An official website of the United States government
Risk Adjustment Factors Included in the SIR
MRSA Bacteremia LabID
MRSA Bacteremia LabID Event Risk Adjustment (CAHs)
The number of predicted MRSA bacteremia LabID events under the 2022 baseline is calculated using a negative binomial regression model and is risk adjusted based on the following variables found to be statistically significant predictors of MRSA bacteremia incidence. Information about the statistical properties of NHSN risk adjustment models, including how the number of predicted events is calculated, is available in NHSN’s Guide to the SIR (2022 baseline).
| Parameter | Parameter Estimate | Standard Error | P-value |
| Intercept | -11.3451 | 0.2029 | <0.0001 |
| Outpatient CO prevalence rate1: >0 per 100 encounters | 0.9991 | 0.2773 | 0.0003 |
| Outpatient CO prevalence rate1: 0 per 100 encounters or no applicable locations | REFERENT | – | – |
| Inpatient CO prevalence rate2: >0 per 100 admissions | 0.8824 | 0.3418 | 0.0098 |
| Inpatient CO prevalence rate2: 0 per 100 admissions | REFERENT | – | – |
Footnotes:
1 Outpatient community-onset (CO) prevalence rate combines MRSA bacteremia data from all emergency departments (EDs) and/or 24-hour observation units into a single, de-duplicated prevalence rate. This rate is calculated as the # of unique community-onset MRSA blood events that occurred in an ED or 24-hour observation unit / total encounters * 100. (i.e., MRSA_EDOBSprevCount / numTotencounters * 100). If the facility does not have an ED or 24-hour observation location that meets the NHSN location definition and thus are not reporting MRSA bacteremia data from these locations, the number of predicted events will be risk adjusted using the referent level of this variable.
2 Inpatient community-onset (CO) prevalence is calculated as the # of inpatient community-onset MRSA blood events, divided by total admissions x 100. (i.e., MRSA_admPrevBldCount / numadms * 100). The prevalence rate for the entire quarter is used in risk adjustment. Unrounded values truly greater than 0 are receiving the correct risk adjustment using the appropriate parameter category for values greater than 0.
