📧 info@waherohealthinstitute.com•📞 (832) 490-4024 Mobile clinical education for your facility

Fall Risk Assessment Tools Compared: Morse, Hendrich II and Johns Hopkins

An older woman in blue uses a walker along a bright hospital corridor with a caregiver walking at her side

Photo: Caleb Oquendo via Pexels

A nurse admits two patients in the same hour. Both come out of the fall risk assessment with the same label, flagged in the same color. One is 79, newly confused, and has never fallen. The other is 54, sharp as a tack, with a bad hip and a walker she has used for years. The tool treated them the same. Their risks have almost nothing in common.

Three instruments dominate fall risk assessment in US hospitals: the Morse Fall Scale, the Hendrich II Fall Risk Model and the Johns Hopkins Fall Risk Assessment Tool. This article covers what each measures, where it was validated, what the accuracy figures mean, what the Agency for Healthcare Research and Quality and The Joint Commission expect of assessment and reassessment, and how nurses are assessed on using a tool. It contains no cut-off scores and no intervention lists; your facility's policy sets those.

Three tools, three origins

The Morse Fall Scale is the oldest and most widely reproduced. Janice Morse and colleagues published its development in 1989 in the Canadian Journal on Aging, from a prospective study across three clinical areas. It scores six items, listed in the AHRQ fall prevention toolkit as history of falling, secondary diagnosis, ambulatory aid, intravenous therapy or heparin lock, gait, and mental status.

The Hendrich II Fall Risk Model came from a different method. Ann Hendrich, Patricia Bender and Allen Nyhuis published its validation in Applied Nursing Research in 2003: a concurrent case-control study in an acute care tertiary hospital, 355 patients who fell and 780 who did not, assessed for more than 600 possible risk factors. Eight parameters survived: confusion, disorientation or impulsivity; symptomatic depression; altered elimination; dizziness or vertigo; male sex; antiepileptics; benzodiazepines; and the rising-from-a-chair element of the Get-Up-and-Go test. The Hartford Institute for Geriatric Nursing's Try This sheet, revised in 2023 and written by Hendrich, names its settings as adult acute care, ambulatory, assisted living, long-term care and population health.

The Johns Hopkins Fall Risk Assessment Tool was developed within The Johns Hopkins Hospital's own falls program with the Johns Hopkins University School of Nursing, with development and pilot work published by Stephanie Poe and colleagues in the Journal of Nursing Care Quality in 2005 and 2007. It is organized around categories: age, fall history, elimination, medications, patient care equipment, mobility and cognition, with an upfront screen for situations that define risk on their own.

What each fall risk tool asks about Three columns listing the items of the Morse Fall Scale, the Hendrich II Fall Risk Model and the Johns Hopkins Fall Risk Assessment Tool, without weights or scores. Morse Fall Scale (6) Hendrich II (8) Johns Hopkins (7 categories) History of falling Secondary diagnosis Ambulatory aid IV therapy or heparin lock Gait Mental status Confusion, disorientation, impulsivity Symptomatic depression Altered elimination Dizziness or vertigo Male sex Antiepileptics Benzodiazepines Get-Up-and-Go (rising from a chair) Age Fall history Elimination Medications Patient care equipment Mobility Cognition plus an upfront screen for defining situations Weights, scores and risk levels are left out on purpose. Your facility's policy sets them.
Items as listed in the AHRQ toolkit (Morse) and in published descriptions of the other two tools. Scoring and thresholds deliberately omitted.

What the accuracy figures mean

Every tool makes a trade. Sensitivity is the share of patients who went on to fall that the tool flagged; specificity is the share who did not fall that it correctly left unflagged. The published numbers describe that trade in a particular hospital, not a league table.

The original validations look strong. Morse's 1989 paper reported sensitivity of 78 percent and specificity of 83 percent, similar results when weights derived from one random half of the data were tested on the other, and interrater reliability of 0.96. Hendrich's 2003 study reported sensitivity of 74.9 percent and specificity of 73.9 percent, and a 2020 electronic health record validation by Hendrich, Bufalino and Groves, also in Applied Nursing Research, covering 214,358 adult inpatients and 625 falls at nine acute-care sites, reported 78.72 and 64.07 percent.

Pooled results are more sobering. A 2013 meta-analysis by Aranda-Gallardo and colleagues in BMC Health Services Research combined 14 studies of acutely hospitalized adults and found pooled sensitivity and specificity of 0.800 and 0.675 for STRATIFY, 0.755 and 0.677 for the Morse Fall Scale, and 0.628 and 0.640 for Hendrich II. Its conclusion: the behavior of these instruments varies considerably with the population and the environment, so their operation should be tested before implementation. The AHRQ toolkit draws the practical lesson: one standard tool across adult units, with unit-specific risk factors added on top.

Pooled sensitivity and specificity of three fall risk tools Grouped bars. STRATIFY sensitivity 0.80 and specificity 0.68; Morse Fall Scale 0.76 and 0.68; Hendrich II 0.63 and 0.64. Source: Aranda-Gallardo et al. 2013. 0 0.5 1.0 0.800.68 0.760.68 0.630.64 STRATIFY Morse Fall Scale Hendrich II Sensitivity Specificity
Pooled values from Aranda-Gallardo and colleagues, BMC Health Services Research 2013;13:122, a meta-analysis of 14 studies in acutely hospitalized adults, rounded to two decimals.

Printable planner

Get the printable 2027 calendar for nurses

The Bedside Dozen: a thirteen-page printable 2027 calendar from Wahero Health Institute, an overview page plus twelve monthly pages, each with one high-yield clinical problem to study that month. US Letter, landscape. A planning aid for spreading competency topics across the year; it does not count as CE contact hours.

The Bedside Dozen: 2027 Nursing Calendar, $9.99

Pay by Cash App, PayPal or Zelle at checkout; the download link is emailed once the payment is confirmed, usually the same day. Study material only: it does not replace the Board's requirements, your employer's policies or the judgment of a licensed clinician.

Why a score is not a plan

The AHRQ toolkit, Preventing Falls in Hospitals (Publication No. 13-0015-EF, January 2013), is unusually blunt. It states that the risk assessment tools are meant to complement clinical judgment, not to replace it. Two patients deemed high risk by score might have different risk factors, one with delirium and the other with impaired gait, and the responses need to be different. The most important application of a tool, it concludes, is to identify the risk factors for which care plans can be developed, and it does not recommend excessive focus on the score.

It also cites David Oliver's 2008 commentary in Age and Ageing: research has shown that scores from fall risk prediction tools do not predict falls any better than a clinician's judgment, and the creator of STRATIFY argues against its score being used for prediction.

The question a score cannot answer. A score tells you that a patient is at risk. It cannot tell you why. The why is the plan, and the why is what an assessor will ask you to say out loud.

The Joint Commission's Sentinel Event Alert 55, Preventing falls and fall-related injuries in health care facilities (September 28, 2015), says the same in accreditation language: its fourth recommended action is an individualized plan of care based on identified fall and injury risks, and a true risk assessment goes deeper than a screening.

What The Joint Commission and AHRQ expect of assessment and reassessment

Sentinel Event Alert 55 reported that hundreds of thousands of patients fall in US hospitals every year, with 30 to 50 percent of falls resulting in injury, and that the average cost of a fall with injury is about $14,000. Of 465 falls with injury reported to the Sentinel Event database since 2009, approximately 63 percent resulted in death. First on its list of contributing factors is inadequate assessment.

Its third recommended action is a standardized, validated tool to identify risk factors, naming the Morse Fall Scale and the Hendrich II Fall Risk Model as examples, preferably integrated into the electronic record, alongside an individualized assessment, with staff trained for inter-rater reliability. The alert cites hospital standard PC.01.02.08, under which the hospital assesses and manages the patient's risk for falls; the current accreditation manual carries the wording in force.

On timing, the AHRQ toolkit suggests assessment on admission, on transfer between units, with a significant change in condition, or after a fall, and at a regular interval for longer stays, while stating honestly that the optimal frequency is unclear and may vary by unit. The alert adds continued reassessment after a fall, because a change in condition after a fall, such as an undiagnosed fracture, is often what precipitates a poor outcome. Your policy sets the interval; the standards set the triggers. This is the logic of trend-based monitoring in our article on recognizing patient deterioration early: a single reading is a snapshot, and the useful information is the change between readings.

The assessment loop as AHRQ and The Joint Commission describe it Four boxes in sequence: admission assessment, identify the risk factors, individualized plan matched to each factor, reassess on transfer, change in condition, after a fall or at a set interval, with an arrow looping back to the second box. Admission risk factor assessment Identify the factors, not just a score Individualized plan matched to each factor Reassess on transfer, change in condition, after a fall, set interval Each reassessment restarts at the factors, because the factors are what changed
The assessment loop, with the triggers suggested in the AHRQ toolkit and Sentinel Event Alert 55. Schematic only; the interval and the plan content are set by facility policy.

Licensing, the EHR and why your hospital chose its tool

The choice of tool was not purely clinical. The Hendrich II Fall Risk Model is a registered mark; materials published by AHI of Indiana state that replication, distribution or use requires written permission, and the Hartford Institute sheet describes the model as built into electronic health records, including through the Epic App Market. The Johns Hopkins tool is distributed through what Johns Hopkins calls flexible licensing for electronic records or paper. The Morse Fall Scale is reproduced in full in the AHRQ toolkit as Tool 3H.

So your hospital's choice reflects its record system, contracts and population as much as the meta-analyses. Assessors expect the facility's build of the tool, and the authoritative copy of its weights and thresholds is in your policy, which is why this article reproduces none.

How nurses are assessed on using the tool

The AHRQ toolkit states that the accuracy of a risk factor assessment tool depends on the person using it, that untrained staff show significant variability even when evaluating the same patient, and that training is needed because untrustworthy results get ignored.

The check-off has four parts. First, scoring a standardized scenario or a real patient the same way the assessor scores it: the inter-rater reliability the alert asks organizations to train for. Items that need an observation rather than a chart review, such as gait and mental status in the Morse scale, are where raters diverge most. Second, translating the result into risk factors in words, not a category. Third, knowing the reassessment triggers in your policy and documenting so the next shift can see what changed. Fourth, communicating the specific risks at handoff, which the alert lists among effective practices.

Our guide to meeting annual competency requirements without chaos explains why observed validation beats attendance, and the gait and cognition findings in the head-to-toe assessment framework are the raw material every tool scores. If you plan your own study year, our printable 2027 nursing calendar, the Bedside Dozen, sets one clinical problem per month.

What assessors are listening for. A risk factor named, not a number recited. A plan element that matches that factor. A reassessment trigger you can state without looking it up. A handoff that carries all three to the next nurse.

Educational use. This article is learning material for nurses and nursing students. It is not clinical advice, and it does not replace your employer's policies, your facility's protocols, or the judgement of a licensed clinician. Always follow the standards and procedures in force where you practise.

Key takeaways

Frequently asked questions

Which fall risk assessment tool is the most accurate? There is no clean winner. The 2013 Aranda-Gallardo meta-analysis found pooled sensitivity and specificity of 0.800 and 0.675 for STRATIFY, 0.755 and 0.677 for Morse and 0.628 and 0.640 for Hendrich II, and concluded that performance varies considerably with population and setting. The best tool is the one your facility has tested on its own patients.

Why does my hospital use a different fall risk tool from the one I learned in school? Tool choice reflects the electronic record, licensing and the patient population as much as the research. Hendrich II is a registered mark, the Johns Hopkins tool is licensed, and the Morse Fall Scale is reproduced in the AHRQ toolkit. Assessors expect your facility's build of its tool.

How often should fall risk be reassessed? The AHRQ toolkit suggests assessment on admission, on transfer, with a significant change in condition or after a fall, plus a regular interval for longer stays, and says the optimal frequency is unclear. Sentinel Event Alert 55 adds reassessment after a fall. Your policy sets the interval.

Does a low fall risk score mean a patient will not fall? No. Every published sensitivity figure is below 100 percent, so some patients who fall were never flagged. The Joint Commission states that any patient of any age or physical ability can be at risk, and AHRQ states that the tools complement clinical judgment rather than replace it.

Practice the assessment, not the arithmetic

The tools score what a nurse observes. Wahero Health Institute's in-person Texas courses, including Head to Toe Physical Assessment and Recognizing Patient Deterioration, build the observation and reasoning that make any fall risk tool worth using, and we validate the competency on site against your own policy.

Head to Toe Assessment Course Talk to Us

Educational content only

This material is published by Wahero Health Institute for professional education and is not individual medical advice, a care protocol, or a substitute for clinical judgment. Always follow your facility's policies, your state's nurse practice act, and your own scope of practice, and confirm medication doses against a current authoritative reference before administration. See our Terms of Use.