Tracking and Reporting Antimicrobial Use: Development of an Antimicrobial Stewardship Dashboard
Bibliographic record
Abstract
Abstract Corresponding author: Kathryn Timberlake, 555 University Ave, Toronto, ON M5G1X8, 416-813-6475, Kathryn.timberlake@sickkids.ca No conflicts of interest. Background Tracking and reporting antimicrobial use (AU) is a core element of the Centers for Disease Control and Prevention recommendations for Antimicrobial Stewardship Programs (ASP). Antimicrobial consumption measured in standardized units such as days of therapy (DOT) per 1000 patient days is a common metric used to identify areas for improvement and assess the impact of antimicrobial stewardship. Additionally, benchmarking between units, services, most responsible physicians (MRPs), and with external hospitals is feasible when AU data is available. Adoption of electronic health records provides an opportunity to leverage available data to generate these metrics. Objective was to design, develop, and validate an AU dashboard for use by an established ASP at a large pediatric academic hospital to facilitate internal and external reporting. Methods The setting was The Hospital for Sick Children, Toronto, Canada. The design of the AU dashboard was developed with input from members of the ASP steering committee, pharmacy, and utilization management committee. Data source was SickKids Enterprise-wide Data in Azure Repository (SEDAR), which receives, filters, curates and validates Epic data daily including medication administration records (MAR). All antimicrobial administrations given to inpatients were identified and classified according to the American Hospital Formulary Service. The first dose of a medication given on a calendar day was assigned one DOT. Antimicrobial administrations were allocated to clinical units and MRPs at specified times each day. The DOT, patient days, and DOT per 1000 patient days were then calculated and visually displayed on a Power BI dashboard. Validation of all data in the report including identification of missing information was conducted. Discrepancies from admission, discharge, transfer, and MRP designation were evaluated. Results We successfully developed the dashboard and validated the data on antimicrobial prescribing by units and MRPs, allowing for comparisons. Units and services with relatively few patient days were often difficult to compare as AU per 1000 patient days lacked precision. Therefore, for comparison, we removed units and services with less than 1000 patient days per year. The dashboard displayed the following: “top 10” units and services in tabular and bar graph formats, AU over time, and cumulative drug use by unit or service in a stacked bar graph. Previous data extracts for annual reporting and benchmarking from the health information system reports took hours to days to compile, but now can be completed in 60-90 minutes. Antimicrobial administration data have been submitted to Public Health Ontario (PHO) for their AU Comparison Tool for benchmarking antimicrobial stewardship and antibiotic use across the province. Additionally, we reported AU to the Canadian Nosocomial Infection Surveillance Program (CNISP) where AU and antimicrobial resistance trends are being explored. Data are now readily available for trending AU to assess the impact of ASP interventions. Conclusion The development and validation of an AU Dashboard using a curated health information system data source (SEDAR) has improved the accuracy, efficiency, and timeliness of AU reporting. Future directions for this work include benchmarking between other pediatric hospitals, developing a dashboard for antimicrobial cost, and identification of future ASP projects. Sharing our findings may help other hospitals who aim to have AU dashboards improve their reports. Future efforts will include reporting AU by MRPs in a de-identified fashion so that physicians can benchmark their own practice. Figure 1. Dashboard Overview
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.047 | 0.088 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.004 | 0.003 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.005 | 0.004 |
| Open science | 0.003 | 0.006 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.006 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".