3.4 NHSN Definitions & Standardized Infection Ratios (SIR)

3.4 NHSN Definitions & Standardized Infection Ratios (SIR)

The National Healthcare Safety Network (NHSN), managed by the CDC Division of Healthcare Quality Promotion, is the nation's most widely used healthcare-associated infection (HAI) tracking system. NHSN provides facilities, states, and federal agencies with standardized surveillance criteria to track infections, measure device safety, evaluate antimicrobial stewardship, and fulfill mandatory reporting requirements for the Centers for Medicare & Medicaid Services (CMS).

Infection Preventionists must understand NHSN's strict surveillance rules, standardized definitions, risk adjustment algorithms, and how to calculate and interpret the Standardized Infection Ratio (SIR).

---

Surveillance Definitions vs. Clinical Diagnostic Criteria

A fundamental concept in infection prevention is the distinction between surveillance definitions and clinical diagnostic criteria:

• Surveillance Definitions: Strict, objective, binary criteria designed to ensure consistent, unbiased population-level benchmarking across thousands of hospitals. They do not incorporate subjective clinical nuance and are not intended to guide bedside patient therapy.
• Clinical Diagnostic Criteria: Flexible clinical judgment used by physicians to treat individual patients. A clinician may diagnose and treat sepsis or urinary tract infection based on non-specific symptoms, even if the patient does not meet strict NHSN surveillance definitions.

---

Key NHSN Time Windows & Attribution Rules

To eliminate variability, NHSN applies standard time window rules to determine whether an infection is healthcare-associated and attributed to a specific location.

1. Infection Window Period (IWP)

The Infection Window Period (IWP) is a 7-calendar-day timeframe surrounding the first positive diagnostic test or specimen collection date. It includes:

• 3 calendar days before the diagnostic specimen collection date
• The calendar day of specimen collection
• 3 calendar days after the diagnostic specimen collection date

All clinical, diagnostic, and laboratory criteria required to satisfy an NHSN site-specific HAI definition must be present within this 7-day window.

2. Date of Event (DOE)

The Date of Event (DOE) is the calendar date on which the first element used to meet the NHSN HAI site criteria occurred within the Infection Window Period.

3. Healthcare-Associated Infection (HAI) Attribution

An infection is attributed to the reporting facility as an HAI if the Date of Event occurs on or after Calendar Day 3 of admission (where Calendar Day 1 is the calendar day of inpatient admission).

• If the Date of Event occurs on Calendar Day 1 or Day 2 of admission, the infection is classified as Present on Admission (POA) and is not counted as a hospital-acquired event.

4. Repeat Infection Location (RIL)

The Repeat Infection Location (RIL) is a 14-calendar-day window starting on the Date of Event (Day 1 of RIL). During these 14 days:

• No additional HAI of the same infection type can be reported for that patient.
• Additional positive cultures or clinical findings of the same type are treated as part of the initial infection episode.
• Organisms isolated during the RIL are added to the initial event report if appropriate.

5. Device Association Rule

A central line, indwelling urinary catheter, or ventilator is considered associated with an infection (e.g., CLABSI, CAUTI, VAE) if:

• The device was in place for more than 2 calendar days on the Date of Event (with Day 1 being the day of device placement), AND
• The device was present on the Date of Event or the calendar day before.

---

Standardized Infection Ratio (SIR) Mechanics & Risk Adjustment

Raw infection rates (e.g., 2.0 CLABSIs per 1,000 line-days) do not account for differences in patient acuity, facility type, or surgical complexity across hospitals. To enable fair comparisons, NHSN uses the Standardized Infection Ratio (SIR).

SIR Formula

$\text{Standardized Infection Ratio (SIR)} = \frac{\text{Observed Number of HAIs}}{\text{Predicted Number of HAIs}}$

Where:

• Observed HAIs: The actual number of infections identified in the facility using NHSN surveillance criteria during a specified timeframe.
• Predicted HAIs: The number of infections calculated by NHSN multivariable logistic regression models based on national baseline data from reference periods.

Risk Adjustment Variables

NHSN multivariable regression models adjust predicted infection counts by controlling for facility-level and unit-level risk factors, including:

• Facility bed size and medical school affiliation
• Unit type (e.g., Medical ICU, Surgical ICU, Cardiothoracic ICU, Med-Surg Ward)
• Patient acuity and device utilization rates
• Surgical risk factors (ASA score, wound class, procedure duration, patient age, BMI)

Interpreting the SIR & 95% Confidence Intervals

• $\text{SIR} = 1.0$: Observed infection count equals the national predicted benchmark.
• $\text{SIR} < 1.0$: Fewer infections observed than predicted (better performance than national benchmark).
• $\text{SIR} > 1.0$: More infections observed than predicted (worse performance than national benchmark).

To determine whether an SIR is statistically significantly different from 1.0, IPs must evaluate the 95% Confidence Interval (CI):

• CI includes 1.0 (e.g., SIR 0.70; 95% CI: 0.45 – 1.15): The observed rate is not statistically significantly different from national baseline.
• Upper CI bound < 1.0 (e.g., SIR 0.50; 95% CI: 0.25 – 0.85): Facility performed statistically significantly better than baseline.
• Lower CI bound > 1.0 (e.g., SIR 1.60; 95% CI: 1.12 – 2.25): Facility performed statistically significantly worse than baseline.

---

Core NHSN HAI Module Definitions Summary