(known) risk factors, which is seen in our study and in the study of Vianio and Bril, who reported that almost 80% of MDRO carriers are unexpected findings (11). However, as shown for MRSA by Lekkerkerk et al., new risk factors can be identified (21). Consequently, it could be worthwhile to investigate the effect of adding additional risk factors to the universal risk assessment, or to identify new risk factors for ESBL-E carriage. Strengths and limitations The main strength of this study is the active sampling of patients upon admission to the hospital, regardless of risk factors or MDRO status. This study also has some limitations. The main limitation of this study is that it is a single center study in a low prevalence country. Therefore, the generalizability of our work is limited, especially to countries with a higher MDRO prevalence. A second limitation is that different cultures were taken for the universal screening compared to the risk-based screening. For example, perianal samples instead of rectal samples were taken for the universal screening. Perianal samples may be less sensitive than rectal samples for detection of MDRO, therefore, the true carriage rate upon admission may be higher than our results indicate. This could also explain why the observed carriage rate of 4.6% for ESBL-E is lower than observed in other studies in the Netherlands (16). Additionally, it could explain why the newly identified VRE carrier was not detected in the universal screening, although it is also known that VRE colonization may be missed when only one culture is taken (22). In general, detection of MDRO is challenging as antibiotic use of the patient or sampling error play a role, which may result in false-negative results. However, for most MDRO, we used enrichment broths to overcome this as much as possible. Another limitation is that we were not able to sample all patients admitted to the hospital. For example, patients admitted during the weekend were not approached for participation. Therefore, our results are not complete. Finally, we only included patients in the study who could speak and read Dutch. Future studies Our results highlight the need for improvement of the universal risk assessment. It should be considered to add questions regarding travel history to the risk assessment, as this is a known risk factor for MDRO carriage (23-25). Other well-known but more general risk factors, such as antibiotic usage, could be of additional value as well. To further identify risk factors and to tailor the questions, a study with MDRO carriers from multiple hospitals in countries with low prevalence of MDRO is needed. Future studies should determine if adding additional questions improves the risk assessment for patients admitted to a hospital in a low-prevalence country and if this strategy is cost-effective. Additionally, costeffectiveness combined with the risk of transmission of identified MDRO, especially ESBLproducing E. coli, should be studied. However, as stated by Van Hout et al. (10) and by Vainio and Bril (11), the universal risk assessment is associated with a high workload for healthcare workers, and adding questions would increase this workload, which needs to be considered. A prediction system, based on data available in EHR of patients across multiple healthcare facilities, including pharmacies and general practitioners, would be a solution to overcome this in the future. 2 85 Comparing universal screening and universal risk-assessment for MDRO
RkJQdWJsaXNoZXIy MTk4NDMw