How to extrapolate a withdrawal time from an EHSLC published detection time: a Monte Carlo simulation appraisal.
Authors: Toutain P-L
Journal: Equine veterinary journal
Summary
# Editorial Summary: Extrapolating Withdrawal Times from Detection Times Veterinarians advising on medication withdrawal intervals face a critical challenge: translating published detection times (DTs) from the European Horserace Scientific Liaison Committee into safe, individualised withdrawal times (WTs) that protect horses from testing positive. Toutain employed Monte Carlo simulations to quantify the safety factor needed when extrapolating from generic DTs to account for natural variation across the equine population. For drugs with low pharmacokinetic variability (coefficient of variation 20%), a multiplication factor of 1.4 provides adequate safety margin covering 90% of horses; however, highly variable drugs (CV 40%) require a substantially larger factor of 2.1–2.2, meaning the WT should be approximately double the published DT. Sensitivity analysis revealed that inter-individual differences in drug half-life drive most uncertainty, whilst factors such as dose, bioavailability and urine concentration have minimal influence on detection time variability—crucially, these biological sources of variation cannot be managed away through procedural changes. For practitioners, this underscores that conservative withdrawal recommendations must account for population variability rather than relying on mean values, and the current EHSLC protocol using six horses produces adequate DTs only about 50% of the time; expanding trials to eight or ten horses would substantially improve the reliability of published guidance.
Read the full abstract on PubMed
Practical Takeaways
- •When extrapolating EHSLC detection times to withdrawal recommendations, multiply by 1.4–2.2 depending on drug variability to safely cover 90% of individual horses and avoid positive drug tests
- •Focus on understanding individual horse metabolism (biological variability) rather than management factors, as half-life variation is the primary source of uncertainty in withdrawal time estimates
- •Be aware that current EHSLC recommendations have inherent uncertainty; using conservative multipliers (especially 2.1–2.2× for variable drugs) provides wider safety margins when advising owners on withholding periods
Key Findings
- •For low variability drugs (CV=20%), a withdrawal time of 1.4× detection time covers 90% of horse population; for high variability drugs (CV=40%), this factor increases to 2.1-2.2×
- •Terminal half-life parameters have the greatest influence on detection time variability, while dose and bioavailability have minimal impact
- •Biological sources of uncertainty cannot be reduced by managerial options and represent the main controllable challenge
- •Current EHSLC 6-horse protocol results in 50% probability of detecting the population 90th percentile; increasing to 10 horses improves this to 90% probability