When the measure comes back for maintenance in 3 years, the developer will have: Explored the possibility of using other all-payer data sources to expand the use of patient-level factors in the risk adjustment model and reduce reliance on facility-level factors.
Annual risk-adjusted standardized infection ratio (SIR) of methicillin-resistant staphylococcus aureus (MRSA) bacteremia LabID Events among adults, children, and neonates hospitalized as inpatients at acute care and oncology hospitals. SIR is reported annually and is calculated by dividing the number of observed MRSA bacteremia LabID Events into the number of predicted MRSA bacteremia LabID Events.
Measure Specs
General Information
The use of this measure will promote Methicillin-resistant Staphylococcus aureus (MRSA) prevention activities that will lead to improved patient outcomes including reduction of MRSA infections, avoidable medical costs, and patient morbidity and mortality through reduced need for antimicrobials and reduced length of stay.
Data is submitted by facilities using the National Healthcare Safety Network (NHSN), web-based application (accessed securely via the Secure Access Management Service).
https://www.cdc.gov/nhsn/psc/cdiff/index.html
https://www.cdc.gov/nhsn/pdfs/pscmanual/12pscmdro_cdadcurrent.pdf
https://www.cdc.gov/nhsn/forms/57.128_LabIDEvent_BLANK.pdf
https://www.cdc.gov/nhsn/forms/57.127_MDROMonthlyReporting_BLANK.pdf
https://www.cdc.gov/nhsn/2022rebaseline/sir-guide.pdf
Numerator
Number of annually observed methicillin-resistant staphylococcus aureus (MRSA) bacteremia LabID Events in hospitalized inpatients.
- Determine the patients who have a positive lab finding from a blood culture for methicillin-resistant staphylococcus aureus (MRSA) Bacteremia LabID event and the date the Lab ID Event was identified.
- Includes Staphylococcus aureus cultured from a blood culture specimen that tests oxacillin-resistant, cefoxitin-resistant, or methicillin-resistant by standard susceptibility testing methods, or by a positive result from molecular testing for mecA and PBP2a; these methods may also include positive results of specimens tested by any other FDA approved PCR test for MRSA.
- Active surveillance testing is excluded.
- Determine if the patient is in an inpatient location.
- Determine if the specimen was collected >3 days after the patient’s admission to the hospital.
- The patient did not have any prior positive MRSA blood specimen LabID events in the previous 14 days in any inpatient location (including IRF/IPF units), emergency department, or 24-hour observation location. Specimen collection date is considered Day 1.
Denominator
Number of annually predicted methicillin-resistant staphylococcus aureus (MRSA) bacteremia LabID Events in hospitalized inpatients.
- Calculate the monthly number of inpatient days by summing the daily count of patients occupying beds, per inpatient location in the facility.
- Calculate the monthly number of inpatient admissions, per inpatient location.
- The number of predicted events in NHSN is calculated based on the 2022 national hospital onset MRSA LabID event aggregate data and is adjusted for each facility using variables found to be significant predictors of MRSA incidence. The number of predicted MRSA LabID Bacteremia Events is calculated using a negative binomial regression model.
- The general formula for the negative binomial regression model is
log (λ) = α + 𝛽𝛽1𝑋𝑋1 + 𝛽𝛽2𝑋𝑋2 + ··· + 𝛽𝛽i𝑋𝑋i , where:
α = Intercept
βi = Parameter estimate
Xi = Value of risk factor (categorical variables: 1 if present, 0 if not present)
i = Number of predictors
The tables below represent the variables found to be statistically significant predictors of MRSA bacteremia LabID events and are used in the negative binomial regression model to calculate the number of predicted healthcare facility-onset MRSA bacteremia LabID events in inpatient hospitals under the 2022 baseline data.
See 7.1 Supplemental Information Attachment Pages 1-2 for the MRSA bacteremia LabID event risk tables.
Exclusions
None
None
Measure Calculation
The National Healthcare Safety Network (NHSN) is a system for tracking healthcare-associated infections (HAIs) using data from US healthcare facilities. NHSN provides facility leaders, state health departments, and the nation with data needed to identify problem areas, measure progress of prevention efforts, and ultimately eliminate HAIs.
NHSN began tracking HAIs in around 300 hospitals and now serves over approximately 38,000 medical facilities. Current participants include acute care hospitals, long-term acute care hospitals, psychiatric hospitals, rehabilitation hospitals, outpatient dialysis centers, ambulatory surgery centers, and nursing homes, with hospitals (over 6,000) and dialysis facilities representing most of the facilities reporting data.
Establishing this system for tracking and preventing HAIs across the county required NHSN to understand key baseline data about facilities and healthcare. Information that allows NHSN to measure the incidence rates of HAIs represented in these baseline data includes:
- Facility demographics (like number of beds and medical school affiliation)
- Units within facilities (like the type of medical services or care provided on a unit)
- Surveillance data about infections (if, when, and where they occur)
The standardized infection ratio (SIR) is a summary metric used by healthcare facilities, CDC, and other public health organizations to track the incidence of HAIs over time. The SIR compares the number of HAIs reported (numerator) to the number that would be predicted (denominator), given the standard population (i.e., national baseline), adjusting for various facility and/or patient-level risk factors that have been found to be significantly associated with differences in HAI incidence. When interpreting the SIR, a value greater than 1.0 indicates that more HAIs were observed than predicted; conversely, an SIR less than 1.0 indicates that fewer HAIs were observed than predicted.
The MRSA Bacteremia LabID Event SIR compares the actual number of MRSA Bacteremia LabID Events reported to the number of MRSA Bacteremia LabID Events that would be predicted. The number of predicted infections is calculated using multivariable regression models generated from nationally aggregated data during a baseline period. These models are applied to a facility’s denominator and risk factor data to generate a predicted number of infections. To enforce a minimum precision criterion, facility SIRs are only calculated when the number of predicted infections is at least 1.0. This rule was instituted to avoid the calculation and interpretation of statistically imprecise SIRs, which typically have extreme values.
SIR = Observed (O) HAIs/Predicted (P) HAIs
1.Total the number of annually observed (numerator) MRSA Bacteremia LabID Events across the facility.
2. Calculate the number of predicted (denominator) MRSA Bacteremia LabID Events for the facility.
The number of predicted infections is the estimated number of MRSA Bacteremia LabID Events for the facility considering several facility factors reported to NHSN. The model is based on aggregated national data reported to NHSN during a specific timeframe (i.e. baseline year 2022). The negative binomial generalized linear model is utilized for MRSA Bacteremia LabID Events. As a national surveillance HAI tracking system that US healthcare facilities must report data to, NHSN must characterize risk of infection in the most efficient way. To minimize the burden of data collection on facilities, NHSN risk models utilize patient location and facility characteristics that are already reported by all facilities. NHSN does not collect additional patient characteristics for inclusion in the risk model because this would create additional burden for facilities.
Negative binomial regression models are used to estimate incidence from a summarized population. The general negative binomial regression formula is:
log (λ) = α + 𝛽𝛽1𝑋𝑋1 + 𝛽𝛽2𝑋𝑋2 + ··· + 𝛽𝛽i𝑋𝑋i , where:
α = Intercept
βi = Parameter estimate
Xi = Value of risk factor (categorical variables: 1 if present, 0 if not present)
i = Number of predictors
3. Divide the number of observed MRSA Bacteremia LabID Events by the number of predicted MRSA Bacteremia LabID Events to obtain the standardized infection ratio (SIR).
•If the SIR is greater than 1.0, then more HAIs were observed than predicted based on the 2022 national aggregate data.
• If the SIR is less than 1.0, then fewer HAIs were observed than predicted based on the 2022 national aggregate data.
• If the SIR equals 1.0, then the same number of HAIs were observed as predicted based on the 2022 national aggregate data.
The tables below represent the variables found to be statistically significant predictors of MRSA Bacteremia LabID Events and are used in the negative binomial regression model to calculate the number of predicted healthcare facility-onset MRSA Bacteremia LabID Events in hospital inpatients under the 2022 baseline data.
The negative binomial generalized linear models for acute care hospitals and critical access hospitals are listed below.
See 1.18a for details.
The measure is not stratified.
N/A
Supplemental Attachment
Measure Record
Point of Contact
NA
Andrea Benin
Atlanta , GA
United States
Paula Farrell
CDC NHSN
Atlanta , GA
United States
Importance
Evidence
A collection of prevention efforts has been identified to reduce the incidence of methicillin-resistant staphylococcus aureus (MRSA) Bacteremia LabID events. These interventions include (i) Appropriate use of antibiotics (ii) Implementing surveillance strategies (iii) Implementing infection control precautions (iv) Use of contact precautions (v) Environmental cleaning.
Clinical practice guidelines for the management of multidrug-resistant organisms (MDROs), including MRSA, have been published. Adherence to the recommendations in the guidelines can result in decreased rates of MDRO transmission and infection. Decreasing rates of infection will result in a lower SIR, which indicates improving performance.
Reference: Siegel JD, Rhinehart E, Jackson M, Chiarello L; Healthcare Infection Control Practices Advisory Committee. Management of multidrug-resistant organisms in health care settings, 2006. Am J Infect Control. 2007 Dec;35(10 Suppl 2):S165-93.
The 2021 National and State Healthcare-Associated Infections Progress Report showed statistically significant increases in hospital-onset methicillin-resistant Staphylococcus aureus (MRSA) (14%) between 2020 and 2021 (Lastinger, L., et al., 2023). Multiple strategies—such as improving catheter insertion techniques, following contact precautions, monitoring hand hygiene, disinfecting caps for intravenous lines, and chlorhexidine baths—have been found to reduce the number of healthcare acquired infections (HAIs) in hospitals. Bathing patients with chlorhexidine gluconate (CHG) wipes has decreased rates of multiple HAIs, including MRSA. A recent study evaluated the use of intranasal mupirocin twice daily and CHG baths daily for 5 days preoperatively in cardiac surgery patients who were colonized with MRSA. These patients also received prophylactic vancomycin and cefazolin with contact isolation precautions. The study showed that preoperative screening for S. aureus and decolonization was associated with a decrease in postoperative colonization (odds ratio 0.73, 95% CI: 0.53 to 1.00, p=0.05) (Saraswat, et. al, 2017). As many of the prevention strategies that facilities implement are nursing driven, given that the tasks are patient-care related another study found that American Nurses Credentialing Center Magnet designated hospitals, compared to non-Magnet hospitals, had a significant and positive coefficient (0.74, P < 0.001) and were associated with a lower MRSA bloodstream infections (Pakyz, A., et al., 2021). A multicenter, randomized, controlled trial where participants were randomly assigned to an education group or decolonization group. Those in the education group received and reviewed an educational binder about MRSA and how it is spread and recommendations for personal hygiene, laundry, and household cleaning (Huang, S., et al., 2019). The decolonization group received and reviewed the identical educational binder and underwent decolonization, which included 4% rinse-off chlorhexidine for daily showering, 0.12% chlorhexidine mouthwash twice daily, and 2% nasal mupirocin twice daily, for 5 days twice monthly for a period of 6 months after hospital discharge (Huang, S., et al., 2019). The study found that participants in the decolonization group who adhered fully to the regimen had 44% fewer MRSA infections than the education group (hazard ratio, 0.56; 95% CI, 0.36 to 0.86) and had 40% fewer infections from any cause (hazard ratio, 0.60; 95% CI, 0.46 to 0.78) (Huang, S., et al., 2019). Topical decolonization led to lower risks of infections and readmissions than hygiene education alone among patients after the transition from hospital to home and other care settings (Huang, S., et al., 2019).
- Huang SS, Singh R, McKinnell JA, Park S, Gombosev A, Eells SJ, Gillen DL, Kim D, Rashid S, Macias-Gil R, Bolaris MA, Tjoa T, Cao C, Hong SS, Lequieu J, Cui E, Chang J, He J, Evans K, Peterson E, Simpson G, Robinson P, Choi C, Bailey CC Jr, Leo JD, Amin A, Goldmann D, Jernigan JA, Platt R, Septimus E, Weinstein RA, Hayden MK, Miller LG; Project CLEAR Trial. Decolonization to Reduce Postdischarge Infection Risk among MRSA Carriers. N Engl J Med. 2019 Feb 14;380(7):638-650.
- Lastinger, L., Alvarez, C., Kofman, A., Konnor, R., Kuhar, D., Nkwata, A., . . . Dudeck, M. (2023). Continued increases in the incidence of healthcare-associated infection (HAI) during the second year of the coronavirus disease 2019 (COVID-19) pandemic. Infection Control & Hospital Epidemiology, 44(6), 997-1001.
- Pakyz AL, Wang H, Ozcan YA, Edmond MB, Vogus TJ. Leapfrog Hospital Safety Score, Magnet Designation, and Healthcare-Associated Infections in United States Hospitals. J Patient Saf. 2021 Sep 1;17(6):445-450.
- Saraswat MK, Magruder JT, Crawford TC, Gardner JM, Duquaine D, Sussman MS, Maragakis LL, Whitman GJ. Preoperative Staphylococcus Aureus Screening and Targeted Decolonization in Cardiac Surgery. Ann Thorac Surg. 2017 Oct;104(4):1349-1356.
Clinical Guideline: Siegel JD, Rhinehart E, Jackson M, Chiarello L; Healthcare Infection Control Practices Advisory Committee. Management of multidrug-resistant organisms in health care settings, 2006. Am J Infect Control. 2007 Dec;35(10 Suppl 2):S165-93.
Healthcare Infection Control Practices Advisory Committee (HICPAC) system for categorizing recommendations in this guideline is as follows:
Category IA. Strongly recommended for implementation and strongly supported by well-designed experimental, clinical, or epidemiologic studies.
Category IB. Strongly recommended for implementation and supported by some experimental, clinical, or epidemiologic studies and a strong theoretical rationale; or an accepted practice (e.g., aseptic technique) supported by limited evidence.
Category IC. Required by state or federal regulations, rules, or standards.
Category II. Suggested for implementation and supported by suggestive clinical or epidemiologic studies or a theoretical rationale.
No Recommendation. Unresolved issue. Practices for which insufficient evidence or no consensus regarding efficacy exists.
Strength of Recommendations
Recommendation
Definition: A Recommendation means that CDC is confident that the benefits of the recommended approach clearly exceed the harms (or, in the case of a negative recommendation, that the harms clearly exceed the benefits). In general, Recommendations should be supported by high- to moderate-quality evidence. In some circumstances, however, Recommendations may be made based on lesser evidence or even expert opinion when high-quality evidence is impossible to obtain and the anticipated benefits strongly outweigh the harms or when then Recommendation is required by federal law.
Implied Obligation: A Recommendation implies that healthcare personnel/healthcare facilities “should” implement the recommended approach unless a clear and compelling rationale for an alternative approach is present.
Conditional Recommendation
Definition: A Conditional Recommendation means that CDC has determined that the benefits of the recommended approach are likely to exceed the harms (or, in the case of a negative recommendation, that the harms are likely to exceed the benefits). Conditional Recommendations may be supported by either low-, moderate- or high-quality evidence when:
- There is high-quality evidence, but the benefit/harm balance is not clearly tipped in one direction.
- The evidence is weak enough to cast doubt on whether the recommendation will consistently lead to benefit.
- The likelihood of benefit for a specific patient population or clinical situation is extrapolated from relatively high-quality evidence demonstrating impact on other patient populations or in other clinical situations (e.g., evidence obtained during outbreaks used to support probable benefit during endemic periods).
- The impact of the specific intervention is difficult to disentangle from the impact of other simultaneously implemented interventions (e.g., studies evaluating “bundled” practices).
- There appears to be benefit based on available evidence, but the benefit/harm balance may change with further research.
- The benefit is most likely if the intervention is used as a supplemental measure in addition to basic practices.
Implied Obligation: A Conditional Recommendation implies that healthcare facilities/ personnel “could,” or could “consider” implementing the recommended approach. The degree of appropriateness may vary depending on the benefit vs. harm balance for the specific setting.
No Recommendation
Definition: No Recommendation is made when there is both a lack of pertinent evidence and an unclear balance between benefits and harms.
Level of Confidence in the Effect Estimate
High: Highly confident that the true effect lies close to that of the estimated size and direction of the effect. For example, confidence in the evidence is rated as “High” when there are multiple studies with no major limitations, there are consistent findings, and the summary estimate has a narrow confidence interval.
Moderate: The true effect is likely to be close to the estimated size and direction of the effect, but there is a possibility that it is substantially different. For example, confidence in the evidence is rated as “Moderate” when there are only a few studies and some have limitations but not major flaws, there is some variation between study results, or the confidence interval of the summary estimate is wide.
Low: The true effect may be substantially different from the estimated size and direction of the effect. For example, confidence in the evidence is rated as “Low” when supporting studies have major flaws, there is important variation between study results, the confidence interval of the summary estimate is very wide, or there are no rigorous studies.
V.A.3. Judicious Use of Antimicrobial Agents
V.A.3.a. In hospitals and LTCFs, ensure that a multidisciplinary process is in place to review antimicrobial utilization, local susceptibility patterns 36 (antibiograms), and antimicrobial agents included in the formulary to foster appropriate antimicrobial use. IB
V.A.3.b. Implement systems (e.g., computerized physician order entry, comment in microbiology susceptibility report, notification from a clinical pharmacist or unit director) to prompt clinicians to use the appropriate antimicrobial agent and regimen for the given clinical situation. IB
V.A.3.b.i. Provide clinicians with antimicrobial susceptibility reports and analysis of current trends, updated at least annually, to guide antimicrobial prescribing practices. IB
V.A.3.b.ii. In settings that administer antimicrobial agents but have limited electronic communication system infrastructures to implement physician prompts (e.g., LTCFs, home care and infusion companies), implement a process for appropriate review of prescribed antimicrobials. Prepare and distribute reports to prescribers that summarize findings and provide suggestions for improving antimicrobial use. II
V.A.4. Surveillance
V.A.4.a.In microbiology laboratories, use standardized laboratory methods and follow published guidance for determining antimicrobial susceptibility of targeted (e.g., MRSA, VRE, MDR-ESBLs) and emerging (e.g., VRSA, MDR-Acinetobacter baumannii) MDROs. IB
V.A.4.b.In all healthcare organizations, establish systems to ensure that clinical microbiology laboratories (in-house and out-sourced) promptly notify infection control staff or a medical director/ designee when a novel resistance pattern for that facility is detected. IB
V.A.4.c. In hospitals and LTCFs, develop and implement laboratory protocols for storing isolates of selected MDROs for molecular typing when needed to confirm transmission or delineate the epidemiology of the MDRO within the healthcare setting. IB
V.A.4.d. Prepare facility-specific antimicrobial susceptibility reports as recommended by the Clinical and Laboratory Standards Institute (CLSI), monitor these reports for evidence of changing resistance patterns that may indicate the emergence or transmission of MDROs. IB/IC
V.A.4.d.i. In hospitals and LTCFs with special-care units (e.g., ventilator-dependent, ICU, or oncology units), develop and monitor unit-specific antimicrobial susceptibility reports. IB
V.A.4.d.ii. Establish a frequency for preparing summary reports based on volume of clinical isolates, with updates at least annually. II/IC
V.A.4.d.iii. In healthcare organizations that outsource microbiology laboratory services (e.g., ambulatory care, home care, LTCFs, smaller acute care hospitals), specify by contract that the laboratory provide either facility-specific susceptibility data or local or regional aggregate susceptibility data in order to identify prevalent MDROs and trends in the geographic area served. II
V.A.4.e. Monitor trends in the incidence of target MDROs in the facility over time using appropriate statistical methods to determine whether MDRO rates are decreasing and whether additional interventions are needed. IA
V.A.4.e.i. Specify isolate origin (i.e., location and clinical service) in MDRO monitoring protocols in hospitals and other large multi-unit facilities with high-risk patients. IB
V.A.4.e.ii. Establish a baseline (e.g., incidence) for targeted MDRO isolates by reviewing results of clinical cultures; if more timely or localized information is needed, perform baseline point prevalence studies of colonization in high-risk units. When possible, distinguish colonization from infection in analysis of these data. IB
V.A.5. Infection Control Precautions to Prevent Transmission of MDROs
V.A.5.a. Follow Standard Precautions during all patient encounters in all settings in which healthcare is delivered. IB
V.A.5.b. Use masks according to Standard Precautions when performing splash-generating procedures (e.g., wound irrigation, oral suctioning, intubation) when caring for patients with open tracheostomies and the potential for projectile secretions and in circumstances where there is evidence of transmission from heavily colonized sources (e.g., burn wounds). Masks are not otherwise recommended for prevention of MDRO transmission from patients to healthcare personnel during routine care (e.g., upon room entry). IB
V.A.5.c. Use of Contact Precautions
V.A.5.C.I. IN ACUTE-CARE HOSPITALS
V.A.5.c.i. Implement Contact Precautions routinely for all patients infected with target MDROs and for patients that have been previously identified as being colonized with target MDROs (e.g., patients transferred from other units or facilities who are known to be colonized). IB
V.A.5.H. DISCONTINUATION OF CONTACT PRECAUTIONS
V.A.5.h.No recommendation can be made regarding when to discontinue Contact Precautions. (See Background for discussion of options.) Unresolved issue
V.A.5.I. PATIENT PLACEMENT IN HOSPITALS AND LTCFS
V.A.5.i.1. When single-patient rooms are available, assign priority for these rooms to patients with known or suspected MDRO colonization or infection. Give highest priority to those patients who have conditions that may facilitate transmission, e.g., uncontained secretions or excretions. IB
V.A.5.i.2. When single-patient rooms are not available, cohort patients with the same MDRO in the same room or patient-care area. IB
* V.A.5.i.3. When cohorting patients with the same MDRO is not possible, place MDRO patients in rooms with patients who are at low risk for acquisition of MDROs and associated adverse outcomes from infection and are likely to have short lengths of stay. II
V.A.6. Environmental Measures
V.A.6.a. Clean and disinfect surfaces and equipment that may be contaminated with pathogens, including those that are in close proximity to the patient (e.g., bed rails, over bed tables) and frequently-touched surfaces in the patient care environment (e.g., door knobs, surfaces in and surrounding toilets in patients’ rooms) on a more frequent schedule compared to that for minimal touch surfaces (e.g., horizontal surfaces in waiting rooms). IB
V.A.6.b. Dedicate noncritical medical items to use on individual patients known to be infected or colonized with MDROs. IB
V.A.6.c. Prioritize room cleaning of patients on Contact Precautions. Focus on cleaning and disinfecting frequently touched surfaces (e.g., bedrails, bedside commodes, bathroom fixtures in the patient’s room, doorknobs) and equipment in the immediate vicinity of the patient. IB
Measure Impact
The Patient Safety Action Network is a coalition of individuals and organizations consisting of patients who have been medically harmed, their loved ones, and concerned patient safety advocates.
“Please accept these comments from the Patient Safety Action Network regarding the following HAI measures; we are commenting on all of them together:
- Catheter-Associated Urinary Tract Infections (CAUTI)
- Central Line Associated Blood Stream Infections (CLABSI
- 30-Day Post-Operative Colon Surgery (COLO) and Abdominal Hysterectomy (HYST) Surgical Site Infection (SSI)
- Methicillin-resistant Staphylococcus aureus (MRSA) Bacteremia LabID Event
- Clostridioides difficile (CDI) LabID Event
- Antimicrobial Use Measure
Fundamentally, each of these measures is important and essential to preventing infections. If we do not measure and publicly report these events in a continuous, standardized way, we cannot truly know or understand when actual progress is made.
There are several target populations for these measures. First, members of the public who may need to use the services of a local hospital at any given point without warning or who have an interest in seeing how their hospital compares to others on hospital acquired infections. The published HAI measures provide that public service. Second, patients being treated at a hospital who are infected might not benefit from the past published HAI measures, but they probably are interested in accountability. One of the first questions many ask is “will my infection be counted?” The next question typically is, “how can we prevent it from happening again to someone else?” To them, these measurements are very important.
The value and meaningfulness of these outcome measures lie in tracking reduction of patient harm over time using individual hospitals’ HAI measures. Progress means fewer infections at each point of measurement with a goal toward no infections. Unfortunately, these measures are rarely presented on a continuum demonstrating whether each hospital has reduced this harm over the years. And they are no longer presented with the actual numbers of infections, which reflect actual infections reported and not an estimate.
We also believe the value of these measures is lowered because of the way they are reported to the public. It appears that the standardization using an SIR of 1.0 as the baseline has established that as the status quo, even though the baseline has been adjusted over time. We wonder how often hospitals accept SIRs of around 1.0 as acceptable. Further, the use of risk adjustment skews the real results in each of these measures, i.e., the patients who got infected. We would rather see a stratified presentation that compares similar hospitals together – without risk adjustments. We believe that would be more meaningful to the public.
Also, the terms used to present the data lead to confusion, such as predicted number of infections and better than/no different/worse than the national benchmark. Many hospitals’ data is “not available,” without context (the hospital failed to report, the hospital does not have enough cases to rate, etc).
Even with these limitations, the measures are important to retain because of their value to patients who expect to be free from preventable harm when hospitalized. You ask about the full meaning of these measures to patients, but that requires some understanding of what happens to them following a hospital acquired infection. These events affect each person in a different way. It can mean a round of antibiotics; a longer stay in the hospital or the need to seek further treatment; continued chronic conditions, including recurrences of the infection; significant medical debt; losing a job due to missing work as a consequence of an infection; losing one’s home due to mounting medical bills and other debts; permanent disability; sepsis that is only survived after intense medical care; and death. This should clearly explain why all these measures are meaningful to patients.
Frankly, we need more infection measures so that all hospital acquired infections are accounted for, like what is done in California. It seems to us that every time federal agencies ask for feedback about these measures, the result is less information to the public.”
Methicillin-resistant Staphylococcus aureus (MRSA) Standardized Infection Ratio serves as a broad, objective measure of healthcare-associated infection (HAI) burden within many patient care locations. HAI reduction has been a national priority set by U.S. Government going back to 2008 with the U.S. Health and Human Services (HHS) National Action Plan to Prevent Health Care-associated Infections: Roadmap to Elimination.1 While there has been overall progress in reducing these specific HAIs, there is room for improvement in both the surveillance and prevention of MRSA.
Measuring MRSA has also been a priority for CMS as indicated by the use of the measure in three CMS Measure Programs, including Hospital-Acquired Condition Reduction Program (HACRP), Hospital Inpatient Quality Reporting Program (HIQR), The Prospective Payment System (PPS)-Exempt Cancer Hospital Quality Reporting (PCHQR) Program, Hospital Value-Based Purchasing Program.
- U.S. Health and Human Services (HHS) National Action Plan to Prevent Health Care-associated Infections: Roadmap to Elimination. Accessed May 2, 2025 at https://www.hhs.gov/oidp/topics/health-care-associated-infections/hai-a…;
Performance Gap
A total of 1,907 Acute Care Hospitals qualified for the measure having at least 1 predicted event. The mean SIR across all hospitals was 0.77 with a range of 0-5.92. A total of 355 hospitals have an SIR=0 meaning they had zero MRSA events. The 10 decile groups represent 190 or 191 hopsitals. The range of mean performance across the 10 groups ranges from 0 to 2.176, indicating a wide range of performance across hospitals.
No Critical Access Hospitals qualified for the metric reporting as all had <1 predicted event.
| Overall | Minimum | Decile_1 | Decile_2 | Decile_3 | Decile_4 | Decile_5 | Decile_6 | Decile_7 | Decile_8 | Decile_9 | Decile_10 | Maximum | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Mean Performance Score | 0.77 | 0.0 | 0.0 | 0.222 | 0.319 | 0.472 | 0.599 | 0.737 | 0.892 | 1.097 | 1.381 | 2.176 | 5.92 |
| N of Entities | 1907 | 190 | 191 | 191 | 190 | 191 | 191 | 190 | 191 | 191 | 191 | ||
| N of Persons / Encounters / Episodes | 150508334 | 8112144 | 10378951 | 17437085 | 16675838 | 18633041 | 16523005 | 17872656 | 18403890 | 14985851 | 11485873 |
Care Gaps
Closing Care Gaps
This criteria is optional for the Fall 2025 cycle.
Feasibility
Feasibility
This is a maintenance measure, and the measure specifications have not changed. Facilities have not notified NHSN of any feasibility issues within the last year.
All required data elements are routinely generated, in structured fields, and used during care delivery. Facilities can choose to submit this data manually via a web form or via submission of CDA electronic files. NHSN has built-in business rules for mandatory data elements and does not allow for the submission of incomplete records.
Addressing NHSN data quality issues is integral to NHSN’s ability to help facilities collect the data needed to identify areas needing prevention efforts, measure progress of prevention efforts, and push toward MRSA elimination. The NHSN team routinely reviews the data reported to NHSN and contacts facilities to resolve confirmed and suspected data quality flags. Data quality checks conducted to help confirm the accuracy of the data being reported include checking MRSA data, implementing business rules within the application, verifying alerts, and confirming that flags are triggered by incomplete data.
NHSN provides facilities with internal validation toolkits, which can be used to audit their internal data to identify any potential inaccuracies or problems. The internal validation toolkit also provides recommendations to facilities for implementing quality control processes to ensure data is accurate and complete.
Additionally, NHSN offers external validation toolkits, which can be used by state or local health departments, or other auditors, to perform checks on the data that facilities submit to NHSN. External validation allows for the auditors to identify gaps in understanding of surveillance definitions or other errors and provide education to ensure data reported to NHSN follows the standardized specifications.
Per the Paperwork Reduction Act (PRA) of 1995, federal agencies cannot conduct or sponsor the collection of information unless the Office of Management and Budget (OMB) has reviewed and approved the proposed data collection. Federal agencies must submit a set of documents known as an Information Collection Request (ICR), to request OMB approval of an information collection. The ICR documents describe what information is needed; why it is needed; how it will be collected; and how much time, money, and effort it will cost the respondents to collect the information.
Multiple data collection forms are utilized to provide surveillance data on MRSA Bacteremia LabID Events. Below are the OMB-approved estimated total annual burden hours and annual cost for all facilities that complete this data collection.
See 7.1 Supplemental Information Attachment Page 3 for burden and cost details.
While CDC can retrieve data by personal identifier, CDC does not, as a matter of practice or policy, retrieve data in this way. Specifically, the primary practice and policy of CDC regarding NHSN data is to retrieve data by the name of the hospital or another non-personal identifier, not an individual patient, for surveillance and public health purposes. Furthermore, patient identifiers are not necessary for NHSN to operate.
An Assurance of Confidentiality is granted for all data collected under NHSN. NHSN’s Assurance of Confidentiality, states the following:
“The voluntarily provided information obtained in this surveillance system that would permit identification of any individual or institution is collected with a guarantee that it will be held in strict confidence, will be used only for the purposes stated, and will not otherwise be disclosed or released without the consent of the individual, or the institution in accordance with Sections 304, 306 and 308(d) of the Public Health Service Act (42 USC 242b, 242k, and 242m(d)).”
This is a maintenance measure, and the measure specifications have not changed.
Proprietary Information
Scientific Acceptability
Testing Data
Reliability Testing: The dataset used for testing is the Center for Disease Control’s (CDC) National Healthcare Safety Network (NHSN), which collects healthcare infection data from facilities throughout the United States. Data are from 1/1/2024 to 12/31/2024.
Risk Adjustment:
The dataset used for the risk adjustment model was derived from the 2022 Center for Disease Control’s (CDC) National Healthcare Safety Network (NHSN), which includes healthcare infection data from facilities reported from 1/1/2022 to 12/31/2022 throughout the United States. The data includes in-plan MRSA bacteremia LabID data and risk factors derived from facility enrollment information and the annual facility survey.
Validity Testing: The dataset used for testing is the Center for Disease Control’s (CDC) National Healthcare Safety Network (NHSN), which collects healthcare infection data from facilities throughout the United States. Data is from 1/1/2024 to 12/31/2024.
Validation Studies:
Date of data used in testing: December 1, 2021, until October 31, 2023
Prascius S, Wells A, Collier AM, Renard A, Hooper D, Stein T. Reduction of hospital-onset methicillin-resistant Staphylococcus aureus (MRSA) bacteremia with the use of twice daily alcohol-based nasal antiseptic in intensive care units. Am J Infect Control. 2025 Aug 19:S0196-6553(25)00504-8. doi: 10.1016/j.ajic.2025.08.006.
Date of data used in testing: January 1, 2013-December 31, 2013.
Pakyz AL, Wang H, Ozcan YA, Edmond MB, Vogus TJ. Leapfrog Hospital Safety Score, Magnet Designation, and Healthcare-Associated Infections in United States Hospitals. J Patient Saf. 2021 Sep 1;17(6):445-450.
Reliability Testing: The dataset used for testing is the Center for Disease Control’s (CDC) National Healthcare Safety Network (NHSN), which collects healthcare infection data from facilities throughout the United States. Data are from 1/1/2024 to 12/31/2024.
Risk Adjustment:
The dataset used for the risk adjustment model was derived from the 2022 Center for Disease Control’s (CDC) National Healthcare Safety Network (NHSN), which includes healthcare infection data from facilities reported from 1/1/2022 to 12/31/2022 throughout the United States. The data includes in-plan MRSA bacteremia LabID data and risk factors derived from facility enrollment information and the annual facility survey.
Validity Testing: The dataset used for testing is the Center for Disease Control’s (CDC) National Healthcare Safety Network (NHSN), which collects healthcare infection data from facilities throughout the United States. Data is from 1/1/2024 to 12/31/2024.
Validation Studies:
Date of data used in testing: December 1, 2021, until October 31, 2023
Prascius S, Wells A, Collier AM, Renard A, Hooper D, Stein T. Reduction of hospital-onset methicillin-resistant Staphylococcus aureus (MRSA) bacteremia with the use of twice daily alcohol-based nasal antiseptic in intensive care units. Am J Infect Control. 2025 Aug 19:S0196-6553(25)00504-8. doi: 10.1016/j.ajic.2025.08.006.
Date of data used in testing: January 1, 2013-December 31, 2013.
Pakyz AL, Wang H, Ozcan YA, Edmond MB, Vogus TJ. Leapfrog Hospital Safety Score, Magnet Designation, and Healthcare-Associated Infections in United States Hospitals. J Patient Saf. 2021 Sep 1;17(6):445-450.
Reliability Testing: The dataset used for testing is the 2024 Center for Disease Control’s (CDC) National Healthcare Safety Network (NHSN) data, which collects healthcare infection data from facilities throughout the United States. Hospitals were excluded from the reporting if they had <1 predicted event.
Risk Adjustment:
The 2022 national aggregate data are reviewed for all potential data quality issues, including outlier values prior to performing the risk adjustment modeling of the SIR denominator for the MRSA bacteremia LabID model. Based on the surveillance protocol for MRSA, data were excluded from modeling consideration if it met the criteria: Inpatient rehabilitation locations and inpatient psychiatric locations that have their own Centers for Medicare and Medicaid Services (CMS) Certification Number (CCN) are excluded.
Validity Testing:
The dataset used for testing is the 2024 Center for Disease Control’s (CDC) National Healthcare Safety Network (NHSN data, which collects healthcare infection data from facilities throughout the United States. Only facilities with both a MRSA and central line associated bloodstream infection (CLABSI) SIR, or a MRSA and Clostridioidesdifficile infection (CDI) LabID Event were included in the analysis (facilities with >=1 predicted event for both event types, respectively, were included).
Validation Studies:
MRSA SIR data was reported to NSHN from December 1, 2021, until October 31, 2023
Prascius S, Wells A, Collier AM, Renard A, Hooper D, Stein T. Reduction of hospital-onset methicillin-resistant Staphylococcus aureus (MRSA) bacteremia with the use of twice daily alcohol-based nasal antiseptic in intensive care units. Am J Infect Control. 2025 Aug 19:S0196-6553(25)00504-8. doi: 10.1016/j.ajic.2025.08.006.
MRSA SIR data was reported to NHSN January 1, 2013 to December 31, 2013.
Pakyz AL, Wang H, Ozcan YA, Edmond MB, Vogus TJ. Leapfrog Hospital Safety Score, Magnet Designation, and Healthcare-Associated Infections in United States Hospitals. J Patient Saf. 2021 Sep 1;17(6):445-450
See 7.1 Supplemental Information Attachment Pages 4-6 for details.
Reliability Testing:
The MRSA risk models used to calculate the predicted number of events were developed using patient care location- and facility-level factors. Since the data collection design did not allow for the capture of patient-level factors such as age or sex, these models are informed by surrogates of patient acuity (e.g., patient care location type, etc.).
Risk Adjustment:
The MRSA bacteremia LabID risk models used to calculate the predicted number of events were developed using facility-level factors and facility-wide inpatient data. The data collection design did not allow for the capture of patient-level factors, such as age or sex.
Validity Testing:
The MRSA risk models used to calculate the predicted number of events were developed using patient care location- and facility-level factors. Since the data collection design did not allow for the capture of patient-level factors such as age or sex, these models are informed by surrogates of patient acuity (e.g., patient care location type, etc.).
Validation Studies:
Prascius S, Wells A, Collier AM, Renard A, Hooper D, Stein T. Reduction of hospital-onset methicillin-resistant Staphylococcus aureus (MRSA) bacteremia with the use of twice daily alcohol-based nasal antiseptic in intensive care units. Am J Infect Control. 2025 Aug 19:S0196-6553(25)00504-8. doi: 10.1016/j.ajic.2025.08.006.
Patient who developed a MRSA LabID event.
Pakyz AL, Wang H, Ozcan YA, Edmond MB, Vogus TJ. Leapfrog Hospital Safety Score, Magnet Designation, and Healthcare-Associated Infections in United States Hospitals. J Patient Saf. 2021 Sep 1;17(6):445-450.
Patient who developed a MRSA LabID event.
Reliability
Signal-to-noise (SNR) reliability testing was performed to distinguish measure scores between facilities (Adams J.L. 2009). The annual standardized infection ratio (SIR) is defined as the sum of observed (O) events at the facility divided by the sum of predicted (P) events calculated from the risk-adjustment model. Signal-to-noise reliability testing denotes between-facility variance and within-facility variance (Adams J.L. 2009). The SNR for each facility SIR is calculated using both the between-facility and within-facility variance across eligible facilities with predicted number ≥1. The between-facility variance is simply the total variance of the SIR facility distribution. However, the within-facility variance of the SIR for each facility was then calculated as Var(O/P) where P is a constant, a nuisance factor with no random variation. The observed (O) was assumed to follow a Poisson distribution with a mean parameter lambda approximated by P. The result is Var(O/P) = Var(O)/P2 = P/P2 = 1/P. Signal to noise reliability scores can range from 0 to 1. A reliability of zero implies that all the variability in a measure is attributable to measurement error. A reliability of one implies that all the variability is attributable to real difference in performance.
References:
- Adams, J. L. (2009). The reliability of provider profiling: a tutorial. RAND.
We calculated the signal-to-noise reliability score for each facility that had at least one predicted MRSA event. Reliability testing was performed on data from 2024, for the Acute Care Hospitals. The mean reliability score was 0.58. There was not sufficient data in the CAH cohort for reliability analysis.
The percentage of facilities with an estimated reliability of >=0.6 was as follows 45% (856/1907).
We calculated the signal-to-noise reliability score for each facility that had at least one predicted MRSA event. Reliability testing was performed on data from 2024, for the Acute Care Hospitals. The mean reliability score was 0.58. There was not sufficient data in the CAH cohort for reliability analysis. The median signal-to-noise reliability score demonstrates moderate reliability.
The percentage of facilities with an estimated reliability of >=0.6 was as follows 45% (856/1907). The decile distribution of reliability measurements can be located in section 5.2.3a above.
Signal-to-Noise reliability scores vary across facilities from zero to one, with a score of zero indicating that all variation is attributable to noise (variation across patients within facilities) and a score of one indicating that all variation is caused by real differences in performance across facilities.
Our interpretation of the results is based on the standards established by Landis and Koch (1977):
< 0 – Less than chance agreement
0 – 0.2 Slight agreement
0.21 – 0.39 Fair agreement
0.4 – 0.59 Moderate agreement
0.6 – 0.79 Substantial agreement
0.8 – 0.99 Almost Perfect agreement
1 Perfect agreement
Landis, J. R., & Koch, G. G. (1977). The measurement of observer agreement for categorical data. biometrics, 159-174.
| | Overall | Minimum | Decile_1 | Decile_2 | Decile_3 | Decile_4 | Decile_5 | Decile_6 | Decile_7 | Decile_8 | Decile_9 | Decile_10 | Maximum |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Reliability | 0.58 | 0.308 | 0.335 | 0.389 | 0.446 | 0.496 | 0.548 | 0.601 | 0.653 | 0.712 | 0.781 | 0.875 | 0.947 |
| Mean Performance Score | 0.770 | 0.766 | 0.751 | 0.790 | 0.690 | 0.715 | 0.776 | 0.758 | 0.802 | 0.828 | 0.822 | ||
| N of Entities | 1907 | 190 | 191 | 191 | 190 | 191 | 191 | 190 | 191 | 191 | 191 | ||
| N of Persons / Encounters / Episodes | 150508334 | 5053570 | 6235125 | 7766428 | 9076812 | 10924702 | 12848588 | 14860455 | 18667779 | 24087685 | 40987190 |
Validity
Validity Testing:
Spearman correlation coefficients were calculated to assess a hypothesized monotonic relationship in the positive direction between the annual MRSA and CDI and MRSA and CLABSI Standardized Infection Ratios (SIR). The annual SIR is defined as the sum of observed (O) events at the facility divided by the sum of predicted (P) events calculated from the risk-adjustment model. Each facility that reported both MRSA and CDI, or MRSA and CLABSI, data for 2024 with at least 1 predicted event for each HAI, respectively, was included. If a facility reported only one of the listed HAI events or did not have at least 1 predicted event for the paired HAIs, they were excluded from the analysis. Correlation coefficients range from -1 to +1, where a coefficient of -1 implies a perfect negative correlation, 0 implies no correlation, and +1 implies a perfect positive correlation. A significance threshold of 0.05 was used to test the result.
We hypothesized that there would be a positive correlation between MRSA and CDI SIRs as well as MRSA and CLABSI SIRs because there is overlap in the infection prevention practices preventing these types of infections (for example, hand hygiene, assessing catheter need and implementing protocols for removal, performing aseptic technique for insertion, performing surveillance in ICU and non-ICU locations, and use of care and maintenance bundles). Thus, we predicted that while the correlations would be positive, they would be weak correlations.
Validation Studies:
A 191-bed hospital within a Midwest health care system experienced an increase in hospital acquired Methicillin-resistant Staphylococcus aureus (MRSA) bacteremia. In response, the hospital sought a targeted intervention for MRSA prevention within the MICU and SICU using an alcohol based nasal antiseptic. In August 2021, Infection Prevention and Control (IPC) partnered with Nursing and the alcohol-based nasal antiseptic product vendor to implement the product in the 2 units. Nasal antiseptic was administered twice daily. Data collected from December 1, 2021, until October 31, 2023 were analyzed using SAS 9.4. To measure the outcome, IPC continued tracking incidence of hospital-onset MRSA (HO-MRSA) bacteremia using the NHSN definitions to generate a SIR using the 2015 baseline model. MRSA rates in 2021, 2022, and 2023 were calculated using total MRSA LabID cases for each year divided by inpatient days. Statistical significance was derived using a two-tailed z-test.
Prascius S, Wells A, Collier AM, Renard A, Hooper D, Stein T. Reduction of hospital-onset methicillin-resistant Staphylococcus aureus (MRSA) bacteremia with the use of twice daily alcohol-based nasal antiseptic in intensive care units. Am J Infect Control. 2025 Aug 19:S0196-6553(25)00504-8. doi: 10.1016/j.ajic.2025.08.006.
Empirical validity testing at the accountable entity level was performed by evaluating published studies from facilities that implemented MRSA prevention activities and hypothesized that these approaches would reduce their MRSA LabID event SIR.
The study focuses on organizational factors such as a hospital’s safety infrastructure (indicated by Leapfrog Hospital Safety Score) or workplace quality (Magnet recognition) to determine whether Magnet and hospitals with better Leapfrog Hospital Safety Scores have fewer hospital associated infections. An ordered probit regression analyses tested associations between Safety Score, Magnet status, and standardized infection ratios (SIR), depicting whether a hospital had a methicillin-resistant Staphylococcus aureus (MRSA) bloodstream infection standardized infection ratio that was “better,” “no different,” or “worse” than a National Benchmark as per CDC NHSN. A total of 1,701 hospitals were included in the study.
Pakyz AL, Wang H, Ozcan YA, Edmond MB, Vogus TJ. Leapfrog Hospital Safety Score, Magnet Designation, and Healthcare-Associated Infections in United States Hospitals. J Patient Saf. 2021 Sep 1;17(6):445-450. doi: 10.1097/PTS.0000000000000378. PMID: 28452915.
Validation Testing:
The acute care hospitals with both a MRSA SIR and a CDI SIR (n=1,906) had a weak but significant positive correlation (rho= 0.05165, p= 0.0241). The acute care hospitals with both a MRSA SIR and a CLABSI SIR (n= 1,869) also had a weak but significant positive correlation (rho= 0.22, p< .0001).
Validation Studies:
From 2021 to 2023, MRSA standardized infection ratios (SIR) declined with an SIR of 1.18 in 2021 to an SIR of 0.90 in 2023.
Prascius S, Wells A, Collier AM, Renard A, Hooper D, Stein T. Reduction of hospital-onset methicillin-resistant Staphylococcus aureus (MRSA) bacteremia with the use of twice daily alcohol-based nasal antiseptic in intensive care units. Am J Infect Control. 2025 Aug 19:S0196-6553(25)00504-8. doi: 10.1016/j.ajic.2025.08.006.
The study found that relative to non-Magnet hospitals, hospitals with a Magnet designation have a significant and positive relationship with MRSA bloodstream infections (0.74, P < 0.001) and are associated with fewer than expected MRSA infections.
Pakyz AL, Wang H, Ozcan YA, Edmond MB, Vogus TJ. Leapfrog Hospital Safety Score, Magnet Designation, and Healthcare-Associated Infections in United States Hospitals. J Patient Saf. 2021 Sep 1;17(6):445-450. doi: 10.1097/PTS.0000000000000378. PMID: 28452915.
Validity Testing:
The significant positive correlations between MRSA and CDI SIR (rho= 0.06775, p= 0.0031) and MRSA and CLABSI SIR (rho= 0.18898, p< .0001) in acute care hospitals demonstrate that the SIRs are valid measures of healthcare quality, as they are all driven by clinically relevant patient care practices and evidence-based infection prevention strategies implemented by the healthcare facilities.
MRSA – CDI SIR: The CDI SIR and MRSA SIR are both laboratory-identified healthcare associated infection outcome measures. Implementation of infection prevention strategies, such as hand hygiene, have been shown to decrease the spread of C.difficile and MRSA. Environmental cleaning and disinfection, and appropriate use of contact/isolation precautions are also important prevention practices that can decrease the transmission of both pathogens.
However, other factors or prevention strategies may differ between the two pathogens. For example, facilities may focus antimicrobial stewardship resources to avoiding or limiting antibiotics associated with high CDI risk (such as fluoroqinolones). Alternatively, some facilities may choose to implement decolonization strategies aimed to reduce MRSA bloodstream infections. For these reasons, we hypothesized that there would be a weak positive correlation between the CDI SIR and MRSA SIR. We predicted only a weak correlation between the two measures because some facilities may choose to focus quality improvement on the prevention of a single HAI (CDI or MRSA) due to resource limitations or other factors.
MRSA – CLABSI SIR: The MRSA standardized infection ratio (SIR) and CLABSI SIR are both healthcare associated infection outcome measures. Implementation of infection prevention strategies, such as prevention bundles or checklists for infection prevention, have been shown to decrease these SIRs.
Some infection prevention strategies are employed to prevent both device-associated infections (CLABSI) as well as pathogen-specific infections (MRSA bacteremia). These include hand hygiene and line insertion and care practices (for both central and peripheral lines), antimicrobial stewardship, and CHG bathing plus intranasal mupiorocin for patients with a central line. Other infection prevention strategies may be targeted at just one of the HAIs, for example nasal decolonization for all ICU patients (not just those with a central line) to decrease MRSA, or daily review of central lines to decrease CLABSI risk.
For these reasons, we hypothesized that there would be a weak positive correlation between the MRSA and CDI SIRs, as well as MRSA and CLABSI SIRs. We predicted only weak correlations between the paired measures because some facilities may choose to focus quality improvement on the prevention of a single HAI due to resource limitations or other factors.
Validation Studies:
From 2021 to 2023, MRSA standardized infection ratios (SIR) declined with an SIR of 1.18 in 2021 to an SIR of 0.90 in 2023. The total number of MRSA cases decreased from 2021 to 2022 with 4 cases and 1 case, respectively. This study supports the premise that the measure score represents an improvable quality measure.
Prascius S, Wells A, Collier AM, Renard A, Hooper D, Stein T. Reduction of hospital-onset methicillin-resistant Staphylococcus aureus (MRSA) bacteremia with the use of twice daily alcohol-based nasal antiseptic in intensive care units. Am J Infect Control. 2025 Aug 19:S0196-6553(25)00504-8. doi: 10.1016/j.ajic.2025.08.006.
The study found that relative to non-Magnet hospitals, hospitals with a Magnet designation have a significant and positive relationship with MRSA bloodstream infections (0.74, P < 0.001) and are associated with fewer than expected MRSA infections. Additionally, the study demonstrates that Magnet designated hospitals have decreased rates of reported MRSA infections. Thus, this study supports the hypothesis that the measure score correctly reflects the quality of care provided and adequately identifies differences in quality.
Pakyz AL, Wang H, Ozcan YA, Edmond MB, Vogus TJ. Leapfrog Hospital Safety Score, Magnet Designation, and Healthcare-Associated Infections in United States Hospitals. J Patient Saf. 2021 Sep 1;17(6):445-450. doi: 10.1097/PTS.0000000000000378. PMID: 28452915.
Risk Adjustment
NHSN follows a highly rigorous process while developing risk adjustment models for its measures. The process begins with a thorough clinical and epidemiological review of all eligible potential risk factors that are currently collected in NHSN. The data available in NHSN are a combination of facility-level, unit-level, and limited patient-level risk factors. Those experts then recommend risk factors to be evaluated statistically. CDC obtains the risk factors considered for the model predicted events (i.e., denominator) by estimating the parameters or probability of risk occurrence. The final model is chosen by finding the optimal parameterizations of all covariates (i.e., risk factors) in linear regression procedures. In other words, risk factors are included in a model if they are determined to significantly impact healthcare associated infection (HAI) incidence. The model is then double-tested by a reverse process that removes non-significant factors. Each best model is fit-tested, calibrated, and validated using industry standard techniques.
References:
- NHSN's Guide to the 2022 Baseline Standardized Infection Ratios. Centers for Disease Control and Prevention website. https://www.cdc.gov/nhsn/2022rebaseline/sir-guide.pdf.
- Obtaining the Number of Predicted Events for the Standardized Infection Ratio (SIR)
- https://wwwdev.cdc.gov/nhsn/2022rebaseline/index.html
See attachment under 5.4.3a below.
Each potential risk factor was tested for association with the outcome using Wald, Likelihood Ratio, and Type III Chi-square tests at significant level for entry ≤ 0.25. This initial analysis was repeated by adding successive model parameters that assess model fit using AIC, BIC, and Deviance; where possible, we evaluated the model’s prediction using the pseudo-adjusted R-squared. Model diagnostics were used to assess potential multicollinearity by variance decomposition and the conditional index. Data points were assessed for high influence and leverage. Linearization and monotonicity were assessed using splines or other regularization methods. Each resulting model from this process was fit using backward elimination (or selection) to detect any possible associations not identified in the former forward stagewise selection process and to seek additional confirmation of any factor associations. Variables were retained in the final model if p<0.05 and confirmed by both forward stagewise and backward selection approaches. Next, the best model was validated via bootstrap sampling that relied on 1,000 replications selected randomly with replacement. If the confidence interval of the beta estimate for a variable contained 0 using the 2.5 and 97.5 percentiles, that variable would be removed from the final model. For the acute-care hospital model, only 1 variable did not meet statistical significance: ED/Obs indicator, every other variable tested was retained in the final model. Finally, the model discrimination was computed with the pseudo-adjusted R-squared.
Similar strategies were used for the critical access hospital model; however, the total number of events from these hospitals was low during model derivation at 53 total events, with most hospitals having zero events. This required a much smaller risk model of degrees of freedom used. The optimal model under these restrictions retained 2 variables: any CO Prev rate in ED/OBS, any CO Prev rate in inpatient. Ultimately, the minimum precision criteria we require hospitals to have (>=1 predicted event) will result in no (zero) hospitals qualifying for measure reporting.
Discrimination of risk models were assessed using the Dispersion-based pseudo R-squared, and calibration was visually investigated by dividing the predicted number of events into deciles and plotting the observed number of events. Additionally, the Root Mean Square Error (RMSE) was calculated between observed and predicted events.
For the acute care hospital model, the dispersion-based pseudo r-square was 40.6%, while it was 4.8% for CAH. The RMSE for ACH was 0.99 and CAH 0.45.The decile calibration plots are attached under 5.4.5a.
The final risk adjustment models demonstrated that differences in facility-level factors were adequately accounted for. Variables were retained based on both statistical significance (p < 0.05) and validation through forward stagewise and backward elimination techniques. For the Acute Care model all but 1 variable that was sent forward for testing was retained in the final model. This is an indication of both, each variables independent association with MRSA events and the number of events we had to model. The Critical Access Hospitals are known to have a much smaller number of events which limit the final risk model. During derivation there were 53 total events in the CAH hospitals leading to 2 variables being retained in the final model with little degree of freedom room for additional covariates. The models were validated using bootstrap sampling, which helped identify and remove any variables with unstable beta estimates, ensuring that the model-maintained generalizability. Overall, the modeling approach demonstrated that the retained risk factors sufficiently captured variation in patient case-mix across facility types. The use of model diagnostics such pseudo-R-squared confirmed good model fit and predictive utility. This indicates that outcome comparisons using the risk-adjusted results are fair and not confounded by underlying differences in population or facility. The retained variables meaningfully explain differences in outcome risk, and the exclusion of non-significant variables and variables that were limited in the model helps to avoid unnecessary model complexity.
See 7.1 Supplemental Information Attachment Pages 7-8 for risk adjustment models.
Use & Usability
Use
NHSN HAI tracking system provides facilities, states, regions, and the nation with data needed to identify problem areas, measure progress of prevention efforts, and ultimately eliminate healthcare-associated infections.
US
Facility hospital
This tool provides a single source search and compare experience, that helps the public choose a Medicare provider.
Over 4,000 Medicare-certified acute-care hospitals, long-term acute care hospitals and over 1,100 acute rehabilitation hospitals across the nation.
Facility, Inpatient/Hospital
Encourages hospitals to improve patients’ safety and reduce the number of conditions people experience from their time in a hospital. The Program encourages hospitals to improve patients’ safety and implement best practices to reduce their rates of infections associated with health care.
General acute-care hospitals across the nation.
Facility, Inpatient/Hospital
CMS collects quality data from hospitals paid under the Inpatient Prospective Payment System, with the goal of driving quality improvement through measurement and transparency by publicly displaying data to help consumers make more informed decisions about their health care.
Over 4,000 Medicare-certified acute-care hospitals across the nation.
Facility, Inpatient/Hospital
Equips consumers with quality-of-care information to make more informed decisions about healthcare options. It is also intended to encourage hospitals and clinicians to improve the quality of inpatient care that is provided to Medicare beneficiaries.
Eleven cancer hospitals across the nation.
Facility, Inpatient/Hospital
The program adjusts payments to hospitals under the Inpatient Prospective Payment System (IPPS), based on the quality of care they deliver.
Over 3,000 hospitals across the country.
Facility, Inpatient/Hospital
Usability
To improve performance on this measure, facilities should review best practices and available guideline recommendations to determine which prevention strategies they can implement. The capability of a facility to implement MRSA LabID Event prevention strategies can vary. Success in reducing rates depends on factors such as available resources, leadership support, and staff engagement.
Prevention strategies can include hand washing, performing routine surveillance, enhanced environmental cleaning, patient and healthcare personnel education, assessing for signs or symptoms of infection, and adherence to clinical guidelines. Conducting root cause analysis of increased prevalence of MRSA or outbreaks helps identify infection control weak points and guide targeted interventions.
The MRSA LabID Event Standardized Infection Ratio (SIR) is an important indicator of MRSA LabID Event prevention effort.
Facilities provide feedback that they are generating Standardized Infection Ratio (SIR) analysis reports, within CDC NHSN monthly, and that they use their SIR to determine if process improvement initiatives should be implemented to reduce MRSA events.
State health departments have advised that they report facilities SIRs publicly, which allows patients and families within the state to select high-quality facilities. State health departments also utilize the MRSA LabID Event Standardized Infection Ratio to target specific facilities with higher SIRs for additional support in initiating prevention activities.
Reporting facilities and state health departments provide feedback on measure performance and implementation through the CDC NHSN Helpdesk. Additionally, during live training such as ‘Ask the Experts’ webinars and educational sessions, an online survey is provided to attendees to share feedback on the measure.
There was no feedback given by users and there was no change to the measure or implementations strategy.
CDC NHSN teams conduct an annual review of each measure protocol. For any measure revision recommendation received, NHSN follows a standard operational procedure designed to ensure thorough evaluation and implementation if appropriate. The process begins with a preliminary discussion and decision making by the NHSN Subject Matter Expert team. User inquiries are then assessed to understand the extent of the concern or improvement. The NHSN team then conducts a literature review to determine whether the recommendations align with current guidelines. If supporting evidence is identified, the NHSN team performs a collaborative review of the findings, followed by input from branch leadership and clinicians. External experts are consulted on an ad hoc basis.
Since 2015, NHSN has released an annual 'Summary of Updates' that outlines changes to the Patient Safety Component protocol based on the review process. These modifications aim to enhance clarity and address feedback received from measured entities. It is important to note that the actual measures themselves are not changed every year.
There was no feedback given by users and there was no change to the measure or implementations strategy.
See 7.1 Supplemental Information Attachment Pages 9-10
Patient medical records and other sources of patient data must be reviewed to determine if the patient meets the necessary criteria for a MRSA LabID Event. It is possible that reviewers may fail to identify that a patient meets criteria thereby under-reporting MRSA LabID Events. Data collectors might also intentionally under-report MRSA LabID Events. Both actions would result in an SIR that is calculated to be lower than actual. Alternatively, patients may be identified as having a MRSA LabID Event, when in fact they do not meet MRSA LabID Event criteria and thereby calculate an SIR that is higher than actual. The NHSN reporting tool includes business logic to minimize misclassification of MRSA LabID Events and the CDC NHSN system generates SIRs automatically, reducing the possibility of manual error in SIR calculation.
Comments
Staff Preliminary Assessment
CBE #1716 Staff Preliminary Assessment
Importance
Strengths
- A clear logic model is provided, depicting the relationships between inputs (e.g., hospital staff, clinical practice guidelines, and health care personnel education), activities (e.g., assessing signs and symptoms for methicillin-resistant Staphylococcus aureus [MRSA] Bacteremia LabID Events, implementing infection control practices to reduce MRSA LabID Events, training hospital staff in infection control practices, and documenting patient symptoms and care provided), and desired outcomes (e.g., reducing standardized infection ratio [SIR], providing optimal patient care, and preventing MRSA Bacteremia LabID Events). This model demonstrates how the measure's implementation will lead to the anticipated outcomes.
The problem this measure addresses presents a significant public health concern with evidence showing a significant increase in hospital-onset MRSA of 14% between 2020 and 2021.
The measure is supported by a comprehensive literature review, including clinical practice guidelines with evidence grading of strong/high and four high quality empirical studies, demonstrating a clear net benefit in terms of reduced MRSA Bacteremia LabID Events.
Data from 1,907 acute care hospitals show a performance gap, with decile ranges from 0 to 2.176, indicating variation in measure performance and less than optimal performance across the target population.
Description of patient input supports the conclusion that the measured outcome is meaningful with at least moderate certainty. Patient input was obtained from the Patient Safety Action Network, a coalition of individuals and organizations consisting of patients who have been medically harmed, their loved ones, and concerned patient safety.
Limitations
- Some of the cited literature are more than 5 years old. For example, the developer cites a 2017 study on the use of intranasal mupirocin and chlorhexidine gluconate (CHG) baths in cardiac surgery patients who were colonized with MRSA.
Additionally, the developer did not provide the years of data used to report the performance gap.
Rationale
- This maintenance measure meets all criteria for 'Met' for importance due to the significance of the problem it addresses, its robust evidence base, a documented performance gap, and a well-articulated logic model, making it essential for addressing MRSA LabID Events.
There is at least moderate confidence that the business case is adequate, i.e., the anticipated impacts of the measure on patient outcomes justify use of the measure.
Closing Care Gaps
The developer did not address this optional domain.
Feasibility Assessment
Strengths
- All required data elements are routinely generated during care delivery, and required elements are available from digital or electronic sources. The developer indicated there have been no changes to the measure specifications. The developer stated that no feasibility issues were found requiring adjustment of the final measure’s specifications.
The developer described the costs and burden associated with data collection and data entry, validation, and analysis. They discussed the potential for data quality issues that could be encountered in implementing or reporting the measure, which include inaccuracies and incomplete data. They also noted mitigation approaches such as routine data reviews by the National Healthcare Safety Network (NHSN) to identify and resolve data quality flags with facilities and providing facilities with internal validation toolkits to audit their internal data to overcome the barriers identified.
The developer described how all required data elements can be collected without risk to patient confidentiality, including retrieving data by the name of the hospital or other non-personal identifiers.
There are no fees, licensing, or other requirements to use any aspect of the measure (e.g., value/code set, risk model, programming code, algorithm).
Limitations
- None identified.
Rationale
- This maintenance measure meets all criteria for 'Met' for feasibility due to its well-documented feasibility assessment, clear and implementable data collection strategy and transparent handling of patient confidentiality, burden, licensing, and fees. These factors collectively ensure that the measure can be implemented effectively and sustainably in a real-world health care setting.
Scientific Acceptability
Strengths
- The developer performed the required reliability testing for this maintenance measure, namely, they conducted accountable entity-level (“measure score”) reliability testing at the level(s) for which the measure is specified. Data sources used for reliability analysis are adequately described and include CDC NHSN data from over 150.5M patients in 1,907 acute care hospitals during the period of 2024.
Limitations
- The developer conducted signal-to-noise reliability testing at the accountable entity-level. Less than 70% of accountable entities were above the expected threshold of 0.6. The developer did not provide an interpretation of nor a rationale for these results. There was not sufficient data in the critical access hospital cohort for reliability analysis.
Rationale
- This maintenance measure is rated as ‘Not Met’ for reliability because the reliability testing results significantly fall below the established thresholds indicating major issues with the consistency and accuracy of the results across different settings and populations. The developer may consider a modification such as increasing the minimum number of predicted events for eligible facilities. By addressing this issue, there is potential to enhance the reliability.
Strengths
- The developer performed the required validity testing for this maintenance measure, namely, they conducted accountable entity-level (“measure score”) validity testing at the level for which the measure is specified. The data source used for validity analysis is the Center for Disease Control’s (CDC) National Healthcare Safety Network (NHSN) during calendar year 2024. The 1,906 ACHs included in the analysis had median bed size of 254, were distributed across all four regions, and were approximately two-thirds major vs. non-major facilities.
The developer conducted empirical validity testing using Spearman’s rank-order correlations between the measure and both the Central Line Associated Blood Stream Infections (CLABSI) SIR and the Clostridioides difficile (CDI) SIR at the accountable entity level. The developer hypothesized positive, weak correlations between the measure and both CLABSI and CDI SIRs, with the rationale that infection prevention practices for different infections overlap, and that entities may prioritize certain practices. The results showed significant, weak correlation coefficients with CLABSI SIR (rho= 0.22) and CDI SIR (rho= 0.05), which are in line with what was hypothesized.
A well-developed logic model and detailed information on relevant, graded clinical recommendations suggest adequate “ruling in” of mechanisms that can explain the measure focus.
The developer conducted statistical risk adjustment, based on a conceptual model, selecting risk factors that have a significant correlation with the outcome. For the acute care hospitals model, the developer reported a pseudo r-square value of 40.6%, indicating acceptable model fit.
Limitations
- While the developer explained how use of different infection prevention practices could yield weak correlations between measures, results from the entity level validity testing could be strengthened with a brief rationale for why the correlation between MRSA SIR and CLABSI SIR was substantially stronger than between MRSA SIR and CDI SIR by referencing the extent of expected overlap in use of such practices for these measures (i.e., a mechanistic explanation).
The developer also cited two studies, one of which used data within the required 5-year time frame (Prascius et al. 2025). This study evaluated implementation of a nasal antiseptic in intensive case units in a 191-bed hospital, comparing MRSA SIR over time in ICUs. This study reported a reduction in MRSA SIR from 1.18 to 0.90 between 2021 and 2023; over the same time period, 4 cases in 2021, 1 case in 2022, and 3 cases in 2023 were reported. The committee should consider the conclusions that can be drawn from the study, given the lack of a comparison group and the small sample of reported MRSA cases from a single hospital.
The measure's low reliability does not support an inference of validity, because it may indicate that observed relationships are not real.
The risk adjustment model includes only facility-level risk factors to minimize the burden of data collection. A robust risk adjustment model for critical access hospitals was not achievable due to sparse event data. Low observed and predicted event incidence results in zero critical access hospitals meeting the minimum precision criterion or qualifying for measure reporting.
Rationale
- This maintenance measure is rated as ‘Not Met But Addressable’ for validity because the validity testing results partially support an inference of validity for the measure, suggesting that the measure somewhat accurately reflects performance on quality and can distinguish good from poor performance to a limited extent.
The risk adjustment methods used demonstrate variation in the prevalence of risk factors across measured entities and that risk factors contribute to unique variation in the outcome. The acute care hospitals model shows acceptable fit and calibration. The critical access hospitals model results in no hospitals qualifying for measure reporting.
Use and Usability
Strengths
- The measure is currently used in the CDC National Healthcare Safety Network (NHSN), CMS Care Compare, CMS Hospital-Acquired Condition Reduction Program (HACRP), CMS Hospital Inpatient Quality Reporting Program (HIQR), CMS Prospective Payment System (PPS)-Exempt Cancer Hospital Quality Reporting (PCHQR) Program, and CMS Hospital Value-Based Purchasing Program.
The developer provided a summary of how accountable entities can use the measure results to improve performance. Specifically, facilities can use best practices and existing practice guidelines to identify and implement MRSA LabID Event prevention strategies including hand washing, conducting routine surveillance, patient and health care personnel education, and enhanced environmental cleaning. Facilities should conduct root cause analysis of increased prevalence of MRSA or outbreaks to help identify infection control weak points and guide targeted interventions.
Facilities provide feedback that they are generating monthly SIR analysis reports within CDC NHSN and that they are using their SIR to determine if improvement initiatives are needed to reduce MRSA events. Additionally, state health departments use the MRSA LabID Event SIR to target hospitals with higher SIRs for additional support in implementing prevention activities. Facilities and state health departments provide feedback on measure and performance via the CDC NHSN Helpdesk and are also able to complete online surveys sharing feedback during live webinars and educational sessions.
The developer noted that no feedback has been received; thus, there were no changes in the measure specifications.
The developer reported changes in performance from 2015 to 2023, in which the overall mean performance score (lower SIR indicates better performance) decreased from 0.998 to 0.755 for acute care hospitals and from 0.994 to 0.675 for critical access hospitals, which supports the argument that this measure is usable.
The developer reported no specific unexpected findings. However, they noted the potential for misreporting MRSA LabID Events and indicated that the NHSN system includes tools to minimize misclassifying MRSA LabID Events and potential manual error in SIR calculation.
Limitations
- None identified.
Rationale
- This maintenance measure is rated ‘Met’ for use and usability because it is actively used in at least one accountability application, with a systematic feedback approach that allows for continuous updates based on stakeholder feedback. The measure also demonstrates a positive trend in performance results, affirming its ongoing usability. The developer reported no unexpected findings.
Committee Independent Review
Support-MRSA-1716
Importance
This is an important measure as there has been a statistically significant increase in hospital-onset methicillin-resistant Staphylococcus aureus (MRSA). Reducing these HAIs (healthcare acquired infections) can be addressed using interventions related to catheter use, contact precautions, hand hygiene, disinfecting caps for IVs, and chlorhexidine baths.
Closing Care Gaps
This was listed as optional.
Feasibility Assessment
As this is a maintenance measure, the specifications have not changed. It is noted on the patient perspective summary that hospitals "have not reported any problems with collecting the needed data".
Scientific Acceptability
For both reliability and validity, the dataset used for testing is the Center for Disease Control’s (CDC).
For both reliability and validity, the dataset used for testing is the Center for Disease Control’s (CDC). As far as risk adjustment, the "dataset used for the risk adjustment model was derived from the 2022 Center for Disease Control’s (CDC)" as well.
Use and Usability
This measure will be used for public reporting, public disease surveillance, accreditation, and quality improvement (internal as well as benchmarks).
Summary
This is still an important measure for improvement over time due to the statistical significance of occurrences mentioned previously.
Support
Importance
Based on my understanding the information seems reasonable
Closing Care Gaps
Based on my understanding the information seems reasonable
Feasibility Assessment
Based on my understanding the information seems reasonable
Scientific Acceptability
Based on my understanding the information seems reasonable
Based on my understanding the information seems reasonable
Use and Usability
Based on my understanding the information seems reasonable
Summary
I am in support of the measure based off my understanding.
Methodological concerns about the importance of this measure
Importance
Recent data showed statistically significant increases in hospital-onset
MRSA between 2020 and 2021. Data from some clinical trials show that prevention efforts can lower the risk of MRSA, and these findings are stronger in MAGNET facilities. Overall, the quality of evidence is mainly rated category IB, suggesting some support from the published evidence.
There still remains variation in performance across acute care hospitals. The mean SIR across 1,907 Acute Care Hospitals was 0.77 (0-5.92), with 355 hospitals (18.7%) having zero MRSA events.
The Patient Safety Action Network had several important concerns about the baseline SIR of 1.0 and data presentation for interpretability. Even with these concerns, they felt measures are important to retain because of their value to patients who expect to be free from preventable harm when hospitalized.
Closing Care Gaps
There are equity concerns of this measure's impact given that there are no patient level factors in the models. Certain groups have higher baseline MRSA colonization and infection risk, independent of hospital quality (e.g., Black patients, those with lower SES, homeless individuals). Failure to account for these differences may attribute population-level risk to hospital performance rather than social risk factors.
Feasibility Assessment
All required data elements are routinely generated, in structured fields, and used during care delivery.
Not a proprietary measure, maintenance measure. Facilities have not notified NHSN of any feasibility issues within the last year. Data quality checks conducted to help confirm the accuracy of the data being reported include checking MRSA data, implementing business rules within the application, verifying alerts, and confirming that flags are triggered by incomplete data. NHSN provides facilities with internal validation toolkits, which can be used to audit their internal data to identify any potential inaccuracies or problems
Scientific Acceptability
The mean reliability score was 0.58. The percentage of facilities with an estimated reliability of >=0.6 was 45% (856/1907), suggesting this measure does not reliably capture facility-level differences in MRSA rates among patients in acute care hospitals. This makes it questionable as to whether the measure can be used to reliably show whether there are true reductions in performance gaps across facilities.
The acute care hospitals with both a MRSA SIR and a CDI SIR (n=1,906) had a weak but significant positive correlation (rho= 0.05165, p= 0.0241). The acute care hospitals with both a MRSA SIR and a CLABSI SIR (n= 1,869) also had a weak but significant positive correlation (rho= 0.22, p< .0001). Magnet hospitals do better on MRSA and have less events. From 2021 to 2023, MRSA SIR declined with an SIR of 1.18 in 2021 to an SIR of 0.90 in 2023, suggesting improvements in MRSA rates. However, given the measure's low reliability it's difficult to determine whether this trend reflects true improvements due to QI efforts, narrowing the performance gap, or is an artifact of imprecise measurement.
Use and Usability
There was no feedback given by users and there was no change to the measure or implementations strategy. The NHSN reporting tool includes business logic to minimize misclassification of MRSA LabID Events and the CDC NHSN system generates SIRs automatically, reducing the possibility of manual error in SIR calculation.
Summary
Low reliability coupled with closing performance gap and lack of patient factors limits the methodologic rigor of this measure, compromising utility
MRSA Bacteremia LabID Event Standardized Infection Ratio (SIR)
Importance
Importance of the measure is adequately described. The steward raises issues on how the terminologies (e.g., predicted, SIR) may be confusing and whether risk-adjustment is the best way to present the measures. They also wonder whether hospitals with SIR of 1.0 may be satisfied with their numbers. All good point to address with the challenges faced with rising MRSA measures in the period 2020-2021.
Closing Care Gaps
Not required for Fall 2025.
Feasibility Assessment
Data obtained through the NHSN system.
Concern: I didn't see any discussion on audits though there was a review of SIRs suggested. This continues to challenge the measure without some audits performed on the quality of data as well as prevention methods implemented.
Scientific Acceptability
Full description provided.
Correlation between MRSA and CDI (and other NHSN measures) used to validate the measure. Literature is also reviewed to reinforce the validity of the measure.
Concern: 1) NHSN measure could all be biased in one direction and their correlations could be positive.
2) Unless I missed the discussion in the document, the empirical study referenced is about the importance of interventions that led to reduction of MRSA. Will such a study be appropriate to use as a validation tool to the measure?
Use and Usability
The steward has adequately addressed use and usability. The challenges with data reporting are also addressed as well as measures taken with NHSN to identify erroneous entries. This is encouraging towards further refining the measure.
Summary
The measure has been in use for several years. The steward has addressed its weak points with data collection as well as reservations in using risk-adjustment. Unless I missed it, not using the stated multiple prevention methods in the risk-adjustment model further challenges the ruse of risk-adjustment when reporting MRSA measures.
Overall evaluation
Importance
Very much so an important topic.
Closing Care Gaps
Did not address.
Feasibility Assessment
Seems feasible as described.
Scientific Acceptability
Agree with concerns noted by staff.
Agree with concerns noted by staff.
Use and Usability
No concerns regarding usability.
Summary
This is an important topic which would benefit from a measure like this. Agree that the acceptability domains could use more details/improvement.
potentially support endorsement
Importance
no comments
Closing Care Gaps
N/A
Feasibility Assessment
no comments
Scientific Acceptability
no comments
Would like to hear developer's rationale for risk model variables to ensure adequate case mix adjustment, especially since variables lean more towards being facility-level variables, vs. patient-level variables
Use and Usability
no comments
Summary
need to determine whether new threshold for reliability can be achieved
Evaluation
Importance
This measure continues to evaluate MRSA prevention activities which address patient safety, outcomes, cost, and LOS.
Closing Care Gaps
Not reported
Feasibility Assessment
All required data elements are routinely generated, in structured fields, and used during care delivery. Facilities can choose to submit this data manually via a web form or via submission of CDA electronic files. NHSN has built-in business rules for mandatory reporting.
Scientific Acceptability
Less than 70% of organizations were above the threshold of 0.6 and no further interpretation was provided.
Testing was performed and detailed documentation was provided for review. Risk adjustment also performed, however, it did not fully account for critical access orgs due to limited data. Limits inference of results and performance levels.
Use and Usability
The measure is used in a variety of programs (NHSN, CMS Care Compare, HACRP, HIQR, PPS, PCHQR, and CMS Hospital VBP programs. Results are intended to drive improvement within organizations and improve patient safety and outcomes.
Summary
Continues to be an important measure
MRSA
Importance
addressed
Closing Care Gaps
Did not address
Feasibility Assessment
Addressed
Scientific Acceptability
addressed
addressed
Use and Usability
One clinical concern is that MRSA Bacteremia SIR is lab automated measures. They do not take into account any clinical judgement. For MRSA bacteremia, the patient might have MRSA at another site POA and again that is not taken into account. It would be ideal to account for outside hospital transfers and if infection was present on admission, which is difficult with lab automated measures.
Summary
One clinical concern is that MRSA Bacteremia SIR is lab automated measures. They do not take into account any clinical judgement. For MRSA bacteremia, the patient might have MRSA at another site POA and again that is not taken into account. It would be ideal to account for outside hospital transfers and if infection was present on admission, which is difficult with lab automated measures.
Overview
Importance
I agree with the staff preliminary assessment
Closing Care Gaps
Developer did not address
Feasibility Assessment
I agree with the staff preliminary assessment
Scientific Acceptability
I agree with the staff preliminary assessment
I agree with the staff preliminary assessment
Use and Usability
I agree with the staff preliminary assessment
Summary
This is an important area of concern but the reliability data are a concern.
Acceptability
Importance
met.
Closing Care Gaps
optional.
Feasibility Assessment
agree with staff assessment comments for feasibility
Scientific Acceptability
Agree with this rationale, "The developer may consider a modification such as increasing the minimum number of predicted events for eligible facilities. By addressing this issue, there is potential to enhance the reliability."
This is best described in the staff assessment details.
Use and Usability
This is met.
Summary
Sci Acceptability needs addressing/reconciliation
Questions on real value of the measure
Importance
I would like to see an explanation of:
- why there is a prediction in the denominator rather than historical data. Predictions can be faulty, and do not necessarily reflect the work the facility is doing.
- How does this measure account for MRSA infections acquired before the hospitalization that may be latent until an immunocompromised moment.
Closing Care Gaps
This is optional and not submitted by developer.
Feasibility Assessment
The developer mentions "facilities have not notified NHSN of any feasibility issues within the last year." But no attempt was done to seek input? Is it because of the Paper Reduction Act of 1995? If so, and we have to accept the OMB information on burden, how recent are these?
Scientific Acceptability
I agree with Battelle's evaluation, but think it is addressable.
Agree with Battelle's evaluation. As well as I am having a hard time seeing how validation of methods to decrease bacteremia is validation of a measure that tracks the incidence of bacteremia compared to predicted cases.
Use and Usability
No concerns.
Summary
Although I agree MRSA infections need to be prevented, I don't see how this measure fairly accounts for HAI and how it can promote improvement when it is based on estimated number of infections rather than actual trending of improving cases.
Public Comments
Methicillin-resistant Staphylococcus Aureus (MRSA) Bacteremia
The developer says, "Reliability testing was performed on data from 2024, for the Acute Care Hospitals. The mean reliability score was 0.58. There was not sufficient data in the CAH cohort for reliability analysis. The median signal-to-noise reliability score demonstrates moderate reliability."
These findings indicate that the measure does not meet basic reliability criteria for Critical Access Hospitals, and should not be used in this context. More than 50% of facilities fall below the reliability threshold of 0.6, so PQM criteria for reliability are "not met." In addition, it isn't clear that the assumptions inherent in the Poisson model used for assessing reliability are supported. I suspect overdispersion relative to Poisson assumptions (mean=variance). The performance period must be lengthened, or the minimum denominator volume must be increased.
See:
https://p4qm.org/sites/default/files/2025-11/Del-3-6-Endorsement-and-Ma…;
Response to MRSA Comment
Thank you for your comment.
Please note that no Critical Access Hospitals qualified for the metric reporting as all had <1 predicted event. The reliability testing you are referencing was performed only on data from acute care hospitals, as a total of 1,907 Acute Care Hospitals qualified for the measure by having at least 1 predicted event.
There are two separate and distinct tasks being conflated in the second comment and in the similar comment that was raised during the E&M Management of Acute Events and Chronic Conditions Advisory Group meeting held December 8, 2025, to review CBE# 1716 and CBE #1717.
First, the task of constructing the model to predict rates of MRSA or CDI incidence used the Negative Binomial model to account for overdispersion. There is a negligible difference between the model parameter point estimates from a Poisson versus Negative Binomial. The rationale for using Negative Binomial model is in the estimation of the standard errors and statistical evaluation of model parameter point estimates. So, when focusing on the second task to estimate Reliability using a signal-to-noise approach, there is need to estimate both between and within variance among the SIR values themselves. To estimate these variances among the point estimates of the SIRs, we chose to apply the Poisson model where the numerator of the SIR is a Poisson random variable with a mean equal to the SIR denominator. This technique does not need to adjust for overdispersion at this point, and thus, is not required to use of the original Negative Binomial model the was necessary for predicting the source incidence rate.