目的: In this study, we use the precise time interval to definite the “long duration RRT”, aiming to explore the effect of long-duration RRT on outcome of critically ill patients with AKI.

方法: This retrospective study was conducted by using the Multiparameter Intelligent Monitoring in Intensive Care II (MIMIC-II) database, version 3.0. The MIMIC-II database (version 3.0) is a publicly available clinical database developed by the Massachusetts Institute of Technology (MIT), Phillips Healthcare, and Beth Israel Deaconess Medical Center (BIDMC). It covers data of patients in ICUs of BIDMC from 2001 to 2012. The total counts of MIMIC-II (version 3.0) contains around 48,000 critically ill patients [6]. Inclusion criteria for the present study were as follows: (1) age≥18 years;(2) admitted to ICU for the first time; (3) received RRT treatment;(4) suffered from AKI. Exclusion criteria: (1) the duration of RRT was less than 24 hours; (2) the diagnosis include end stage renal disease (ESRD) or chronic kidney disease(CKD) without AKI. AKI and CKD are defined according to the International Classification of Diseases, 9th Edition ICD-9 codes. We analyzed the duration of RRT among all patients in this study, the median is 100.5 hours. Therefore, long-duration RRT is defined as the time over or equal to 120 hours (5 days). The primary outcome measure of this study was the survival rate at 28-day after initiation of RRT. The secondary outcomes measure were the survival rates at 60-day,90-day and 180-day after initiation of RRT. Meanwhile, ICU mortality and hospital mortality, the length of ICU stay and length of hospital stay are compared between the two groups as well.

结果: We extract 1020 patients in total who received RRT treatment from the MIMIC-Ⅱdatabase. After the elimination of patients coincide with the exclusion criterion, there are 528 patients with AKI. Finally, 308 AKI patients in the non-long-duration RRT group, while other 220 AKI patients in the long-duration RRT group. There are no differences in demographic characteristics and vital signs among these two groups, including the SAPS score [24(20,28) vs 24.5(21.3,28), p=0.341]. From the survival analysis, it is obvious that the long-duration RRT could not only improve the primary outcome of 28-day mortality, but also could reduce the mortality at 60-day,90-day and 180-day (Figure 1). However, as displayed by Table 1, except the 28-day all-cause mortality, the difference is nonexistent between non-long-duration and long-duration RRT groups at 60-day,90-day and 180-day all-cause mortality. This is a very interesting result. Based on the current statistical results, we analyzed the causes of this difference. Compared with all-cause mortality, the survival analysis takes survival time and patients’ outcome into the effect to AKI patients’ mortality, and the former only focused on the mortality at a specific point of time. Obviously, the long-duration RRT could prolong the survival time of AKI patients and increase the 28-day survival rate, but the final long-term outcome could not be improved, such as 60-day,90-day and 180-day mortality. In addition, the ICU mortality and hospital mortality, the length of ICU stay and length of hospital stay were not improved by long-duration RRT treatment.

结论: Long-duration RRT is a useful treatment to prolong the survival time and reduce mortality of AKI patients at 28-day, 60-day, 90-day and 180-day. It also could improve the short-term all-cause mortality (28-day), but the final long-term outcome could not be improved. Therefore, the long-duration RRT may could not change the final outcomes of AKI patients.