Study population
During the study period, 9,079 patients were admitted consecutively. Of them, patients were excluded because of the following reasons: 5,725 patients had an ICU LOS of less than 24 hours, 110 patients were younger than 18 years old, one patient received renal transplantation during the past 3 months, 95 patients had received RRT before admission to the ICU, and 11 had insufficient clinical recordings. Thus, 3,107 patients were enrolled in the BAKIT study. Of these patients, 194 were excluded because of incomplete data for calculating OASIS, and finally, 2,954 patients were included in our study (Fig. 1).
The characteristics of the entire cohort are shown in Table 1. The median age was 64 (IQR:51–76) years, and 61.6% were men. The all-cause 28-day mortality rate was 17.0%, and the median ICU LOS was 4 (IQR:2–9) days. Among the included patients, the median OASIS was 28 (IQR:23–36), the median APACHE II score was 14 (IQR:10–20), the median SAPS II was 6 (IQR:3–8) and the median SOFA score was 6 (IQR:3–8). MV was used in 1,960 (66.4%) patients, 1,230 patients (41.6%) received vasopressors, 1,506 (51.0%) patients developed AKI as defined by the KDIGO criteria, and 252 patients (8.5%) underwent RRT.
Table 1
Patient characteristics by 28-day mortality
Characteristic
|
All patients (n = 2954)
Median(IQR)
Number (%)
|
Survivors
(n = 2453)
Median(IQR)
Number (%)
|
Nonsurvivors
(n = 501)
Median(IQR)
Number (%)
|
P
value
|
Age(years)
|
64(51–76)
|
63(50–75)
|
72(59–81)
|
< 0.001
|
Male sex
|
1819(61.6)
|
1518(61.9)
|
301(60.1)
|
0.751
|
ICU course
|
|
|
|
|
Vasoactive therapy
|
1230(41.6)
|
1033(42.1)
|
197(39.3)
|
0.457
|
MV
|
1960(66.4)
|
1591(64.9)
|
369(73.7)
|
< 0.001
|
Sepsis
|
848(28.7)
|
540(22.0)
|
308(61.5)
|
< 0.001
|
AKI
|
1506(51.0)
|
1099(44.8)
|
407(81.2)
|
< 0.001
|
RRT
|
252(8.5)
|
127(5.2)
|
125(25.0)
|
< 0.001
|
Severity of illness
|
|
|
|
|
OASIS
|
28(23–36)
|
27(22–33)
|
38(31–45)
|
< 0.001
|
APACHEII
|
14(10–20)
|
13(9–18)
|
22(17–28)
|
< 0.001
|
SAPSII
|
33(25–44)
|
31(24–40)
|
50(39–63)
|
< 0.001
|
SOFA
|
6(3–8)
|
5(3–7)
|
9(6–11)
|
< 0.001
|
Admission category
|
|
|
|
|
Urgent
|
427(14.5)
|
367(15.0)
|
60(12.0)
|
< 0.001
|
Emergency
|
1068(36.2)
|
732(29.8)
|
336(67.1)
|
Elective
|
1459(49.4)
|
1354(53.2)
|
105(20.9)
|
Comorbidities
|
|
|
|
|
Hypertension
|
1176(39.8)
|
949(37.3)
|
227(45.3)
|
|
Coronary heart disease
|
569(19.3)
|
427(17.4)
|
142(28.3)
|
|
Congestive heart failure
|
188(6.4)
|
113(4.6)
|
75(15.0)
|
|
COPD
|
158(5.3)
|
120(4.9)
|
38(7.6)
|
|
Diabetes
|
511(17.3)
|
418(17.0)
|
93(18.6)
|
|
Chronic kidney disease
|
151(5.1)
|
108(4.4)
|
43(8.6)
|
|
Liver disease
|
82(2.8)
|
63(2.6)
|
19(3.8)
|
|
Cancer
|
407(13.8)
|
352(14.3)
|
55(11.0)
|
|
Hematological disease
|
24(0.8)
|
13(0.5)
|
11(2.2)
|
|
Category of ICU admission diagnosis
|
|
|
|
|
Cardiovascular
|
820(27.8)
|
733(28.8)
|
87(17.4)
|
|
Respiratory
|
516(17.5)
|
356(14.0)
|
160(31.9)
|
|
Neurologic
|
436(14.8)
|
337(13.3)
|
99(19.8)
|
|
Trauma
|
225(7.6)
|
203(8.0)
|
22(4.4)
|
|
Gastrointestinal
|
578(19.6)
|
485(19.1)
|
93(18.6)
|
|
Metabolic
|
66(2.2)
|
54(2.1)
|
12(2.4)
|
|
Outcomes
|
|
|
|
|
ICU LOS(days)
|
4(2–9)
|
4(2–7)
|
6(3–13)
|
< 0.001
|
Hospital LOS(days)
|
19(12–29)
|
19(12–28)
|
21(11–34)
|
0.002
|
Data are expressed as the median (interquartile range, IQR), and number (percentage). MV, mechanical ventilation; AKI, acute kidney injury; RRT, renal replacement therapy; OASIS, the Oxford Acute Severity of Illness Score; APACHE II, Acute Physiology and Chronic Health Evaluation II; SAPS II, Simplified Acute Physiology Score II; SOFA, Sequential Organ Failure Assessment; COPD, chronic obstructive pulmonary disease; LOS, length of stay. |
There were statistically significant differences in age, MV, sepsis, AKI, RRT, OASIS, APACHE II, SAPS II, SOFA, admission category, ICU LOS and hospital LOS between survivors and nonsurvivors.
Comparison of characteristics between the survival and nonsurvival groups of AKI patients
AKI patient characteristics according to 28-day mortality are shown in Table 2. Nonsurviving AKI patients were older (P < 0.001), had higher illness severity scores and were more likely to be diagnosed with sepsis. Positive fluid balance in the first 24 hours was more common among nonsurvivors.
Table 2
AKI patient characteristics by 28-day mortality
Characteristic
|
AKI patients
(n = 1506)
Median(IQR)
Number (%)
|
Survivors
(n = 1099)
Median (IQR)
Number (%)
|
Non-survivors
(n = 407)
Median (IQR)
Number (%)
|
P
value
|
Age (years)
|
67(53–78)
|
64(51–77)
|
74(59–81)
|
< 0.001
|
Male gender
|
918(61.0)
|
674(61.3)
|
244(60.0)
|
0.886
|
Baseline creatinine (µmol/L)
|
84.0(71.6–97.0)
|
83.4(71.0–97.0)
|
85.0(75.0-97.3)
|
0.685
|
Severity of illness
|
|
|
|
|
APACHEII
|
17(12–23)
|
15(10–20)
|
23(18–29)
|
< 0.001
|
SAPSII
|
39(30–52)
|
35(27–45)
|
52(41–65)
|
< 0.001
|
SOFA
|
7(4–10)
|
6(4–9)
|
9(6–12)
|
< 0.001
|
OASIS
|
31(24–39)
|
28(23–35)
|
39(32–46)
|
< 0.001
|
ICU course
|
|
|
|
|
Vasoactive therapy
|
622(41.3)
|
451(41.0)
|
171(42.0)
|
0.942
|
MV
|
1052(69.9)
|
748(68.1)
|
304(74.7)
|
0.045
|
sepsis
|
603(40.0)
|
335(30.5)
|
268(65.8)
|
< 0.001
|
Positive fluid balance first 24 hours
|
1083(71.9)
|
743(67.6)
|
340(83.5)
|
< 0.001
|
Use of diuretics on the first day of admission
|
510(33.9)
|
368(33.5)
|
142(34.9)
|
0.876
|
Staging of AKI 1
|
699(46.4)
|
592 (53.9)
|
107(26.3)
|
|
2
|
357(23.7)
|
260(23.7)
|
97(23.8)
|
< 0.001
|
3
|
450(29.9)
|
247(22.5)
|
203(49.9)
|
|
RRT
|
241(16.0)
|
121(11.0)
|
120(29.5)
|
< 0.001
|
Outcomes
|
|
|
|
|
Hospital LOS(days)
|
20(11–30)
|
22(14–34)
|
13(6–23)
|
< 0.001
|
ICU LOS(days)
|
5(3–11)
|
5(3–11)
|
6(4–12)
|
0.030
|
Data are expressed as the median (interquartile range,IQR), and number (percentage). AKI, acute kidney injury; OASIS, the Oxford Acute Severity of Illness Score; SAPS II, Simplified Acute Physiology Score II; SOFA, Sequential Organ Failure Assessment; APACHE II, Acute Physiology and Chronic Health Evaluation II; MV, mechanical ventilation; LOS, length of stay; RRT, renal replacement therapy. |
The 28-day Mortality Of Aki Patients According To Oasis
The distribution of OASIS in AKI patients is shown in Fig. 2. OASIS ranged from 6 to 64, and the median value was 31 (IQR: 24, 39). The distributions of the OASIS with corresponding 28-day mortality are also presented in Fig. 2. As each score increased, the 28-day mortality of AKI patients increased accordingly, indicating more serious illness and worse prognosis.
Comparison of ROC curve and AUCs of the four scoring systems in evaluating the 28-day mortality of AKI patients
In Fig. 3, OASIS had the highest discriminatory power in predicting the prognosis of AKI patients. The AUC values of OASIS, APACHE II and SAPS II in predicting 28-day mortality were 0.771 (95% CI, [0.742, 0.799]), 0.764 (95% CI, [0.735, 0.792]), and 0.767 (95% CI, [0.739, 0.796]), respectively, which were higher than that of SOFA (0.686; P < 0.001). Table 3 shows the pairwise comparison of the ROC curves, and there were no statistically significant differences between the AUC values of OASIS, APACHE II, and SAPS II in predicting 28-day mortality.
Table 3
Pairwise comparison of ROC curves for predicting 28-day mortality in AKI patients
Variables
|
Difference between areas
|
Standard Error
|
Z
|
P value
|
APACHEII∼OASIS
|
0.00706
|
0.0115
|
0.612
|
0.5408
|
APACHEII∼SAPSII
|
0.00377
|
0.0112
|
0.336
|
0.737
|
APACHEII∼SOFA
|
0.0774
|
0.0155
|
4.995
|
< 0.0001
|
OASIS∼SAPSII
|
0.00329
|
0.0125
|
0.264
|
0.7917
|
OASIS∼SOFA
|
0.0845
|
0.0172
|
4.91
|
< 0.0001
|
SAPSII∼SOFA
|
0.0812
|
0.0167
|
4.864
|
< 0.0001
|
OASIS, the Oxford Acute Severity of Illness Score; APACHE II, the Acute Physiology and Chronic Health Evaluation II; SOFA, the Sequential Organ Failure Assessment; SAPS II, the Simplified Acute Physiology Score II. |
The predictive ability of OASIS, APACHE II, SAPS II and SOFA score for poor outcomes
The ROC curves for the prediction of 28-day mortality, ICU mortality and in-hospital mortality by each severity scale are shown in Table 4. The sensitivities, specificities, positive predictive values, and negative predictive values of the optimal cut-off values (from the Youden index) for each scale to predict the three outcomes are listed in Table 5. The cut-off value of OASIS for the prediction of 28-day mortality was 33 with a sensitivity of 87.75%and specificity of 46.26%, as calculated by the ROC curve analysis. OASIS ≥ 33 predicts poor short-term prognosis in patients with AKI.
Table 4
Area under the curve of various parameters for predicting poor outcomes in AKI patients
Severity of illness
|
AUC
|
Standard error
|
P-Value
|
95%confidence interval
|
Lower limit Upper limit
|
28-day mortality
|
|
|
|
|
OASIS
|
0.771
|
0.015
|
0.000
|
0.742 0.799
|
APACHEII
|
0.764
|
0.014
|
0.000
|
0.735 0.792
|
SAPSII
|
0.767
|
0.015
|
0.000
|
0.739 0.796
|
SOFA
|
0.686
|
0.017
|
0.000
|
0.653 0.719
|
ICU mortality
|
|
|
|
|
OASIS
|
0.804
|
0.014
|
0.000
|
0.777 0.832
|
APACHEII
|
0.800
|
0.014
|
0.000
|
0.773 0.827
|
SAPSII
|
0.801
|
0.014
|
0.000
|
0.774 0.829
|
SOFA
|
0.689
|
0.018
|
0.000
|
0.654 0.724
|
Hospital mortality
|
|
|
|
|
OASIS
|
0.783
|
0.014
|
0.000
|
0.756 0.811
|
APACHEII
|
0.776
|
0.014
|
0.000
|
0.748 0.804
|
SAPSII
|
0.784
|
0.014
|
0.000
|
0.757 0.811
|
SOFA
|
0.676
|
0.017
|
0.000
|
0.643 0.709
|
AUC, area under the receiver operating characteristic curve; OASIS, the Oxford Acute Severity of Illness Score; APACHE II, the Acute Physiology and Chronic Health Evaluation II; SOFA, the Sequential Organ Failure Assessment; SAPS II, the Simplified Acute Physiology Score II. |
Table 5
Performance of multivariable models for predicting poor outcomes in AKI patients
Severity of illness
|
YI
|
Cutoff
|
Sen%
|
Spe%
|
+LR
|
-LR
|
PPV
|
NPV
|
28-day mortality
|
|
|
|
|
|
|
|
|
APACHEII
|
0.3976
|
17
|
63.02
|
76.73
|
2.71
|
0.48
|
87.2
|
45.3
|
SAPSII
|
0.4142
|
39
|
61.37
|
80.06
|
3.08
|
0.48
|
88.5
|
45.2
|
SOFA
|
0.2885
|
7
|
60.15
|
68.70
|
1.92
|
0.58
|
82.8
|
40.7
|
OASIS
|
0.4305
|
33
|
87.75
|
46.26
|
2.55
|
0.40
|
86.5
|
49.6
|
ICU mortality
|
|
|
|
|
|
|
|
|
APACHEII
|
0.4562
|
16
|
57.95
|
87.67
|
4.70
|
0.48
|
94.0
|
38.4
|
SAPSII
|
0.4794
|
39
|
60.62
|
87.33
|
4.78
|
0.45
|
94.1
|
39.9
|
SOFA
|
0.2877
|
7
|
58.56
|
70.21
|
1.97
|
0.59
|
86.8
|
33.7
|
OASIS
|
0.4930
|
33
|
69.85
|
79.45
|
3.40
|
0.38
|
91.9
|
44.1
|
Hospital mortality
|
|
|
|
|
|
|
|
|
APACHEII
|
0.4200
|
16
|
59.53
|
82.47
|
3.40
|
0.49
|
89.4
|
45.2
|
SAPSII
|
0.4499
|
39
|
62.53
|
82.47
|
3.57
|
0.45
|
89.8
|
47.1
|
SOFA
|
0.2677
|
7
|
59.65
|
67.12
|
1.81
|
0.6
|
81.8
|
40.2
|
OASIS
|
0.4636
|
33
|
71.84
|
74.52
|
2.82
|
0.38
|
87.4
|
51.7
|
YI, Youden’s index; Sen, sensitivity; Spe, specificity; LR+, positive likelihood ratio; LR−, negative likelihood ratio; NPV, negative predictive value; PPV, positive predictive value; OASIS, the Oxford Acute Severity of Illness Score; APACHE II, the Acute Physiology and Chronic Health Evaluation II; SOFA, the Sequential Organ Failure Assessment; SAPS II, the Simplified Acute Physiology Score II. |
Logistic Regression Analyses Of 28-day Mortality In Aki Patients
Logistic regression model was used to test the efficacy of OASIS in predicting 28-day mortality in patients with AKI (Table 6). Because OASIS is collinear with APACHE II and SAPS II, variables considered for multivariable analysis included age, sex, OASIS, SOFA, use of vasoactive drugs, MV, RRT and underlying diseases. OASIS was entered as a continuous variable and a categorical variable (the cut-off value of OASIS was 33), respectively. Multivariable analysis showed that 28-day mortality increased by 8.5% (95% CI, 1.065–1.106) for every point increase in the OASIS, and the 28-day mortality of patients with high OASIS was 3.826 times higher than that of patients with low OASIS. In addition to OASIS, sepsis (OR, 1.823; 95% CI, 1.339–2.481), RRT (OR, 1.802; 95% CI, 1.263–2.570), old age (OR, 1.012; 95% CI, 1.003–1.021), higher SOFA score (OR, 1.091; 95% CI, 1.042–1.141) and MV (OR, 2.016; 95% CI, 1.416–2.871) were significantly associated with a higher risk of death in multivariable analysis.
Table 6
Logistic regression analyses of 28-day mortality in AKI patients
variable
|
P-value
|
OR (95CI%)
|
variable
|
P-value
|
OR (95CI%)
|
Sepsis
|
0.000
|
1.823(1.339–2.481)
|
Sepsis
|
0.000
|
2.002(1.479–2.711)
|
RRT
|
0.000
|
1.802(1.263–2.570)
|
RRT
|
0.002
|
1.755(1.232-2.500)
|
Age
|
0.013
|
1.012(1.003–1.021)
|
Age
|
0.001
|
1.016(1.007–1.025)
|
OASISa
|
0.000
|
1.085(1.065–1.106)
|
OASISb
|
0.000
|
3.826(2.724–5.326)
|
SOFA
|
0.000
|
1.091(1.042–1.141)
|
SOFA
|
0.000
|
1.123(1.075–1.173)
|
MV
|
0.000
|
2.016(1.416–2.871)
|
MV
|
0.001
|
1.779(1.251–2.529)
|
Multivariable logistic regression to assess the association of OASIS with 28-day mortality. |
a. OASIS was entered as a continuous variable. |
b. OASIS was entered as a categorical variable, the cut-off value of OASIS was 33. |
AKI, acute kidney injury; RRT, renal replacement therapy; OASIS, the Oxford Acute Severity of Illness Score; SOFA, Sequential Organ Failure Assessment; MV, mechanical ventilation; OR, odds ratio; CI, confidence interval. |
Subgroup Analyses
According to the KDIGO criteria, AKI patients were divided into stage 1, stage 2 and stage 3. Patient characteristics by AKI stage are shown in Table S2. The ROC curves of OASIS, APACHE II, SAPS II and SOFA score for predicting of 28-day mortality in each subgroup are shown in Table 7. OASIS had a good predictive effect in each subgroup. Table 7 shows the calibration of the risk scores. OASIS had good calibration in each subgroup, except the stage 3 subgroup. To verify the predictive effect of OASIS on the 28-day mortality among different populations of AKI patients, we divided the AKI patients into elective surgery, nonelective surgery, MV, non-MV, Sepsis, nonsepsis, RRT, non-RRT, over 65 years and up to 65 years groups, as shown in Table S3. OASIS had a good prediction effect in most subgroups.
Table 7
Receiver operating characteristic curves of risk scores for predicting of 28-day mortality of each subgroup according to KDIGO criteria in AKI patients. * Denotes P > 0.05 for Hosmer-Lemeshow test, reflecting good calibration. The risk score column in bold has the highest AUC value for each group.
Group
|
APACHEII
|
SAPSII
|
SOFA
|
OASIS
|
Stage1
|
0.780*
|
0.803*
|
0.657*
|
0.767*
|
Stage2
|
0.728*
|
0.707*
|
0.619*
|
0.765*
|
Stage3
|
0.675*
|
0.709
|
0.649*
|
0.730
|
AKI, acute kidney injury; AUC, area under the receiver operating characteristic curve; APACHE II, the Acute Physiology and Chronic Health Evaluation II; SAPS II, the Simplified Acute Physiology Score II; SOFA, the Sequential Organ Failure Assessment; OASIS, the Oxford Acute Severity of Illness Score. |