A SCRAM alcohol curve can peak well after breath alcohol peaks. The two instruments sample different signals, and the delay depends on what part of each curve you compare.
In a controlled study of 30 participants assigned to alcohol or a nonalcoholic beverage, Fairbairn and Kang reported that the SCRAM peak occurred 120 minutes after the breath alcohol peak in the analyzed data. A separate time-series analysis estimated a 69-minute average lag across the breath alcohol curve. Peak-to-peak delay and average curve lag answer different questions.
Mark the first sustained rise, the measured peak and the last reading separately. Place them beside any independently documented drink times or breath samples. A population-average lag does not give you a reliable fixed correction for an individual event.
Fairbairn and Kang, 2019 laboratory study of transdermal alcohol timing