Lifetime Citizen Portal Access — AI Briefings, Alerts & Unlimited Follows
CDC webinar outlines 2022 NHSN rebaseline for CDI LabID SIR and how facilities can run updated reports
Loading...
Summary
CDC presenters Kevin and Karen explained the 2022 rebaseline standardized infection ratio (SIR) models for CDI LabID, the seven risk factors in the acute-care model, a worked example of the SIR calculation, and steps facilities should take to run updated NHSN reports and perform data-quality checks.
Two presenters in a CDC webinar explained the National Healthcare Safety Network's (NHSN) new 2022 rebaseline standardized infection ratio (SIR) models for Clostridioides difficile infection (CDI) LabID and showed how facilities can access and validate the updated NHSN reports.
Kevin, a presenter from CDC's statistics team, summarized the metric: "It's a simple ratio, simple division of the of a hospital's observed HAIs divided by their number of predicted HAIs," and described how the bottom (predicted events) is generated by statistical models rather than reported counts.
The explanation matters because NHSN now hosts the 2022 baseline models for facility and group users to run internal analyses, while CMS has said it plans to adopt the 2022 models for quality reporting in 2026. That means the choice of baseline and model parameters can change the expected number of events against which facilities are compared.
Kevin walked through the modeling approach: NHSN limited the baseline to a single year (2022) for the rebaseline, used a negative binomial regression suitable for count data, screened candidate facility characteristics in univariate tests, combined selected variables into a parsimonious multivariable model, and ran diagnostics to verify assumptions. For the acute-care hospital (ACH) CDI LabID model, the final model includes seven risk factors: inpatient community-onset prevalence rate, CDI test type, combined ED/observation community-onset prevalence rate, facility type (per NHSN enrollment), number of ICU beds, medical-school affiliation category, and average length of stay.
Using a fictitious example, Kevin demonstrated the calculation steps: sum the model parameter estimates (including the intercept), exponentiate that sum, multiply by the facility's patient days to get the predicted number of events (the transcript example produced 4.826 predicted events), then divide the observed event count (example: 2) by that predicted value to obtain the SIR (example SIR = 0.414). He also showed the associated p value (0.1869) and 95% confidence interval and noted that if the CI includes 1 or p>0.05, the SIR is not statistically different from 1.
Karen, returning to report navigation, described how the 2022 baseline reports appear in NHSN (new dataset names such as BS3/PS3, default quarterly cadence with an option to generate monthly results after regenerating datasets, and display/filter choices). She emphasized that "the CDI SIR for ACH or acute care hospitals is calculated at the facility wide inpatient, or you may hear me refer to it as FacWideIN level, and it cannot be calculated for any individual location," and noted that certain facility types (inpatient psychiatric facilities and Department of Veterans Affairs hospitals) are excluded from the 2022 reports.
Karen reviewed common reasons the SIR may not be calculated for a facility: the number predicted is less than 1; any required risk factor is missing (for example, missing annual-survey fields such as medical type, ICU bed count, or length of stay); or the CDI test method was not reported for the quarter. She explained NHSN displays a "CDI data months excluded from SAR due to a missing risk factor" table to help facilities identify omitted months and recommended specific data-quality checks: verify numerator events via the CDI LabID line list (indicator variable = 1 when an event is counted), review the risk-factor (table 2) and rates tables for prevalence rates, and confirm patient-day counts either via summary line lists or by examining the raw data-entry fields.
Resources Karen pointed to include the NHSN rebaseline web page, the updated 2022 SIR Guide, the NHSN SIR Model Explorer (to view parameter estimates), an implementation guide and change log listing inclusion/exclusion rules and differences from the 2015 models, and forthcoming TAP reports updated for the 2022 baseline. For support, users with SAMS access should submit questions through the ServiceNow portal; other users may email helpdesk@nhsncdc.gov. Media inquiries were directed to the CDC Media Office (mediacdc@cdc.gov).
Next steps noted in the webinar: facilities can begin running the 2022 baseline SIR reports in NHSN for internal analysis (regenerate datasets and use the group-by options to produce monthly or quarterly outputs), verify risk-factor and patient-day inputs when SIRs fail to calculate, and consult the model explorer and implementation guide for parameter-level details. CMS's formal adoption for quality reporting remains scheduled for 2026, per the presentation.
