First draw water samples for Legionella culturing of distal water sites?

Our approach to culturing environmental samples in our hospital is to collect 1 liter water samples from a tap after the water has been run until it gets “hot”–some sites may take several minutes before the sample is collected. After looking at our numbers, and reading about different ways to collect the sample, I realize that we probably are NOT getting a good representation of the distal site, but more of what is present centrally. We prefer to collect water rather than swab so we can get a quantitative number (I did read your paper recommending the swab as the more sensitive though!–I am concerned that there may be a great deal of variability in how a swab would be collected). Would it be better for us to collect the sample immediately after turning on the tap, and collecting a smaller volume? We really can’t do an immediate and a post flush– we have a difficult enough time having the state lab to handle one sample/site!

One liter samples are not necessary. Our experience and studies have has shown that the first draw sample is a more sensitive sample than a “post-flush” sample. In addition, we sampled approximately 100ml, and plated 0.1 mL directly and after concentration by filtration (100mL filtered and resuspended in 10mL). We validated this method by performing a multi-center study and correlated site positivity with disease incidence using this method (article in press). You are correct that variability in quantitative counts exist for swabs, but variability also exists for water. So, quantitation is simply too variable to be useful.