AUC Curves Harmony Data
Giovanna Saracino
28 May, 2024
1 AUC Curves
I used Harmony Data (from output 2x2 matrices)
I calculated AUC Curves using True Positive(TP) True Negative(TN) False Positive(FP) False Negative(NP)
2 AUC GRAIL Q10
##
## Detailed Sensitivity and Specitivity Output
##
## Input Matrix:
## 1=Event 0=No-Event
## ≤30 197 1319
## >30 27 306
##
## The sample sensitivity is: 87.9%
## 95% Confidence Limits for true sensitivity are: [83.7, 92.2]
##
## The sample of specificity is: 18.8%
## 95% Confidence Limits for true specificity are: [16.9, 20.7]
##
## The sample value of Youden's J is: 6.8
## 95% Confidence Limits for Youden's J are: [2.1, 11.4]
##
## Sample value for Percent Agreement (PA) is: 27.2%
## 95% Confidence Limits for PA are: [25.2, 29.2]
## ## [1] "AUC - Grail (Q10):"## [1] 0.533886
3 AUC GRAIL (Median)
##
## Detailed Sensitivity and Specitivity Output
##
## Input Matrix:
## 1=Event 0=No-Event
## ≤30 172 927
## >30 52 698
##
## The sample sensitivity is: 76.8%
## 95% Confidence Limits for true sensitivity are: [71.3, 82.3]
##
## The sample of specificity is: 43%
## 95% Confidence Limits for true specificity are: [40.5, 45.4]
##
## The sample value of Youden's J is: 19.7
## 95% Confidence Limits for Youden's J are: [13.7, 25.8]
##
## Sample value for Percent Agreement (PA) is: 47.1%
## 95% Confidence Limits for PA are: [44.8, 49.3]
## ## [1] "AUC - Grail (Median):"## [1] 0.5986978
4 CKD-Epi RF
##
## Detailed Sensitivity and Specitivity Output
##
## Input Matrix:
## 1=Event 0=No-Event
## ≤30 142 630
## >30 82 995
##
## The sample sensitivity is: 63.4%
## 95% Confidence Limits for true sensitivity are: [57.1, 69.7]
##
## The sample of specificity is: 61.2%
## 95% Confidence Limits for true specificity are: [58.9, 63.6]
##
## The sample value of Youden's J is: 24.6
## 95% Confidence Limits for Youden's J are: [17.9, 31.4]
##
## Sample value for Percent Agreement (PA) is: 61.5%
## 95% Confidence Limits for PA are: [59.3, 63.7]
## ## [1] "AUC - CKD-Epi:"## [1] 0.6231181