1230. Estimating Effective Point-of-Use Biocide Levels for Legionella Control in Building Water Systems Using a Large, Real-World Dataset

Abstract

Background

Legionnaires’ disease is a potentially life-threatening illness often associated with Legionella growth in water sources. Oxidizing biocides, such as chlorine (CL), monochloramine (MC) and chlorine dioxide (CD), can reduce Legionella contamination. However, limited guidance exists regarding optimal target biocide levels in building water systems to prevent Legionella growth. We examined Legionella and biocide data collected by Department of Veterans Affairs medical facilities nationally to estimate effective biocide levels.

Methods

Water samples collected at point of use for routine surveillance purposes between 2015 and 2017 were used for this analysis. Samples were limited to those with reported biocide being CL, MC or CD. Samples with biocide levels above safe drinking water maximums and from nonpotable water sources were excluded. Samples were stratified by hot and cold water and univariate logistic generalized additive models were used to assess nonlinear associations of probability of Legionella positivity and biocide level.

Results

The dataset included 144,458 samples (cold: 72,674; hot: 71,784) from 789 buildings at 168 hospitals, with 99,419 samples with reported biocide as CL, 40,922 as MC, and 4,117 as CD. For CL, cold water analysis showed a minimum probability of positivity at approximately 0.5 parts per million (ppm), but with a second minimum at 2 ppm. Hot water showed an inflection point around 0.6 ppm, but the likelihood of positivity continued to decrease until plateauing beyond 2 ppm (Figure 1). Cold water with MC showed a minimum probability of positivity at 0.3 ppm followed by a second minimum at 1.7 ppm with plateau beyond that concentration. Hot water showed a similar graph with initial minimum at 0.25 ppm and a second minimum at 1.6 ppm (Figure 2). CD graphs for both hot and cold showed a decrease at 0.2 ppm (Figure 3).

Conclusion

The variability in the dynamics of Legionella inhibition by different biocides as seen in our analysis indicates minimum biocide targets for the different agents. For CL and MC, biocide levels >2 ppm at point of use in building water systems may not provide added benefit for suppression of Legionella.

Disclosures

All authors: No reported disclosures.