History of stewardship at the Minneapolis VA Health Care System
The Veterans Health Administration (VHA) has been at the forefront of antimicrobial stewardship programs (ASPs) for nearly 2 decades, beginning with the formation of its national Antimicrobial Stewardship Initiative in 2010, and then mandating that all VA medical facilities must participate in antimicrobial stewardship in a 2014 VHA directive (VHA Directive 1031). This directive required each facility to create an ASP policy and implementation plan, designate a facility stewardship clinician, and pharmacy champion, perform an annual evaluation of ASP activities, and report antimicrobial use to the National Healthcare Safety Network. This mandate preceded similar requirements for non-VA facilities from The Joint Commission and Centers for Medicare and Medicaid Services in 2017 1 , a time when 97% of VA hospitals reported having an ASP, Reference Vaughn, Saint and Greene2 compared to 48% on non-VA facilities. Reference O’Leary, van Santen, Webb, Pollock, Edwards and Srinivasan3 This earlier start and the ability for local ASPs to focus their efforts on the most pressing needs of their facility created a climate of innovation that continues today.
One such innovation at the Minneapolis VA Health Care System (MVAHCS) was a computer decision support system (CDSS) providing guidance for and orders to preferred therapies for common infectious syndromes. As CDSS use increased, the link between a correct diagnosis and optimal antimicrobial use was observed to be a key factor driving appropriate antimicrobial use. Evaluation of the CDSS demonstrated that it improved the appropriateness of antimicrobial courses, and that when use was inappropriate, the most common reason was that the initial diagnosis was incorrect. With an incorrect initial diagnosis, antimicrobial use was far less likely to be appropriate compared to when the initial diagnosis was correct (11% vs 62%; P < .001). Reference Filice, Drekonja and Thurn4 With the establishment and maturation of a formal ASP at the MVAHCS, the early connection between the correct diagnosis and optimal antimicrobial use has been carried forward, especially after the concept of a collaborative effort to appropriately use laboratory testing to guide patient management, or “diagnostic stewardship,” was formally defined in 2018 Reference Patel and Fang5 . Here we review how the ASP team at the MVAHCS has implemented diagnostic stewardship to improve patient care, specifically focusing on areas where modifying how diagnostic tests are ordered or reported improved the management of infections.
High-yield targets for diagnostic stewardship and efforts undertaken
Data from the MVAHCS and from the literature demonstrated areas where antimicrobial use was often inappropriate or excessive. These areas include asymptomatic bacteriuria, excessive methicillin-resistant Staphylococcus aureus (MRSA) coverage, and Clostridioides difficile colonization. In seeking to optimize antimicrobial therapy for actual infections and minimize broad-spectrum coverage, different aspects of diagnostic stewardship have been developed and built into the daily workflow. Areas for intervention were chosen based on both systematic data (ie, high rates of nosocomial Clostridioides difficile infection [CDI], excessive use of anti-MRSA agents relative to other facilities, etc.), and by anecdotal observation (ie, antimicrobials for asymptomatic bacteriuria being a recurring theme in Morbidity and Mortality conference).
Urinary tract infections
Bacteriuria often leads to inappropriate antimicrobial treatment, particularly among patients without symptoms of a urinary tract infection (UTI), where treatment provides no clinical benefit but contributes to antimicrobial resistance, adverse drug events, CDI, and unnecessary healthcare costs. Reference Nicolle, Gupta and Bradley6 A previous multi-site VA study found that an evidence-based algorithm combined with case-based clinician education can reduce inappropriate urine culture (UC) orders and reduce days of antimicrobial therapy, but this approach is resource-intensive. Reference Grigoryan, Naik and Lichtenberger7
To address this issue the MVAHCS introduced a revised UC ordering menu within the Computerized Patient Record System (CPRS), the Department of Veterans Affairs’ legacy electronic medical record, providing concise, point-of-care guidance on appropriate and inappropriate indications for UC ordering in both inpatient and outpatient settings (Figure 1). Ten UC laboratory quick orders tied to separate indications and containing specimen-collection instructions for staff, were placed within the menu, and all UC ordering was routed to it. The intervention requires clinicians to select a specific indication before a UC can be ordered, encouraging reconsideration of testing in asymptomatic patients, unless screening for bacteriuria is specifically indicated (ie, in pregnant individuals or prior to urologic procedures).
Screen capture of Urine Culture ordering menu at the Minneapolis VA Health Care System. Screen capture by D. Drekonja, July 31 2026.

Evaluation of the intervention demonstrated a significant reduction in UCs (12,190 preintervention, 8,996 postintervention; 26%; P < .001), while patient volume (measured in clinic visits and inpatient bed days) remained unchanged. Reference Stensgard, Masoud, Gravely and Drekonja8 The revised UC order menu represents a sustainable, scalable solution that requires minimal dedicated staffing hours for implementation and maintenance. This intervention has been adopted at 37 other VA medical centers, consistently associated with similar results to those observed at the MCAHCS. Specific evaluation of other outcomes such as antimicrobial use, appropriateness of diagnosis, and impact on catheter-associated UTI rates has not been performed.
Clostridioides difficile infection
Diagnosis of CDI has undergone significant change from the era of relatively insensitive enzyme-linked immune assay (ELISA) for the preformed toxins of C. difficile, when clinicians often attempted to mitigate the low sensitivity by obtaining serial testing, a practice that resulted in more false-positive results than initially missed false-negatives. Reference Peterson and Robicsek9 The introduction of highly sensitive polymerase chain reaction (PCR) testing seemed to simplify CDI testing, with a single sensitive, specific, and rapid test available for the detection of the gene for toxin B production. However, as PCR testing became widely adopted, a consistent pattern was observed: moving to PCR testing led to an increase in CDI cases, some of which in patients with minimal or transient symptoms. Polage et al noted that when PCR-positive patients were subsequently tested for the presence of toxin via ELISA testing, a large proportion of patients were PCR-positive but toxin negative (PCR+/Toxin-), indicating presence of the gene for toxin B production but with either no or little actual toxin present. Reference Polage, Gyorke and Kennedy10 Clinical outcomes of these patients similarly suggested a different disease process, with outcomes (mortality, duration of diarrhea, and recurrent CDI) all being similar to patients who tested PCR negative, and significantly different from those who were PCR+/Toxin+. Guidelines from the Infectious Diseases Society of America (IDSA) in 2021 endorsed 2 testing strategies for CDI: stand-alone PCR testing or a two-stage strategy starting with an initial highly sensitive test (PCR or glutamate dehydrogenase testing), followed by a confirmatory toxin test for those initially positive. Reference Johnson, Lavergne and Skinner11
The MVAHCS initially used a stand-alone PCR strategy, combined with educational efforts to limit CDI testing to patients with 3 or more liquid stools daily, no laxative or tube feed use, and no other cause for diarrhea. After observing high CDI rates, and with case review of nosocomial CDI cases suggesting that inappropriate testing was common, a decision was made to transition to a two-stage testing strategy, whereby all PCR + samples received confirmatory toxin testing, with results typically reported 1–2 hours later. PCR+/Toxin- samples were reported with a comment of “C. difficile present, toxin NOT detected,” with ASP team members available to help with test interpretation, especially among hospitalized patients. Stewardship team members performed prospective audit and review and discussed interpretation of two-stage testing in the context of the specific clinical case, focusing on patient care for the initial case and education of the provider for future cases. A retrospective analysis of three years of data using the two-stage testing strategy demonstrated that 65% of PCR + patients were Toxin-, with 51% of PCR+/Toxin- patients receiving treatment, compared to 99% of PCR+/Toxin+ patients. Among the PCR+/Toxin- patients who were not treated, there was no discernable signal of harm when compared to the PCR+/Toxin- patients who did receive treatment. Reference Marker and Drekonja12
Anti-MRSA therapy
Empiric therapy for MRSA is often initiated in hospitalized patients, many of whom have traditional risk factors for MRSA colonization/infection, or a history of MRSA isolation in the past. However, it is increasingly recognized that many patients receiving anti-MRSA therapy (at the MVAHCS, predominantly vancomycin) have no documented indication for ongoing MRSA therapy, and that such use confers no benefit but increases costs, toxicity, and may affect bacterial resistance. The MVAHCS ASP team developed a diagnostic stewardship intervention to facilitate cessation of unnecessary anti-MRSA therapy. Specifically, an order for an MRSA nares PCR was embedded with all new intravenous vancomycin orders. With a negative test having a negative predictive value of >90% for pneumonia, Reference Parente, Cunha, Mylonakis and Timbrook13 and similar data for diabetic foot infections, Reference Mergenhagen, Croix, Starr, Sellick and Lesse14 this strategy allows the ASP team to provide the primary team with reassurance from objective data that cessation of vancomycin was appropriate and safe. This practice, combined with other stewardship efforts, have allowed the MVAHCS to consistently rank in the lower quartile of anti-MRSA therapy use among VA centers of similar size and complexity, reducing nephrotoxicity, selection of antimicrobial resistance, and costs.
Obtaining an MRSA nares PCR with all new vancomycin starts provides similar information to the ASP team with other infections. Combined with lack of MRSA isolated from clinical cultures, the lack of nasal MRSA colonization helps the ASP team to overcome the common desire to continue vancomycin in the setting of clinical improvement. Whether MRSA nares testing has affected other test ordering or a change in clinical diagnoses made is not known.
Diagnostic stewardship team and resources
Staffing for an ASP in the VA system is guided by VHA directive 1,031, which has been periodically updated, most recently in September 2023. Recommended staffing includes a baseline for all VA medical facilities and additional staffing based on number of acute and long-term care beds, and there is also guidance regarding informatics, infection prevention, laboratory and facility leadership support. The needs of diagnostic stewardship specifically require the support and close collaboration of laboratory leadership, since some interventions result in increased testing (ie, MRSA nares PCR with all intravenous vancomycin starts, and the addition of confirmatory toxin testing for CDI), whereas others (revised UC order menu) reduce testing volume. Informatic support is also a key requirement for diagnostic stewardship, ideally from a dedicated person or team that can become familiar with the goals, processes, and workflows of the overall ASP team, and collaborate to make user-friendly menus that are intuitive to clinicians and guide them to appropriate orders.
Keys to success
Recognizing that there are many models for successful ASPs in general and diagnostic stewardship in particular, we have observed the following keys to success. 1) Institutional support. This is manifested in many ways and cannot come solely from the Chief of Staff or other high-placed leaders in the organization but rather must come from a wide array of respected clinician leaders in the medical center. When the attendings in the intensive care unit recognize that cessation of anti-MRSA therapy in patients with a negative MRSA nares PCR makes their management simpler, when hospitalists appreciate input on CDI colonization vs infection, and when the emergency department leadership adopts new testing algorithms to minimize unnecessary antibiotic courses, all of these efforts demonstrate to other clinicians in the facility that diagnostic stewardship is important even among the most acutely ill patients in the hospital, and that they should incorporate it into their practice too. 2) Local independence. Having the ability to focus on issues of importance to the hospital brings a sense of importance and immediacy to stewardship. When clinicians see that changes to urine testing and CDI testing reduces the number of times they need to wrestle with the conundrum of colonization versus infection, they recognize that diagnostic stewardship is making life simpler and reducing the cognitive load they need to spend on extraneous issues. 3) Informatic support. The ability to rapidly move from observation of a potential problem, to gathering data, to building a solution in CPRS has been a key to diagnostic stewardship at the MVAHCS. Querying local and national data warehouses using relational databases has been vital to gathering data to identify clinical quality issues and to measure the impact of solutions. A skilled pharmacy informaticist with specialized knowledge in antimicrobial use is crucial to this and is supported by VA ASP staffing guidance. Maintaining this ability to rapidly implement tools during the transition to a new Federal Electronic Health Record in 2027 is a priority that the ASP team is working through. 4) Time and resources. Staffing concerns are a reality in most of the United States Health Care system, and the MVAHCS is not immune to these. Diagnostic stewardship requires sufficient staffing, and for staff to have the time needed to devote to this work and not to competing priorities. 5) People. Ultimately an ASP is charged with optimizing antimicrobial use and diagnostic testing in the entire system, from ICU to clinic, and the interpersonal skills needed to build productive working relationships with clinicians in these areas are key for the physician and pharmacy champions. Because stewardship involves interactions with leaders from so many different clinical settings of all acuity types, it is important to find and develop ASP leadership with a combination of outstanding clinical and interpersonal skills.
Relationship to CDC core elements of diagnostic excellence
The US Centers for Disease Control and Prevention has developed core elements of hospital diagnostic excellence, 15 which they describe as encompassing “multiple practices to improve diagnosis, including ensuring that diagnostic tests are ordered, interpreted, communicated, and acted upon appropriately.” These elements are: (1) Hospital Leadership Commitment and Accountability, (2) Multidisciplinary Expertise, (3) Patient, Family, and Caregiver engagement, (4) Actions, (5) Education, and (6) Tracking and Reporting. The diagnostic stewardship efforts at the MVAHCS rely heavily on several of these core elements, especially hospital leadership commitment, multidisciplinary expertise, and tracking and reporting. The diagnostic stewardship interventions described above represent the actions to move towards diagnostic excellence. Parts of the core elements which the MVAHCS ASP can more fully explore include engaging patients, families, and caregivers, and improving education on diagnostic stewardship to a wider audience. Whereas current diagnostic stewardship education focuses on clinicians, there are other healthcare professionals who would benefit from education on aspects of stewardship, as would patients and their caregivers. For example, giving healthcare professionals and patients a greater understanding that urine color or odor is not necessarily indicative of infection would be a paradigm shift for urine testing, with benefits that would extend beyond the hospital and into nursing homes, home care, and beyond.
Conclusions
The VA was an early adopter of antimicrobial stewardship, and that legacy has been the foundation on which diagnostic stewardship at the MVAHCS has been built. Institutional support, the ability to prioritize local issues, and the capability to rapidly develop and implement tools in CPRS have all been instrumental in growing a program that uses modifications in diagnostic testing to improve antimicrobial use and optimize care. Transition to a new Electronic Health Record will potentially provide new opportunities for diagnostic stewardship, but changes to informatics and having protected time for stewardship efforts will be a challenge.
Author contribution
All authors have contributed and meet authorship criteria.
Financial support
This work was supported by the resources of the MVAHCS.
The views expressed in this article are those of the authors and do not necessarily represent the views of the US Department of Veterans Affairs or the US government.
Competing interests
All authors report no conflicts of interest relevant to this article.