Background: Accurate microbiological surveillance in haemodialysis centres is important as end-stage renal patients can suffer from pyrogenic reactions due to bacterial contamination of dialysis fluids. To evaluate the microbiological quality of haemodialysis fluids, special nutrient-poor culture techniques are necessary. Although the Association for the Advancement of Medical Instrumentation (AAMI) recommends Tryptic soy agar (TSA) as the standard agar, several studies have resulted in a general preference for Reasoner's 2A (R2A) agar, as it appeared to be more sensitive in demonstrating contamination of typical haemodialysis associated bacteria. In the Netherlands TSA is still used for culturing dialysate, while dialysis water is cultured on R2A. Therefore, the aims of our study were to evaluate bacterial yields of dialysis fluids on both media, and to qualify their use in routine microbiological monitoring within our haemodialysis centre.
Methods: Between April 1995 and March 1996, 229 samples of pre-treated and final purified dialysis water, and samples of dialysates were collected. The specimens were aseptically taken from the tap, various points of the reverse osmosis (RO) water-treatment system, and the effluent tubes of 32 bicarbonate haemodialysis machines. Samples of 0.1 ml were inoculated in duplicate on spread plates with TSA and R2A agars. After 10 days of incubation at 25+/-2 degrees C, the numbers of colonies were quantified. The ranges of spread were taken 0-100 and 0-200 colony-forming units per milliliter (c.f.u./ml).
Results: The R2A agar had significantly higher colony counts than TSA agar for both dialysis water and dialysates. Considering 100 c.f.u./ml as the upper allowable bacterial limit for all dialysis fluids, microbiological non-compliance (bacterial growth) would be missed in 16% when using only TSA media (TSA < or =100 c.f.u./ml and R2A >100 c.f.u./ml), while this was 3% when using only R2A (TSA >100 c.f.u./ml and R2A < or =100 c.f.u./ml, P<0.0001). Considering 200 c.f.u./ml as the upper limit, non-compliance would have been missed in 10% when using only TSA (TSA < or =200 c.f.u./ml and R2A >200 c.f.u./ml), and 2% when using R2A (TSA > 200 c.f.u./ml and R2A < or =200 c.f.u./ml, P = 0.0011).
Conclusions: Microbiological surveillance of haemodialysis fluids, including pre-treated dialysis water samples collected from RO treatment systems, can be performed more precisely with R2A media than TSA, when incubated at 25+/-2 degrees C for 10 days.