Using wastewater-based epidemiology to estimate drug consumption-Statistical analyses and data presentation

Area of Science:

• Environmental chemistry
• Forensic toxicology
• Epidemiology

Background:

• Wastewater analysis is a valuable tool for assessing population drug use.
• Statistical challenges, particularly with censored data (results below quantification or detection limits), hinder accurate epidemiological interpretation.
• Unobserved or non-quantifiable data are common and expected to persist with broader application.

Purpose of the Study:

• To demonstrate the utility of censored data techniques in wastewater analysis for drug use estimation.
• To account for measurement errors and explore distributions of excreted drugs per capita.
• To provide more complete and valid annual estimates of population drug excretion.

Main Methods:

• Analysis of daily wastewater samples from 19 municipalities over 56 days in 2009.
• Quantification of methamphetamine, benzoylecgonine, MDMA, methadone, oxycodone, and hydrocodone.
• Application of four statistical approaches (censored data techniques, Maximum Likelihood Estimation, Kaplan-Meier, complete data) to estimate annual averages, accounting for data censoring.

Main Results:

• Significant variation in censored data proportions (4% to 94%) was observed across drugs and municipalities, with MDMA showing the widest range.
• The choice of statistical approach was dependent on the level of data censoring.
• Confidence bounds facilitated visualization and comparison of drug excretion estimates across municipalities.

Conclusions:

• Wastewater results below detection or quantification limits contain crucial information.
• Incorporating censored data techniques leads to more complete and valid estimates of drug excretion.
• This methodology enhances the epidemiological utility of wastewater-based drug monitoring.