LAUSR

Patient infections with Non-tuberculous Mycobacterium (NTM) have been attributed to some reusable medical devices (1, 2), such as heater cooler devices (3–5), dental unit waterlines (6), bronchoscopes (7), and automated endoscope reprocessors (8, 9). Such incidents can be related to insufficient reprocessing or growth of resistant organisms. For example, NTM infections can arise from patient exposure to contaminated water from established biofilms in water systems, and in some cases, aerosolization of the contaminated water (10). These medical devices are regulated by the U.S. Food and Drug Administration (FDA) and the Agency seeks to better understand the mechanism bywhich devices can transmit NTM. SomeNTM-specific challenges include potentially years-long incubation period to clinical infection, subsequent difficulty identifying the bacteria to the species-level and extended duration of treatment (11). While general awareness around clinically relevant, rapidand slow-growing NTM and infections appears to have increased (11), a literature review of Mycobacterium research reveals the trend of testing rapid-growing Mycobacterium spp. as surrogates and extrapolating data to their slow-growing counterparts (12–14). However, with more than 150 known NTM species (15), questions about the applicability of surrogate data have been voiced for decades (11, 13). Comparative analyses between slowand rapid-growing NTM tend to be limited and in some cases, have demonstrated notable variation, as well as intra-species differences (13, 14, 16, 17). This raises questions about the applicability of using one species of NTM (herein referred to as surrogate NTM) in place of another for medical device testing. The topic of appropriate surrogates can be considered in many ways. However, in the context of this manuscript we will discuss the applicability of NTM surrogates for testing intermediateand high-level disinfection1 of, and aerosolization from, reusable medical devices in the context of specific NTM outbreaks. Note that this manuscript does not address or suggests any changes in wellestablished disinfection validation methods that are routinely used for medical device testing.