Clinical characteristics and prognosis of acute kidney injury in elderly patients with sepsis / 中华危重病急救医学

ABSTRACT

Objective To analyze the incidence of acute kidney injury (AKI) in elderly patients with sepsis, compare the clinical characteristics and prognosis between AKI and non-AKI elderly patients with sepsis, and to investigate the impact of classification of AKI and renal replacement therapy (RRT) on the outcome of elderly patients with sepsis. Methods The clinical data of 490 septic patients over 65 years old, admitted to intensive care unit (ICU) of Tianjin First Center Hospital from April 1st, 2016 to December 31st, 2018 were retrospectively analyzed. The patients were divided into two groups according to those with or without AKI. The clinical characteristics of patients were compared, and subgroup analysis of elderly septic patients with AKI was performed according to Kidney DiseaseImproving Global Outcomes (KDIGO) staging criteria and whether RRT was performed, to observe the effects of AKI staging and RRT on the prognosis of elderly septic patients with AKI. Multivariate Cox regression analysis was used to screen the risk factors of death in elderly patients with sepsis associated AKI. Results ① A total of 490 septic elderly patients were enrolled, including 249 patients with AKI and 241 patients without AKI, with the AKI incidence of 50.8%. Compared with non-AKI group, the patients in AKI group were older (years old 72.0±7.2 vs. 68.8±5.1), acute physiology and chronic health evaluation Ⅱ (APACHEⅡ) score and sequential organ failure assessment (SOFA) score were evidently higher (23.1±6.1 vs. 22.0±3.7, 9.4±3.8 vs. 6.1±3.5); the duration of mechanical ventilation [days 7.0 (5.0, 10.0) vs. 6.0 (3.0, 9.0)], length of ICU stay [days 12.0 (7.0, 15.0) vs. 7.0 (4.0, 13.0)] and total length of hospital stay [days 15.0 (10.0, 21.5) vs. 12.0 (7.0, 15.0)] were longer, and ICU mortality and 28-day mortality were evidently higher [22.9% (57/249) vs. 14.1% (34/241), 36.1% (90/249) vs. 24.5% (59/241), all P < 0.05]. ② According to KDIGO staging, 93 patients were in stage 1, 70 in stage 2 and 86 in stage 3 of AKI. The rate of RRT was increased with increase in KDIGO staging [14.0% (13/93), 30.0% (21/70), 88.4% (76/86)], the duration without mechanical ventilation within 28 days was shortened [days 20.0 (0, 23.0), 8.0 (0, 20.5), 8.0 (0, 13.0)], the rate of kidney recovery was decreased [71.0% (66/93), 51.4% (36/70), 37.2% (32/86)], meanwhile, the ICU and 28-day mortality was increased [12.9% (12/93), 38.6% (27/70), 20.9% (18/86), and 26.9% (25/93), 35.7% (25/70), 46.5% (40/86), all P < 0.05]. ③ 110 elderly septic patients with AKI were treated with RRT, and 139 without RRT. Compared with non-RRT group, the ratio of mechanical ventilation in RRT group was lowered [46.4% (51/110) vs. 68.3% (95/139)], the duration without mechanical ventilation within 28 days [days 18.0 (0, 23.0) vs. 10.0 (0, 13.0)], the length of ICU stay [days 13.0 (12.0, 17.9) vs. 10.0 (6.0, 14.0)] and the total length of hospital stay [days 22.5 (15.0, 46.0) vs. 16.0 (12.0, 23.0)] were prolonged, and the 28-day mortality was evidently increased [50.0% (55/110) vs. 25.2% (35/139), all P < 0.01], however, no significant difference in ICU mortality was found [27.3% (30/110) vs. 19.4% (27/139), P > 0.05]. ④ Cox regression analysis showed that SOFA score [relative risk (RR) = 1.214, 95% confidence interval (95%CI) = 1.117-1.319], KDIGO stage (RR = 4.077, 95%CI =1.850-8.982), vasoactive substance usage (RR = 2.896, 95%CI = 1.502-5.584), and mechanical ventilation (RR = 5.787, 95%CI = 1.512-22.156) were the risk factors of 28-day mortality in elderly septic patients with AKI (all P < 0.05). Conclusions The incidence of AKI for elderly septic patients with AKI was about 50%, who had a worse prognosis as compared with non-septic AKI patients. The higher the stage of KDIGO, the higher the mortality in elderly septic patients with AKI was. RRT can decrease the rate of mechanical ventilation, whereas, it may not improve the prognosis of elderly septic patients with AKI.