Cui Wei1,2, Gao Hongxia3, Fang Hui1,2, Qin Xianhui4, Jin Danqun5, Liu Haipeng6,7,8. 1. Department of Scientific Research and Education, Anhui Provincial Children's Hospital, Hefei, Anhui Province, China. 2. Anhui Institute of Pediatric Research, Hefei, Anhui Province, China. 3. Department of Gynecology and Obstetric, The First Affiliated Hospital of Anhui University of Chinese Medicine, Hefei, Anhui Province, China. 4. National Clinical Research Center for Kidney Disease, Guangzhou, Guangdong Province, China. 5. Pediatric Intensive Care Unit, Anhui Provincial Children's Hospital, Hefei, Anhui Province, China. 6. Department of Scientific Research and Education, Anhui Provincial Children's Hospital, Hefei, Anhui Province, China. itishaipeng@yeah.net. 7. Anhui Institute of Pediatric Research, Hefei, Anhui Province, China. itishaipeng@yeah.net. 8. Department of Scientific Research and Education, Anhui Provincial Children's Hospital, 39, Wangjiang Road, Baohe District, 230051, Hefei, Anhui, China. itishaipeng@yeah.net.
Abstract
BACKGROUND: The definition of pediatric AKI continues to evolve. We aimed to find a better AKI definition to predict outcomes and identify risk factors for AKI in a Chinese PICU. METHODS: This study consisted of 3338 patients hospitalized in a Chinese PICU between 2016 and 2018. AKI was defined and staged using pROCK criteria, which were compared with KDIGO criteria. AKI outcomes, including mortality, daily cost and length of stay (LOS), were assessed. Risk factors for AKI were also estimated. RESULTS: The incidence of AKI in the PICU was 7.7% according to pROCK criteria. The characteristics of patients with KDIGO-defined AKI who did not meet the pROCK were similar to those without AKI. pROCK outperformed KDIGO in predicting mortality with a higher c index in the Cox models (0.81 versus 0.79, P = 0.013). AKI, as well as AKI stages, were associated with higher mortality (HR: 10.5, 95%CI: 6.66-19.5), daily cost (β = 2064, P < 0.01) and LOS (β = 2.30, P < 0.01). Age, comorbidities, mechanical ventilation (MV), pediatric critical illness score (PCIS) and exposure to drugs had significant influence on AKI occurrence. CONCLUSIONS: The mortality predictability of pROCK was slightly greater than that of KDIGO. Older age, underlying comorbidities, MV, decreased PCIS and exposure to drugs were potential risk factors for AKI. IMPACT: Two AKI criteria, pROCK and KDIGO, were significantly associated with an increased risk of mortality and pROCK was slightly greater than that of KDIGO. Older age, comorbidities, mechanical ventilation, decreased PCIS and exposure to drugs were potential risk factors for AKI. This study first used the pROCK criteria to provide an epidemiologic description of pediatric AKI in Chinese PICU. This study compared the AKI outcomes across the pROCK and KDIGO AKI criteria, indicating the prior utility for AKI classification in Chinese children. This study indicated that the potential risk factors for AKI were older age, comorbidities, mechanical ventilation, decreased PCIS and exposure to drugs.
BACKGROUND: The definition of pediatric AKI continues to evolve. We aimed to find a better AKI definition to predict outcomes and identify risk factors for AKI in a Chinese PICU. METHODS: This study consisted of 3338 patients hospitalized in a Chinese PICU between 2016 and 2018. AKI was defined and staged using pROCK criteria, which were compared with KDIGO criteria. AKI outcomes, including mortality, daily cost and length of stay (LOS), were assessed. Risk factors for AKI were also estimated. RESULTS: The incidence of AKI in the PICU was 7.7% according to pROCK criteria. The characteristics of patients with KDIGO-defined AKI who did not meet the pROCK were similar to those without AKI. pROCK outperformed KDIGO in predicting mortality with a higher c index in the Cox models (0.81 versus 0.79, P = 0.013). AKI, as well as AKI stages, were associated with higher mortality (HR: 10.5, 95%CI: 6.66-19.5), daily cost (β = 2064, P < 0.01) and LOS (β = 2.30, P < 0.01). Age, comorbidities, mechanical ventilation (MV), pediatric critical illness score (PCIS) and exposure to drugs had significant influence on AKI occurrence. CONCLUSIONS: The mortality predictability of pROCK was slightly greater than that of KDIGO. Older age, underlying comorbidities, MV, decreased PCIS and exposure to drugs were potential risk factors for AKI. IMPACT: Two AKI criteria, pROCK and KDIGO, were significantly associated with an increased risk of mortality and pROCK was slightly greater than that of KDIGO. Older age, comorbidities, mechanical ventilation, decreased PCIS and exposure to drugs were potential risk factors for AKI. This study first used the pROCK criteria to provide an epidemiologic description of pediatric AKI in Chinese PICU. This study compared the AKI outcomes across the pROCK and KDIGO AKI criteria, indicating the prior utility for AKI classification in Chinese children. This study indicated that the potential risk factors for AKI were older age, comorbidities, mechanical ventilation, decreased PCIS and exposure to drugs.