Effects of holding time and measurement error on culturing Legionella in environmental water samples
| Autor: | W. Dana Flanders, Kimberly H. Kirkland, Brian G. Shelton |
|---|---|
| Rok vydání: | 2013 |
| Předmět: |
Environmental Engineering
Time Factors Mean squared error Legionella Sample processing Sample (statistics) Environment Measurement error Sensitivity Environmental water Statistics Humans Legionella monitoring Waste Management and Disposal Water Science and Technology Civil and Structural Engineering Holding time Legionella culture Observational error biology Ecological Modeling Environmental engineering Replicate biology.organism_classification Pollution Ecological Modelling Environmental science Shipping effects Sample holding time Water Microbiology |
| Zdroj: | Water research. 62 |
| ISSN: | 1879-2448 |
| Popis: | Outbreaks of Legionnaires' disease require environmental testing of water samples from potentially implicated building water systems to identify the source of exposure. A previous study reports a large impact on Legionella sample results due to shipping and delays in sample processing. Specifically, this same study, without accounting for measurement error, reports more than half of shipped samples tested had Legionella levels that arbitrarily changed up or down by one or more logs, and the authors attribute this result to shipping time. Accordingly, we conducted a study to determine the effects of sample holding/shipping time on Legionella sample results while taking into account measurement error, which has previously not been addressed. We analyzed 159 samples, each split into 16 aliquots, of which one-half (8) were processed promptly after collection. The remaining half (8) were processed the following day to assess impact of holding/shipping time. A total of 2544 samples were analyzed including replicates. After accounting for inherent measurement error, we found that the effect of holding time on observed Legionella counts was small and should have no practical impact on interpretation of results. Holding samples increased the root mean squared error by only about 3–8%. Notably, for only one of 159 samples, did the average of the 8 replicate counts change by 1 log. Thus, our findings do not support the hypothesis of frequent, significant (≥= 1 log10 unit) Legionella colony count changes due to holding. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect