Skip to main content

Table 5 Numerical results for the single-feature logistic regression model

From: A deep-learning model to continuously predict severe acute kidney injury based on urine output changes in critically ill patients

Window (h) Threshold (ml/h/kg) Sensivity Specificity Precision LR+ LR− auROC
2 0.251 0.733 0.733 0.008 2.745 0.364 0.786
3 0.288 0.733 0.733 0.008 2.745 0.364 0.786
4 0.311 0.739 0.741 0.009 2.853 0.352 0.798
5 0.341 0.738 0.741 0.01 2.849 0.354 0.804
6 0.362 0.739 0.741 0.011 2.853 0.352 0.810
7 0.372 0.755 0.759 0.012 3.133 0.323 0.813
8 0.407 0.743 0.741 0.012 2.869 0.347 0.817
9 0.427 0.749 0.75 0.012 2.996 0.335 0.815
10 0.457 0.744 0.741 0.013 2.873 0.345 0.817
11 0.471 0.75 0.75 0.012 3 0.333 0.818
12 0.487 0.753 0.75 0.013 3.012 0.329 0.817