Spiked Virus Level Needed To Correctly Assess Enteric Virus Recovery in Water Matrices

Quantitative microbial risk assessment (QMRA) identifies human enteric viruses in municipal wastewater as the pathogen group requiring the highest log reductions for various reuse applications. However, the performance of methods for estimating virus concentration is not well understood, and without performance assessment, actual risks are likely severely underestimated. To evaluate the efficiency of virus recovery from water, a water sample is often spiked with “known” amounts of virus, and the virus is then recovered after a series of analytical procedures. Yet for water matrices such as wastewater, due to the unknown background concentrations of targeted viruses in the matrix and the variable recovery efficiency between individual processes, only an approximation of the recovery efficiency may be obtained from such spike-and-recovery experiments. In this study, we demonstrated theoretically that for two widely used approximations, the error in estimating virus recovery should be less than the ratio of the amount of target virus in the background sample to that in the spike. Furthermore, we developed an applicable method, based on this new understanding, for deciding on the amount of virus for spiking before conducting a spike-and-recovery experiment, so that the approximation error is restricted to an acceptable level for each individual process. Finally, we applied the method to a set of experimental data for viruses in wastewater, demonstrating its utility and noting its general applicability to other pathogens or water matrices. IMPORTANCE The performance of procedures for pathogen log reduction is at the heart of new risk-based guidance/regulation globally, yet the methods for undertaking assessments of pathogen recovery are not standardized despite their fundamental impacts on assessing log reductions. Here we describe the level of spiking agent(s) that is necessary to correctly assess spiked pathogen/surrogate recovery with whatever method is deployed. The significance of our research lies in identifying the importance of the amount of spiking agents for reducing uncertainty in recovery estimates, which will allow the development of a recommendation for spiking experiments, proactively applying this understanding.