Optimizing Hydrotherapy Water Quality in a Burn Unit

Background:   A new 20-bed burn unit opened in our 820-bed academic medical center with hydrotherapy hoses installed in the ceiling of all procedure and patient rooms. A process for disinfecting and maintaining hydrotherapy hoses in wound care could not be found in the literature or as standard of practice with other burn centers.

Objective:   To assess different procedures for the maintenance of hydrotherapy hoses to ensure low bacterial counts in the water for safe wound care.

Methods:   Three separate methods were used to test for effectiveness in minimizing water bacterial counts. First, we employed hyperchlorination to disinfect the entire hose system. Next, we tested flushing of the hoses for different lengths of time, with and without the use of disinfection. Last, we attached an in-line water filter at the end of the hose and cultured with and without disinfection. All samples were collected in a sterile container with sodium thiosulfate to neutralize chlorine. The containers were packaged in thermal freezer bags with ice packs and shipped overnight to a reference laboratory for total heterotrophic plate counts. 

Results:   Baseline water samples for all methods were >300,000 cfu/ml. Hyperchlorination decreased the bacterial count, but the results were inconsistent (1100-33,000 cfu/ml). Flushing the hoses also decreased the bacterial counts inconsistently (540-6400 cfu/ml). The in-line water filters produced the lowest bacterial counts (<1-2 cfu/ml). However, bacterial counts with the filter increased over several days (280->300,000 cfu/ml). Special nozzles were attached to the hoses for effective wound care technique. Due to manufacturing limitations, the nozzles cannot be included in the disinfection or flushing process. Despite low bacterial counts with the in-line filter, the addition of a nozzle increased the counts (>300,000 cfu/ml). Changing to a single use, sterile nozzle consistently produced the lowest bacterial counts (<1-32 cfu/ml).

Conclusions: The hyperchlorination process was not effective, time consuming, and had a corrosive side effect on the hoses. Flushing the hoses for thirty minutes reduced the total bacterial count, but the results were inconsistent. The lowest bacterial count was obtained with the combination of a thirty minute flush, a new in-line water filter attachment, and a single use, sterile nozzle.