Part 2: Identifying Research MethodologiesAfter reading each of the four peer-reviewed articles you selected, use the Matrix Worksheet template to analyze the methodologies applied in each of the four

Major Article A qualitative, interprofessional analysis of barriers to and facilitators of implementation of the Department of Veterans Affairs’Clostridium difficileprevention bundle using a human factors engineering approach Eric Yanke MD a,Helene Moriarty PhD, RN b,c,Pascale Carayon PhD d, Nasia Safdar MD, PhD e,f,* aDepartment of Medicine, William S. Middleton Memorial Veterans Hospital, Madison, WIbVillanova University College of Nursing, Villanova, PAcDepartment of Nursing, Corporal Michael J. Crescenz Veterans Affairs Medical Center, Philadelphia, PAdDepartment of Industrial and Systems Engineering, Center for Quality and Productivity Improvement, University of Wisconsin-Madison, Madison, WIeDepartment of Medicine, William S. Middleton Memorial Veterans Hospital and Division of Infectious Diseases, Madison, WIfUniversity of Wisconsin Medical School and Infection Control Department, University of Wisconsin-Madison, Madison, WI Key Words:Clostridium difficile infection prevention human factors engineering focus groups bundle Background:Clostridium difficileinfection (CDI) is increasingly prevalent, severe, and costly. Adherence to infection prevention practices remains suboptimal. More effective strategies to implement guidelines and evidence are needed.

Methods:Interprofessional focus groups consisting of physicians, resident physicians, nurses, and health technicians were conducted for a quality improvement project evaluating adherence to the Department of Veterans Affairs’ (VA) nationally mandatedC difficileprevention bundle. Qualitative analysis with a visual matrix display identified barrier and facilitator themes guided by the Systems Engineering Initiative for Patient Safety model, a human factors engineering approach.

Results:Several themes, encompassing both barriers and facilitators to bundle adherence, emerged. Rapid turnaround time ofC difficilepolymerase chain reaction testing was a facilitator of timely diagnosis. Too few, poorly located, and cluttered sinks were barriers to appropriate hand hygiene. Patient care work- load and the time-consuming process of contact isolation precautions were also barriers to adherence.

Multiple work system components serve as barriers to and facilitators of adherence to the VA CDI pre- vention bundle among an interprofessional group of health care workers. Organizational factors appear to significantly influence bundle adherence.

Conclusion:Interprofessional perspectives are needed to identify barriers to and facilitators of bundle implementation, which is a necessary first step to address adherence to bundled infection prevention practices.

Published by Elsevier Inc. on behalf of Association for Professionals in Infection Control and Epidemiology, Inc. Clostridium difficileinfection (CDI) is an increasingly prevalent, severe, and costly health care–associated infection worldwide. 1 Recent surveillance data indicateC difficileis responsible for nearly 500,000 infections and 29,000 deaths per year in the United States. 2 The economic burden of CDI in the United States is significant and likely exceeds $3 billion per year. 3 Although effective infection control practices are crucial for pre- ventingC difficiletransmission, 4health care worker (HCW) adherence remains suboptimal. 5Many health care institutions have created bundled infection control interventions to prevent CDI. However, these bundles can be difficult and complex to implement, even in the context of highly integrated health care systems. 6 Lack of HCW adherence to infection prevention processes is a complex issue. Previous research using focus groups suggests clin- ical guideline ambiguity (ie, uncertainty or vagueness in guidelines that prevents a system from achieving its purpose) is a prominent * Address correspondence to Nasia Safdar, MD, PhD., Department of Medicine, William S. Middleton Memorial Veterans Hospital, UWMF Centennial Building, 1685 Highland Ave, Madison, WI 53705.

E-mail address: [email protected] (N. Safdar).

Funding/support: N.S. and H.M. are supported by the VA Patient Safety Center of Inquiry.

Conflicts of interest: None to report. 0196-6553/Published by Elsevier Inc. on behalf of Association for Professionals in Infection Control and Epidemiology, Inc.https://doi.org/10.1016/j.ajic.2017.08.027American Journal of Infection Control 46 (2018) 276-84 Contents lists available at ScienceDirect American Journal of Infection Control journal homepage: www.ajicjournal.org American Journal of Infection Control theme when attempting to implement evidence-based practices to reduce health care–associated infections. 7Given this gap between knowledge and implementation, effective strategies for translat- ing evidence and guidelines into effective practice are needed. 8 The Systems Engineering Initiative for Patient Safety (SEIPS) model represents an innovative human factors engineering approach to patient safety. The SEIPS model has been applied extensively in the health care field, 9including in infection prevention. 6,10 At the core of the SEIPS model is the work system that encompasses multiple interacting components: a person, tasks, tools and technologies, the physical environment, and organizational conditions ( Fig 1 ). These 5 components are interrelated and influence care processes, such as implementation of a CDI bundle in health care settings. The De- partment of Veterans Affairs (VA) mandated implementation of a national CDI bundle at every VA hospital in early 2012, and imple- mentation of such CDI bundles likely reduces CDI rates.

11 Guided by the SEIPS model, we conducted focus groups to perform a quality improvement, work system analysis of the VA’s nationally mandated CDI prevention bundle relevant to health care providers’ adherence to the CDI bundle (testing and diagnosis, hand hygiene, and contact isolation precautions [CIP]). 12Antimicrobial stewardship is addressed under a separate VA initiative and not a component of this bundle.

METHODS Design In this qualitative, descriptive project, 4 focus groups were con- vened over a 5-month period to identify work system barriers and facilitators to implementation of the VA CDI bundle. In contrast with individual interviews, focus groups promote conversations about a range of perceptions and experiences, and provide opportuni- ties for group members to refine their comments based on feedback from others. 13,14 In accordance with our institution’s institutional review board exemption policy and self-certification tool, this project did not constitute research as defined under 45 CFR 46.102(d).Therefore, this quality improvement project was exempt from in- stitutional review board review.

Setting and participants The convenience sample consisted of attending hospitalist phy- sicians, internal medicine resident physicians, and registered nurses (RNs) and health technicians (HTs) employed at our VA hospital, an 87-bed facility. Eligibility criteria included the following: regular contact with inpatients on the general medicine units and ability to understand English. E-mails were sent to all attending physi- cians, resident physicians, and RNs and HTs working on the general inpatient medical units to briefly introduce the project and invite participation.

Procedure Four focus groups were conducted—1 with attending physi- cians, 1 with resident physicians, and 2 with RNs and HTs—between July and November 2013. The focus groups with attending physi- cians and resident physicians occurred during regular conference times. The RN and HT groups occurred outside their scheduled work hours; therefore, RNs and HTs received an hour of compensation time for their participation. No other compensation was provided, but light refreshments or lunch was provided. The attending phy- sician group had 7 participants, the resident physician group had 8 participants, and the RN and HT groups had 7 participants total.

The group facilitator (E.Y.) reviewed ground rules for confiden- tiality of the discussion and again reviewed the group’s purpose—to identify barriers to and facilitators of use of the VA nationally man- dated CDI prevention bundle. All groups were audio recorded with a digital recorder. The facilitator posed a series of open-ended ques- tions ( Appendices 1 and 2), guided by the SEIPS work system components and general literature on guideline implementation, and followed with probes to elicit elaboration. Another author (N.S.) recorded field notes during the groups to document nonverbal be- haviors and track the flow of communication. Duration of the focus WORK SYSTEM External Environment OUTCOMES PROCESS Fig 1.Systems Engineering Initiative for Patient Safety model. 277 E. Yanke et al. / American Journal of Infection Control 46 (2018) 276-84 groups averaged 45 minutes. Of note, the group facilitator (E.Y.) and other author (N.S.) did not have any direct, supervisory responsi- bilities for the participants, therefore avoiding any potential source of bias that could have influenced responses.

Qualitative analysis Digital recordings of focus groups were transcribed verbatim by a professional transcription service, after any identifying informa- tion was removed. The first author listened to the audio recordings while reviewing the transcripts to confirm accuracy of transcription.

Final transcript versions were uploaded to NVivo 8 (QSR Interna- tional, Melbourne, Australia) to facilitate data organization.

A template organizing approach 15organized data in a priori cat- egories corresponding to the interview questions derived from the 5 SEIPS work system components: person, organization, tools, tasks, and environment. Two investigators (E.Y. and H.M.) began open line- by-line coding of transcripts using in vivo terminology (actual words) to classify text units within each work system component, using a visual matrix display. 16This was followed by data reduction to group key text units into subthemes based on their similarity and fit with the 5 predetermined categories. 17,18 Our coding process ( Table 1 ) maximized the trustworthiness and efficiency of coding.

Trustworthiness and rigor Criteria for rigor in qualitative research—credibility, dependabil- ity, and confirmability—were followed to insure trustworthiness of the data. 19-23 Credibility (confidence in truth of the findings) was enhanced through the use of focus groups with different health pro- fessionals; multiple perspectives from these groups revealed a more complete picture of factors contributing to implementation of the CDI bundle. A member check also established credibility; the first author reviewed findings with 2 participants in each group, and all 8 persons confirmed that the preliminary findings and salient points accurately captured their ideas. Dependability of findings(consistency of findings) was enhanced through the use of 2 coders for analyses. One coder had minimal clinical expertise in infection control and no involvement with development of the interview tool or actual data collection; this added to dependability of data because it limited the potential for the coder’s prior experience to influ- ence the analysis. To support confirmability of the data, the 2 coders created an audit trail by recording theoretical and operational memos, thereby documenting the decisions they made through- out the analytic process.

FINDINGS Attending physician findings Testing and diagnosis Attending physicians identified many organizational features that facilitate CDI testing and diagnosis: more frequent testing; empir- ical contact isolation precautions (CIP) whenC difficiletesting is ordered; and near universal testing of symptomatic, newly admit- ted patients ( Table 2 ). One attending physician highlighted an institutional culture that assumes new diarrhea representsC difficile until proven otherwise: “Well, honestly, I think that at the moment they have diarrhea, they have C. diff.” In addition, attending phy- sicians recognized and appreciated nurses’ role (person factor) in recognizing earlyC difficilesymptoms, initiating early testing, and placing patients in CIP immediately when testing was ordered. At- tending physicians also reported that laboratory policies of only testing stool once per week and rejecting stool if nonliquid (despite other clinical indicators) were organizational barriers. Another person barrier that attending physicians recognized was their own un- awareness of many laboratory guidelines forC difficiletesting, particularly guidelines related to possible, atypical presentations.

Regarding theC difficilepolymerase chain reaction (PCR) test itself as a tool, attending physicians reported many positive aspects of the current test: more expedient (faster), efficient (1 test instead of 3), and more available (performed on weekends). Attending phy- sicians further described theC difficilePCR test as highly sensitive but also expressed concern about the cost and false positives that have occurred.

Hand hygiene From a person perspective, attending physicians reported un- certainty whether they should wash their hands inside or outside a patient’s room (task interacting with physical environment). They also expressed concern for how patients might react to the loca- tion and number of staff performing hand hygiene: “I wonder what the patient would think if a full team of five people all go in and out of his or her bathroom and wash their hands there.” Regard- ing tools for handwashing, attending physicians noted that sink water was often too hot; however, they described soap supplies as usually adequate. Environment barriers included clutter around patient sinks and the need to touch room curtains after washing hands in the room, possibly causing recontamination.

CIP Person factors that were barriers or facilitators to CIP were most commonly cited. Attending physicians reported a lack of clarity around many aspects of CIP: gown reuse, the level of patient contact requiring gowns, gown and gloving donning and doffing sequence, and when to discontinue CIP. They observed noncompliance of family members and food service workers with gowning. Attending phy- sicians also reported that HCWs engaged in brief tasks in patients’ room would not don gowns. However, they acknowledged the pro- active nursing practice of placing patients in CIP when testing is Table 1 Protocol for coding 1. Development of a matrix display for each of the 5 work system factors by group (attending physicians, resident physicians, and RNs and HTs).

2. Independent reading of each group transcript to achieve a deeper understanding of the whole gestalt.

3. Review and discussion of operational definitions and descriptions of the 5 work system factors in select articles by model developer and studies using the model.

4. Independent coding of 1 focus group, followed by comparison to evaluate consistency in coding text units under work system factors in the matrix.

5. Refinement of agreement around operational definitions for each work system factor to increase consistency in coding; addition of other interesting findings category to the matrix.

6. Independent coding of first focus group again by 2 coders with follow-up comparison to evaluate consistency.

7. After consistency in coding is achieved, independent coding of 3 other focus groups by 2 coders, who will memo and annotate liberally.

8. Follow-up comparisons to assess consistency between coders in coding text units in 3 other focus groups.

9. Review of matrix data display for all focus groups; text units under each work system factor will be condensed into distinct themes by each coder.

10. Discussion of themes under each work system factor to reach agreement between coders; agreed upon themes were then categorized as either barriers or facilitators to implementation of CDI bundle.

11. Examination of similarities and differences of themes within each group (attending physicians, resident physicians, and RNs and HTs) and among groups.

NOTE. Through discussion among 3 research team members, we developed a pro- tocol for coding the focus group transcripts that was followed by the 2 coders (2 coinvestigators).

CDI,Clostridium difficileinfection;HT, health technician;RN, registered nurse.

278E. Yanke et al. / American Journal of Infection Control 46 (2018) 276-84 ordered and reported that environmental services appears particularly effective and organized as it relates to CIP. Attending physicians also identified several organizational barriers: the emer- gency department (ED) and outpatient clinics do not routinely follow CIP, the hospital does not share compliance data with staff, and the policy and enforcement of family compliance with CIP are unclear; however, some attending physicians felt it was nursing’s respon- sibility to enforce CIP with families. Attending physicians specifically articulated their desire for data on the impact of the VAC difficile bundle to increase staff motivation for compliance. Organizational facilitators included consistent use of CIP whenC difficiletesting is ordered and the impression that environmental services is effec- tive despite attending physicians’ lack of awareness of their actual processes forC difficileroom cleaning: “I don’t think I have a clue what they do. The room gets empty, someone comes, they clean it. . .” Environment factors observed by attending physicians in- cluded barriers both inside and outside patient rooms. Gown and glove dispensers were in different locations (ie, gown dispenser outside patient rooms, glove dispenser inside patient rooms). Iso- lation stethoscopes were often missing inside patient rooms, and signs outside patient rooms were confusing and easy to miss: “I’ll be honest, if I’m entering and the patient is in precautions, I wear the gown. But have I entered without wearing a gown when I was supposed to? Yes, because I didn’t see the sign. I just totally missed the sign.” However, attending physicians acknowledged that the gown dispensers present outside all patient rooms were visible and accessible. Regarding the tools component, attending physicians de- scribed the supply of clean gowns as usually adequate; however, gown dispensers were sometimes empty. They expressed concern that bags of gowns in the intensive care units were not clearly la- belled if clean or dirty. Further, attending physicians described the overall task of CIP compliance as time-consuming and expressed concern that they spend less time talking to and examining pa- tients under CIP.Resident physician findings Testing and diagnosis Resident physicians identified a preponderance of person factors related to testing and diagnosis, particularly the variable testing thresholds among attending physicians and nurses ( Table 3 ). This resulted in “task ambiguity” regarding when exactly to orderC difficile testing and who makes the final decision to order testing as exem- plified by one resident physician’s comments: “[Resident physicians] get asked [by nurses], ‘Do you want to run a C. diff?’ and [resident physicians] say, ‘Um, maybe?’ I don’t know. Some attendings go one way, and some attendings say ‘Do not run a C. diff.’” Regarding the tool component, resident physicians feltC difficilePCR test was being performed more often, but this was unlikely to cause significant harm. Relevant to the task component, resident physicians re- ported occasional difficulty in actually obtaining a stool sample (eg, stool flushed in toilet, patient unable to produce stool) after order- ing testing.

Hand hygiene Resident physicians reported lacking education (person com- ponent) on the details of effective handwashing (eg, where to wash hands, handwashing technique). They also described the task of handwashing as time-consuming and the tools available for hand- washing as inadequate (eg, too few sinks, sink water too hot), as reflected in one resident physician’s comment:

I think it’s an access and time thing too. If you’ve got two resi- dents and then two medical students and an attending and all of you go in there and you’ve got 12 patients to round on post- call. Half your team is wandering down the hall continuing rounds, and half are still washing their hands. I don’t know if people are doing a good job then, either, or if it’s like slap some soap on, throw your hands under the water, and run. Table 2 Classification of themes identified in focus group for attending physicians (n=7) SEIPS component Testing and diagnosis Hand hygiene CIP Person •Unaware of laboratory guidelines for stool and atypical presentations •Identify and appreciate nurses’ role in testing •Lack of clarity around where handwashing should occur •Concern about patients’ reactions to staff handwashing location (eg, patient bathroom) •Lack of clarity around many aspects of CIP•Observation of lack of gowning by family, food service, and also by HCWs during brief patient contacts •Nurses are proactive in initiating CIP before testing results back •ES is effective with CIP Task•Adherence is time-consuming•Less time spent with patients under CIP Tools •Many positive aspects of newer PCR test (expedient, efficient, available, and highly sensitive) •Sinks (water too hot)•Soap supplies adequate •Supplies usually adequate•ICU gown bags without clear labels as to dirty or clean status Organization •Barriers to testing related to laboratory policy •Facilitators to testing (increased frequency, almost universal testing of new admissions, assumption of CD with diarrhea until testing proves otherwise) •Patient placed in CIP once CD test ordered •ER and outpatient clinics do not institute CIP•Hospital does not share bundle compliance data•Desire data on impact of bundle to increase motivation for bundle adherence •Family compliance and enforcement responsibilities not clear •Nursing with responsibility to enforce with families•CIP begin when testing ordered to reduce transmission•Physicians see ES as effective, but lack knowledge of processes for cleaning room Environment •Sinks cluttered•Potential contamination from room after handwashing •Problems with location of equipment (gowns and gloves) and containers •Problems with signs•Stethoscopes missing in CIP rooms•Visible and accessible wall gown dispensers CD,Clostridium difficile;CIP, contact isolation precautions;ER, emergency room;ES, environmental services;HCW, health care worker;ICU, intensive care unit;PCR, poly- merase chain reaction;SEIPS, Systems Engineering Initiative for Patient Safety. 279 E. Yanke et al. / American Journal of Infection Control 46 (2018) 276-84 CIP Many person factors were identified as barriers to CIP. Resi- dent physicians were unaware of the VAC difficilebundle despite being very familiar with the individual bundle components (eg, gowning, gloving, hand hygiene). Resident physicians reported they lacked education regarding CIP details (eg, how to remove gowns, how long to keep patients in CIP). In addition, they voiced concern about the adherence of other HCWs but also acknowledged they do not consistently adhere to CIP themselves. Resident physicians were particularly concerned about family members’ lack of CIP ad- herence and resulting risk of CDI spread, and they reported it was unclear who should be educating and enforcing CIP with families.

Tool factors were also commonly reported. Inappropriate person- al stethoscope use, low-quality isolation stethoscopes, and lack of Electronic medical record functionality were described as signifi- cant barriers to resident physician adherence to CIP, whereas consistently stocked gowns and consistent use of CIP for any pos- sible enteric infection (eg, viral gastroenteritis) were facilitators.

Resident physicians noted that the environment outside the pa- tients’ rooms did not effectively and consistently alert providers to patients under CIPs (eg, signs were not highly visible), and there was a lack of adequate equipment (eg, gown hampers, isolation stethoscopes) in the environment inside the room. From an orga- nization standpoint, resident physicians described a culture of equality and teamwork that facilitates effective CIP use: “I feel like people take it pretty seriously. I appreciate that if they notice someone is leaving and only uses hand sanitizer, I think anybody is comfortable pointing it out, regardless of what your role is on the health care team. So, I think that is a positive as an institution.” RN and HT findings Testing and diagnosis RNs and HTs recognized multiple person factors as facilitators ofC difficiletesting and diagnosis ( Table 4 ). They reported strong knowledge ofC difficilerisk factors, symptoms, and possible mimickers and good RN and HT collaboration, which allowed ini- tiation of earlier testing. RNs reported orderingC difficiletesting despite external resistance (from providers and other facilities) but also described difficulty in obtaining a stool sample from a patient, which could delay diagnosis (related to task and organization com- ponents). Overall, RNs believed the risk-benefit ratio favored more frequentC difficiletesting: “I’ve given a call and said, ‘Hey, this is what’s going on. Do you mind if we do a C. diff?’ I’ve never hadanyone say no because I think they trust our judgment at the bedside and know what we are seeing or if there was a big change.” Organization themes were next most common factors identi- fied by RNs and HTs. RNs felt they usually had institutional support for ordering testing independently. However, they also described variable RN-provider communication when actually orderingC difficiletesting. Other organizational barriers they recognized were laboratory guidelines that only allow testing of liquid stool and prob- lems with the ED (pressure from the ED to admit patients, missed CDI diagnoses in the ED). Regarding theC difficilePCR test, a tool, RNs noted that rapid test results facilitate CDI diagnosis and reduce unnecessary room changes, and RN concern forC difficiletransmis- sion motivates their initiation of earlyC difficilePCR testing.

Hand hygiene Tools were most commonly identified as barriers to effective hand hygiene: problems with soap dispensers (empty or malfunction- ing), problems with sinks (water too hot or not automatic), and clutter on patient sinks. Signs on hand sanitizers directing staff to wash hands and the recent addition of an extra sink in the hallway were viewed as facilitators of hand hygiene. Environment barriers were also cited: inadequate number of sinks and poor placement of sinks. Broken soap dispensers were described as being fixed quickly from an organization standpoint. However, the task of hand- washing itself was reported as time-consuming, and RNs and HTs expressed desire for a faster method of hand hygiene, as reflected in one comment: “It would be great to have a different product. Of course, everyone wants a product that’s easy, that’s quick, you know, that would do the same job as hand washing. That’s hard.” CIP RNs and HTs described a culture of institutional support for CIP compliance and support for independent RNC difficiletesting and decision-making. For example, RNs would independently start CIP whenC difficiletesting was ordered. RNs and HTs felt most staff were compliant with CIP, and they usually felt comfortable pointing out lapses in compliance; however, this was harder with physician non- compliance. They also reported appreciation of environmental services’ work in preventingC difficiletransmission. However, unclear policies surrounding universal gown use in rooms and disposal of C difficilepatient feces were depicted as organization barriers. Also related to organization culture, RNs expressed their perception of nursing staff being primarily responsible for preventingC difficile transmission: “If there [were] a transmission, I guarantee most of Table 3 Classification of themes identified in focus group for resident physicians (n=8) SEIPS component Testing and diagnosis Hand hygiene CIP Person •Threshold for testing varies•Ambiguity regarding who orders testing and when •Lack education regarding how and where •Residents aware of individual bundle components, but not bundle as a whole•Resident lack of knowledge and education•Staff with inconsistent adherence to CIP•Concern about CD transmission with lack of family member adherence to CIP•Unclear who is educating and enforcing CIP with families Task •Challenges in obtaining specimen for testing •Time-consuming Tools •Risk-benefit ratio for CD testing •Sinks (number, characteristics, and water too hot) •Stethoscopes (low quality or inappropriate use)•EMR functionality•Adequate gown supply•Consistent CIP use for enteric infections Organization•Culture of equality and teamwork•Inconsistent adherence to bundle•Lack knowledge of ES role Environment•Inside room (lack of equipment)•Outside room (visual cues for isolation and CD status lacking) CD,Clostridium difficile;CIP, contact isolation precautions;EMR, Electronic medical record;ES, environmental services;SEIPS, Systems Engineering Initiative for Patient Safety. 280E. Yanke et al. / American Journal of Infection Control 46 (2018) 276-84 us would feel like it’s our fault.” RNs and HTs described several person facilitators of CIP: (1) RNs independently review patients’ history and symptoms and order CIP if indicated; (2) the infection control nurse is a significant resource who facilitates timely CIP ini- tiation and discontinuation; (3) environmental services are knowledgeable and willing to educate other HCWs on CIP room cleaning; and (4) food service and visitors are increasingly com- pliant with CIP. RNs and HTs were largely unaware of the VAC difficile bundle but felt the VA has been doing a good job of reducing CDI transmission. However, RNs and HTs also observed variable family member compliance with CIP despite RN education: “We will explain to the family members why they should be wearing a gown. A lot of times, they won’t. They say, ‘I have been around him for the last month, if I am going to have it, I have it.’ And then we just make sure we push hand washing, going in, going out. We can’t force them to wear a gown.” Multiple tool factors were viewed as barriers: (1) isolation gowns fell off easily, (2) personal provider stethoscopes were used inap- propriately in CIP rooms, (3) isolation stethoscopes were occasionally inappropriately removed from CIP rooms, (4) lack of bleach wipes, (5) overloaded and poorly maintained soiled linen bags, (6) lack of pens in CIP rooms, and (7) lack of EMR functionality (linkingC difficile testing to CIP ordering). RNs and HTs identified task barriers of CIPs, in particular their increased workload when caring for multiple CIP patients, leading to their inadvertent noncompliance with CIP when nurses were very busy: “[You] spend half your day gowning up.”They also acknowledged their awareness of the increased work- load for environmental services in cleaning a CIP room, which typically requires an hour. Environment factors noted included overall good accessibility of CIP supplies (eg, gowns, gloves, isolation stetho- scopes) and clear CIP signs outside of patient rooms.

DISCUSSION To our knowledge, this is the first qualitative, interprofessional analysis of the VA nationally mandatedC difficileprevention bundle using the SEIPS model. Our analysis suggests all components of the SEIPS work system function as both facilitators of and barriers to compliance with the VAC difficileprevention bundle.

Several common themes, encompassing both barriers and fa- cilitators, emerged among interprofessional groups related to CDI testing and diagnosis. Both RNs and HTs and attending physicians described the rapid turnaround time of theC difficilePCR test (tool) as a facilitator of timely diagnosis. Highlighting the value of mul- tiple perspectives with interprofessional focus groups, RN and HTs specifically noted the benefit of rapid test results in avoiding un- necessary room changes (eg, in the setting of double patient rooms), a theme absent in resident or attending physician groups. This is consistent with previous, multidisciplinary focus groups analyses that have documented differences in perception of value of patient safety practices between nurses and physicians. 24 Table 4 Classification of themes identified in RN and HT focus groups (n=7) SEIPS component Testing and diagnosis Hand hygiene CIP Person •RNs identify CDI risk factors, symptoms, and mimics, which allow them to initiate earlier testing •HTs collaborate with RNs for identification of CDI •RNs order CD testing despite external resistance •Delay in obtaining stool sample delays diagnosis •RNs believe that risk-benefit ratio supports more frequent testing •Many RNs unaware of bundle •RNs review patient history and symptoms and order CIP independently •Resource of infection control nurse facilitates initiating and discontinuing CIP •Variable family adherence to CIP despite RN education•ES is knowledgeable and willing to education others on CIP room cleaning •Adherence to CIP by food service and visitors has improved Task •Handwashing time-consuming•Desire for faster methods •Increased RN and HT time and workload•RNs aware CIP room cleaning is time-consuming (approximately 1 h) •Inadvertent noncompliance with CIP when nurses very busy Tools •Concern for CDI transmission drives initiation of early testing •Rapid PCR testing facilitates CDI diagnosis and reduces unnecessary room changes •Soap dispenser problems (empty or broken) •Sink problems (hot water, manual faucets, or clutter on sink) •Signs on hand sanitizers direct staff to wash hands •Addition of sink in hallway •Adequate gown supplies•Gowns fall off easily•Equipment and supply problems•Personal stethoscopes are inappropriately used by providers•Isolation stethoscopes meant for patient rooms are taken outside •No automatic EMR ordering of CIP when ordering CDI testing Organization •RNs usually have organizational support for independently ordering test •Variable RN-provider communication when ordering CDI testing •Laboratory guidelines only allow for testing loose stools •Problems with ED (pressure and missed diagnosis) •Broken soap dispensers fixed quickly •Unclear policies on always wearing gowns in CIP rooms and disposal of CIP patient feces •Most staff adherent to CIP; RNs usually comfortable pointing out lapses in CIP, although harder to do with MDs •Culture of institutional support for CIP•RNs have organizational support for independent testing and decision-making •RNs appreciate role and workload of ES in preventing CDI transmission •RNs think an EMR prompt screening for CDI symptoms would expedite early testing •Ordering CD testing prompts nurses to start IP Environment •Problems with sinks (too few or poor location) •Accessibility of CIP supplies (eg, gowns, gloves, stethoscopes)•CIP signs are clear CD,Clostridium difficile;CDI,Clostridium difficileinfection;CIP, contact isolation precautions;EMR, Electronic medical record;ES, environmental services;HT, health tech- nician;IP, infection prevention;MD, medical doctor;PCR, polymerase chain reaction;RN, registered nurse;SEIPS, Systems Engineering Initiative for Patient Safety. 281 E. Yanke et al. / American Journal of Infection Control 46 (2018) 276-84 Attending physicians were generally appreciative of RNs’ proactive role in ordering and sending CDI testing. However, resident physi- cians described ambiguity regarding who enters the order for testing and who ultimately decides whether the test should be run, an example of responsibility ambiguity related to guideline adherence. 7 The issue of RN- versus physician-initiatedC difficiletesting is also particularly relevant in light of recent evidence suggesting discor- dance between physician and RN perception of CDI risk factors. 25 Identification of barriers to hand hygiene was remarkably con- sistent across all professional groups. Members from all groups described sink-related (tool) barriers: too few sinks and water that was too hot. Closely related were the environmental barriers of poor sink location and cluttered sinks and the time-consuming task of handwashing (as opposed to alcohol-based hand sanitizer). Resi- dent physicians and attending physicians both reported a lack of knowledge (person) regarding the exact technique and location for handwashing (eg, inside patient’s room vs outside patient’s room).

The prominence of these physical, tangible themes, particularly related to sink use, is consistent with previous qualitative analysis of HCWs’ perception of barriers to prevention of methicillin- resistantStaphylococcus aureustransmission. 26However, studies of increasing the number of available sinks have been underwhelm- ing regarding improved handwashing compliance. 27This further underscores the need for evaluation of multiple work system factors (beyond the environment) when pursuing health care system design and argues in favor of an interdisciplinary design process involv- ing both architectural and clinical staff.

CIP themes were more diverse and wide-ranging. However, the barrier of workload and time (task) was prominent and consistent among most groups, which has been previously well described in the infection prevention literature. 26,28 Several RNs reported inad- vertent noncompliance with CIP in the setting of a heavy workload with multiple patients under CIP. Attending physicians similarly ad- mitted they probably spend less time with CIP patients because of time constraints. The phenomenon of health care providers spend- ing less time with patients in CIP has been objectively documented previously and is considered one of several significant adverse con- sequences of CIP. 29Given the prominence of workload and time barriers associated with providing health care to CIP patients in our project and others, the data as a whole suggest consideration of modifying health care providers’ workload and number of pa- tients assigned related to infection prevention needs (ie, limiting the number of patients in CIP, assigning fewer total patients).

An additional RN and HT identified barrier was the challenge of obtaining, appropriately cleaning, and disposing of common sup- plies or tools needed in patient rooms (eg, pens, whiteboard markers, stethoscopes). The lack of clear policy on this issue was also raised, reflecting the organization component of this barrier as well. Con- cerns regarding how to implement theC difficilebundle in multiple contextualized patient situations (eg, how to correctly use a pen with CIP) suggest that simulation may be a useful technique to prepare clinicians for safely making difficult decisions in complex patient care environments. In light of recent evidence suggesting HCW gown and glove adherence is generally poor, simulation may become crucial to ensure infection prevention practice adherence in high- stakes circumstances, particularly given the recent emergence of the Ebola virus. 30,31 RNs and HTs reported overall good staff adherence to CIP and also described a culture of institutional support for CIP, both facili- tators of CIP adherence within the organization.RNs further stated they were generally comfortable pointing out lapses in CIP but were less comfortable pointing out lapses to physicians. Reluctance of nursing staff to remind physicians of infection prevention prac- tices lapses has been described in the literature, but overall has received little attention. 28Previous studies also suggest veteranpatients are similarly reluctant to identify lapses in hand hygiene of VA physicians. 32Our findings indicate that further research on the dynamics and perceived hierarchy among nurses, physicians, and patients could offer insight into potential improvement in com- munication processes contributing to HCW compliance with infection prevention practices.

All professional groups reported poor visitor (usually family member) adherence to CIP as a significant barrier. Both physician groups reported observing visitors not following CIP and ex- pressed uncertainty regarding who is responsible for visitors’ education and CIP enforcement. RNs reported taking responsibili- ty for visitor CIP education, but they described frequent visitor CIP nonadherence despite providing education. A recent literature review found visitor adherence to isolation precautions (including hand hygiene and CIP) to be variable but generally poor. 33To date, no study has specifically evaluated visitor adherence to CIP in patients with CDI. Recent survey data also suggest that CIP education is not con- sistently provided to visitors, in either written or in-person formats. 33 Our findings reinforce the need for further research on the role of hospital visitors in infection prevention and the need to identify better methods to proactively engage patients and visitors in in- fection prevention practices.

A prominent global theme among all professional groups was a lack of knowledge of the VA CDI bundle as a whole while dem- onstrating knowledge of many, but not all, individual elements of the CDI bundle. Most participants were aware of the role of hand hygiene and CIP, but knowledge of timely testing and diagnosis as a bundle element was low. This finding has implications for the success of the CDI bundle overall: effective bundles are composed of multiple elements, all of which are necessary, sufficient, and must be consistently and uniformly executed. 34Lack of knowledge of the complete bundle jeopardizes its effective execution by HCWs.

Our project had several strengths. The qualitative, interprofessional nature of our project guided by a human factors and systems en- gineering model provided detailed insight into the current state of the VA nationally mandated CDI bundle. Consistent with previous infection prevention focus group research, 7our experience sup- ports the utility and value of focus groups in illuminating work system factors from multiple perspectives in a safe and candid manner. These multiple HCW perspectives are essential to identify barriers to bundle implementation and adherence, which often differed significantly among HCW groups. Identification of facilitators is equally impor- tant for development of protocols and interprofessional initiatives to modify work system factors with the ultimate goal of improving bundle implementation, adherence, and effectiveness.

Our project also had limitations. Addition of laboratory person- nel would have further strengthened our interprofessional approach, particularly given the frequency with which laboratory policy and procedure was referenced during focus group interviews. Similar- ly, focus groups with family members, hospital visitors, and patients would provide an even broader perspective. We also did not include direct, visual observations of HCWs using CIP as part of this project; however, we have reported such findings previously. 6 Our focus groups were somewhat smaller than typically found in social science research, but such smaller groups are often ideal for addressing sensitive issues (eg, bundle adherence) in health research. 35,36 This descriptive project was a necessary first step to better understand work system factors related to bundle imple- mentation. Future research should include larger samples from multiple VA and non-VA sites to support generalizability of find- ings to other samples and settings.

All groups recognized the role of environmental services in re- ducing CDI transmission, and we conducted environmental services focus group interviews because of the importance of their role. These findings will be reported in a future article. 282E. Yanke et al. / American Journal of Infection Control 46 (2018) 276-84 In conclusion, multiple work system components serve as barriers to and facilitators of adherence to the VA nationally mandated CDI prevention bundle among an interprofessional group of HCWs. Rapid C difficiletest results were a common facilitator of timely CDI di- agnosis, whereas the increased time required for handwashing and CIP were prominent barriers among all groups. Organizational and environmental factors influencing bundle adherence identified in this project also point to important directions for systems and environ- mental redesign. Taken as a whole, our findings underscore the need for interprofessional perspectives and collaboration to identify bar- riers to and facilitators of adherence to infection prevention practices.

References 1.Dubberke E. Clostridium difficile infection: the scope of the problem. J Hosp Med 2012;7(Suppl):S1-4.

2.Lessa FC, Mu Y, Bamberg WM, Beldavs ZG, Dumyati GK, Dunn JR, et al. Burden of Clostridium difficile infection in the United States. N Engl J Med 2015;372:825-34.

3.O’Brien JA, Lahue BJ, Caro JJ, Davidson DM. The emerging infectious challenge of clostridium difficile-associated disease in Massachusetts hospitals: clinical and economic consequences. Infect Control Hosp Epidemiol 2007;28:1219-27.

4.Vonberg RP, Kuijper EJ, Wilcox MH, Barbut F, Tüll P, Gastmeier P, et al. Infection control measures to limit the spread of Clostridium difficile. Clin Microbiol Infect 2008;14(Suppl):2-20.

5.Dhar S, Marchaim D, Tansek R, Chopra T, Yousuf A, Bhargava A, et al. Contact precautions: more is not necessarily better. Infect Control Hosp Epidemiol 2014;35:213-21.

6.Yanke E, Zellmer C, Van Hoof S, Moriarty H, Carayon P, Safdar N. Understanding the current state of infection prevention to prevent Clostridium difficile infection:

a human factors and systems engineering approach. Am J Infect Control 2015;43:241-7.

7.Gurses AP, Seidl KL, Vaidya V, Bochicchio G, Harris AD, Hebden J, et al. Systems ambiguity and guideline compliance: a qualitative study of how intensive care units follow evidence-based guidelines to reduce healthcare-associated infections.

Qual Saf Health Care 2008;17:351-9.

8.Yanke E, Carayon P, Safdar N. Translating evidence into practice using a systems engineering framework for infection prevention. Infect Control Hosp Epidemiol 2014;35:1176-82.

9.Carayon P, Schoofs Hundt A, Karsh BT, Gurses AP, Alvarado CJ, Smith M, et al.

Work system design for patient safety: the SEIPS model. Qual Saf Health Care 2006;15(Suppl):i50-8. 10. Carayon P, Wetterneck TB, Rivera-Rodriguez AJ, Hundt AS, Hoonakker P, Holden R, et al. Human factors systems approach to healthcare quality and patient safety.

Appl Ergon 2014;45:14-25. 11. Barker AK, Ngam C, Musuuza JS, Vaughn VM, Safdar N. Reducing clostridium difficile in the inpatient setting: a systematic review of the adherence to and effectiveness of C. difficile prevention bundles. Infect Control Hosp Epidemiol 2017;38:639-50. 12. U.S. Department of Veterans Affairs. Implementation of guideline for prevention of clostridium difficile infection (CDI). In: Veterans Health Administration (VHA) acute-care facilities (VAIQ # 7210267). Washington (DC): Department of Veterans Affairs; 2012. 13. Krueger R, Casey MA. Focus groups: a practical guide for applied research. 3rd ed. Thousand Oaks (CA): SAGE Publications; 2000. 14. Johnson B, Turner LA. Data collection strategies in mixed methods research. In:

Tashakkori A, Teddlie C, editors. Handbook of mixed methods in social & behavioral research. Thousand Oaks (CA): SAGE Publications; 2003. pp. 297-319. 15. Miller WL, Crabtree BF. The dance of interpretation. In: Crabtree BF, Miller WL, editors. Doing qualitative research. 2nd ed. Thousand Oaks (CA): Sage Publications; 1999. pp. 127-44. 16. Averill JB. Matrix analysis as a complementary analytic strategy in qualitative inquiry. Qual Health Res 2002;12:855-66. 17. Graneheim UH, Lundman B. Qualitative content analysis in nursing research:

concepts, procedures and measures to achieve trustworthiness. Nurse Educ Today 2004;24:105-12. 18. Hsieh HF, Shannon SE. Three approaches to qualitative content analysis. Qual Health Res 2005;15:1277-88. 19. Gearing RE. Bracketing in research: a typology. Qual Health Res 2004;14:1429-52. 20. Guba EG. Criteria for assessing the trustworthiness of naturalistic inquiries. ECTJ 1981;29:75-91. 21. Morse JM. Designing funded qualitative research. In: Denzin NK, Lincoln YS, editors. Handbook of qualitative research. Thousand Oaks (CA): Sage Publications; 1994. pp. 220-35. 22. Sandelowski M. The problem of rigor in qualitative research. ANS Adv Nurs Sci 1986;8:27-37. 23. Walsh D, Downe S. Appraising the quality of qualitative research. Midwifery 2006;22:108-19. 24. McDonald R, Waring J, Harrison S, Walshe K, Boaden R. Rules and guidelines in clinical practice: a qualitative study in operating theatres of doctors’ and nurses’ views. Qual Saf Health Care 2005;14:290-4. 25. Blakney R, Yanke E, Fink C, Wigton R, Safdar N. Optimizing diagnostic testing for Clostridium difficile: the perceptions of physicians and nurses on when to order testing for C difficile. Am J Infect Control 2015;43:889-91. 26. Seibert DJ, Speroni KG, Oh KM, Devoe MC, Jacobsen KH. Preventing transmission of MRSA: a qualitative study of health care workers’ attitudes and suggestions.

Am J Infect Control 2014;42:405-11. 27. Whitby M, McLaws ML. Handwashing in healthcare workers: accessibility of sink location does not improve compliance. J Hosp Infect 2004;58:247-53. 28. Jang JH, Wu S, Kirzner D, Moore C, Youssef G, Tong A, et al. Focus group study of hand hygiene practice among healthcare workers in a teaching hospital in Toronto, Canada. Infect Control Hosp Epidemiol 2010;31:144-50. 29. Dashiell-Earp CN, Bell DS, Ang AO, Uslan DZ. Do physicians spend less time with patients in contact isolation?: a time-motion study of internal medicine interns.

JAMA Intern Med 2014;174:814-5. 30. Mitchell R, Roth V, Gravel D, Astrakianakis G, Bryce E, Forgie S, et al. Are health care workers protected? an observational study of selection and removal of personal protective equipment in Canadian acute care hospitals. Am J Infect Control 2013;41:240-4. 31. Zellmer C, Van Hoof S, Safdar N. Variation in health care worker removal of personal protective equipment. Am J Infect Control 2015;43:750-1. 32. Lent V, Eckstein EC, Cameron AS, Budavich R, Eckstein BC, Donskey CJ. Evaluation of patient participation in a patient empowerment initiative to improve hand hygiene practices in a Veterans Affairs medical center. Am J Infect Control 2009;37:117-20. 33. Munoz-Price LS, Banach DB, Bearman G, Gould JM, Leekha S, Morgan DJ, et al.

Isolation precautions for visitors. Infect Control Hosp Epidemiol 2015;36:747- 58. 34. Resar R, Griffin FA, Haraden C, Nolan TW. Using care bundles to improve health care quality. Cambridge (MA): Institute for Healthcare Improvement; 2012. 35. Côté-Arsenault D, Morrison-Beedy D. Maintaining your focus in focus groups:

avoiding common mistakes. Res Nurs Health 2005;28:172-9. 36. Tausch AP, Menold N. Methodological aspects of focus groups in health research.

Glob Qual Nurs Res 2016;3:2333393616630466. APPENDIX 1.

Interview guide for focus groups with attending physicians and resident physicians Person What do you know about the VA nationalC difficilebundle?

Do you believe the currentC difficilebundle reduces transmis- sion ofC difficileat the VA, that is, do you think the bundle is effective?

Organization Do you feel the VA as a whole or our VA hospital makes reduc- ing the transmission ofC difficilean organizational priority?

How/why?

Tools What are barriers to testing patients forC difficilediarrhea in the hospital?

What are barriers to diagnosing patients withC difficilediar- rhea in the hospital?

Tasks What are barriers to placing patients in isolation precautions for C difficilediarrhea in the hospital?

What are barriers to using contact isolation precautions for pa- tients withC difficilediarrhea in the hospital?

Environment What are barriers to performing appropriate hand hygiene after caring for patients withC difficilediarrhea in the hospital? 283 E. Yanke et al. / American Journal of Infection Control 46 (2018) 276-84 APPENDIX 2.

Interview guide for 2 focus groups with RNs and HTs Person What do you know about the VA nationalC difficilebundle (en- vironmental management, hand hygiene, contact precautions, and cultural transformation)?

How well do the currentC difficilepolicies reduce transmission ofC difficileat the VA, that is, do you think the bundle is effective?

Organization How well does the Madison VA make reducing the transmis- sion ofC difficilean organizational priority?

How well do all members of the patient care team (medical doctors, RNs, certified nursing assistants, patient transport, food service, and environmental services) appropriately share and take responsibility for reducingC difficiletransmission?

How important do you think environmental services and room cleaning is to reducing the transmission ofC difficile?

Tools What are barriers to testing and diagnosing patients withC difficile diarrhea in the hospital?How often do you suggestC difficiletesting to a physician (res- ident or staff medical doctor)? How well is this suggestion received?

What will make you more likely to suggestC difficiletesting to the primary team (eg, stool features, patient features)?

Tasks What are barriers to placing patients in isolation precautions for C difficilediarrhea in the hospital?

What are barriers to using contact isolation precautions for pa- tients withC difficilediarrhea in the hospital?

How comfortable do you feel pointing out when other members of the patient care team do not adhere toC difficileprecautions (eg, handwashing, gowning, gloving)?

Environment What are barriers to performing appropriate hand hygiene after caring for patients withC difficilediarrhea in the hospital?

How long do you typical see patients kept inC difficileprecau- tions? Do medical doctors ever discontinue isolation precautions on a patient withC difficile? Do you ever ask medical doctors to dis- continueC difficileprecautions?

How could the layout of the rooms be changed to make contact isolation precautions easier to use?

284E. Yanke et al. / American Journal of Infection Control 46 (2018) 276-84