Cross validation of gas chromatography-flame photometric detection and gas chromatography–mass spectrometry methods for measuring dialkylphosphate metabolites of organophosphate pesticides in human urine

We report two analytical methods for the measurement of dialkylphosphate (DAP) metabolites of organophosphate pesticides in human urine. These methods were independently developed/modified and implemented in two separate laboratories and cross validated. The aim was to develop simple, cost effective, and reliable methods that could use available resources and sample matrices in Thailand and the United States. While several methods already exist, we found that direct application of these methods required modification of sample preparation and chromatographic conditions to render accurate, reliable data. The problems encountered with existing methods were attributable to urinary matrix interferences, and differences in the pH of urine samples and reagents used during the extraction and derivatization processes. Thus, we provide information on key parameters that require attention during method modification and execution that affect the ruggedness of the methods. The methods presented here employ gas chromatography (GC) coupled with either flame photometric detection (FPD) or electron impact ionization-mass spectrometry (EI-MS) with isotopic dilution quantification. The limits of detection were reported from 0.10 ng/mL urine to 2.5 ng/mL urine (for GC-FPD), while the limits of quantification were reported from 0.25 ng/mL urine to 2.5 ng/mL urine (for GC–MS), for all six common DAP metabolites (i.e., dimethylphosphate, dimethylthiophosphate, dimethyldithiophosphate, diethylphosphate, diethylthiophosphate, and diethyldithiophosphate). Each method showed a relative recovery range of 94–119% (for GC-FPD) and 92–103% (for GC–MS), and relative standard deviations (RSD) of less than 20%. Cross-validation was performed on the same set of urine samples (n = 46) collected from pregnant women residing in the agricultural areas of northern Thailand. The results from split sample analysis from both laboratories agreed well for each metabolite, suggesting that each method can produce comparable data. In addition, results from analyses of specimens from the German External Quality Assessment Scheme (G-EQUAS) suggested that the GC-FPD method produced accurate results that can be reasonably compared to other studies.