OBJECTIVE: To conduct a process evaluation of a respiratory culture diagnostic stewardship intervention.

DESIGN: Mixed-methods study.

SETTING: Tertiary-care pediatric intensive care unit (PICU).

PARTICIPANTS: Critical care, infectious diseases, and pulmonary attending physicians and fellows; PICU nurse practitioners and hospitalist physicians; pediatric residents; and PICU nurses and respiratory therapists.

METHODS: This mixed-methods study was conducted concurrently with a diagnostic stewardship intervention to reduce the inappropriate collection of respiratory cultures in mechanically ventilated children. We quantified baseline respiratory culture utilization and indications for ordering using quantitative methods. Semistructured interviews informed by these data and the Consolidated Framework for Implementation Research (CFIR) were then performed, recorded, transcribed, and coded to identify salient themes. Finally, themes identified in these interviews were used to create a cross-sectional survey.

RESULTS: The number of cultures collected per day of service varied between attending physicians (range, 2.2-27 cultures per 100 days). In total, 14 interviews were performed, and 87 clinicians completed the survey (response rate, 47%) and 77 nurses or respiratory therapists completed the survey (response rate, 17%). Clinicians varied in their stated practices regarding culture ordering, and these differences both clustered by specialty and were associated with perceived utility of the respiratory culture. Furthermore, group "default" practices, fear, and hierarchy were drivers of culture orders. Barriers to standardization included fear of a missed diagnosis and tension between practice standardization and individual decision making.

CONCLUSIONS: We identified significant variation in utilization and perceptions of respiratory cultures as well as several key barriers to implementation of this diagnostic test stewardship intervention.

<p><strong>BACKGROUND AND OBJECTIVES: </strong>Inappropriate vancomycin use is common in children's hospitals. We report a quality improvement (QI) intervention to reduce vancomycin use in our tertiary care PICU.</p>

<p><strong>METHODS: </strong>We retrospectively quantified the prevalence of infections caused by organisms requiring vancomycin therapy, including methicillin-resistant Staphylococcus aureus (MRSA), among patients with suspected bacterial infections. Guided by these data, we performed 3 QI interventions over a 3-year period, including (1) stakeholder education, (2) generation of a consensus-based guideline for empiric vancomycin use, and (3) implementation of this guideline through clinical decision support. Vancomycin use in days of therapy (DOT) per 1000 patient days was measured by using statistical process control charts. Balancing measures included frequency of bacteremia due to an organism requiring vancomycin not covered with empiric therapy, 30-day mortality, and cardiovascular, respiratory, and renal organ dysfunction.</p>

<p><strong>RESULTS: </strong>Among 1276 episodes of suspected bacterial infection, a total of 19 cases of bacteremia (1.5%) due to organisms requiring vancomycin therapy were identified, including 6 MRSA bacteremias (0.5%). During the 3-year QI project, overall vancomycin DOT per 1000 patient days in the PICU decreased from a baseline mean of 182 DOT per 1000 patient days to 109 DOT per 1000 patient days (a 40% reduction). All balancing measures were unchanged, and all cases of MRSA bacteremia were treated empirically with vancomycin.</p>

<p><strong>CONCLUSION: </strong>Our interventions reduced overall vancomycin use in the PICU without evidence of harm. Provider education and consensus building surrounding indications for empiric vancomycin use were key strategies.</p>