Skip to main content
. 2023 Feb 20;177(4):373–383. doi: 10.1001/jamapediatrics.2022.5975

Table 2. Sensitivity and Specificity of M-CHAT(-R/F) Overall and by Study Characteristics.

Characteristic No. Sens Spec DOR (95% CI) −2LL P value
Overall 49a 0.83 0.94 75.40 (35.56-159.87) NA NA
M-CHAT version
M-CHAT 31 0.83 0.94 63.47 (25.38-158.73) 1.35 .72
M-CHAT-R 18 0.83 0.95 102.69 (27.87-378.26)
Likelihood level
Low 27 0.83 0.99 334.78 (156.69-715.31) 58.28 <.001b
High 22 0.82 0.70 10.93 (4.45-26.83)
Case confirmation strategy
Concurrent 40 0.86 0.93 85.84 (37.56-196.16) 17.66 .001b
Prospective 9 0.57 0.97 38.71 (8.38-178.82)
Sample size
<500 20 0.83 0.78 17.31 (6.48-46.21) 38.26 <.001b
500-5000 17 0.80 0.96 88.94 (29.58-267.43)
>5000 12 0.85 0.99 553.41 (159.34-1922.06)
Follow-up
Initial only 22 0.82 0.84 24.61 (9.01-67.23) 11.29 .01b
Follow-up 27 0.83 0.97 182.87 (71.12-470.22)
Language
English or primarily English 26 0.78 0.85 20.19 (7.85-51.91) 23.96 <.001b
Other 23 0.89 0.98 361.76 (145.80-897.58)

Abbreviations: DOR, diagnostic odds ratio; M-CHAT, Modified Checklist for Autism in Toddlers; M-CHAT-R, Modified Checklist for Autism in Toddlers, Revised; NA, not applicable; Sens, sensitivity; Spec, specificity; −2LL, −2 log likelihood difference.

a

Studies classified as “mixed” or “unknown” in any category were excluded from the analysis (n = 2).57,58 When the overall model was run with the full set of 51 studies, sensitivity was only slightly different, specificity was identical, and DOR was slightly higher (78.71 [95% CI, 38.11-162.57]). Therefore, for comparison across the subanalyses, the common set of 49 studies is reported.

b

P < .05.