Lifetime Citizen Portal Access — AI Briefings, Alerts & Unlimited Follows
NHSN training walks facilities through generating and interpreting AR analysis reports
Loading...
Summary
Virgie Phils, senior public health analyst on the NHSN AUR module team, demonstrated how to generate datasets and run antimicrobial resistance (AR) analysis reports, explained key report types and interpretation rules (including thresholds for percent tested and percent susceptible), and pointed attendees to cheat sheets and contact resources.
Virgie Phils, senior public health analyst on the NHSN antimicrobial use and resistance (AUR) module team, led a virtual training demonstrating how facilities should generate datasets and run the NHSN AR option analysis reports.
Phils said generating reporting datasets is the required first step: newly uploaded data will not appear in NHSN analysis outputs until a facility selects Analysis → Generate datasets and completes the process. "Generating datasets is the first step to performing analysis in NHSN," she said.
The training outlined three report families that use AR event (numerator) data — an event line list, an AR event bar chart, and the facility-wide antibiogram with percent tested — and three that use AR organism phenotype definitions (line list, frequency table, rate table). Phils also described AR summary (denominator) line lists, STIX incidence/prevalence rate tables, and risk-adjusted measures (the SRIR and PSIR), noting that SRIR/PSIR and incidence/prevalence have separate trainings and were not covered in depth in this session.
On practical use: Phils showed how to run the default HTML report, modify reports (filter by time period or location such as ED or a ward), and export datasets for external analysis in Excel or SAS. She recommended exports when the HTML line listing reaches the display row limit because the AR Event Line List displays one row per drug susceptibility test for each event — meaning a single event ID can produce multiple rows.
Phils reviewed visualization and decoding: the AR Event bar chart displays counts by location over time, defaults to monthly pathogen-by-location view, and can show up to 12 time periods. Pathogen codes shown in charts can be decoded via the report modification screen or by consulting the Information Data Model (IDM) spreadsheet in the AR CDA toolkit.
On interpretation rules that affect decision-making, Phils highlighted two numeric thresholds facilities must know: percent susceptible (%S) is calculated only when at least 30 isolates were tested for that organism–antimicrobial pairing in the chosen time period, while percent tested is calculated when at least one isolate has a susceptibility result. "%s is only calculated if at least 30 isolates are tested for the specific antimicrobial during the time period of your antibiogram," she said, and explained that rolling up the time period (to half-year or year) can increase isolate counts to meet the 30-isolate threshold.
Phils described phenotype-based reports: NHSN uses phenotype definitions (listed in Appendix I of the AUR protocol) to flag events that meet definitions such as MRSA or CRE. These flagged events populate AR organism reports (line list, frequency table, rate table) that report counts and percentages for defined phenotypes. She walked through a rate-table example: in January 2025, 46 Acinetobacter species events were reported, 44 were tested and 40 were nonsusceptible or resistant, yielding a reported percentage of about 91% (95% CI 79.5–97) for that phenotype and month.
On denominator data, Phils covered the AR summary line list, which shows patient days and admissions for FacWideIN and encounter counts for outpatient locations. She flagged a user interface error in the product: a header label reads 'survey patient days' though the data are from the AR summary record, and she said that bug has been submitted for a future NHSN release. She also clarified how the 'no AR events' field appears in output ("no" when events were reported, "yes" when the 'report no events' box was checked, and blank when neither applies).
Phils closed by offering practical scenarios and knowledge checks to help attendees pick the right report for tasks such as graphing isolates over the last 12 months (use the bar chart), counting phenotype-defined MDR isolates (use the frequency table), calculating percent susceptible for a pairing (use the facility-wide antibiogram), validating monthly AR events (use the AR Event Line List), and reviewing denominator validation (use the AR summary line list). She pointed attendees to the AUR module web page, quick reference guides, the CDA toolkits, and recorded trainings, and invited questions to the annual training community or via the NHSN help desk (ServiceNow or nhsn@cdc.gov).
The training is intended to help facility reporters validate data, select appropriate NHSN AR reports for surveillance and stewardship work, and correctly interpret output for local infection prevention and antimicrobial stewardship decisions.
