Screening
In many infection control guidelines, swab-based screening for MDROs is recommended to identify asymptomatic carriers [
7,
8]. Typical screening sites include nostrils, throat, and groin for MRSA and the perineal region for Gram-negative bacteria and VRE. However, important details concerning the best methodology are always lacking, e.g., which kind of swab to use, as there are many different materials available, e.g., rayon, polyurethane foam, or flocked nylon, which have a major impact on the detection rate. Warnke et al. compared different types of swabs for nasal MRSA screening in an artificial nose model. The best results were achieved for flocked nylon swabs with a sensitivity of 100%. Two kinds of rayon-made swabs only reached a sensitivity of 13 and 0%, respectively [
9]. For Gram-negative bacteria, swabs made of nylon flocks or polyurethane foam were found to be superior to conventional rayon swabs (
p < 0.001). In addition, swabbing deep intra-anally resulted in significantly higher recovery than perianal swabbing only (
p < 0.001) [
10]. Thus, the material and the structure of the swab itself have a major influence on the detection rate. However, neither are mentioned in infection control guidelines nor described in detail in most studies. In addition, the optimal frequency of the screening is also still under debate, e.g., entry screening, daily, weekly, or exit screening. Furthermore, different laboratory techniques can also be applied, e.g., normal culture, selective agars (e.g., chromogenic agar), or molecular techniques, and in the majority of the cases, the kind of technique is also not known to the clinician—although this can once again influence the detection rate.
Although all these limitations and pitfalls are obvious and known within the health care community, still infection control instruments, such as contact precautions with or without single room isolation, rely only on the screening results. Moreover, it is also still under debate which patient population should be screened: some medics demand universal screening, others try to define populations at risk. Finally, especially with a low detection rate of approximately only 1–2%, the cost–benefit ratio of screening is questioned in the literature [
11].
In conclusion, due to the many different aspects, screening is very difficult to standardize and to the best of our knowledge so far not standardized. Therefore, false negative as well as false positive results are frequent. Thus, we advocate that screening should not be the major instrument used to assess a patient’s pathogen-specific colonization status.
Isolation
In most guidelines concerning the management of MDRO, contact precautions (CPs), including standard precautions plus isolation in a single room and usage of gowns and gloves are considered as essential infection control instruments to prevent transmission in a hospital setting. However, since most clinical studies investigating the effectiveness of these infection control measures have major deficiencies in design and reporting, evidence to support this approach is rather limited. On the other hand, one cannot exclude that well-designed prospective randomized controlled clinical studies might reveal a favorable effect.
In a prospective interrupted time series analysis in three general medical-surgical ICUs of two teaching hospitals in London, Cepeda et al. investigated the effect of moving or not moving MRSA-positive patients into single rooms to prevent MRSA transmission [
12]. A cohort with 443 MRSA-positive patients was put into single room, and thus isolated; whereas, the other cohort with 423 patients was not. Both groups were screened regularly over a 6-month period. Interestingly, MRSA acquisition rates were similar for both cohorts. In detail, isolated and non-isolated patients had colonization rates of former negative patients of 12 and 10%, respectively. Compliance with hand hygiene was monitored during the study period and was estimated to be a very low 21% only. The authors conclude that in a setting of similar patient characteristics, MRSA acquisition rates could not be reduced by CPs alone, especially with a low compliance with hand hygiene.
In a systematic review and meta-analysis of nine studies, including a total of 30,949 patients, De Angelis et al. analyzed the effect of different infection control measures aimed at reducing the spread of VRE in a hospital setting. Notably, the rate of VRE acquisition was not reduced by CPs (RR = 1.08) at all; in contrast, the introduction of alcohol-based hand dispensers in place and an intensive educational program to improve compliance with hand hygiene reduced VRE acquisition significantly by 47%. [
13]. In general, the overall quality of the single studies retracted was low and thus the results achieved by the meta-analysis is of limited evidence as well. Hence, a prospective randomized clinical trial is required to shed light on this particular matter.
Kullar et al. reviewed 15 studies questioning routine use of CPs especially for MRSA and conclude that there are only a few preliminary data to support this approach in endemic settings [
14]. For example, Morgan et al. observed a total of 7743 health care worker (HCW) activities over 1989 h in four acute care facilities and documented that patients subjected to CPs had fewer HCW visits (− 36.4%,
p < 0.001) as well as direct contacts (− 17.7%,
p = 0.02) per hour [
15]. In addition, significantly more alcoholic hand rubs (+ 15.8%,
p = 0.001) were performed by HCWs on exiting an isolation room. The authors conclude that not CPs themselves but probably fewer visits and less direct contact of HCW with patients and especially better hand hygiene practices might be the reasons for the reduced risk of transmission.
Bardossy et al. investigated the effect of discontinuation CP in a retrospective study in an 800 bed teaching hospital. There were no significant differences in infection rates with MRSA and VRE catheter-associated urinary tract infections, ventilator-associated pneumonia, central-line associated bloodstream infections, surgical site infections and hospital-acquired MRSA bacteremia during the two 12-month periods including more than 76,000 patients [
16]. However, asymptomatic transmission of MRSA and VRE were not investigated by routine screening. Martin et al. compared the laboratory-identified clinical culture rates of MRSA and VRE after cessation of CP and introduction of 2% chlorhexidine bathing in all units of two hospitals [
17]. There was no significant difference in screening culture rates for MRSA (
p = 0.09) and VRE (
p = 0.14) before and after discontinuing of CP. Calculating the costs for nursing time spent with donning and gloving, costs for personal protective equipment and the chlorhexidine washing solution, annual savings of 643.776$ were achieved.
Furthermore, there is an increasing body of evidence showing that cessation of CPs does not lead to an increase in infection rates. In a quasi-experimental before-and-after study, Edmond et al. examined the effect of discontinuing CPs for MRSA and VRE in an 865-bed academic medical center [
18]. During the study, CPs were replaced by hand hygiene promotion and daily antiseptic baths with chlorhexidine, and a bare-below-the-elbows protocol was implemented. Compliance with hand hygiene was high with over 85%; however, the compliance with the antiseptic bathing was not monitored. Comparing both strategies, no change was noticed in the rates of MRSA or VRE device-associated infections—in the ICUs as well as on the normal wards. In addition, no changes in catheter-associated urinary tract infections, central-line associated bloodstream infections, and ventilator-associated pneumonia were observed with all other pathogens.
Almyroudis et al. studied the effect of discontinuing systematic surveillance and CPs on the incidence of VRE bacteremia in a 125-bed hematology–oncology unit [
19]. Between 2008 and 2011, the incidence of VRE bacteremia for patients under active surveillance and CPs was 2.32/1000 patient days (PD). Interestingly, from 2011 to 2014, surveillance and CPs were stopped and the incidence further decreased to 1.87/1000 PD. In 2013, daily bathing with chlorhexidine-impregnated washcloths was additionally implemented for all patients.
In addition, there are distinct disadvantages associated with the medical care of patients in isolation and these are of major concern and reported frequently. For example, Zahar et al. compared 170 patients in isolation with 980 non-isolated patients in two ICUs in France. Preventable adverse events, e.g., hypo- and hyperglycemia, errors in anticoagulant prescription, and ventilator-associated pneumonia due to resistant bacteria occurred significantly more frequent in isolation [
20]. Patients reported more discomfort, depression, and anxiety when undergoing CPs and HCWs were less likely to visit patients as well as have less contact [
21]. Compliance of HCWs with basic infection control guidelines is low, especially with high proportions of patients undergoing CPs [
22].
In conclusion, with respect to the limited qualities and the diversity of the cited studies, active surveillance and CPs did not seem to prevent MRSA, VRE and ESBL transmission and infections in a susceptible patient population.
Eradication
In a prospective cohort study, Mattner et al. investigated the persistence of MRSA in 1032 MRSA-positive patients of a German university hospital between 2002 and 2005 [
23]. Topical decolonization with mupirocin nasal ointment and antiseptic body wash with either octenidine or chlorhexidine for 5 days was performed, respectively. The overall half-time of MRSA persistence was 549 days and was even prolonged if multiple body sites were affected.
Ammerlaan et al. performed a systematic review concerning eradication of MRSA including 23 studies with a total of 2114 patients [
24]. Nasal application of mupirocin for up to 7 days was most effective for MRSA eradication with an estimated success of 90% 1 week after treatment. However, in the long-term follow-up (14–365 days), recurrence of MRSA could be observed in about 40% of patients. The effectiveness of topical mupirocin treatment was further reduced when multiple body sites were colonized.
Eradication of VRE has been studied with different regimens of antimicrobial agents with high intraluminal concentrations, e.g., bacitracin, gentamicin or doxycycline; however, the results never showed a sustained decolonization due to these regimes [
25,
26].
A novel approach with application of oral linezolid, daptomycin, and
Lactobacillus rhamnosus following bowel preparation with polyethylene glycol achieved VRE clearance only in two of four liver transplant patients [
27]. The duration of intestinal carriage of VRE is not well investigated, but is suspected to be even longer than MRSA.
The eradication of intestinal carriage of Gram-negative bacteria is even more difficult if not impossible. In a double-blind, randomized, placebo-controlled single center study, Huttner et al. investigated the efficacy of an oral decolonization regimen consisting of colistin, neomycin, and nitrofurantoin vs. placebo in a total of 54 patients with intestinal carriage of ESBL [
28]. The primary outcome was detection of ESBL in rectal swabs after 28 days with additional cultures taken on day 6 of treatment and on days 1 and 7 after treatment. Regarding the primary outcome, there was no significant difference between both treatment groups [14/27 (52%) vs. 10/27 (37%),
p = 0.27]. During treatment and on day 1 after treatment, intestinal carriage of ESBL was significantly lower in the treatment group. However, this effect was not observed on day 7 after treatment. Therefore, no long-term effect could be demonstrated. Thus, the usage of colistin for the purpose of decolonization as one of the last remaining effective antibiotics for highly resistant Gram-negative bacteria has to be questioned critically.
Finally, of note is that all other trials aiming to eradicate intestinal carriage of Gram-negative bacteria failed to show sustainable success [
29].
Therefore, it is well accepted in the medical community that these pathogens cannot be successfully eradicated.
Summary of the classical approach
The complex pathogen-specific screening system as implemented in many medical institutions has many limitations since the swabbing material to be used, the frequency and the location patients should be screened are not well defined. In addition, neither the optimal quantity of the medical sample is known, nor is there any consistent laboratory standard to screen for the different MDROs. In summary, the current screening systems for complex pathogens that are in operation today are not standardized, nor can they be since multiple factors, especially sample size are not possible to standardize. In addition, the patient population that needs to be swabbed is not well defined, and thus, varies between medical institutions. So far, studies investigating the transmission rate of MDRO from patients undergoing CPs failed to find a reduction. However, it should also be considered that the study design was not always comprehensive in terms of patient numbers or pathogen monitoring, and therefore, the overall evidence is still preliminary at this stage. In contrast, disadvantages of single room isolation concerning patient discomfort, preventable adverse events, and reduced physician and HCWs contact time are well described. Finally, eradication, even for MRSA, is a major medical challenge, and currently, impossible in patients harboring VRE or multidrug-resistant Gram-negative bacteria. Thus, alternatives to current screening, isolating, and eradicating procedures are urgently required.