Discussion Our results show that in a low endemic setting, a universal screening strategy identifies significantly more MDRO carriers than through the currently implemented universal risk assessment combined with risk-based screening. The result that a universal screening strategy identifies more carriers in our low endemic setting is not surprising. This could be explained by the fact that more patients are microbiologically screened than through the risk-based screening. Secondly, the universal risk assessment only includes questions regarding a limited number of risk factors, and some MDRO carriers do not have any of the predefined risk factors, as shown for MRSA (15). Since we did not identify new MRSA, CPE, highly resistant P. aeruginosa or A. baumannii, our discussion will be focused on ESBL-E. The question remains what the added benefit of identifying these MDRO (in our setting all ESBL-E) carriers would be. Several studies have studied the effect of isolation practices for known ESBL-E carriers (16, 17). Kluytmans-van den Bergh et al. showed that transmission from index patients was higher for patients with unprotected ward stay, compared to patients who were cared for under contact precautions directly upon admission, although not significantly. This highlights the importance of timely identification and isolation (16). Our previous study showed that most ESBL-E carriers also remain unidentified through clinical cultures throughout their hospitalization (12). This, in combination with the high percentage of unidentified carriers upon admission, raises concern for unidentified transmissions throughout the hospital and the potential clinical implications. While we did not identify CPE, the study of Phee et al. highlighted the key role universal screening has in identifying the true prevalence of carbapenemase-producing organisms (7). Some of the sequence types (ST) that were found among ESBL-E. coli in our study have been reported to spread in hospitals with a blaNDM gene (ST10 in Mexico and ST167 in Denmark) (18, 19). Our findings regarding the yield of the universal risk assessment and risk-based screening were in agreement with the study by Van Hout et al. (10) and Vainio and Bril (11). Both our results and the results of Van Hout et al. identified that the currently installed strategy is unable to identify most carriers, and that the highest yield is through the question “are you a known MDRO carrier?”. Consequently, Van Hout et al. proposed abandoning the riskbased screening, and only installing transmission-based precautions for (previously) known carriers (10). Vainio and Bril identified that the question about hospitalization abroad substantially contributed to the yield, and consequently they suggest a simplified risk assessment, only asking about known MDRO carrier status and recent hospitalization abroad (11). However, according to our results, most known carriers do not report they are a known carrier. This could be deliberate, to prevent being cared for in isolation, or it could be that the patient is not aware of or does not completely understand their own MDRO status (20). Also, due to frequent inter-hospital patient transfers, communication on the current MDRO status of a patient may be delayed. This could lead to a delay in installing isolation practices, and consequently could lead to transmission to other patients and to the hospital environment. It is important to notice that due to transmission in the population of ESBL-E, it is difficult to implement an effective risk factor screening strategy. Not all ESBL-E carriers have 84 Chapter 2.3
RkJQdWJsaXNoZXIy MTk4NDMw