Endotracheal aspirates

Background state prior to diagnostic stewardship: Endotracheal aspirate cultures (EACs) are commonly used in the evaluation of pediatric patients with artificial airways suspected to have pneumonia or tracheitis.^{Reference Sick-Samuels, Koontz and Xie9} However, EACs may demonstrate bacteria that are present in the airway but not causing infection, limiting their overall specificity in diagnosis.^{Reference Venkatachalam, Hendley and Willson10} Adult centers may use bronchoalveolar lavage (BAL) or mini-BAL more often than pediatric settings however, across the age spectrum, national guidelines promote EACs over invasively obtained samples.^{Reference Kalil, Metersky and Klompas11} Ordering, processing, and reporting of EACs are highly variable,^{Reference Sick-Samuels, Koontz and Xie9} and contribute to antimicrobial prescribing.^{Reference Prinzi, Parker, Thurm, Birkholz and Sick-Samuels12,Reference Prinzi, Parker, Curtis and Ziniel13} Processing these samples requires laboratory time and resources, as these cultures often grow multiple organisms that would require significant resources to perform speciation and susceptibility testing. Typically, most bacterial growth in cultures is reported, even when the result likely represents colonization. Compounding this issue, many electronic medical records (EMRs) report any positive test result with emphasized text (eg, red, capitalized, highlighted, use of exclamation marks) and may add wording indicating an “abnormal result.” These results and signals reinforce perceived diagnostic value of the test result, nudging clinicians toward treatment. Respiratory infections have overlapping features with other pathology, and even a seasoned clinician who understands the low specificity of an EAC may have difficulty ignoring an “abnormal” result, particularly when the patient is very ill, there exists diagnostic uncertainty, and the result is released to the family.^{Reference Willson, Hoot and Khemani14,Reference Morrison, Kono and Rush15}

Optimal state after diagnostic stewardship: There are multiple moments of differing understanding and priorities by healthcare workers along the process of a diagnostic test that allow opportunity for collaboration to assure the test adds value (Table 1). Two clinician-based targets for quality initiatives targeting EACs with success include education to consider pre-test probability and standardization of specimen collection.^{Reference Sick-Samuels, Fackler, Berenholtz and Milstone16–Reference Ormsby, Conrad and Blumenthal18} A laboratory-based approach is pre-screening specimens for quality to reduce the processing of specimens that are likely to be negative or uninterpretable. Laboratories can enact conditional culturing based on Gram stain features, and/or require order clinical indications for testing.^{Reference Jelic, Weber and Birkholz19–Reference Tripathi, Kenaa and Claeys21} Children’s Hospital Colorado’s laboratory enacted selective culture criteria, and now does not culture specimens that are likely to be low diagnostic value (ie, specimens with a Gram stain that is negative for organisms, positive for three or more bacterial morphologies, and/or positive for epithelial cells). These scenarios exemplify the role of the laboratory to decrease test results that may be non-specific and risk misinterpretation by clinicians. AS can support these types of changes, by listening attentively for stakeholder concerns, liaising between the laboratory and clinicians, educating about the enacted changes, and assessing results, including consideration of impact on antibiotic prescribing and balancing measures (eg, ventilator-associated events, ventilator-days).

Blood cultures

Background state prior to diagnostic stewardship: Blood cultures are often overutilized for common pediatric hospital conditions requiring admission.^{Reference Combs, Liberman and Lee22} The time for a blood culture to grow, for an organism to be identified, and initial susceptibility data to be known is significantly less than in the past.^{Reference Biondi, Mischler and Jerardi23,Reference MacBrayne, Williams, Prinzi, Pearce, Lamb and Parker24} Clinician understanding of the laboratory methods in this fast-moving field are sometimes outdated; this results in clinicians 1) holding on to the 48–72 h “rule out” (which is reinforced by automated results appearing in the EMR only at 24-h intervals), 2) distrusting results of multiplex polymerase chain reactions (PCRs) or not considering them “final” enough to target antimicrobial therapy, and 3) misinterpreting reported results, particularly when the laboratory reports a list of gene targets without an interpretation (eg, “detected: enterobacterales, Klebsiella pneumoniae, ctx-m” rather than the interpreted “ESBL-producing K. pneumoniae”). The antimicrobial steward can bring laboratorian observations to clinicians, and bring clinicians’ reflections to the laboratory, work with quality assurance teams to explore optimal processes, reinforce appropriate indications for blood culture collection, and optimize EMR interfaces to embrace the full potential of new platforms for rapid blood culture identification.

Optimal state after diagnostic stewardship: In terms of DS in clinician decision-making, the literature demonstrates the low utility of blood cultures in various clinical syndromes, including uncomplicated community acquired pneumonia, skin and soft tissue infections, surveillance for thermoregulated patients, and repeat cultures in patients with gram-negative bacteremia of urinary source.^{Reference Combs, Liberman and Lee22,Reference Elerian, Sparks and Meyer25,Reference Woods-Hill, Koontz and Voskertchian26} In addition, clinical algorithms can be used to decrease blood culture testing when the pre-test probability is low, particularly during blood culture bottle shortages²⁷, and thus avoid significant cost, resource utilization, patient discomfort, and the adverse implications of a contaminant for a patient.^{Reference Woods-Hill, Koontz and Colantuoni28,Reference Theophanous, Ramos and Calland29} AS ideally works with clinicians and quality assurance teams to implement better ordering practices up front, but even after a blood culture grows, the steward is an essential resource in supporting the interpretation and reaction to a test through real-time decision support and up-to-date guidance. Laboratories play an essential role in optimizing reporting of results into language that is easily interpreted by clinicians. Notably, increased impacts of novel rapid molecular diagnostics such as multiplex PCR and MALDI-TOF are realized by partnering with AS,^{Reference Banerjee, Teng and Cunningham30–Reference Messacar, Hurst and Child32} as reinforcement and education within specific patient-contexts may be key to avoiding unnecessary antibiotic prescribing. Opportunities for the care triangle to explore include AS relaying positive results in real-time, including interpretation of results and resistance markers, antimicrobial recommendations, explanation of unfamiliar species names, and prevention of readmissions^{Reference Messacar, Hurst and Child32} and treatment of likely contaminants. Antimicrobial stewards may be the first to realize trends in culture results/contaminants that should be reported to IPC for investigation. To garner local support, local evaluation is often needed, either to enact change, or as follow-up to change for reassurance regarding patient safety.

Gastrointestinal molecular panels

Background state prior to diagnostic stewardship: Prior to multiplex PCR testing of stool, gastroenteritis was a clinical diagnosis or based on less sensitive, more complicated testing (eg, stool culture, ova and parasite testing, viral electron microscopy, ELISA). With the implementation of multiplex PCRs (gastrointestinal panel, GIP), identification of organisms increased, but in many circumstances, this added uncertainty around the clinical significance of positive testing.^{Reference Cotter, Thomas, Birkholz, Ambroggio, Holstein and Dominguez33,Reference Cotter, Thomas and Birkholz34} Clostridioides difficile testing highlights this conundrum, as pediatric patients are known to be colonized early in life.^{Reference Tougas, Lodha and Vandermeer35} Although some clinicians understand this epidemiology, others may still interpret a positive C. difficile result as indication for treatment. Infection preventionists are required to track and report HAIs, which can be falsely inflated due to non-selective testing criteria and result reporting. Other GIP results, such as the various targets for Escherichia coli spp., are also observed to cause confusion.

Optimal state after diagnostic stewardship: Creative DS solutions are needed to curb unnecessary multiplex PCR testing and resultant antimicrobial overuse. Potential DS solutions include patient risk stratification,^{Reference Ho, Cotter, Thomas, Birkholz and Dominguez36} EMR-based conditional ordering, and laboratory specimen selective testing based on stool quality, recency of testing, and laxative use. One quality initiative coupled education with EMR conditional ordering to reduce rates of low-value stool testing.^{Reference Cotter, Stokes, Tong and Dominguez37} Selective reporting of results from the laboratorian side may be preferable to trying to teach clinicians to “ignore” certain results. For example, C. difficile can be selectively reported only when specifically ordered, and/or only in children over one year of age. In some adult centers, C. difficile is only reported if a reflex toxin assay is positive.^{Reference Qutub, Govindan and Vattappillil38,Reference Vasoo, Stevens and Portillo39} These potential solutions all require significant collaboration by those in the care triangle (Figure 1), particularly for de-implementation efforts if testing has become a very common institutional practice contributing to over-diagnosis and treatment.^{Reference Singh, Claeys and Advani40}

Urine testing

Background state prior to diagnostic stewardship: In pediatrics, there are significant DS opportunities around the diagnosis and treatment of urinary tract infections. Progress is hampered by lack of gold standard definitions, lack of high-quality data, and propagation of certain dogma. To avoid the need to obtain another sample, many pediatric clinicians send a culture in tandem with a urinalysis (UA), rather than sending the culture only after an abnormal UA.^{Reference Claeys, Zhan and Pineles41} For laboratory diagnostics, the definition of a positive UA, or even a positive culture, is not historically well developed,^{Reference Doern and Richardson42,Reference Shaikh, Lee, Krumbeck and Kurs-Lasky43} leading clinicians to question the utility of a UA in neonates in particular. Recent work challenges this belief,^{Reference Schroeder, Chang, Shen, Biondi and Greenhow44–Reference Roberts46} including supporting the strong positive and negative predictive value of a UA for bacteremic pyelonephritis.^{Reference Schroeder, Chang, Shen, Biondi and Greenhow44} Another misconception is that urine is a sterile fluid,^{Reference Wald and Eickhoff47} and identifying bacteriuria is abnormal. Converse to adult patients,^{Reference Advani, Turner and North48} efforts to prevent treatment of asymptomatic bacteriuria (ASB) in pediatric patients are lacking, despite studies demonstrating ASB may be present in 1%–10% of healthy children^{Reference Mohammed, Abdelfattah, Ibraheem and Younes49–Reference Wettergren, Jodal and Jonasson51} and 95% of cases may resolve without treatment,^{Reference Wettergren and Jodal52} consistent with adult data.^{Reference Carey, Vaughn, Mann, Townsend, Chopra and Patel53–Reference Gágyor, Bleidorn, Kochen, Schmiemann, Wegscheider and Hummers-Pradier55} Although we have some understanding how many children with a negative UA may have a positive culture,^{Reference Shaikh56} our understanding of the number of such patients needed to treat to benefit one (eg, prevent progression to serious infection) vs number needed to treat to harm one is paramount to DS of urine cultures in pediatrics. Laboratorians are often asked to culture, identify, and perform susceptibility testing on specimens that had a negative or no UA, were obtained via nonsterile collection (eg, cotton ball, indwelling catheter), and/or are growing multiple colony types suggestive of contamination. Hospitals may be required to report these results as HAIs. Stewards observe unnecessary treatment of these “infections” and downstream adverse effects.

Optimal state after diagnostic stewardship: The UA should be used in all age groups to guide decisions on whether to culture a urine specimen based on high negative predictive value.^{Reference Schroeder, Chang, Shen, Biondi and Greenhow44,Reference Tzimenatos, Mahajan and Dayan45} In children with a negative UA, it is likely appropriate to recommend watchful waiting in the majority of cases. Clinicians and laboratorians can collaborate on DS efforts to reduce inappropriate urine culturing. Using the UA to determine which specimens should be cultured could be done at the level of clinician or, alternatively, at the level of the laboratory; reflecting from a workflow perspective, the second option has many benefits in pediatrics, given the need to perform catheterization to obtain a sample.^{Reference Advani, Turner and North48,Reference Fakih, Advani and Vaughn57} A clinician-driven option is to only catheterize a child if the UA is positive.^{Reference Lavelle, Blackstone and Funari58–Reference Ben-David, Carroll and Kornitzer60} A collaborative solution could include conditional ordering in the EMR, eg, “culture only if UA abnormal,” or “culture even if UA abnormal” (to accommodate special situations like neutropenic hosts). A lab-based initiative could require the clinician to order a culture, with the laboratory performing the UA and doing the culture only if it is positive. All would decrease culturing of UA-negative specimens. Such changes are likely to benefit from AS and IPC to aid in enactment, understanding of result implications, and monitoring of impact and balancing measures.

Harnessing other potential opportunities and following through

Although these examples illustrate the collaborations between clinicians, laboratorians, and quality assurance teams, additional opportunities for antimicrobial stewards in DS require a system of prioritization. In all cases, local stakeholders are vital, and the antimicrobial steward promotes observation, reflection, exploration, enactment and evaluation while pivoting between members of the care triangle. For example, if there is a clear opportunity for an improvement, but there are no obvious invested stakeholders, the steward may need to focus on laying groundwork for change and seeding the idea across the hospital until another interested party emerges and the project can move to the Exploration and Enactment stages. Conversely, if stakeholders approach the antimicrobial steward seeking change, supporting the interest quickly is key to harnessing momentum.

Stewardship of certain tests may be particularly high value such as 1) uncommon but expensive tests (eg, untargeted DNA sequencing^{Reference Gaston, Chiang, Dee, Dulek, Banerjee and Humphries61,Reference Jhaveri, Weiss, Winkler, Pyden, Basu and Pecora62} ), 2) tests developed for adult patients with unclear application for pediatric patients (eg, bone and joint infection multiplex PCRs designed for infection of prosthetic joints in adults), and 3) tests ordered very commonly that lack diagnostic specificity for the infectious process or do not benefit management decisions. In infectious diseases, there are many examples that fall into this third category: eg, inflammatory markers, chest radiographs, viral blood PCRs, IgM tests with low specificity, and fungal markers. Beyond focus on a specific test, consideration of particular clinical scenarios where testing may be over utilized can improve value, for example, pan-cultures in febrile intensive care unit (ICU) patients, surveillance cultures in thermoregulated patients, optimal testing for patients with respiratory symptoms (eg, shift from broad multiplex PCR panels to narrower PCR panels/no PCRs and point of care testing).^{Reference Hitchcock, Gomez, Pozdol and Banaei63–Reference Otto and Chen67}

The final moment of AS-DS synergy is following through on initiatives to evaluate efficacy and potential harm (Table 1).^{Reference Toma, Davey, Marwick and Guthrie68} Desired outcomes will vary by initiative. Balancing measures (potential harm) can include length of stay, return or escalation to intensive care, readmission, changes in HAI incidence, and progression to more serious infection. Non-clinical outcome measures may be considered as well, such as an increase in use of an alternative tests, clinician time, and patient satisfaction (eg, patients may expect testing and it takes more time to explain why they may not benefit from a test). Measurement of both desired and undesired outcomes is important to solidify change, iterate unsuccessful initiatives, and reassure clinicians to further improve patient care and disseminate optimal practices.

Cost

DS programs will require personnel and system resources that incur costs. Although this may require financial support, the benefits of avoiding problematic diagnostic testing will justify the cost. With the overabundance of information a clinician must sort through every day, AS and DS need to synergize to enhance the utility of diagnostic tests. The overriding goal of this DS-AS synergy is to increase the value of care, through optimizing test ordering, implementation, and reporting. Although some tests may add cost, if they improve outcomes this may add value; conversely, some tests do not improve outcomes, result in actionable care, may add harm, and are costly and/or frequent, lowering value. There are costs to the hospital, the patient, and society. Hospital costs may be measurable, or more indirect. Examples include costs of supplies and equipment, costs for person power to collect, process and report testing, costs of treatment^{Reference MacBrayne, Williams and Obermeier69} and costs of extended hospital stays. The downstream impacts of unfavorable HAI metric reporting can include losses to reimbursement, or fewer admissions due to impacts on reputation. The additive values are substantial; as an example, a single pediatric C. difficile case may cost $93,000.^{Reference Sammons, Localio, Xiao, Coffin and Zaoutis70} Potentially avoidable patient and family costs include the direct cost of the testing and treatment and longer time away from work or school, and risk for experiencing an adverse drug reaction and associated care needs. Another potential harm to families that is difficult to measure is the psychologic strain or confusion when patients see positive results but may not understand how to interpret them. Costs to society include healthcare expenses, antimicrobial resistance, and environmental pollution.^{Reference Fan, Lu, Fernandez and Jaggi71}

As an example of the latter, microbiology laboratories dispose of a large amount of waste daily from routine cultures; the costs, in terms of monetary, energy, and greenhouse gases, for disposal of this waste are considerable.^{Reference Hofmeister, Gonzalez, Hildreth, Savage, Leber and Jaggi72} Sources include autoclaving or incineration prior to disposal, transportation to processing areas, and methane emissions occurring from landfills. Adding to this is the antibiotic waste disposal associated with discarded unused anti-infectives, and the water contamination from administered, then excreted antibiotics. Understanding the true costs of diagnostic and antimicrobial overuse to our hospitals, our patients, and society provides the context needed for the best care and support for our programs.^{Reference Doshi, Vuppula and Jaggi73–Reference Freifeld, Todd and Khan75} The downstream implications may be notable. As an example, if unnecessary blood cultures are prevented, this leads to less contaminant blood cultures in patients without true bacteremia, fewer repeated confirmatory blood cultures, fewer reported central line-associated bloodstream infections (CLABSIs), less treatment for bacteremia that was not real, fewer adverse impacts from unnecessary antimicrobials, like C. difficile, fewer resources to hospitalize and care for the patient, fewer supplies and personal protective equipment gear utilized, and less waste and environmental contamination. The recent commercial blood culture bottle shortage in 2024 may serve as a natural experiment to study the impact of reduced blood culturing from a patient outcomes, IPC, and environmental lens.

Identifying low value processes to target for improvement takes multidisciplinary observation and identification of common goals. The antimicrobial steward is uniquely placed to add to these observations, connect and reflect with interested parties, facilitate stakeholder exploration, and synergize the processes of enactment and evaluation. Although true for all ASPs, a large portion of the success of Handshake Stewardship stems from collaborative diagnostic stewardship efforts^{Reference Hurst, Child, Pearce, Palmer, Todd and Parker76–Reference Zembles, MacBrayne, Mitchell and Parker78} originating from in-person discussions during microbiology and stewardship rounds. As AS evolves, there is opportunity to expand AS programs to include DS in collaboration with clinicians, laboratorians, IPC, and quality assurance teams.