NHSN webinar explains 2022 CAUTI baseline models, report changes and steps to run updated SIR reports

Irene Khan, subject matter expert with the NHSN Acute Care Analytics Team, and Rob Wild, statistician on the NHSN Statistics Team, led a webinar walkthrough of the 2022 baseline Standardized Infection Ratio (SIR) models for Catheter-Associated Urinary Tract Infections (CAUTI) and the updated analysis reports now available in NHSN.

The session focused on how the SIR is calculated and used. "The Standardized Infection Ratio or SIR is a risk‑adjusted summary measure for Healthcare‑Associated Infections," Rob said, and explained the SIR is computed by dividing the facility's observed events by the model‑predicted number of events. He demonstrated a worked example in which a fictitious acute care unit produced a predicted events value of 3.294 and an SIR of 0.911, and emphasized that p‑values and 95% confidence intervals are reported alongside the SIR to judge whether a facility's rate differs statistically from the national benchmark of 1.

Presenters described the modeling approach used for the 2022 baseline: NHSN fitted negative binomial regression models to 2022 facility data (with pre‑specified exclusion criteria) and evaluated covariates (location type, bed counts, medical school affiliation, average length of stay and proportion of beds that are ICU in the ACH model) to produce parameter estimates. Rob pointed attendees to the NHSN model explorer and SIR guide for full parameter tables and diagnostics.

Irene walked through operational changes and report navigation in NHSN. The 2022 baseline analytic reports are organized under a "Baseline Set 3" (bs3) folder and will produce SIRs only for 2022‑onward data; they are not compatible with pre‑2022 data because of changes in model structure and risk adjustments. She noted an important change for inpatient rehabilitation facilities (IRFs): IRF units embedded in hospitals now receive a separate IRF SIR under the 2022 baseline. She also flagged that facilities enrolled as public health emergency facilities and psychiatric hospitals/units are excluded from the 2022 CAUTI SIR calculations; those facilities may still run 2015 baseline reports if needed.

On data quality, Khan recommended a standard checklist before running SIR reports: confirm the monthly reporting plan includes all relevant locations; verify that CAUTI events meet NHSN surveillance criteria and are entered correctly (the caucount); ensure urinary catheter day denominators (numucathdays or numddays) are complete; resolve outstanding alerts; and regenerate datasets so reports incorporate any corrections. She explained the new Table 5 output ("CAUTI data not included in the SIR") lists records excluded from the SIR calculation and why (missing risk factor data, missing device dates, unresolved alerts), enabling facilities to troubleshoot gaps.

Khan also explained that when the model's predicted count (numPred) is less than one, the SIR, p‑value and confidence interval are not generated; facilities can expand the reporting window or use rate reports to track trends in those cases. Attendees were shown how to locate and modify reports (title, output format, time period, filters) and were directed to quick reference guides, the SIR guide, the implementation guide/change log, and the NHSN model explorer for parameter details.

The presenters reiterated external reporting and timing: "The HAI measures using this updated 2022 standard population will be publicly reported on the CMS compare tool beginning in fall 2026 using four quarters of the calendar year 2025 data," the webinar noted, and presenters advised facilities to familiarize themselves with 2022‑baseline SIRs for internal surveillance and benchmarking.

The webinar closed with links to consolidated resources, talking points to help explain the rebaseline to leadership and external partners, and contact routes for NHSN support.

Next procedural step: facilities should run the line listings and summary data line listing, resolve any alerts or missing survey fields flagged in Table 5, regenerate datasets (bs3), and then re-run SIR reports to ensure the 2022 baseline SIRs reflect corrected data.