To elucidate the relationship between antofloxacin (AT) plasma concentration and QT interval prolongation, compare the effects of different correction and analytical methods on conclusions, and estimate the possible false-positive rate in thorough QT (TQT) studies.

Twenty-four healthy Chinese volunteers from a four-period crossover TQT study orally received 200 mg/d AT, 400 mg/d AT, 400 mg/d moxifloxacin, and a placebo in a random order for 5 d for each. QT interval samples were collected on d 1 and d 5. Population models were established describing the relationship between QT and AT concentration. The yardstick from ICH E14 guidelines was used to measure the effect of drugs on QT prolongation both in biostatistical and modeling analyses. A possible false-positive rate was estimated by constructing a 1000-time bootstrap to obtain the rate-of-difference values between d 1 and d 5 over 5 ms in the placebo period.

In the modeling analysis, the QT prolongation estimate at the mean maximal concentration of AT (4.51 μg/mL) was 3.84 ms, and its upper bound of the one-sided 95 % CI was 7.04 ms, which showed a negative effect on QT interval prolongation. The estimation for the false-positive rate was 31 % in this study.

The effect of AT on QT interval prolongation may not have been significant at the dosage of 400 mg. Baseline and placebo adjustments were necessary in TQT studies. Population modeling has demonstrated clear superiority in making full use of data to accurately analyze the relationship between drugs and QT intervals.