The Difference between the Two Representative Kampo Formulas for Treating Dysmenorrhea: An Observational Study

Figure 2

Model calibration using a receiver operating characteristic curve. The full model (broken line) included 9 predictive variables, and the final model (black line) included 7 predictive variables. Each univariate model was drawn using dotted lines. The final model had an area under the curve of 0.88 (95% CI: 0.81–0.93), and the full model had an area under the curve of 0.88 (95% CI: 0.82–0.94). The univariate models had areas under the curve of 0.56–0.71 (95% CI: not shown).