Adenosine Triphosphate Quantification Correlates Poorly with Microbial Contamination of Duodenoscopes

OBJECTIVE. The aim of this study was to quantify the correlation between adenosine triphosphate (ATP) measurements and bacterial cultures from duodenoscopes for evaluation of contamination following high-level disinfection. DESIGN. Duodenoscopes used for any intended endoscopic retrograde cholangiopancreatography (ERCP) procedure were included. Microbiologic and ATP data were collected concomitantly and in the same manner from ERCP duodenoscopes. SETTING. A high-volume endoscopy unit at a tertiary referral acute-care facility. METHODs. Duodenoscopes were sampled for ATP and bacterial contamination in a contemporaneous and highly standardized fashion using a "flush-brush-flush" method for the working channel (WC) and a dry flocked swab for the elevator mechanism (EM). Specimens were processed for any aerobic bacterial growth (colony-forming units, CFU). Growth of CFU > 0 and ATP relative light unit (RLU) > 0 was considered a contaminated result. Frequency of discord between among WC and EM measurements were calculated using 2 X 2 contingency tables. The Spearman correlation coefficient was used to calculate the relatedness of bacterial contamination and ATP as continuous measurements. RESULTS. The Spearman correlation coefficient did not demonstrate significant relatedness between ATP and CFU for either a WC or EM site. Among 390 duodenoscope sampling events, ATP and CFU assessments of contamination were discordant in 82 of 390 WC measurements (21%) and 331 of 390 of EM measurements (84.9%). The EM was frequently and markedly positive by ATP measurement. CONCLUSION. ATP measurements correlate poorly with a microbiologic standard assessing duodenoscope contamination, particularly for EM sampling. ATP may reflect biological material other than nonviable aerobic bacteria and may not serve as an adequate marker of bacterial contamination.