 (124, 340), all P < 0.01]. The mortality and RRT incidence within 28 days in ICU of AoCKD patients were significantly higher than those of PAKI patients [39.13% (9/23) vs. 12.11% (39/322), 26.09% (6/23) vs. 4.04% (13/322), both P < 0.01], while no significant difference was found in the length of ICU stay. Kaplan-Meier survival curve analysis showed that the 28-day cumulative survival rate in ICU in AoCKD patients was significantly lower than PAKI patients (Log-Rank: χ = 5.939, P = 0.015). Multivariate Cox regression analysis showed that admission to ICU due to respiratory failure [hazard ratio (HR) = 4.458, 95% confidence interval (95%CI) was 1.141-17.413, P = 0.032], vasoactive agents treatment in ICU (HR = 5.181, 95%CI was 2.033-13.199, P = 0.001), and AoCKD (HR = 5.377, 95%CI was 1.303-22.186, P = 0.020) were independent risk factors for ICU death within 28 days.

Conclusions: Further detailed classification (PAKI, AoCKD) based on KDIGO-AKI standard combined with eGFR is related to ICU mortality in critical patients within 28 days.