Screening tool
Figure S1 illustrates the shift in test classification measures at
varying cut-off probabilities for the test positivity criterion. The
optimal combination of sensitivity and specificity was at a cut-off
point of 0.700. Application of the cut-off to the study population
resulted in the following classification measures: sensitivity 80.9%,
specificity 77.1%, PPV 96.8%, NPV 62.2% (Table 4). We created a
graphically designed screening tool with color coded risk scores based
on the prediction model (Figure 4). Green cells indicate low
probabilities of having an inadequate calcium intake, orange cells
indicate medium high probabilities, red cells indicate high
probabilities. User instructions are enclosed with the screening tool.