Standardized Antimicrobial Administration Ratio (SAAR)

The Standardized Antimicrobial Administration Ratio (SAAR) is CDC’s risk-adjusted metric for comparing a hospital’s antibiotic and antifungal use against a national predicted baseline. It is reported through the National Healthcare Safety Network (NHSN) Antimicrobial Use and Resistance (AUR) Module’s Antimicrobial Use (AU) Option, and it is the primary benchmarking tool antimicrobial stewardship programs use to answer a question infection rates alone cannot: not just whether a facility has a resistance or Clostridioides difficile problem, but whether it is prescribing more of a given drug class than comparable hospitals typically do.

SAAR is easy to confuse with the Standardized Infection Ratio (SIR), and the two are frequently discussed side by side in stewardship and infection-prevention reporting — but they measure different things entirely. SIR is an observed-to-predicted ratio of infection events (CLABSI, CAUTI, SSI, and so on). SAAR is an observed-to-predicted ratio of drug administration. A facility can have a normal SIR and an elevated SAAR, or the reverse; neither ratio substitutes for the other, and a stewardship program that tracks only one is missing half of what NHSN’s own methodology was built to surface.

What the SAAR actually measures

The SAAR is calculated as:

SAAR = observed antimicrobial use ÷ predicted antimicrobial use

A SAAR of exactly 1.0 means a facility’s antimicrobial use matches what NHSN’s risk-adjustment model predicts for a comparable population. Above 1.0 suggests more use than predicted; below 1.0 suggests less. Observed use is the number of antimicrobial days (also called days of therapy — any amount of a specific agent administered to a patient in a calendar day, as documented in the electronic medication administration record or bar-code medication administration system) that a facility reports for a defined agent category in a defined group of patient care locations.

Predicted use comes from CDC’s own predictive models, built from nationally aggregated NHSN data (2017 for adult and pediatric locations, 2018 for neonatal locations) and risk-adjusted for factors including patient-care location type, facility type, teaching status, bed size, ICU bed count, and average length of stay. CDC develops a separate predictive model for each antimicrobial agent category, which is why a facility gets multiple SAARs rather than one composite score.

Where the numerator and denominator come from

The AU Option’s primary metric is antimicrobial days per 1,000 days present — the aggregate count of patients housed in a location or facility on any given day of a calendar month. A secondary metric, antimicrobial days per 100 admissions, is reported for facility-wide inpatient (FacWideIN) data only. Both numerator and denominator must be submitted electronically via an HL7 Clinical Document Architecture file generated from the facility’s pharmacy and/or eMAR system — NHSN does not accept manual data entry for the AUR Module, which is one of the reasons AU Option participation lags behind HAI-event reporting: it requires a working electronic interface, not just staff time.

The 22 SAAR agent categories

NHSN generates SAARs for 22 antimicrobial agent categories — 7 adult, 8 pediatric, and 7 neonatal — combined with 17 eligible NHSN location types (8 adult, 5 pediatric, 4 neonatal), for a maximum of 47 possible SAARs per facility. The categories were defined by CDC with input from external infectious-disease physicians and pharmacists, grouped around stewardship-relevant clinical targets rather than pharmacologic class alone:

Adult categories: all antibacterial agents; broad-spectrum agents predominantly used for hospital-onset infections; broad-spectrum agents predominantly used for community-acquired infections; agents predominantly used for resistant Gram-positive infections (e.g., MRSA); narrow-spectrum beta-lactam agents; agents posing the highest risk for C. difficile infection (not mutually exclusive with the other categories); antifungal agents predominantly used for invasive candidiasis.

Pediatric categories: the same seven as adult, plus a dedicated azithromycin category.

Neonatal categories: all neonatal antibacterial agents; vancomycin predominantly used for late-onset sepsis; broad-spectrum agents predominantly used for hospital-onset infections; third-generation cephalosporins; ampicillin predominantly used for early-onset sepsis; aminoglycosides predominantly used for early- and late-onset sepsis; fluconazole predominantly used for candidiasis.

Because none of the 17 eligible location types covers every inpatient unit in a hospital, NHSN is explicit that no SAAR is a “facility-wide” figure — each SAAR reflects only the specific location types and agent category it was built for.

Interpreting a SAAR value correctly

NHSN’s own guidance is deliberately cautious about what a SAAR does and doesn’t prove. A SAAR statistically significantly above 1.0 may indicate overuse; a SAAR not statistically different from 1.0 indicates use in line with the referent population; a SAAR statistically significantly below 1.0 may indicate underuse. But NHSN states plainly that a SAAR alone is not a definitive measure of the appropriateness or judiciousness of antimicrobial use — a facility can have a SAAR near 1.0 while still prescribing inappropriately, and a SAAR above 1.0 that hasn’t reached statistical significance can still reflect a meaningful excess worth investigating. The categories themselves were built on expert consensus about stewardship priorities, not a claim that every deviation has an immediate clinical consequence such as C. difficile incidence or a specific resistant infection — those outcomes are typically multifactorial and lagged, and a stewardship program should treat the SAAR as a screening tool that flags where to look, not a verdict.

SAARs can be run by month, quarter, half-year, year, or a cumulative period, and can be filtered to a specific location or subset of eligible locations.

From SAAR to a stewardship action: AU-CAD and Targeted Assessment reports

NHSN’s Targeted Assessment for Prevention (TAP) reports use the SAAR alongside a companion metric, the Antimicrobial Use Cumulative Attributable Difference (AU-CAD) — the number of antimicrobial days a facility would need to reduce (or could add) to reach a chosen SAAR target. A facility running a TAP report with a SAAR of 1.30 and an AU-CAD of 75, for example, would need to reduce antimicrobial days by roughly 75 across the reporting period to bring that category’s SAAR down to a target of 0.95. Because CDC does not set a universal SAAR target — the metric isn’t a definitive appropriateness measure, so no single cutoff is clinically justified for every facility or category — TAP reports are built to rank locations and agent categories by AU-CAD instead, letting a stewardship program prioritize the interventions likely to have the largest measurable effect rather than chasing a fixed number.

SAAR vs. SIR: two different observed/predicted ratios

Both metrics share the same underlying logic — observed divided by predicted, risk-adjusted against national NHSN benchmark data — which is exactly why they get conflated. The distinction that matters operationally:

• SIR (see Standardized Infection Ratio) measures observed healthcare-associated infection events — CLABSI, CAUTI, SSI, VAE, and other definitions covered in Healthcare-Associated Infection Definitions — against a predicted event count.
• SAAR measures observed antimicrobial administration against predicted administration. It says nothing directly about infection rates, and a facility’s SIR and SAAR for a related organism or drug class can move independently of each other.

A stewardship program that only watches SIR can miss overuse that hasn’t yet produced a measurable outbreak or resistance signal; a program that only watches SAAR can miss the infection-control side of the same clinical picture. NHSN built the AUR Module and the HAI-event modules as companion, not interchangeable, surveillance systems.

Why SAAR reporting now carries federal weight beyond stewardship

AUR Module reporting isn’t only an internal stewardship exercise. Under the CY 2026 Public Health and Clinical Data Exchange objective (42 CFR 495.24(f)), CMS’s Promoting Interoperability Program for eligible hospitals requires attestation on eight measures, six of which are mandatory — and Antimicrobial Use Surveillance and Antimicrobial Resistance Surveillance are two of the six. A hospital’s AU Option submission to NHSN is what that measure is built on. See Promoting Interoperability Program: Objectives, Security Risk Analysis, and the Payment Penalty for the full measure set, the active-engagement options, and the payment-adjustment consequence of non-compliance. This is a separate reporting obligation from the NHSN HAI measures used in CMS’s Hospital IQR and Hospital-Acquired Condition Reduction programs — see NHSN Reporting Requirements for Hospitals for how those two federal reporting tracks relate.

Where SAAR fits inside an antimicrobial stewardship program

CDC’s Core Elements of Hospital Antibiotic Stewardship Programs list tracking and reporting as one of seven core elements, and SAAR (alongside antimicrobial days per 1,000 days present and facility-specific rate tables) is the standard NHSN-native way to do that tracking with a national comparator built in. See Antimicrobial Stewardship Program: CDC Core Elements and Requirements for the full seven-element framework and how SAAR data feeds a program’s action element.

SAAR data is also a natural input for prospective audit and feedback, the stewardship intervention most directly tied to prescribing review: a location or agent category with a persistently elevated SAAR is exactly where a structured antibiotic time-out or a targeted audit-and-feedback cycle tends to have the most room to move the number. Similarly, because several SAAR categories track agents used against resistant Gram-positive organisms and multidrug-resistant organisms specifically, elevated SAARs in those categories are a signal worth cross-checking against a facility’s MDRO prevention program data, since the two surveillance streams (drug use and resistant-organism transmission) often point at the same units.

Who runs and reports the AU Option in practice

Because the AU Option requires an electronic HL7 CDA feed from the pharmacy or eMAR system rather than manual entry, standing up SAAR reporting is typically a joint project between pharmacy informatics, infection prevention, and IT — not something an infection preventionist can configure alone. All NHSN-enrolled inpatient facility types are eligible to participate (general acute care, critical access, children’s, long-term acute care, military/veterans, oncology, orthopedic, psychiatric, rehabilitation, surgical, and women’s/children’s hospitals, among others), but ambulatory surgery centers, long-term care facilities, and outpatient dialysis facilities cannot currently submit AU Option data at all.

Frequently asked questions

What counts as a “good” SAAR?

There is no CDC-defined target. A SAAR near 1.0 that isn’t statistically different from the national referent is the closest thing to a neutral signal, but CDC deliberately declines to set universal targets because the SAAR alone doesn’t establish whether use is clinically appropriate. TAP reports and AU-CAD values exist precisely because ranking by improvement opportunity, not chasing a fixed cutoff, is the intended use.

Is SAAR the same as the Standardized Infection Ratio (SIR)?

No. Both are observed-over-predicted, risk-adjusted NHSN ratios, but SIR measures infection events and SAAR measures antimicrobial administration. See the comparison above.

Can every hospital generate a SAAR for every unit?

No. SAARs are only produced for the 17 NHSN-eligible adult, pediatric, and neonatal location types that have submitted AU Option data, and none of the resulting SAARs represents facility-wide use — NHSN is explicit that no current SAAR covers all inpatient locations in a facility.

How often can a facility pull a SAAR report?

SAAR reports can be generated by month, quarter, half-year, year, or a custom cumulative period, and can be filtered down to a specific eligible location or subset of locations.

Does SAAR reporting satisfy a federal reporting requirement on its own?

AU Option submission underlies the Antimicrobial Use Surveillance measure inside CMS’s Promoting Interoperability Program (42 CFR 495.24(f)), which is separate from the NHSN HAI-event measures used in Hospital IQR and the HAC Reduction Program. A facility working toward both needs to track them as distinct obligations, not one combined requirement.