Comparison of techniques for culture of dialysis water and fluid

Introduction: Microbiological culture of dialysis water and fluid is a routine safety measure. In the United States (U.S.), laboratories perform these cultures on trypticase soy agar at 35–37°C for 48 h (TSA-48h), not on the tryptone glucose extract agar or Reasoner's 2A agar at 17–23°C for 7 days (TGEA-7d and R2A-7d, respectively) recommended by international standards. We compared culture methods to identify samples exceeding the accepted action level of 50 CFU/mL. Methods: Dialysis water and fluid samples collected from 41 U.S. dialysis programs between 2011 and 2014 were cultured at two U.S. laboratories. Each sample was cultured using (1) either TGEA-7d or R2A-7d and (2) TSA-48h. We compared proportions exceeding the action level by different methods and test characteristics of TSA-48h to those of TGEA-7d and R2A-7d. Findings: The proportion of water samples yielding colony counts ≥50 CFU/mL by TGEA-7d was significantly different from the proportion by TSA-48h (P = 0.001; difference in proportion 4.3% [95%CI 1.3–7.3%]). The proportions of dialysis fluid samples ≥50 CFU/mL by TGEA-7d and TSA-48h were not significantly different; there were no significant differences for comparisons of R2A-7d to TSA-48h. Discussion: In dialysis fluid, TSA-48h was comparable to TGEA-7d and R2A-7d in identifying samples as having bacterial counts ≥50 CFU/mL. In dialysis water, TSA-48h was comparable to R2A-7d in identifying samples ≥50 CFU/mL, but TGEA-7d did yield significantly more results above 50 CFU/mL. Nonetheless, the negative predictive value of a TSA-48h result of <50 CFU/mL in dialysis water exceeded 95%.