Introduction

Infectious diseases represent a leading cause of mortality worldwide. Antimicrobial stewardship (AMS) programs mitigate the threats posed by antimicrobial resistance. Interventions implemented by AMS programs have resulted in 30% reduction in deaths associated with multidrug-resistant organisms (MDROs) in hospitals from 2012 to 2017. ¹ Much of the progress was halted during the first year of the COVID-19 pandemic, with a rise of 15% in MDRO infection associated mortality. This stark reversal highlights the ongoing urgency to strengthen AMS programs, across the continuum of care.

Healthcare systems provide an opportunity to bolster AMS across a myriad of settings. Yet, building capacity presents unique challenges. ^{Reference Buckel, Stenehjem, Hersh, Hyun and Zetts2–Reference Stenehjem, Braun and Chitavi7} The complexity and scale of these systems, encompassing diverse patient populations, care settings, and healthcare professionals, necessitate adaptable strategies. Balancing the need for dedicated AMS personnel and technology with the overall resource allocation of the system can be challenging. Additionally, fostering a culture of AMS across a large, geographically diverse workforce requires sustained effort and engagement. Finally, collecting and analyzing comprehensive data across a vast network of facilities requires robust data infrastructure.

We present Ascension, a system with 100+ hospitals and 2,600 outpatient sites, where we have established a system AMS program since 2015. Our system is a compilation of geographically distinct markets in 11 states. Here we address eight pillars that systems can implement to provide support for stewardship across their sites (Figure 1). We describe the structure that was built from system, to market, and local level to cascade stewardship efforts, provide concrete examples of processes to standardize and enhance outcomes, and emphasize real-time data access and accountability for sustainability.

Figure 1.

Eight pillars for an antimicrobial stewardship program in a large healthcare system.

Implementing key program processes

A cornerstone of our system’s program is the strong collaboration between various healthcare professionals. The system AMS committee is composed of infectious disease physicians and pharmacists, non-infectious disease physicians and pharmacists, nurses, microbiologists, infection preventionists, quality leaders, and clinical informaticists. The multidisciplinary representation in the market and local committees also mirrors the system level. This collaboration fosters knowledge sharing amongst thought leaders and key stakeholders, facilitates the implementation of evidence-based guidelines, and ensures effective rollout of standardized initiatives. With the goal of reducing clinical variation across the system, this multidisciplinary collaboration helps standardize practices across the continuum of care.

We use a robust data infrastructure to monitor and analyze antimicrobial use and resistance patterns, in conjunction with other infectious diseases-associated quality metrics (Table 1). ^{Reference Guharoy, Fakih and Seggerman13–Reference Guharoy, Daragjati and Groves15} Teams are able to access the data at the local, market, and system levels, and compare their performance to other Ascension sites or markets and the national level. The dashboards allow for all stakeholders to have real-time visibility into their progress and opportunities with numerous filters and drill-down functionalities. Outside of standard Antimicrobial Utilization and Resistance (AUR) reporting to the National Healthcare Safety Network, additional metrics are reported which include but are not limited to: antibiotic utilization, antibiotic resistance, process-based and outcomes-based metrics (Table 1). ^16,17 Performance is monitored and trended using the dashboards, and reports are shared regularly with leadership, clinicians, and other stakeholders, supporting informed decision-making and continuous program improvement. The dashboards are built to be flexible to updates and the development and refinement of new metrics. Further, clear lines of accountability are established for AMS activities, including defining roles and responsibilities for prescribers, pharmacists, infection preventionists, and microbiology. Nurses also have clear responsibilities such as obtaining appropriate cultures prior to starting antibiotics, identifying sepsis early, suggesting when to convert from intravenous to oral antibiotics, and documenting a detailed allergy history. Case Management as well is involved with clarifying antibiotic orders for discharge to optimize patient compliance. Leaders from each respective discipline conduct regular performance reviews and feedback mechanisms to monitor progress and identify areas for improvement with escalation pathways to the system-level champions and physician market leaders. The system AMS team also reports current AMS data on monthly Quality meetings with market leaders and conducts in-person market visits to strengthen relationships with the patient-facing teams and address opportunities and barriers.

Table 1.

Standardized system-wide reported antimicrobial stewardship metrics

NHSN - National Health Safety Network; EHR - Electronic Health Record; AUR - Antimicrobial Use and Resistance Reporting; SAAR - Standardized Antimicrobial Administration Ratio ^{Reference Cosgrove, Ahn, Dullabh, Gordon, Miller and Tamma6} ; SRIR - Standardized Resistant Infection Ratio ^{Reference Cosgrove, Ahn, Dullabh, Gordon, Miller and Tamma6} ; CBC - Complete Blood Count; CMP - Complete Metabolic Panel; PCR - Polymerase Chain Reaction.

^a Includes Medicare Severity Diagnosis Related Groups (MS-DRGs): 94, 95, 96, 152, 153, 177, 178, 179, 193, 194, 195, 288, 289, 290, 371, 372, 373, 485, 486, 487, 539, 540, 541, 548, 549, 550, 602, 603, 689, 690, 853, 854, 855, 856, 857, 858, 862, 863, 867, 868, 869, 870, 871, 872.

^b Value-based purchasing criteria for antimicrobial utilization reporting applied where applicable ^{Reference Stenehjem, Braun and Chitavi7} .

Not Listed: Infection prevention metrics reported by Ascension.

The system AMS program annually identifies a few high-impact priority goals. Specific evidence-based recommendations are developed to optimize antimicrobial prescribing and build upon the existing framework as a standardized effort. ⁹ These targeted interventions include: prospective audit and feedback, antibiotic stewardship rounds, antibiotic preauthorization/restriction, development and implementation of local treatment guidelines (Table 2). Technology is also leveraged to hardwire standardized best practices, such as automatic stop dates for antibiotics, indication selection requirements, and cascade reporting for susceptibility results. The use of CDS tools (Table 3), disease state-specific order sets, and electronic-health record alerts further guide care across all settings. In the outpatient setting AMS is also supported by strategies such as watchful waiting, delayed prescribing, and payer-driven value-based reimbursement initiatives. This last component relates to a variety of measures that payers track and incorporate into scoring for systems to incentivize best practices, such as reducing antimicrobial prescribing for acute respiratory tract infections. ¹⁸ These approaches have consistently maintained our Standardized Antimicrobial Administration Ratio (SAAR) for all antimicrobials below 1 with around 670, 140, and 86 Days of Therapy per 1,000 Days Present for all antimicrobials, broad-spectrum beta-lactams, and vancomycin, respectively. ^{Reference Cosimi, Mupanomunda and Garrett-Ray19} All the while keeping risk-adjusted mortality less than expected for infectious disease diagnoses, with observed/expected 0.746 and 0.674 for sepsis and pneumonia, respectively. ^{Reference Fakih, Daragjati and Sturm20} Additional efforts to improve remdesivir prescribing for high-risk COVID-19 patients resulted in significantly less remdesivir use for Ascension compared to peers with consistently significantly less mortality than peer comparators. ^{Reference Daragjati, Cosimi and Ghosh21} Antimicrobial prescribing in the ambulatory setting is also lower than payer-reported national averages for respiratory infections. Bronchitis being an example where percentage antimicrobial prescribing per encounter was reported as 48% for Ascension and payor comparators performing around 65%–80%. ^{Reference Mackey, Cosimi and Daragjati22,23}

Table 2.

Standardized system-wide AMS initiatives

Table 3.

Standardized system-wide AMS clinical decision support pharmacy rules