Written and maintained by CASRAI Editorial Board
Last updated
A nursing-sensitive indicator is a quality measure whose result is affected, provided, or influenced by nursing personnel — but for which nursing is not the exclusive driver. That’s not a loose description; it’s close to the working definition used in the nursing-quality literature (AHRQ’s own patient-safety handbook states it as “processes and outcomes that are affected, provided, and/or influenced by nursing personnel, but for which nursing is not exclusively responsible”). For hospital patient-safety officers, quality directors, and risk managers, that distinction matters operationally: it determines which measures belong on a nursing-quality dashboard, which belong to an ANCC Magnet® application, and which are just general hospital quality metrics that happen to involve nurses somewhere in the care pathway.
This guide covers the National Quality Forum (NQF)-endorsed nursing-sensitive measure set, how it’s organized, and a practical test for deciding whether a given metric is genuinely nurse-sensitive or not. For the specific benchmarking database most U.S. hospitals actually use to collect and compare this data, see NDNQI: What It Collects, How Unit-Type Benchmarking Works, and What Magnet Actually Requires — this guide covers the underlying measure concept and origin; that one covers the specific commercial database.
What Actually Makes an Indicator “Nurse-Sensitive”
Not every measure a nurse touches is nurse-sensitive. A hospital’s 30-day readmission rate, for example, is shaped by physician discharge decisions, medication reconciliation, social determinants, and post-acute follow-up — nursing plays a role, but it’s one input among many, and the measure doesn’t reliably move when nursing staffing or practice changes on its own. That’s the practical test:
- A plausible, direct causal pathway from nursing decisions or actions to the outcome — not a contributing factor buried among several stronger ones.
- Documented sensitivity to nursing structure and process — research literature actually shows the rate moving with nurse staffing levels, skill mix, or care-delivery model, not just a plausible story.
- Nursing is a primary driver, even if not the only one — the definition above is explicit that nursing doesn’t have to be exclusively responsible, only substantially so.
This is why certain hospital-acquired infections split the way they do. CAUTI and CLABSI are both on the classic nursing-sensitive list because insertion technique, maintenance-bundle adherence, and timely catheter/line removal are substantially nursing-driven activities. Surgical site infection, by contrast, is shaped more heavily by operative technique, prophylactic-antibiotic timing (often a pharmacy/anesthesia workflow), and OR environmental controls — nursing has a role, but not the primary one — which is a large part of why it doesn’t appear on the nursing-sensitive list even though it’s every bit as much a patient-safety priority.
Origin: From the ANA Nursing Report Card to NQF Endorsement
The nursing-sensitive indicator concept traces to the American Nurses Association’s Safety and Quality Initiative in the 1990s, built on the idea that nursing’s contribution to patient outcomes was real but largely invisible in hospital quality reporting, which was (and still mostly is) organized around medical and institutional measures rather than nursing-specific ones. That initiative evolved into the National Database of Nursing Quality Indicators (NDNQI), launched in 1998, with the California Nursing Outcomes Coalition (CalNOC) developing a parallel state-based nursing-outcomes program around the same period.
The National Quality Forum then took the concept through its formal multi-stakeholder consensus-endorsement process, publishing National Voluntary Consensus Standards for Nursing-Sensitive Care: An Initial Performance Measure Set in 2004 — commonly referred to as the “NQF-15.” That endorsement gave the concept standing beyond a single vendor’s product: it meant a specific, named committee of payers, providers, and consumer representatives had agreed these 15 measures met NQF’s evidence and importance criteria for nursing-sensitive care, not just that one database happened to collect them.
A naming note worth flagging for anyone citing NQF today: NQF’s measure-endorsement function has since been folded into what CMS’s eCQI Resource Center now calls the Consensus-Based Entity (CBE) — on current CMS electronic clinical quality measure (eCQM) listings, a “CBE ID” has replaced what used to be labeled an “NQF number.” The original 15-measure nursing-sensitive set is still commonly cited by its original NQF name in the nursing-quality literature, but don’t be surprised to see CBE terminology on newer, digitally-specified measures descended from the same concept.
The NQF-Endorsed Nursing-Sensitive Measure Set (15 Measures)
The original set organizes cleanly into the three classic Donabedian quality-measurement categories — structure, process, and outcome — even though NQF’s own framing at the time grouped them slightly differently. Mapped that way:
| Category | Measure | What it captures |
|---|---|---|
| Structure (nursing capacity/environment) |
Nursing hours per patient day | Total nursing care hours (RN, LPN/LVN, and unlicensed) delivered per patient day, by unit |
| Skill mix | The proportion of direct-care hours provided by RNs versus other nursing personnel | |
| Practice Environment Scale–Nursing Work Index (PES-NWI) | A validated RN survey instrument measuring the work environment’s support for professional nursing practice | |
| Voluntary turnover | The rate at which RNs voluntarily leave a unit or organization | |
| Process (what nursing does) |
Smoking-cessation counseling | Documented counseling for patients admitted with AMI, heart failure, or pneumonia — a holdover from that era’s condition-specific process measures rather than a nursing-specific action, but retained in the original NQF-15 set |
| Outcome (what happens to the patient) |
Failure to rescue | Deaths among patients who develop a recognized complication, as a proxy for how well the care team detects and responds to clinical deterioration |
| Pressure ulcer prevalence | The predecessor terminology to today’s hospital-acquired pressure injury (HAPI) measures | |
| Falls | All patient falls on a unit, regardless of injury | |
| Falls with injury | The clinically weightier subset of falls that result in patient harm — tracked separately from the falls rate because prevention strategies and severity differ | |
| Restraint prevalence | Use of vest and limb restraints | |
| CAUTI | Catheter-associated urinary tract infection | |
| CLABSI | Central-line-associated bloodstream infection | |
| Ventilator-associated pneumonia | Now more often tracked under the broader NHSN ventilator-associated event (VAE) surveillance framework — see VAE: NHSN Surveillance Definition, VAC/IVAC/PVAP Criteria, and Reporting |
Falls, pressure ulcers, and restraint use are worth dwelling on specifically, since they’re the three most commonly cited “textbook” nursing-sensitive outcomes and the reason: all three are directly shaped by unit-level nursing decisions — fall-risk assessment and rounding frequency, turning and skin-assessment schedules, and the clinical judgment call on whether a restraint is truly necessary versus a less restrictive alternative — in a way that’s much harder to argue for measures further from the bedside. See Restraint and Seclusion Under the CMS Conditions of Participation (42 CFR 482.13) for the regulatory mechanics governing restraint use specifically.
How These Indicators Get Risk-Adjusted and Benchmarked
A raw rate is close to meaningless without a comparison baseline: a 3.5% pressure-ulcer prevalence means something different on a step-down unit than on a med-surg floor. The dominant current benchmarking vehicle, NDNQI, risk-adjusts by categorical stratification into like unit types (adult critical care, step-down, medical, surgical, med-surg combined, rehabilitation, psychiatric, pediatric, PICU, NICU, obstetric/postpartum) rather than by patient-level covariate modeling — which is the structural reason NDNQI’s unit-level nursing-sensitive rates aren’t directly comparable to CMS’s hospital-level, patient-level-risk-adjusted public measures (Hospital-Acquired Condition Reduction Program, Value-Based Purchasing) even when both are nominally measuring “falls” or “pressure injuries.” Full detail on how that benchmarking works, and what an NDNQI subscription actually delivers, is in the NDNQI guide linked above.
Where the Original Set Shows Up in Today’s Reporting
Most hospitals no longer collect the original NQF-15 as a standalone chart-abstracted set; the concept lives on in two separate, only loosely-connected places:
- Voluntary benchmarking — NDNQI (now owned and operated by Press Ganey) and similar databases, which quality and nursing-leadership teams use for unit-level trending and Magnet documentation.
- Mandatory federal reporting — several of the original outcome concepts now have digitally-specified successors in CMS’s Hospital Harm eCQM family: Hospital Harm—Pressure Injury (CMS826v3) and Hospital Harm—Falls with Injury (CMS1017v2) are both risk-adjusted, EHR-derived measures reported through the CMS Hospital Inpatient Quality Reporting Program — conceptual descendants of the original NQF pressure-ulcer and falls-with-injury measures, but independently specified and reported, not the same data pipeline.
The practical implication for a quality director building a measurement strategy: “nurse-sensitive indicator” today is a category of measure, not one single reporting requirement with one submission deadline. A given hospital may be tracking the concept simultaneously through an NDNQI subscription (for Magnet and internal trending), CMS eCQMs (for federal IQR reporting), and its own state-mandated nurse staffing/outcomes reporting where applicable — three different data pipelines measuring conceptually related but not identical things.
Why This Distinction Matters for Patient-Safety and Quality Teams
Treating every hospital quality metric as equally “about nursing” dilutes the case for nursing-specific interventions — if falls, pressure injuries, and nurse staffing data get lost inside a general dashboard alongside metrics nursing barely influences, it’s harder to make an evidence-based staffing or workflow argument to leadership. Conversely, treating genuinely multidisciplinary metrics (overall mortality, readmissions, most surgical-site infections) as if they were nursing-owned problems misattributes accountability and can lead a quality team to intervene on the wrong lever entirely. The NQF-endorsed set exists specifically to draw that line with some methodological rigor behind it, rather than leaving the question to instinct. For the broader patient-safety and infection-prevention measurement landscape this fits into, see the Patient Safety & Infection Prevention hub, and for how a serious adverse event on any of these measures gets investigated once it occurs, see Sentinel Event: What It Means, and What Happens Next and Root Cause Analysis for CAPA: Choosing 5 Whys, Fishbone, or Fault Tree.
Frequently Asked Questions
What’s the difference between the NQF nursing-sensitive measure set and NDNQI?
The NQF-endorsed set is the underlying concept and consensus-endorsed measure definitions; NDNQI is one specific commercial database (now owned by Press Ganey) that hospitals subscribe to for collecting, risk-adjusting, and benchmarking that data at the unit level. See the NDNQI guide for what a subscription actually includes.
Are CAUTI and CLABSI nursing-sensitive indicators or infection-prevention indicators?
Both. They sit on the original NQF-15 list because catheter and line care are substantially nursing-driven activities, and they’re also core NHSN healthcare-associated infection surveillance categories that infection preventionists track and report independently. The same event can legitimately belong to both measurement frameworks at once — see CAUTI and CLABSI for the NHSN surveillance-definition side.
Is every hospital required to report the NQF-15 nursing-sensitive measures?
No. The original set is a voluntary consensus-endorsed measure set, not a federal reporting mandate. What is mandatory is the overlapping-but-distinct set of CMS eCQMs (including the Hospital Harm falls-with-injury and pressure-injury measures) reported through the Hospital IQR Program, and any state-specific nurse staffing or outcomes reporting laws that apply.
How is nurse staffing itself measured as a nursing-sensitive indicator?
Primarily through nursing hours per patient day (broken out by RN, LPN/LVN, and unlicensed staff) and skill mix (the RN share of those hours), both reported at the unit level rather than as a hospital-wide average — a hospital-wide figure can mask a critical-care unit that’s dangerously understaffed while a step-down unit is over-staffed.
