Table 2.

Performance characteristics of computable phenotype algorithm across PEDSnet population with three or more nephrology encounters

CenterSample Cases/NoncasesSensitivity (95% CI)Specificity (95% CI)PPV (95% CI)NPV (95% CI)Accuracy (95% CI)AUC (95% CI)
187/10096 (89 to 99)98 (93 to 100)98 (92 to 100)96 (90 to 99)97 (93 to 99)97 (94 to 99)
249/10098 (88 to 100)96 (90 to 99)92 (80 to 98)99 (95 to 100)97 (92 to 99)97 (94 to 100)
374/100100 (94 to 100)85 (78 to 91)77 (66 to 86)100 (96 to 100)90 (85 to 94)93 (90 to 96)
460/10096 (86 to 100)89 (82 to 94)80 (68 to 89)98 (93 to 100)91 (86 to 95)93 (89 to 97)
574/10097 (90 to 100)93 (87 to 97)91 (81 to 96)98 (93 to 100)95 (90 to 98)95 (92 to 98)
683/9997 (90 to 100)89 (82 to 94)86 (76 to 92)98 (93 to 100)92 (87 to 96)93 (90 to 97)
746/9983 (70 to 92)98 (92 to 100)96 (85 to 99)91 (83 to 96)92 (87 to 96)90 (85 to 96)
875/10097 (90 to 100)95 (89 to 98)93 (85 to 98)98 (93 to 100)96 (92 to 98)96 (93 to 99)
All548/79896 (94 to 97)93 (91 to 94)89 (86 to 91)97 (96 to 98)94 (92 to 95)94 (93 to 95)
  • AUC, area under the receiver operating characteristics curve.