Skip to main content
. 2018 Jun 28;214(1):4–10. doi: 10.1192/bjp.2018.122

Table 2.

Model performance in the test data-set

Time Outcome Model AUC Accuracy Sensitivity Specificity PPV NPV
Week 6 Response Baseline 0.60 0.59 0.67 0.51 0.59 0.59
Total improvement 0.73 0.70 0.81 0.59 0.67 0.74
Item improvement 0.77 0.70 0.74 0.66 0.70 0.71
Item interactions 0.77 0.70 0.72 0.67 0.70 0.70
Remission Baseline 0.64 0.68 0.09 0.98 0.69 0.68
Total improvement 0.74 0.71 0.30 0.92 0.64 0.72
Item improvement 0.78 0.74 0.44 0.89 0.66 0.76
Item interactions 0.78 0.74 0.43 0.89 0.66 0.76
Week 12 Response Baseline 0.62 0.69 1.00 0.00 0.69 N/A
Total improvement 0.67 0.72 0.93 0.27 0.73 0.62
Item improvement 0.71 0.71 0.90 0.29 0.73 0.57
Item interactions 0.70 0.70 0.90 0.29 0.73 0.56
Remission Baseline 0.62 0.59 0.61 0.57 0.60 0.58
Total improvement 0.68 0.64 0.69 0.59 0.64 0.64
Item improvement 0.74 0.67 0.69 0.64 0.67 0.66
Item interactions 0.73 0.66 0.65 0.68 0.68 0.65

AUC, area under the (receiver operating characteristic) curve; N/A, not applicable (undefined); NPV, negative predictive value; PPV, positive predictive value.