Variable | N^{a} | Ratio (95% CI) | z Test, P | TOST | |
---|---|---|---|---|---|

z_{1} P | z_{2}P | ||||

PPV^{b} | .001 | .94 | <.001 | ||

M-CHAT | 98 | 0.57 (0.47–0.67) | |||

PCP M-CHAT/F | 98 | 0.78 (0.69–0.86) | |||

PPV^{c} | .30 | .29 | .004 | ||

PCP M-CHAT/F | 98 | 0.78 (0.69–0.86) | |||

AC M-CHAT/F | 97 | 0.71 (0.62–0.80) | |||

Sensitivity^{c} | .38 | 0.30 | .01 | ||

PCP M-CHAT/F | 98 | 0.55 (0.46–0.65) | |||

AC M-CHAT/F | 97 | 0.49 (0.39–0.59) | |||

Specificity^{c} | .43 | .20 | .008 | ||

PCP M-CHAT/F | 98 | 0.79 (0.70–0.87) | |||

AC M-CHAT/F | 97 | 0.74 (0.65–0.83) | |||

Accuracy^{c} | .24 | .40 | .005 | ||

PCP M-CHAT/F | 98 | 0.60 (0.56–0.75) | |||

AC M-CHAT/F | 97 | 0.57 (0.47–0.67) |

Table includes suspected ASD/new phenotype. Comparisons were made with 2-tailed

*z*test and TOST, Δ = 0.1 unit of proportion. Only patients with a positive screen on the M-CHAT are included in this sample. CI, confidence interval.↵a One patient did not have a complete AC M-CHAT/F (Fig 1).

↵b Relevant difference.

↵c Equivalence.