Abbreviations: CI = confidence interval; EHS-Net = Environmental Health Specialists Network; FDA = Food and Drug Administration; OR = odds ratio. * Disassembled, cleaned, and sanitized. † California, Minnesota, New York, New York City, Rhode Island, and Tennessee. § The reference level is the second category listed. Thus, the odds ratio is for the first category listed compared to the second category listed. ¶ Numbers vary because of missing data. ** P-values for the overall ORs: p = 0.001 and p<0.001 for the manager and worker models, respectively. † † Certification defined as having taken and passed a food safety test and been issued a certificate. § § Somewhat easy, neither easy nor difficult to clean, somewhat difficult, or difficult. ¶ ¶ P-values for the overall ORs: p = 0.803 and p = 0.856 for the manager and worker models, respectively. *** P-values for the overall ORs: p = 0.441 and p = 0.445 for the manager and worker models, respectively. chubs daily reported more frequent slicer cleaning than did delis with fewer slicers, serving fewer customers, or selling fewer chubs daily. These characteristics are likely indicators of deli size, and these data are consistent with other findings suggesting that both chain and larger establishments' food safety practices tend to be better than those of independent and smaller establishments (9,10). Compared with both independent and smaller delis, chain and larger delis might have more resources, more or better trained staff, or more standardized cleaning procedures.The association of required manager food safety training and certification with more frequent reported slicer-cleaning is consistent with other findings indicating that training and certification are important in retail food safety (9,10), and highlights the important role that management can play in food safety. The finding that delis with workers with more food safety knowledge and experience had more frequent reported slicer cleaning suggests that workers also play an important role in food safety.Simple logistic regression findings suggest other characteristics that might improve cleaning frequencies. Written slicercleaning policies and worker ratings of slicers as being easy to clean were both associated with more frequent reported cleaning, suggesting that workplace policies and slicer design can affect cleaning frequency. Finally, delis with a food safetycertified manager had better reported cleaning frequencies, again pointing to the importance of training and certification.Because slicer-cleaning frequency and disassembly guidance are presented separately from each other in the FDA Food Code, some deli managers might be unaware that cleaning should include disassembly, and might clean and sanitize slicers without disassembling them. It is also possible that some slicers included in this study, especially newer ones, do not need to be disassembled to be fully cleaned.The findings in this study are subject to at least three limitations. First, the interview data might be affected by social desirab...