We oversampled facilities reporting actively screening for CRE/CP-CRE (9/16) and those with more CRE positive cultures (Quartiles 1 and 2) as shown in Table 3. Forty-three interviews were conducted with microbiology laboratory staff (N=20), MPCs (N=15), infection control nurses (N=5), and physicians (N=3) as shown in Table 4.
CFIR domains and constructs: Categorizing Coded Responses
In Figure 2, we present the 841 interview coded responses categorized by CFIR domains related to system-wide CRE guideline implementation: 1) Inner Setting (n=429, 51%), 2) Implementation Process (n=186, 22.1%), 3) Intervention Characteristics (n=133, 15.9%), 4) Characteristics of Individuals (n=49, 5.8%), and 5) Outer Setting (n=44, 5.2%) (see Appendix 2 for CFIR domain and construct definitions).
CFIR domains and constructs: Examining Associations
In Table 2, we show results of selected quantitative analyses comparing CFIR constructs and open codes by whether the sites were actively screening for CRE (vs. non-screening) and any (vs. no) CRE positive cultures. Our analysis identified the following constructs as being significantly associated with CRE screening or presence of CRE: Leadership Engagement, Relative Priority (“CRE is as important as other Hospital Acquired Infections (HAIs)”), Available Resources (e.g., IT support, staffing), Team Communication (operationalized as communication breakdown), and Access to Knowledge & Information (e.g., laboratory/clinical staff and patient/family educational material).
VAMCs that were actively screening for CRE reported significantly more leadership engagement in implementing CRE policies than those VAMCs that were not screening, 100% vs. 68.2%, (p=0.002). Sites screening for CRE reported that it was treated as seriously as other HAIs (e.g., with Methicillin Resistant Staphylococcus aureus (MRSA)) when compared to sites not screening for CRE, 91.7% vs. 55.6% (p=0.01). Sites not screening for CRE were significantly more likely to report a lack of available resources (compared to sites screening for CRE) 81.7% vs. 45.1% (p<0.0001).
VAMCs with any CRE cases (vs. no CRE cases) were less likely to report team communication breakdowns than sites with no CRE, 100% vs. 80.8% (p=0.02). Sites with any CRE reported better access to knowledge and information compared with sites with no CRE, 88.9 % vs. 36.8% (p=0.02).
Understanding the Multi-Level Aspects of CRE Guideline Implementation
Results of qualitative analyses include findings from representative responses coded by CFIR domains and constructs (Figure 2) as well as emergent open-coded responses identified during analysis, including responses categorized as “best practices” (defined as an innovative approach to implementation or guideline recommendations).
Within the Intervention Characteristics domain, responses categorized within the Intervention Source construct described which CRE guidelines participants reported using (e.g., VA, CDC, state), with most reporting using the VA’s CRE guidelines. We coded responses that described adherence to CDC or state guidelines (not VA) as negative. Best practices examples involved sites adapting or modifying current MDRO policies to include recommendations from the VA CRE guideline.
Within the Outer Setting domain, Patient Needs & Resources construct was defined as strategies and/or materials used to educate or engage Veterans and their families. Positive responses focused on the adequacy of and access to adequate educational materials to inform patients and their families about CRE (including the availability of materials in various languages and literacy levels). Negative responses described sites lacking knowledge of or access to handouts for educating laboratory and/or clinical staff and patients and families. An MPC described it this way:
It would be nice to have [more educational] materials [and staff release time for CRE-related training] (MPC).
Best practices described strategies to promote staff training, development of lab/clinical staff, and patient and family educational materials, including educational brochures targeting those with low literacy.
Within the Characteristics of Individuals domain, Self-Efficacy was defined as the participant’s confidence in understanding and implementing CRE guidelines. Most participants’ self-efficacy was high, reporting feeling “very” or “fully” confident in implementing the guidelines. The negative response described existing challenges in cohorting MRSA patients and being unsure of how to address CRE patients. Best practices included protected time for a dedicated person to monitor CRE at their VAMC.
Within the Implementation Process domain, the Formally Appointed Internal Implementation Leaders construct was defined as participant’s existing infection control structure (units/departments/committee/resources) and/or laboratory testing of other MDROs. Negative responses centered on gaps in existing MDRO reporting strategies (e.g., overreliance on verbal vs. electronic communication). Best practices described a strong infection control structure based on good communication that enables sites to easily notify infection control team members of a CRE positive patient. Commercially available software (e.g., Theradoc®) was also described as facilitating staff communication.
Table 5 describes CFIR Inner Setting domains, definitions and representative positive, negative as well as best practices responses. Just over half of all coded responses addressed constructs within the Inner Setting domain (e.g., factors promoting successful CRE guideline implementation). Because these factors are potentially modifiable, and are important for addressing effective implementation, four of the top six Inner Setting constructs are highlighted (Table 5): 1) Available Resources (n=122, 28.4%), 2) Networks and Communication (n=67, 15.6%), 3) Leadership Engagement (n=51, 11.8%), and 4) Culture (n=30, 6.9%).
Inner Setting Constructs
Successful implementation of the guidelines begins with stakeholders’ perceptions of the immediacy of the problem. One interviewee described trying to implement the guidelines as very challenging in the absence of any CRE:
It’s like waiting for it to snow in Florida. You’re doing preparedness training [but] people don’t listen and … [they don’t] think it’s real. Because they haven’t seen it … People don’t begin to be proficient unless they come in contact with it. [Laboratory Staff]
Participants at sites with CRE cases indicated that successful implementation of the CRE guidelines required access to and/or deployment of various resources (e.g., staffing, acquisition of new laboratory equipment to conduct polymerase chain reaction [PCR] testing, and/or information technology). One MPC described obtaining necessary resources in the following way:
[Our] Chief of Staff… did go to [the] chiefs of various services and said to them, you need to take the CRE screening seriously. Leadership certainly pushed all the physicians to implement the guidelines. If we … had more positive CRE [cases], I’m positive they [leadership] will be on top of it. [MPC]
At sites where the guidelines were not fully implemented, participants described inefficient or overly complicated processes for obtaining new laboratory equipment and/or addressing staff turnover, insufficient contact isolation rooms, and/or dependence on unreliable strategies (e.g., word-of-mouth) for communicating patient CRE status.
Best practices for the Available Resources construct included strategies to enhance dissemination of CRE educational materials to providers (e.g., posters and train-the-trainer in-services), timely access to laboratory reports to inform clinical decision-making, and more formal strategies to systematically communicate patients’ CRE test results (positive or negative) to all relevant parties (e.g., admission/discharge templates in the electronic health record [EHR]).
To confirm CP-CRE, all VAMCs who did not have the available equipment were required to purchase new laboratory equipment or to identify an outside laboratory with PCR-testing capacity. Once a potential case of CP-CRE was identified, participants described ongoing monitoring of confirmation test results of CP-CRE and developing strategies to ensure adequate isolation space to keep the identified individual in contact precautions until CP-CRE status was confirmed.
Networks and Communication:
Coded responses described positive and negative aspects regarding CRE guideline dissemination. Sites reporting good communication described adapting a robust pre-existing infection control infrastructure to successfully address CRE. Other sites cited overly complicated and/or hierarchical bureaucracy that hindered dissemination, contributing to delays in development or approval of local policies as impeding CRE guideline implementation. Best practices for this construct included employing strategies, tools, and innovations to facilitate local guideline implementation. One such strategy was to include CRE in existing MDRO policies versus developing a CRE specific policy, which often delayed guideline implementation. Participants most commonly reported learning about the VA CRE Guidelines through one or more national, regional, and local communication mechanisms, including hospital and discipline-specific (e.g., laboratory, infectious disease, pharmacy) email listservs. Guidelines were also discussed and disseminated at local in-service trainings, new employee orientation, and/or ad hoc trainings (e.g., daily patient rounds), as well as during local interdisciplinary committee meetings (e.g., antibiotic stewardship committee meetings).
Leadership engagement was defined in terms of local leaders’ commitment to, involvement in, and accountability for CRE guideline implementation. Sites successfully implementing the CRE guidelines commonly cited local (VAMC, laboratory, and/or Infection Control [IC]) leadership engagement in facilitating timely procurement of laboratory testing equipment. Negative responses described local (VAMC, laboratory and/or IC) leadership as “uninvolved” in guideline implementation and/or unwilling to address adequate isolation space, staffing to conduct CRE testing, release time for staff education, or funding to purchase laboratory equipment. Best practices described active leadership engagement in guideline implementation and participation by all relevant clinical and laboratory staff.
Culture responses were classified as “positive” when they described strong local infection control practices, a proactive approach to CRE prevention and/or management and/or heightened awareness of infection control and prevention for emerging pathogens (e.g., active screening for CRE in high-risk patients or a plan to incorporate active screening into current practices if or when CRE incidence increased). “Negative” culture responses described local knowledge about CRE as lacking and/or poor staff compliance with hand hygiene and/or use of personal protective equipment (PPE) (gowns or gloves):
It’s very challenging, I’ve been here a year. When I do environmental rounding and corrections, if someone is not [using] PPE for example, [I get] a lot of push back. I think [staff] think CRE is serious, because I’ll go talk to them face-to-face, if someone has a history of it. But MRSA, isn’t treated as seriously. [MPC]
Best practices examples for culture addressed interdisciplinary collaboration, education, and/or cross-training of laboratory staff on microbiology/ infection control practices.