Skip to main content
. 2021 Jun;147(6):2249–2262.e7. doi: 10.1016/j.jaci.2021.01.025

Table II.

Proportions of participants with a change (or no change) in threshold, (overall and cohorted into those with lower reaction thresholds to peanut), by IPD meta-analysis results

Cumulative reaction threshold at initial challenge (mg of peanut protein) At IPD meta-analysis, the proportion of participants (and 95% CI) with
Increase in threshold
No change in threshold Decrease in threshold
±Max half-log change ± Max 1-log change
> Half-log Any Any > Half-log
Any (n = 534) 18.3% (11.0-28.9) 35.5% (26.1-46.3) 32.2% (25.3-40.0) 30.3% (24.5-36.9) 8.2% (4.5-14.5) 71.2% (56.2-82.6) 91.2% (84.1-95.3)
>150 mg (ie, no objective symptoms in response to ½ peanut) (n = 113) 3.2% (0.5-16.5) 15.6% (10.3-23.9) 28.3% (20.8-37.3) 59.6% (42.1-75.0) 11.9% (6.2-21.6) 85.0% (77.1-90.4) 96.5% (90.4-98.8)
<150 mg (n = 423) 22.5% (14.5-33.2) 40.1% (30.7-50.2) 34.5% (25.6-44.7) 23.0% (19.3-27.3) 7.0% (3.5-13.6) 68.1% (53.9-79.7) 89.9% (82.9-94.2)
<50 mg (n = 213) 35.7% (28.1-44.1) 56.6% (47.0-65.8) 28.7% (21.8-36.7) 15.0% (10.8-20.5) 6.6% (3.9-10.9) 56.9% (47.7-65.6) 81.7% (75.9-86.3)
≤15 mg (n = 101) 50% (38-62) 72% (63-80) 21% (12-35) 5.0% (2.1-11) 2.8% (0.7-11) 44% (30-59) 68% (59-77)
≤5 mg (n = 56) 63% (47-77) 82% (70-90) 10% (2.6-31) 5.4% (1.7-15) 5.4% (1.7-15) 24% (8-53) 61% (47-73)

max, Maximum.

A half-log or 1-log change in threshold is equivalent to a shift in reaction threshold by 1 or 2 dosing increments when a challenge protocol based on PRACTALL is used. Boldface highlights "no change in threshold" and "+/- max 1/2-log change" which are arguably the most important outcome measures.