Establishing an Antimicrobial Stewardship Collaborative Across a Large, Diverse Health Care System

The Joint Commission Journal on Quality and Patient Safety 2019; 45:591–599

Establishing an Antimicrobial Stewardship Collaborative Across a Large, Diverse Health Care System Andrea Y. Logan, PharmD, BCPS; Julie E. Williamson, PharmD, BCPS; Emily K. Reinke, PhD; Steven W. Jarrett, PharmD; Michael S. Boger, MD, PharmD; Lisa E. Davidson, MD

Background: Alarming trends in antibiotic resistance sparked a National Action Plan endorsing antimicrobial stewardship programs (ASPs) in health care facilities. Atrium Health consists of 28 acute care facilities with varying levels of ASP maturity. The organization sought to establish an ASP collaborative across a diverse network by uniting local resources with a central advisory team. Methods: In fall 2015 each facility chose a pharmacist, a physician, and an administrative ASP champion. Broad-spectrum antibiotic use was tracked monthly using days of therapy (DOT) per 1,000 patient-days as a standard metric. A gap analysis survey of Centers for Disease Control and Prevention (CDC) core elements for ASPs was conducted to stratify facilities into one of three tiers, with Tier 1 having the most comprehensive ASP. Baseline antibiotic usage data were collected, and DOT reduction goals were set for each facility. Site visits were conducted in winter 2016, and a post-visit summary outlining major goals was provided. Pharmacists held monthly facility meetings to assess progress and a bimonthly virtual meeting for sharing best practices networkwide. In addition, curriculum for an ASP symposium was developed based on identified educational needs. Results: Almost all hospitals (25/28) fully implemented the CDC core elements for ASPs within the first year of establishing the systemwide collaborative. Most facilities (78.6%) achieved their DOT reduction goal ranging from 1%–2.5% to 5%–10%. Conclusion: Despite many challenges, building a unified ASP collaborative across a diverse system enabled many hospitals to adopt best practices and improve antimicrobial use.

A

n alarming rise in infections caused by multidrugresistant organisms (MDROs), coupled with a decline in the development of new antibiotics to address this threat, has created a global public health crisis. Antibiotic-resistant infections have been associated with increased morbidity and mortality, and if current trends continue, estimates predict that MDROs will kill more patients than cancer by 2050. 1 Antimicrobial prescribing patterns are a major contributing factor to the emergence of antibiotic resistance. A staggering 55% of patients receive at least one antibiotic during a hospital admission, but data suggest that 30%–50% of antibiotics prescribed in the acute care setting are inappropriate or unnecessary. 2,3 Further compounding this issue has been the trend toward use of more broad-spectrum antibiotics (for example, carbapenems, vancomycin) in US hospitals. 3 Fluoroquinolones, a class of drugs highly associated with Clostridium difficile infection (CDI) and adverse effects, also remain commonly prescribed. 3–6 To preserve the effectiveness of currently available antibiotics, optimal and judicious use through antimicrobial stewardship programs (ASPs) is essential. Government and regulatory bodies advocate for the implementation of ASPs in the acute care setting as a principal strategy in combatting antimicrobial resistance. In 2015,

the National Action Plan for Combatting AntibioticResistant Bacteria was developed, calling for a 20% reduction in antibiotic use in acute care settings by 2020 through ASPs. 7 The Centers for Disease Control and Prevention (CDC) has long recommended that all hospitals have an ASP and proposed seven core elements for hospital ASPs (Table 1). 8 In 2016, The Joint Commission mandated that all acute care and critical access hospitals establish ASPs whose activities encompassed all of CDC’s core elements. 9 Full adoption of these core elements would be no small task, considering the 2014 National Healthcare Safety Network (NHSN) survey findings showing that only 39% of US hospitals reported achievement of all seven core elements. 10 Smaller hospitals were even less likely to adopt these core elements into their ASPs. While 56% of larger institutions (more than 200 beds) reportedly met all seven core elements, only 22% of critical access hospitals (50 beds or fewer) and 39% of small hospitals (51–200 beds) reported full implementation. 10 Application of these elements is challenging in small and critical access hospitals that lack infectious disease (ID) consultation services and dedicated resources for an ASP. 11 In addition, small hospitals consume broad-spectrum antibiotics at rates similar to those of larger hospitals despite a lower case mix index, which further

1553-7250/$-see front matter © 2019 The Joint Commission. Published by Elsevier Inc. All rights reserved. https://doi.org/10.1016/j.jcjq.2019.03.002

592

Andrea Y. Logan, PharmD, BCPS, et al.

Table 1. Description of the CDC ASP Core Elements for Acute Care Hospitals Core Element

Description

1. Leadership Dedicating necessary human, financial, and incommitment formation technology resources 2. Accountability Appointing a single leader, preferably a physician, to be responsible for program outcomes, with support from a multidisciplinary team 3. Drug expertise Appointing a single pharmacist leader responsible for improving antibiotic use 4. Action Implementing at least one recommended action, such as systemic evaluation of ongoing treatment need after a set period of initial treatment 5. Tracking Monitoring antibiotic prescribing and resistance patterns 6. Reporting Regular reporting information on antibiotic use and resistance to physicians, nurses, and relevant staff 7. Education Educating clinicians about resistance and optimal prescribing CDC, Centers for Disease Control and Prevention; ASP, antimicrobial stewardship program.

emphasizes the importance of establishing ASPs within these care settings. 12,13 With these unique challenges, the formation of multihospital collaboratives is increasingly recognized as a strategy to facilitate building ASPs at institutions with limited resources. 11,14,15 This article describes the first year of forming a systemwide stewardship collaborative across a diverse health care system.

METHODS Setting

Atrium Health (AH), formerly known as Carolinas HealthCare System, consists of 28 acute care facilities providing care in North Carolina, South Carolina, and Georgia. Hospital services range from critical access hospitals to large academic medical centers with solid organ and bone marrow transplant services. Of the 28 hospitals within the AH network, 19 (67.9%) are classified as small hospitals (SHs, 51–200 beds) or critical access hospitals (CAHs, 50 beds or fewer). The majority of facilities (57.1%; 16/28) exist within one of three regional networks (RNs), which share select resources, polices, and protocols. RNs consist of 3–10 hospitals ranging in size from 47 to 827 beds, and most smaller hospitals within these RNs have access to in-person or remote (that is, telehealth) ID consultation services. In contrast, among the 35.7% (10/28) of hospitals without ID consultation services, almost all (8/10) are SHs or CAHs not associated with an RN. Four different electronic medical record (EMR) systems are used across the network. Appendix 1 (available in online article) summarizes AH acute care facility characteristics.

Establishing an Antimicrobial Stewardship Collaborative

In late 2015, foundational efforts to create an ASP framework began across the AH network. Initial steps included identifying local resources and champions and pairing them with an advisory team of content experts led by a medical director, two clinical pharmacists, and a data analyst. Positions on the advisory team were funded by the Centers for Medicare & Medicaid Services (CMS) Hospital Engagement Network (HEN) 2.0 contract. One full-time equivalent (FTE) was dedicated for each of the two clinical pharmacist and the data analyst positions. The medical director position was shared with the AH infectious diseases division with 0.5 FTE funded by CMS. Gap Analysis

A gap analysis survey assessing implementation of the CDC core elements for ASPs was conducted. Appendix 2 summarizes achievement of the core elements for the 15 RNs or facilities within AH (3 RNs and 12 individual facilities listed representing a total of 28 facilities). Many facilities (8/15 RNs or facilities), particularly CAHs, lacked demonstrated leadership commitment, with no formal, written support statement and/or budgeted and financial support for ASP. Only a few RNs or facilities (3/15) had a formal ASP at baseline led by a designated ID physician and ID clinical pharmacist(s) who performed daily stewardship. At some facilities (5/15), hospitalists or staff pharmacists without budgeted time incorporated stewardship into daily responsibilities on an ad hoc basis. Many facilities (8/15) had already implemented at least one ASP action, ranging from basic activities such as intravenous to oral (IV to PO) conversion to more advanced activities (for example, formal audit with feedback performed daily). Unless associated with an RN, no CAHs and only some SHs had basic stewardship activities at baseline. CAHs and SHs associated with an RN were able to meet all CDC core elements through access to a a centralized stewardship team that performed services such as daily prospective audit and feedback or creation of stewardship initiatives and polices. Some facilities tracked antibiotic consumption, but only 6 facilities or RNs had a formal reporting mechanism. Almost all facilities disseminated an antibiogram as a form of stewardship education, but few (6/15) provided routine education on stewardship initiatives. Overall, only 4 facilities or RNs (representing 17 out of 28 facilities) met all seven of the CDC core elements for ASPs. Based on this gap analysis, facilities were stratified into three tier levels, with Tier 1 signifying mature ASPs with implementation of most core elements, and Tier 3 representing hospitals taking the first steps of program establishment (see Table 2). Overall, baseline antimicrobial stewardship efforts were found to vary markedly across our system. Establishing a Measure

AH established reductions in antibiotic use as a yearly systemwide quality goal in 2016. Antibiotic consumption was monitored monthly using the standard days of therapy

Volume 45, No. 9, September 2019

593

Table 2. Landscape of ASPs at Atrium Health by Tier Tier*

Number of Facilities or Regional Networks

Description

1

5

2

3

3

7

• Receive budgeted support • Engaged in targeted ASP activities • Engaged in baseline ASP activities • Receive some budgeted support • Engaged in baseline ASP activities • Program lacks formal structure • Engaged in baseline ASP activities (for example, IV to PO conversion, drug dosing and monitoring) • Building support for program with no budgeted or formal administrative support

ASP, antimicrobial stewardship program; IV to PO, intravenous to oral; CDC, Centers for Disease Control and Prevention. Tiers represent varying levels of baseline ASP maturity, ranging from Tier 1 (mature ASPs with implementation of most CDC core elements) to Tier 3 (hospitals in foundational efforts of building an ASP with minimal implementation of CDC core elements).

*

(DOT) per 1,000 patient-days metric applied to 17 select, broad-spectrum antibiotics. These select antimicrobials were chosen based on spectrum of activity and propensity for collateral damage. A list of target antimicrobials is provided in Appendix 3. A DOT was defined as the administration of a single dose of a specified drug on a given calendar day, even if multiple doses were given on that day. Because AH does not have a unified EMR platform, the source of antibiotic utilization data (administrative, charge, or order data) was facility dependent. Both numerator and denominator data were obtained manually or through generated reports from local facility stakeholders. As required for the systemwide quality metric, baseline usage data from September 2014 to August 2015 were collected. Baseline data were obtained from all facilities except one CAH. This hospital did not have an EMR system until May 2015, thus six months of DOT data were collected and annualized to determine the baseline. Although no reference existed to guide the assignment of DOT reduction goals, a step-wise approach to obtaining the National Action Plan goal of a 20% reduction in inpatient antibiotic use by 2020 was used. Target and stretch DOT reduction goals for the 2016 calendar year were set for each facility (see Appendix 4), ranging from 1%–2.5% for facilities engaged in baseline ASP activities (for example, Tier 3 hospitals), to 5%–10% for facilities with mature, structured programs that received budgeted support (for example, Tier 1 hospitals). In summary, comprehensive ASPs were given more aggressive DOT reduction goals with the assumption that these programs had the resources required to achieve these goals; while basic ASPs were assigned lower reduction goals for the initial year, allowing efforts to be focused on building the foundational resources necessary to establish a successful program. Site Visits

Each facility or RN identified a pharmacist, a physician, and an administrative ASP champion. Site visits to each hospital or RN were conducted by the central ASP advisory team from January to March 2016. These visits allowed the advisory team to gain insight into daily work flow, resources,

EMR and clinical decision support (CDS) system capabilities (for example, report generation, abstraction of DOT data), and to establish three to four major goals for improvement with recommendations. Several key themes emerged from these site visits: Communication. The site visits revealed the need for many facilities, particularly Tier 3 facilities, to coordinate efforts to spearhead ASP initiatives by forming a multidisciplinary team. Various departments, such as the microbiology lab or infection prevention, were often found to be working in silos. The central advisory team suggested the establishment of a multidisciplinary stewardship committee that met at least monthly or to incorporate stewardship into another standing committee meeting at facilities with limited time or resources. Maximizing or Creating Resources. Dedicated ASP time and resources were a significant barrier for many facilities. Emphasis at low-resource hospitals was placed on maximizing existing resources rather than the addition of dedicated time for stewardship. At many Tier 2 and 3 facilities, pharmacists already participated in daily interdisciplinary rounds as part of their daily responsibilities. As such, a recommendation at these facilities was to use interdisciplinary rounds as a platform for pharmacists to make stewardship interventions to optimize antibiotic prescribing such as deescalation, bug-drug mismatch, or IV to PO conversion. At Tier 1 facilities, the physician and pharmacy champion often did not have protected stewardship time. Therefore, the advisory team recommended that some facilities develop a business plan justifying additional resources for a dedicated stewardship pharmacist. Optimizing Technology. Many Tier 1 facilities were found to routinely use (or were in the process of adopting) CDS systems such as Sentri7® or VigiLanz® that generate customized alerts or reports to facilitate targeted ASP interventions. Some of these institutions were also in the process of implementing rapid identification technology (for example,

594

Andrea Y. Logan, PharmD, BCPS, et al.

BioFire®). One institution had a well-established pharmacistdriven, rapid identification program and served as a valuable resource to other hospitals adopting this technology. The advisory team made suggestions for integration of rapid identification and/or CDS systems into daily work flow. Education and Training. The need for ID and stewardship training for staff pharmacists was evident at Tier 3 facilities where a lack of provider buy-in combined with the discomfort of pharmacists in making high-level ASP recommendations was a common barrier. The development of empiric antibiotic therapy guidelines for common disease states such as pneumonia and urinary tract infection (UTI) was recommended at these facilities. Antimicrobial Stewardship Targets. Common targets that emerged from the site visits included the need to shift empiric prescribing away from fluoroquinolone treatment to alternatives for UTI and pneumonia, revising protocols to discourage the practice of using urinalysis results to automatically reflex to a urine culture, education to stop screening and treatment of asymptomatic bacteriuria, overprescribing of vancomycin and unnecessary use of double gram-negative coverage for pneumonia, and decreasing use of ertapenem for surgical prophylaxis. Establishment of a Systemwide ASP Collaborative

Table 3 summarizes the services provided to individual facilities by the centralized systemwide ASP collaborative. This collaborative used a variety of forums to provide stewardship-related education. The bimonthly systemwide meetings occurred via videoconference and provided opportunities to assess progress toward systemwide DOT reduction goals, share best practices, and disseminate education on relevant topics such as asymptomatic bacteriuria, Staphylococcus aureus bacteremia, and strategies for delabeling patients with a penicillin allergy label, among others. The collaborative also started to host an annual symposium featuring national and local stewardship experts as speakers and included breakout sessions to tailor education to facilities that are building an ASP (for example, microbiology basics) vs. mature ASPs (for example, implementation of rapid diagnostic technology). A newsletter, Bug Bytes, was also regularly distributed to the collaborative providing stewardship pearls, brief literature reviews, and guideline updates. Newsletter topics have included interpretation of DOT data, criteria for fluoroquinolone prescribing, and tips for de-escalating antibiotics during influenza season. The clinical pharmacists from the central advisory team set up monthly coaching calls with each facility’s stakeholders to review trends in monthly DOT data, discuss targets for intervention, and assess application of ASP core elements. Monthly DOT reports included antibiotic use data presented by drug class and the percentage that each class contributes to overall DOT. For example, fluoroquinolones were the top drug class prescribed at several of the Tier 3

Establishing an Antimicrobial Stewardship Collaborative

Table 3. Summary of Atrium Health Systemwide ASP Collaborative Services Core Systemwide ASP Collaborative Services • Educational Programming B Bimonthly systemwide meetings B Annual symposium on antimicrobial stewardship • Bug Bytes Newsletter B Bimonthly newsletter distributed via stewardship collaborative listserv B Topics include infectious disease and stewardship pearls, interpretation of antibiotic use data, and regulatory requirements • Monthly Facility-Specific Coaching Calls B Review monthly trends in antibiotic consumption data, including trends by antibiotic class B Track progress of local stewardship implementation goals B Assist with the creation of local policies and protocols (for example, empiric guidelines, order sets, rapid diagnostic protocols) B Provide on-site stewardship education for providers and other clinicians as requested • Data Support and Analysis B Monthly tracking of antimicrobial use B Summary analysis provided monthly B Review trends and develop action plans as needed on monthly coaching calls • Clinical Expertise B ID pharmacist and physician resources available to review and develop protocols, provide educational programming, participate in local meetings, and/or assist with provider or provider group–related ASP issues • The Joint Commission Preparation B Gap analysis of regulatory requirements conducted for all network hospitals B Provided examples and templates of documentation required to demonstrate various elements of an active stewardship program (for example, letter of support from senior leadership, minutes from multidisciplinary ASP committee meetings, antibiotic consumption data reports, patient and provider education handouts) ASP, antimicrobial stewardship program; ID, infectious disease.

facilities. Thus, a recommendation was made to implement focused antibiotic time-outs targeting these drugs. The central advisory team also assisted with the development of empiric guidelines and order sets as a strategy to discourage the first-line prescribing of fluoroquinolones for diagnoses like UTI and pneumonia. Tier 1 and 2 facilities received guidance on developing rapid diagnostic protocols and CDS– generated targets for intervention as well as the integration of these tools into pharmacist work flow. In addition, the central team provided support in the development of a business plan to justify additional ASP pharmacist positions at two facilities (Tier 1 and 2). Across all hospitals, the clinical pharmacists and ID physician from the central advisory team also frequently attended local facility meetings (for example, Pharmacy and Therapeutics Committee, medical staff meetings) to provide stewardship education as requested. Following publication of The Joint Commission standard for ASPs at acute care and critical access hospitals in summer 2016, the central advisory

Volume 45, No. 9, September 2019

team identified the need for a formal mechanism to provide stewardship education at the local facility level to providers, pharmacists, nurses, and other clinicians, as well as patients and families. As a result, the collaborative created both a patient education handout and a clinician education slide deck for dissemination.

595

DOT/1,000 patient-days decreased by 11.5% from baseline (baseline = 476.5; 2016 = 421.5). The number of facilities or RNs meeting all seven CDC core elements for ASPs increased from 4 to 12 (representing 25 out of 28 facilities; see Appendix 5). This collaborative helped 9 facilities achieve the Leadership and/or Drug Expertise elements. All facilities also implemented the Education element, using materials created by the central advisory team.

RESULTS

As illustrated in Figure 1, 78.6% (22/28) of facilities exceeded their target goal (green bars) or stretch goal (blue bars) for DOT reduction. Almost all facilities (89.3%; 25/ 28) decreased broad-spectrum antibiotic consumption from baseline levels. A few facilities decreased antibiotic use but did not reach their target goal for DOT reduction (10.7%; 3/28; orange bars) or increased use from baseline (10.7%; 3/28; red bars). Some facilities reduced DOT/1,000 patient-days by greater than 10% from baseline. In 2016, cumulative consumption of broad-spectrum antibiotic

DISCUSSION

All AH facilities, regardless of hospital size and stewardship program maturity, reduced broad-spectrum antibiotic use from baseline except for 3 hospitals (1 large, 1 small, 1 critical access) during the first year of the collaborative. Although the 3 hospitals whose antibiotic use increased from baseline were able to implement all the CDC stewardship core elements, many of the changes occurred later in the project year due to leadership and staffing transitions (for

Figure 1: Bar graph of cumulative antibiotic days of therapy (DOT) for 2016 by facility in relation to baseline (red line), target goal (green line), and stretch goal (blue line). Bar graphs are color coded by goal achievements (blue: stretch goal achieved; green: target goal achieved; orange: decreased DOT from baseline but target not attained; red: increased DOT from baseline). Hospitals are listed in order of monthly patient census (lowest to highest). LH, large hospital (N 200 beds); SH, small hospital (51–200 beds); CAH, critical access hospital (≤ 50 beds); *3-hospital regional network; **3-hospital regional network; ***10-hospital regional network.

596

Andrea Y. Logan, PharmD, BCPS, et al.

example, changes in stewardship leadership and staffing model, competing responsibilities for physician champion). Thus, progress in antibiotic use reduction by these programs was not reflected until the following year. Antibiotic use at CAH3 appeared to substantially increase from baseline. In 2016, this hospital managed an increasing number of patients on long-term IV antibiotics compared to the baseline period. Also, the calculated baseline for CAH3 (annualized from six months of data) did not account for high antibiotic-use seasons. With the adjusted baseline likely being falsely low, antibiotic use appears to have almost doubled during the 2016 year. Administrative leadership and stakeholders at this hospital were aware of these limitations and still proceeded with the assigned baseline and DOT reduction goals to participate in the collaborative. Interestingly, a few facilities (CAH2, CAH5, SH7) met target or stretch DOT reduction goals despite not fully implementing the CDC stewardship core elements. The small community hospital transitioned to a new hospitalist group affiliated with a large academic medical center, which appeared to favorably decrease antibiotic prescribing compared to the baseline period. Although the CAHs implemented only minimal initiatives specified by the Action core element and did not meet the Drug Expertise element, passive educational interventions (for example, handouts displayed in prescribing areas) provided by the central advisory team appeared to be moderately successful in changing prescribing behavior at a hospital with a small provider pool. Overall, 22 of 28 hospitals met target and stretch antibiotic DOT reduction goals ranging from 1%–2.5% to 5%–10%. Establishing systemwide antibiotic reduction goals endorsed by senior leadership was critical for motivating local stakeholders to engage in the ASP collaborative and implement change. The importance of leadership commitment

Establishing an Antimicrobial Stewardship Collaborative

has been demonstrated by other ASP networks and is a core element of ASPs. 8,16 Buy-in from senior leadership was particularly important at select facilities requiring increased resources for a successful ASP, such as the adoption of rapid diagnostic technology or creation of a dedicated stewardship pharmacist position. Another key to the success of this collaborative was direct contact with each facility to better understand local work flow, needs, and challenges related to ASP. Ongoing communication allowed the central ASP advisory team to customize implementation of best practices and tailor education according to facility resources and goals. The recommendation for each hospital to identify an administrative and physician champion was also essential. Strong leadership spearheaded local ASP initiatives by engaging and influencing the coordination of multidisciplinary teams. Empowering local stakeholders to select a few, meaningful ASP interventions (for example, antibiotic time-out, order set modification) was imperative for successfully establishing a program and effectively reducing antibiotic use, as demonstrated by other large health care networks. 8,16,17 Table 4 summarizes (by hospital tier) examples of achievements during the first year of this collaborative. Building ASPs at smaller hospitals with limited resources and no access to ID consultation services presented unique challenges. Almost all Tier 3 facilities in the AH collaborative had significant technical limitations related to their EMR system and needed improved surveillance tools to identify patients for targeted ASP intervention. These EMR limitations were compounded by the lack of dedicated time for assigned physician and pharmacy champions to conduct daily stewardship activities and spearhead initiatives. Pharmacists at smaller facilities also lacked sufficient stewardship training needed to make high-level interventions. In addition to human and technology needs, many facilities required a significant culture change within their hospitals for providers

Table 4. ASP Achievements by Tier Tier 1

2

3

• New ASP pharmacist full-time equivalent (FTE) created to oversee stewardship at a system level • Successful decrease of vancomycin and linezolid due to implementation of a negative MRSA polymerase chain reaction (PCR) protocol • Multifaceted strategy resulting in substantial fluoroquinolone reduction • Establishment of a multidisciplinary ASP committee • Implementation of antibiotic time-out into daily work flow • Integration of rapid diagnostic technology • Approval of a new ASP pharmacist FTE • Integration of a surveillance system to aid with identification of ASP interventions for prospective audit with feedback • Letter of commitment to antimicrobial stewardship signed by key leadership • Establishment of a multidisciplinary ASP committee • Implementation of antibiotic time-out into daily work flow • Modification or removal of urinalysis reflex triggers • Approval of empiric antibiotic guidelines for key disease states

ASP, antimicrobial stewardship program; MRSA, methicillin-resistant Staphylococcus aureus; CDC, Centers for Disease Control and Prevention. * Tiers represent varying levels of baseline ASP maturity, ranging from Tier 1 (mature ASPs with implementation of most CDC core elements) to Tier 3 (hospitals in foundational efforts of building an ASP with minimal implementation of CDC core elements).

Volume 45, No. 9, September 2019

to be receptive to recommendations that optimize antimicrobial therapy. Physician buy-in at these facilities was made even more challenging by the lack of ID consultation services that often provides additional support and expertise. A cluster randomized trial comparing three ASP strategies applied to small hospitals revealed that these institutions require a high degree of centralized support that provide services such as ID– controlled antibiotic restrictions and ID review of select cultures to maximize antibiotic reduction. 18 The application of such a model would be challenging in a nonintegrated network because providing remote ID services to many SHs and CAHs within AH would require a unified EMR system. The approach to smaller facilities with limited resources was to provide access to central expertise with routine touchpoint calls with clinical pharmacists and an ID physician. This collaborative structure allowed the successful implementation of interventions that decrease antibiotic consumption, such as prospective audit with feedback and IV-to-PO conversion protocols. 19,20 Overall, the formation of a systemwide collaborative provided benefit for hospitals regardless of baseline ASP maturity and offered services comparable to other ASP multihospital networks in the nation. Examples include determination of baseline ASP maturity, establishment of key ASP stakeholders, dissemination of stewardship education, development of antibiotic utilization reports, and routine site visits and/or conference calls to discuss strategies and recommendations for local ASP initiatives. 17,18,21–25 Within a year of establishing the collaborative, almost all facilities (25/28) fully adopted the CDC core elements for ASPs. Institutions with limited resources gained access to expertise required to create disease state–specific protocols and to educate local providers achieving the Action and Education elements, respectively. In addition, SHs and CAHs that actively engaged in the collaborative could meet the Accountability and Drug Expertise CDC core elements by using collaborative physician and pharmacist leaders from the central advisory team through routinely scheduled touchpoint meetings. Otherwise, these low-resource facilities would not have been able to achieve these core elements due to competing physician and pharmacist responsibilities. The creation of a monthly DOT report allowed all facilities to meet the Tracking and Reporting core elements for ASPs as well. The few facilities that did not meet all core elements were unable to do so partially due to significant transitions within their hospitals to integrate with new health systems and are currently no longer part of AH. Larger hospitals optimized their ASPs through sharing best practices, such as implementation of rapid diagnostic technology and the creation of CDS alerts. The various educational platforms in this collaborative were also beneficial for all tiers of hospitals. As with other ASP networks, this collaborative drew success from a centralized advisory team, which created the drive and provided support in the local implementation of ASP initiatives. 16,21,24

597

Following the first year of establishing a systemwide ASP collaborative, key changes were made to ensure continued engagement of vital local stakeholders. Goals for the following year remained focused on measuring a reduction in broad-spectrum antimicrobial use through measuring DOT monthly. An analysis of our antibiotic consumption throughout the first year of this collaborative revealed patterns comparable to national trends 3,12,24; with high use of fluoroquinolones, the first-line antipseudomonal betalactams (cefepime and piperacillin/tazobactam), and vancomycin across all facilities, despite ASP maturity. Therefore, all facilities were tasked with a single, unified DOT reduction goal of 5%–7.5% following the first year of the collaborative focusing on initiatives such as asymptomatic bacteriuria and beta- lactam allergy assessments to target antibiotic agents most highly used. The central advisory team continued to provide the same core services to system facilities. However, services expanded to include the development of clinical toolkits and data collection to assess appropriateness of antibiotic prescribing for systemwide initiatives. The systemwide stretch reduction goal of 7.5% was met in 2017, and all hospitals continue to decrease prescribing of broad-spectrum antibiotics, highlighting sustained success of the program. There are several limitations with spearheading antimicrobial stewardship through a collaborative approach in a heterogenous health system that measures progress primarily by tracking DOT/1,000 patient-days of broad-spectrum antibiotics. Although the use of a unified, systemwide quality metric was important for leadership buy-in from individual hospitals, we recognize that the DOT reduction is a volume-based metric, which does not necessarily reflect appropriateness of antibiotic prescribing. Reductions in broad-spectrum DOT were seen within the first year of establishing the collaborative; however, the possibility of a preexisting downward trend of historical DOT/1,000 patientdays was not assessed. Also, the annual DOT reduction goals assigned for facilities were not selected based on objective standards because no such standard exists, and the literature surrounding this remains limited. Facilities with mature ASPs selected more aggressive reduction goals (for example, 5%–10%), while those building ASPs selected lower reduction goals (for example, 1%–2.5%). However, those with newer programs likely required greater DOT reduction goals because they were starting from a higher baseline and had more opportunities to focus on low-hanging fruit ASP interventions. Another weakness with the AH DOT measure was the exclusive focus on reduction of broad-spectrum antibiotics, which does not consider shifts in prescribing to narrower-spectrum antibiotics without a reduction in overall antimicrobial use. Unlike other networks that track use of all antibiotic agents, 12,24,26 AH was able to monitor only 17 antibiotics due to the EMR system limitations at many facilities, which require time-consuming manual abstraction of DOT data. In addition, similar to other ASP networks, 16

598

Andrea Y. Logan, PharmD, BCPS, et al.

generation of real-time DOT data was difficult due to delays in data reporting. This data lag limits the ability of local stakeholders to implement timely changes in response to trends in antibiotic use. Another major challenge specific to a nonintegrated network was obtaining consistent DOT data from facilities with different clinical information or decision support platforms. Data heterogeneity has made benchmarking antibiotic use data challenging, a problem not faced by other multihospital networks with a uniform data platform. 12,24,26 Finally, this collaborative focused only on DOT reduction as a measure of antimicrobial stewardship; outcome measures such as CDI reduction were not considered. Although DOT reduction does not necessarily correlate with improved patient outcomes, this metric is endorsed by national organizations to track effectiveness of ASPs. 8,15,20 CONCLUSION

We believe it was important to establish an antibiotic use metric early in the formation of the systemwide ASP collaborative. Although examining broad-spectrum DOT exclusively was not ideal, focusing on 17 broad-spectrum, highuse antibiotics was a sufficient initial measure that allowed the inclusion of all hospitals with various EMR systems in the systemwide quality metric. Formalizing a systemwide goal that was routinely reported to senior leadership was key to propelling an effective ASP collaborative across a heterogenous system. The next priority of the AH ASP network is standardization of DOT data and participation in the NHSN Antimicrobial Use module. Funding. Funding was provided by the Centers for Medicare & Medicaid Services Partnership for Patients Hospital Engagement Network (HEN) 2.0 contract. Conflicts of Interest. All authors report no conflicts of interest.

Andrea Y. Logan, PharmD, BCPS, is Clinical Pharmacist Coordinator, Antimicrobial Stewardship, Atrium Health, Charlotte, North Carolina. Julie E. Williamson, PharmD, BCPS, is Clinical Pharmacist Coordinator, Antimicrobial Stewardship, Atrium Health. Emily K. Reinke, PhD, is Research Program Leader, Sports Medicine and Orthopedic Surgery, Duke University Medical Center. Steven W. Jarrett, PharmD, is Medication Safety Officer, Atrium Health. Michael S. Boger, MD, PharmD, is Infectious Disease Physician, Atrium Health. Lisa E. Davidson, MD, is Medical Director of the Antimicrobial Support Network, Atrium Health. Please address correspondence to Andrea Y. Logan, [email protected].

SUPPLEMENTARY MATERIALS

Supplementary data to this article can be found online at https://doi.org/10.1016/j.jcjq.2019.03.002 .

REFERENCES 1. Review on Antimicrobial Resistance. Antimicrobial Resistance: Tackling a Crisis for the Health and Wealth of Nations. O’Neill J, et al. https://amr-review.org/sites/default/files/AMR%

Establishing an Antimicrobial Stewardship Collaborative

20Review%20Paper%20-%20Tackling%20a%20crisis%20for% 20the%20health%20and%20wealth%20of%20nations_1.pdf 2. Dellit TH, et al. Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America guidelines for developing an institutional program to enhance antimicrobial stewardship. Clin Infect Dis. 2007 Jan 15; 44:159-177. 3. Baggs J, et al. Estimating national trends in inpatient antibiotic use among US hospitals from 2006 to 2012. JAMA Intern Med. 2016 Nov 1; 176:1639-1648. 4. Donskey CJ. Fluoroquinolone restriction to control fluoroquinolone-resistant Clostridium difficile. Lancet Infect Dis. 2017; 17:353-354. 5. Wieczorkiewicz JT, et al. Fluoroquinolone and macrolide exposure predict Clostridium difficile infection with the highly fluoroquinolone- and macrolide-resistant epidemic C. difficile strain BI/NAP1/027. Antimicrob Agents Chemother. 2015 Nov 2; 60:418-423. 6. See I, et al. NAP1 strain type predicts outcomes from Clostridium difficile infection. Clin Infect Dis. 2014; 58:13941400. 7. The White House. National Action Plan for Combating Antibiotic-Resistant Bacteria. https://www.cdc.gov/drugresistance/ pdf/national_action_plan_for_combating_antiboticresistant_bacteria.pdf 8. Centers for Disease Control and Prevention. The Core Elements of Hospital Antibiotic Stewardship Programs. https:// www.cdc.gov/antibiotic-use/healthcare/pdfs/core-elements. pdf 9. The Joint Commission. Approved: new antimicrobial stewardship standard. Jt Comm Perspect. 2016; 36 1, 3–4, 8. 10. Pollack LA, et al. Antibiotic stewardship programs in US acute care hospitals: findings from the 2014 National Healthcare Safety Network Annual Hospital Survey. Clin Infect Dis. 2016 Aug 15; 63:443-449. 11. Stenehjem E, et al. Antibiotic stewardship in small hospitals: barriers and potential solutions. Clin Infect Dis. 2017 Aug 15; 65:691-696. 12. Stenehjem E, et al. Antibiotic use in small community hospitals. Clin Infect Dis. 2016 Nov 15; 63:1273-1280. 13. Magill SS, et al. Prevalence of antimicrobial use in US acute care hospitals, May–September 2011. JAMA. 2014 Oct 8; 312:1438-1446. 14. Centers for Disease Control and Prevention. Implementation of Antibiotic Stewardship Core Elements at Small and Critical Access Hospitals. https://www.cdc.gov/antibiotic-use/ healthcare/pdfs/core-elements-small-critical.pdf 15. National Quality Forum. National Quality Partners Playbook: Antibiotic Stewardship in Acute Care. http://www. qualityforum.org/Publications/2016/05/National_Quality_ Partners_Playbook__Antibiotic_Stewardship_in_Acute_Care. aspx 16. Srinivasan A, Davidson LE. Improving patient safety through antibiotic stewardship: the Veteran’s Health Administration leads the way, again. Infect Control Hosp Epidemiol. 2017; 38:521-523. 17. Daragjati F, et al. What antimicrobial stewardship interventions are we focusing on? The experience of a large healthcare system. Open Forum Infect Dis. 2016; 3(Suppl 1): S201. 18. Stenehjem E, et al. Impact of implementing antibiotic stewardship programs in 15 small hospitals: a cluster randomized intervention. Clin Infect Dis. 2018 Aug 1; 67:525-532.

Volume 45, No. 9, September 2019

19. Langford BJ, et al. Assessing the impact of antimicrobial stewardship program elements on antibiotic use across acute-care hospitals: an observational study. Infect Control Hosp Epidemiol. 2018; 39:941-946. 20. Barlam TF, et al. Implementing an antibiotic stewardship program: guidelines by the Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America. Clin Infect Dis. 2016 May 15; 62:e51-e77. 21. Guharoy R, et al. Universal implementation of antimicrobial stewardship programs leads to reduction in antibiotic use and cost savings: pilot results from a large United States health system. Open Forum Infect Dis. 2016; 3(Suppl S1): S201. 22. Hawkins MR, et al. Characteristics of antimicrobial stewardship activities in community hospitals upon enrollment in the Duke

599

23.

24.

25.

26.

Antimicrobial Stewardship Outreach Network (DASON). Open Forum Infect Dis. 2014; 1(Suppl 1):S96. Fakih MG, et al. Health systems can play a pivotal role in supporting antimicrobial stewardship. Clin Infect Dis. 2016 Nov 15; 63:1391. Kelly AA, et al. A report of the efforts of the Veterans Health Administration National Antimicrobial Stewardship Initiative. Infect Control Hosp Epidemiol. 2017; 38:513-520. Burgess LH, et al. Phased implementation of an antimicrobial stewardship program for a large community hospital system. Infect Control. 2019; 47:69-73. Jones T, et al. Benchmarking antimicrobial use in emergency departments among community hospitals in the Duke Antimicrobial Stewardship Outreach Network. Open Forum Infect Dis. 2017; 4(Suppl 1):S248.