Introduction

There is a continuing rise in the incidence of infections caused by multidrug resistant Gram-negative organisms in acute-care hospitals, outpatient healthcare facilities, and even in purely community settings [1–5, 6••]. These organisms can be resistant to all commercially available antibiotics, including carbapenems (e.g. imipenem, meropenem) and polymixins (e.g. colistin). The elderly are specifically prone to infection caused by resistant Gram-negative organisms, and many of these pathogens are endemic to United States’ (US) long-term care facilities (LTCF) [3, 7–10]. This worrisome trend is evident not just in the US, but worldwide [11, 12]. Multidrug resistant organisms (MDRO) lead to devastating outcomes among individuals [7, 13], and on a broader level, pose an eminent threat to public health [14, 15]. Older adults are particularly prone to poor outcomes following MDRO infections [6••, 8, 9, 16–18]. The Infectious Disease Society of America (IDSA) has listed groups of pathogens that pose the highest threat due to increments in prevalence and lack of effective therapeutics [1]. These are known as the "ESKAPE" pathogens, i.e. pathogens that currently cause the majority of hospital infections and effectively “escape” the effects of available antibacterial drugs [1].

The ESKAPE pathogens are composed mostly of Gram-negative bacilli: i.e. Pseudomonas aeruginosa, Acinetobacter baumannii, and various beta-lactamases (including carbapenemases) producing Enterobacteriaceae (Klebsiella pneumoniae, Enterobacter species, Escherichia coli) [1]. Their mechanisms of resistance include not only hydrolyzing enzymes, but also porin mutations and hyper-expression of efflux pumps (Fig. 1). Gram-positives pathogens are also listed as part of the ESKAPE pathogens, and include methicilllin-resistant Staphylococcus aureus (MRSA) and vancomycin-resistant enterococci (VRE); however, Gram-negatives pose a greater threat [7, 9]. This is in part due to the lack of success in bringing new drugs to the market with activity against Gram-negative MDROs [1]. This is in contrast to new and novel agents that have been developed for treating MRSA and VRE infections [7].

Fig. 1
figure 1

Mechanisms of beta-lactam resistance among Gram-negative bacilli. The mechanisms of beta-lactam resistance among Gram-negative bacteria include: 1) enzymes that destroy beta-lactams (e.g. beta-lactamases, carbapenemases), 2) porin mutations which preclude the entry of antibiotics into bacteria, 3) hyper-expression of efflux pumps that expel antibiotics outside the bacteria, and 4) mutations of penicillin binding proteins (PBP). From Munoz-Price LS, Weinstein RA. Acinetobacter infection. N Engl J Med 2008 Mar 20;358(12):1271–81. Copyright © 2008 Massachusetts Medical Society. Reprinted with permission from Massachusetts Medical Society

Full size image

Pseudomonas aeruginosa and Acinetobacter baumannii

P. aeruginosa and A. baumannii, are common human pathogens associated with high rates of morbidity and mortality [19–22]. Both pathogens are frequent causes of healthcare-associated pneumonia and hospital-acquired pneumonia (including ventilator-associated pneumonia), as well as other infectious syndromes [23]. Age is an independent predictor for carriage of these pathogens in the hospital settings, and older age is associated with increased mortality from these infections [19, 20, 24, 25]. The recent major epidemiological shifts of these infections are related to the development of antibiotic resistance [26] (Table 1) and their detection outside of the intensive-care unit (ICU) setting. Infections are now being observed in lower acuity (e.g., step-down units) locations in hospitals and in ambulatory healthcare settings, including LTCFs [16, 26, 27]. The prevalence of carbapenem-resistant A. baumannii increased from <1 % of all A. baumannii isolations in 2003 to 58 % in 2009 in one US acute care facility [26]. This trend has been reported from multiple additional facilities in the US [23] and worldwide [12]. Frequently, the only available agents to treat these carbapenem-resistant isolates (both A. baumannii and P. aeruginosa) are 1) polymixins, i.e. colistin and polymixin b, which are highly toxic [28] and poorly determined efficacy; 2) some aminoglycosides, which are also toxic and inappropriate as single agents for most infectious clinical syndromes [29]; and 3) tigecycline (only for A. baumannii, not for P. aeruginosa), which is unsuitable for severe invasive bloodstream infections due to its low serum levels [30] and resistance rates [7, 22, 31–33] (Table 1).

Table 1 Classification of multidrug resistant phenotypes among Gram-negative organisms: features and management considerations
Full size table

The Enterobacteriaceae: Escherichia coli, Klebsiella pneumoniae, and Enterobacter Species

Extended-Spectrum Beta-Lactamases (ESBLs)

The spread of common enteric pathogens such as E. coli, K. pneumoniae, and Enterobacter species, which are resistant to a wide range of antimicrobials, is an additional serious threat that has emerged in the past decade [1, 7]. ESBL-producing Enterobacteriaceae (Table 1) were initially isolated from ICU patients of acute-care facilities [34], but now have spread to ambulatory settings [5, 35] and LTCFs. Furthermore, they have been isolated from previously healthy patients who had no documented contact to healthcare environments or procedures, and who had not recently consumed antibiotics [36, 37]. However, advanced age has been identified as an independent risk factor for ESBL infection in multiple reports [5, 35, 38].

ESBL-producing Enterobacteriaceae are frequently co-resistant to additional classes of antimicrobials, e.g. fluoroquinolones and trimethoprim-sulfamethoxazole, and therefore there is frequently no oral therapeutic options to treat these infections [39] (Table 1). ESBL infections are frequently being managed only with parenteral agents, with all their associated complications and patients' safety issues [5]. Fosfomycin might be a potential oral therapeutic option for ESBL-producing isolates [40]. However, this agent is only approved for simple urinary tract infections [41], and there is lack of controlled data pertaining to its efficacy for ESBL infections [42•, 43]. Despite conflicting recent data [44••], carbapenems are still considered by most as the agents of choice for treating ESBL infections [39, 45] (Table 1).

Carbapenem-Resistant Enterobacteriaceae (CRE)

An important issue contributing to resistance among Enterobacteriaceae is the emergence and spread of strains resistant to carbapenems [4]. The elderly are particularly prone to develop CRE infections [46–48]. Therapeutics for CRE are scarce and are limited to polymixins, tigecycline, and sometimes aminoglycosides [49, 50] (Table 1). The mortality rate among bacteremic cases is ~70 % [51], and the median time to initiation of appropriate therapy among these patients is ~100 hours, i.e. these patients suffer from significant delays in initiation of appropriate antimicrobial therapy [29]. Due to the frequent transfer of patients among facilities within a region, multidrug resistant Gram-negative organisms have spread to LTCFs and long-term acute care hospitals (LTACHs). These facilities constitute important part of the continuum of care and seem to play an important role in the further spread of organisms within a geographical area [16, 52••].

The most prevalent carbapenemase in the US [53] and abroad [54] is Klebsiella pneumoniae carbapenemase (KPC), mainly produced by K. pneumoniae [55], E. coli [55] and Enterobacter species [56]. Additional carbapenemases reported in the US, Asia, and Europe are the metallo-β-lactamases [57]. Recently, the New-Delhi metallo-β-lactamase (NDM), which originated in the Indian sub-continent, has spread worldwide [58], including sporadic cases in the US [59]. Antibiotic exposure is one of the major independent predictors for acquisition of CRE [47]. Given the paucity of therapeutic options, infection control measures and antimicrobial stewardship efforts are our main line of defense against these highly resistant strains.

Risk Factors for Acquisition of Multidrug Resistant Gram-negatives Among the Elderly

As previously stated, the elderly are prone to infection by multidrug resistant Gram-negative organisms [60, 61], including P. aeruginosa [17, 62–65], A. baumannii [19, 66, 67], ESBL-producing Enterobacteriaceae [5, 68], and CREs [46, 48, 69]. However, these data are likely biased, and it would be more reasonable to assume that older age is a confounder rather than a real predictor for acquisition of MDROs; this is because elderly are more often 1) institutionalized (permanently or for prolonged periods), 2) have higher severity and amount of co-morbidities [65, 70, 71], 3) have declines in functional status (i.e. incontinence), 4) have deteriorated consciousness, 5) have external foreign invasive devices (e.g. central lines, urinary catheters, mechanical ventilation) [65], 6) more often are going through invasive procedures [65], 7) suffer from acquired immunosuppressive states (apart from the immunosenescence associated with normal aging [65, 72–75]), and 8) receive relatively more antibiotics, which leads to acquisition of MDROs through induction of selective pressure [9, 10, 47, 65, 76, 77]. Since these aforementioned variables represent independent predictors for acquisition of MDROs [19, 20, 47, 48, 61], advanced age might represent a confounder in these analyses. Although age is not a modifiable variable, studies that aim to prevent acquisition of these strains among the elderly are warranted as they represent a population at risk for acquisition and infection by MDROs [19, 47, 78].

Infection Control Issues at LTCFs

For the most part, infection control research is geared to acute care hospitals, and thus targets groups of patients with higher case mix indexes and shorter lengths of stay. Therefore, infection control at LCTFs has primarily constituted an extrapolation of interventions implemented at acute care hospitals, with some modifications. However, the resources available to acute care hospitals, for both personnel and assets, are far superior than the limited means available to the majority of nursing homes. This precarious situation is further aggravated by the current healthcare circumstances, which encourage acute care hospitals to discharge patients earlier in their hospitalization, subsequently increasing the severity of illness of the patients admitted to chronic settings, including nursing homes [10].

In a previous review on the topic, Mody described the challenges of infection control programs at LCTFs as: 1) structural (sub-optimal nursing staffing, high turnover of staff, variable acuity of illness among residents, and limited availability to patient’s laboratory data), 2) process factors (variable staff education, limited use or availability of diagnostic tests, and constrained use of quality indicators and regional databases), and 3) host factors [79]. Nursing home residents are at a particular high risk of colonization with MDROs due to their age, comorbidities, and frequent transfers to and from acute care hospitals with resulting high antibiotic exposures [80•]. However, the significance of acquisition of MDROs among nursing home residents has been previously questioned [81]. Nevertheless, nursing home residents, due to their frequent healthcare facility transfers, can carry and spread MDROs throughout different healthcare facilities of a geographical region [6••]. In 2008, a series of acute care hospitals, long-term acute care hospitals, and nursing homes were involved in the regional spread of KPC-producing K. pneumoniae [6••]. Transfer of unknown colonized patients among facilities was one of the main reasons for the rapid emergence of this organism within the region. A similar experience was recently described in Israel; in this country, a multicenter, national survey, detected a large unrecognized proportion of nursing home residents as carriers of carbapenem-resistant K. pneumoniae. Interestingly, but not surprisingly, this study showed that different long-term care facilities had different prevalence rates of colonization, i.e. rehabilitation centers had the lowest prevalence (3 %), and skilled wards had the highest prevalence rates (25 %) [82•].

Due to host factors among their residents and the prolonged length of stays at LTCFs, infection control practices usually done at acute care settings are particularly challenging to implement in chronic settings. Nevertheless, infection control programs at LCTFs are needed and required by the Centers for Medicare and Medicaid. According to the 2008’s SHEA/APIC infection control guidelines at LCTFs, infection control programs should have an Infection Control Committee (with core and ad hoc members) and an infection preventionist; these guidelines also delineate the roles of the infection control program at LTCFs [83]. In general, infection control programs serve several functions, including surveillance of hospital acquired infections and acquisition of MDROs, prevention of infections, control of outbreaks, creation and implementation of policies, and education of both staff and patients on infection control practices [79].

Even though the roles of infection control programs have been clearly identified by the above guidelines, not all facilities comply with these recommendations. Castle and colleagues reviewed the rate of deficiency citations given to nursing homes over the period of 2000 and 2007 [84]. They found that in average 15 % of all nursing homes received citations during the study period, although most of these citations were not critical. More importantly, the rate of citations was noted to increase from 12.8 % in 2000 to 17.3 % in 2007 [84]. Therefore, LTCFs constitute healthcare institutions with major opportunities for improvement in infection control.

In acute care hospitals, infection control interventions geared to preventing transmission of resistant Gram-negative organisms across patients include contact precautions, increased hand hygiene, identification of asymptomatic carriers, physical separation (i.e. cohorting) of MDRO carriers from non-carriers (in some cases using dedicated nursing staff), and environmental disinfection (Table 2). These interventions are geared to affect the three major pathways of transmission of MDROs from carriers to non-carriers: healthcare worker hands, shared equipment, and contaminated environment (Fig. 2). However, these interventions are challenging to translate to LTCFs due to their unique characteristics; mainly long length of stays (months or years) among patients with high mobility (not bedridden) and with multiple opportunities of social interactions in common areas (e.g. dining room, lounge areas, gym).

Table 2 Infection control interventions
Full size table
Fig. 2
figure 2

Main pathways of patient-to-patient transmission of multidrug resistant organisms (MDRO)

Full size image

There are some infection control topics pertaining to resistant Gram-negative organisms that deserve special attention:

  • Contact Precautions in Long-Term Care Facilities: Contact precautions is usually defined as the use of gowns and gloves upon entry to a patient’s room. Contact precautions constitute one of the main interventions for preventing transmission of resistant organisms among institutionalized patients. However, most of the literature on MDROs in nursing homes focuses on prevention of Gram-positive organisms (e.g. MRSA) rather than Gram-negatives. Furthermore, data on the effectiveness of contact precautions at LTCFs is currently lacking. Moreover, the mental and psychological consequences of long-term isolation have not been studied [85, 86].

    In 2004, Trick and collaborators compared routine glove use versus contact isolation precautions based on MDRO status. These investigators found no difference in acquisition of neither MRSA nor VRE between the intervention and control groups [87]. The authors suggested that universal glove use might provide a feasible alternative to the usual contact isolation practices practiced in acute care settings. However, there are no such controlled data pertaining to multidrug resistant Gram-negative organisms.

    The 2008 SHEA/APIC guidelines for infection control in LTCFs recommends the use of contact precautions only among residents colonized or infected with MDROs that are ill and fully dependent on staff for daily activities or if secretions cannot be contained [83]. These guidelines also encourage the use of single rooms for MDRO colonized individuals; alternatively, cohorting of rooms with other residents (ideally colonized with the same MDRO) is recommended. As a last choice, a room could be shared between an MDRO colonized resident and a “low risk” MDRO negative resident [83]. An interesting and novel approach was postulated in 2011 in which LTCF residents would be targeted for enhanced contact precautions based on the presence of indwelling catheters rather than on colonization with MDROs [88].

    In summary, even though there is a potential risk for patient-to-patient transmission of resistant Gram-negative organisms, confinement of mobile residents to their rooms due to their MDRO status should be discouraged. Additionally, contact precautions should be reserved for a small subset of MDRO colonized residents.

  • Hand Hygiene: Hand hygiene is one of the most effective and least expensive infection control interventions in healthcare. However, compliance with hand hygiene among health care workers is far from ideal [89]. Most studies looking at hand hygiene interventions have been performed in acute care settings and only few had been performed at LTCFs. A Canadian study measured the hand hygiene compliance of the staff at two LTCFs, finding an overall compliance rate of 14.7 % out of 459 hand hygiene opportunities. Not having a sink in the room significantly decreased compliance with hand hygiene (0 % vs. 15.7 %, p = 0.02) [90]. Similarly, an Italian study found an adherence with proper hand hygiene practices of only 17.5 % (54 out of 308 opportunities) [91].

    In 2003, Mody and colleagues evaluated the hand hygiene theoretic knowledge levels of LTCFs' staff before and after the introduction of a waterless alcohol-based hand rub and an education program [92]. The authors found increased hand hygiene frequency in one of the two wards tested. Additionally, hand cultures performed among the health care workers found Gram-negative organisms in 66 % of the hands cultured [92]. Ashraf and collaborators evaluated the knowledge, attitudes, and barriers related to hand hygiene at 17 LTCFs. Among 1,143 respondents, only a third was familiar with the CDC hand hygiene guidelines, being the main barrier to hand hygiene the absence of alcohol-based hand rub or sinks [93].

    In summary, promotion of hand hygiene at LTCFs constitutes an area with great potential for improvement. Furthermore, hand hygiene interventions should concentrate on both educational programs for the staff and increasing the availability of hand disinfectants.

  • Transitions of Care: An important problem among the elderly derives from their higher frequency of interfacility transfers. In 2009, the rate of hospitalizations from nursing homes was 204.5 per 1,000 persons aged 65 years and older [94]. Transitions of care between nursing homes and other facilities (e.g. acute care hospitals, long-term acute care hospitals, rehabilitation centers) make elderly patients susceptible to care fragmentation, with a subsequent impact on their quality of care [80•]. This less than ideal communication between facilities has been detected as a contributing factor in the spread of resistant Gram-negatives within a region [6••, 9]. Additionally, the variability on isolation practices between facilities creates confusion in the management of the patient.

    Therefore, transitions of care constitute another important area where major improvements can be made. Interventions at this level should include clear documentation upon interfacility transfers of any cultures with multidrug resistant Gram-negative organisms—bodily sources and dates—not just from the most recent admission, but also from previous healthcare stays. Obviously, this tracking of MDROs would be vastly facilitated by an integrated electronic microbiology database across healthcare facilities. In the meantime, we should concentrate in optimizing the documentation of MDRO carriage status across the continuum of care within a region [9].

  • Environmental Disinfection: The healthcare environment constitutes a reservoir for MDROs in the acute care setting [95]. Furthermore, heightened cleaning of hospital surfaces has been associated to decreased acquisition of MDROs among patients [96]. There are several ways to determine the degree of cleaning in a facility. An easy and inexpensive way is by using ultraviolet markers in the form of a powder [97] or gel [98] (Fig. 3). After inconspicuous application on high touch surfaces (e.g. bed rails, bedside tables, drawers), the markers should be inspected using an ultraviolet lamp after at least 48 hours from application. Markers found present at 48 hours represent surfaces that have not been cleaned. These regular evaluations and feedback of cleaning should be used as part of a quality improvement process within the institution.

    Fig. 3
    figure 3

    Ultraviolet gel used to monitor environmental cleaning (A) Ultraviolet marker under ambient light; (B) ultraviolet marker under ultraviolet light; (C) ultraviolet gel

    Full size image

    Murphy and colleagues used ultraviolet markers to evaluate the degree of cleaning at ten different LTCFs [99]. Only 22 % of 577 ultraviolet marks were removed (cleaned). The same group performed environmental cultures, although only looking for MRSA, finding 16 % of surfaces cultured positive for this organism. The rate of surface positivity in each particular nursing home was associated with the prevalence of MRSA among residents and time spent cleaning the rooms. Interestingly, infection control practices at these nursing homes were not associated with the degree of surface positivity for MRSA [99].

    At this time, data on the degree of environmental contamination at LTCFs with multidrug resistant Gram-negative organisms is lacking. Further research is necessary to determine the effect of heightened environmental disinfection on the patient-to-patient transmission of Gram-negatives in LTCFs.

Role of Antimicrobial Stewardship Programs in Reducing Acquisitions of Multidrug Resistant Gram-Negative Organisms

The IDSA defines stewardship as “coordinated interventions designed to improve and measure the appropriate use of antimicrobial agents by promoting the selection of the optimal antimicrobial drug regimen, including dosing, duration of therapy, and route of administration” [100]. The primary goal of stewardship is to maximize clinical outcomes while minimizing unintended consequences (e.g. emergence of resistance). Stewardship programs prevent the emergence of resistant Gram-negatives among patients, by reducing selective pressure [101–103]. Furthermore, establishing these programs in every healthcare facility, including LTCFs, is now recommended, and even mandated and regulated in many countries [104, 105]; although, translating this into common practices has been proven challenging [100, 106].

Antibiotics are frequently misused among older adults. In a point survey analysis conducted in Australia, the prevalence and appropriateness of antibiotic use was studied among 257 residents of residential aged care facilities [107]. The age of 28 % of the residents that were surveyed was greater than 85 years and 71 % had been in these facilities for more than a year. Sixty-seven percent were incontinent, and 80 % had some degree of cognitive impairment. Among the residents, 23 (9 %) were receiving antibiotics at the time of the survey: 17 were prescribed antibiotics for treatment of active infections, while six were given prophylactic antibiotics. During the preceding 26 months, there were 988 antibiotic courses administered; of these, 392 (39.7 %) did not fulfill the criteria for bacterial infection. This high antimicrobial utilization—frequently unwarranted—is what makes stewardship programs fundamental in both acute-care and LTCFs [107]; the latter is particularly important, given that residents at LTCFs might already be colonized with multidrug resistant Gram-negative organisms [3, 9, 52••, 78, 82•, 107, 108].

Hospitals and LTCFs that implemented stewardship programs have experienced improved outcomes such as reduced antibiotic consumption, decreased rates of Clostridium difficile infection rates, reduced overall expenditures for antibiotic treatment, reduced overall healthcare-related costs, decreased length of antibiotic therapy for various syndromes (e.g. community-acquired pneumonia), reductions in rates of inappropriate antimicrobial therapy, reductions in delays of initiation of appropriate therapy, and fewer antibiotic adverse events [109–115, 116••, 117–119, 120••, 121]. Several studies focused on the impact of implementation of antibiotic stewardship programs on acquisition of multidrug resistant Gram-negative organisms, but none measured the impact specifically among elderly adults [122].

There are numerous reports of interventions conducted at various types of LTCFs in order to reduce the rate of new MDRO acquisitions [6••, 27, 116••, 123–126, 127••, 128–130, 131•, 132–135]. Many were conducted in nursing homes consisting primarily of older adults [136•, 137, 138]. Some of the interventions were conducted as part of an outbreak investigation [16, 27] and some were conducted in order to decrease the endemicity and basal rate of acquisition of these deadly pathogens [116••]. Part of these interventions was establishment and implementation of a comprehensive stewardship program. Despite the fact that the direct impact of these programs on acquisition of MDROs among elderly has not yet been quantified, there is no reason to assume that the observed positive impact on the general population [115, 139], and LTCF residents [16, 116••, 118, 120••, 123, 124, 126, 133, 136•, 137], would not also be seen among the elderly.

Conclusions

The prevalence of multidrug resistant Gram-negative organisms is increasing worldwide. Elderly patients constitute important epidemiological links due to their frequent admissions and transfers across different healthcare facilities within a region. In addition, elderly patients are subjected to worse outcomes following the acquisition of multidrug resistant Gram-negative organisms. Given that LTCFs concentrate this type of patient population, more emphasis should be placed on implementing interventions targeting antimicrobial stewardship programs, hand hygiene among the staff, improving transitions of care (with better communication and documentation of colonization status) and environmental disinfection practices.