Effectiveness of ATP bioluminescence assay for presumptive identification of microorganisms in hospital water sources

Abstract

Background: Laboratory analysis of organisms in water include arduous methods, such as the multiple tube and membrane filter. The ATP bioluminescence system, proposes a new way of measuring cellular material in water by measuring adenosine triphosphate (ATP) levels, which are expressed in relative light units (RLU). The ATP bioluminescence assay has been increasingly used to assess the microbiological safety of the hospital environment. However, there are few studies investigating the use of this methodology to evaluate the microbiological quality of water. The objective of the present study was to verify whether ATP, as measured by the 3 M (TM) Clean-Trace Water (TM) ATP test, can be used as an alternative tool for presumptive testing for the presence of microorganisms in hospital water. Methods: Water samples (N = 88) were collected from faucets (74) and water purifiers (14) in a university hospital. The sample were filtered by the membrane filter technique (100 mL for bacterial analysis and 100 mL for fungal analysis) and then submitted to ATP bioluminescence assay to the determine quantity of RLU in each sample. In order to compare RLU and the presence of microorganisms, a receiver operating characteristic (ROC) curve was used to calculate sensitivity and specificity (levels higher than 90% were considered significant). In addition, control tests were conducted to compare RLU to the quantities of bacterial and fungal organisms added to distilled water (ANOVA and Tukey's tests; p <= 0.05). This inoculum was compared to RLU emission, and the data were analyzed by calculating the Pearson's correlation coefficient, with a 95% confidence interval. Results: In the present study, 94.3% of the water samples presented bacterial growth. Of these, 15.6% showed heterotrophic bacteria above recommended levels and fungal contamination was detected in 55.6% of samples. Sensitivity and specificity of the samples were not significant (<90%), and the correlation between ATP and the presence of these microorganisms in the samples (hospital water) was not significant, whereas, in distilled water, the results revealed a significant difference (p < 0.0001). Conclusions: These results demonstrated that the ATP test cannot be used as an alternative tool for presumptive assessment of the presence of microorganisms in water.