Three-Step Falls Prediction Model

Abstract

The Three-Step Falls Prediction Model is a compact, robust clinical tool designed to estimate the probability of a patient experiencing a fall incident within a six-month timeframe. This instrument achieves its predictive power by integrating the scores derived from three widely recognized and validated component measures: the Vragenlijst valgeschiedenis (Fall History Questionnaire), the New Freezing of Gait questionnaire (NFOG-Q), and the 10-Meter Walk Test (10MWT). The model itself consists of three primary scoring components, yielding a total score that ranges from 0 to 11. Based on this composite score, the patient’s risk of falling is clinically classified into three distinct categories: small (low risk), moderate (medium risk), or large (high risk).

Keywords

Falls Prediction Model, Parkinson’s disease, NFOG-Q, 10MWT, Mobility, Geriatrics, Fall risk assessment, Lower extremity, Performance measure.

Authors

Paul SS, et al. (2013), KNGF-richtlijn Ziekte van Parkinson (2016) (Dutch Adaptation).

Purpose

The primary purpose of the Three-Step Falls Prediction Model is to provide clinicians, particularly physical therapists and geriatric specialists, with a quick and reliable method for identifying individuals at high risk of future falls. This proactive identification is crucial for implementing targeted preventative interventions.

By synthesizing self-reported history (fall incidents), subjective perception of motor performance (Freezing of Gait), and objective physical capability (gait speed), the model offers a comprehensive, yet efficient, assessment. It is specifically useful in managing chronic neurological conditions, such as Parkinson’s disease, where gait instability and fall risk are significant concerns affecting quality of life and morbidity.

Construct

The scale measures the multi-factorial construct of future fall probability, which is a key indicator of functional decline and reduced independence. This construct is operationalized through three distinct, weighted factors:

• Factor 1: Fall History: Assessing prior incidence of falls, which is historically one of the strongest predictors of future fall risk.
• Factor 2: Freezing of Gait (FOG): Measuring perceived or reported episodes of FOG using the NFOG-Q, a crucial symptom of gait instability often associated with advanced Parkinson’s disease.
• Factor 3: Objective Gait Performance: Quantifying maximum walking speed via the 10-Meter Walk Test (10MWT), providing an objective measure of mobility and functional capacity.

The integration of these three factors allows the model to capture the complex interplay between historical risk, subjective impairment, and objective physical performance that contributes to the overall risk of falling.

Validity

The validity of the Three-Step Falls Prediction Model is primarily rooted in its predictive validity, demonstrated in the original research by Paul SS et al. (2013). The model was specifically designed to maximize sensitivity and specificity in predicting falls over a six-month period, demonstrating its ability to accurately classify individuals into low, moderate, and high-risk categories.

Given that the model relies on the scores of three pre-existing, well-validated instruments—the Fall History Questionnaire, the NFOG-Q, and the 10MWT—it inherits a degree of established construct and criterion validity from these component measures. The synthesis method (the three-step scoring process) provides incremental validity over assessing the components in isolation, proving its efficacy as a composite predictive tool.

Reliability

Detailed psychometric data specific to the reliability (e.g., test-retest reliability, inter-rater reliability) of the scoring algorithm of the Three-Step Model itself are derived from the original validation study. Because two of the three components (Fall History and the 10MWT) are highly objective or standardized performance measures, they contribute significantly to the model’s overall reliability when administered correctly.

The NFOG-Q component, while self-reported, is also known to exhibit acceptable internal consistency and temporal stability. The aggregated score system ensures a reliable categorization of risk, provided that the administration protocols for the 10MWT are strictly followed and the patient history is accurately obtained.

Factor Analysis

The Three-Step Falls Prediction Model is not structured as a conventional psychometric scale requiring exploratory or confirmatory factor analysis to determine underlying latent variables. Instead, it functions as a clinical decision rule or algorithm where three predefined, theoretically distinct factors (history, subjective motor deficit, and objective performance) are weighted and combined linearly based on established risk literature.

The model’s structure is therefore based on clinical relevance and predictive power rather than latent factor structure. The three steps are treated as independent predictors, whose combined additive effect determines the final risk stratification (0–11 range).

Instrument

Test Type: Performance, Questionnaire

Format: Structured combination of self-report questionnaire data and objective performance testing.

Language Available: Original version (English); Dutch adaptation (Nederlandse versie: KNGF-richtlijn Ziekte van Parkinson).

Population Group: Adults, Elderly (Geriatrics)

Age Group: Typically utilized with older adults where fall risk is clinically relevant.

Population Details: Specifically validated and widely used in patient populations diagnosed with neurological conditions that impact mobility, most notably Parkinson’s disease.

Test Methodology: The assessment requires the completion of three steps: 1) Gathering fall history data; 2) Administering the NFOG-Q; and 3) Conducting the 10MWT. Scores from these three instruments are summed according to the model’s algorithm (total range 0–11) to categorize the 6-month fall risk.

Keywords

Prediction, Mobility assessment, Geriatric risk, Freezing of Gait, 10MWT, Physical therapy, Rehabilitation, Psychometrics, Clinical algorithm, Neurological assessment.

Authors

Author ORCID Identifier: Information not provided in source material.

Affiliation Email addresses: Information not provided in source material.

Correspondence Address: Information not provided in source material.

Permissions & Fee and Test Year

The original research outlining the model was published by Paul SS, et al. in 2013. The Dutch version was formalized and adopted by the KNGF-richtlijn Ziekte van Parkinson in 2016. Since the model relies on combinations of standard clinical tests, its use is generally free for clinical and research applications, though the specific component scales may have their own usage guidelines.

The original explanatory form (Toelichtingsformulier) for the Dutch adaptation can be downloaded here: Toelichtingsformulier PDF.

The instrument itself (Meetinstrument) for the Dutch adaptation can be downloaded here: Meetinstrument PDF.

Reference’s

• Paul SS, Canning CG, Sherrington C, Close JCT, Paul M, Paul M, Lord SR. (2013). The Three-Step Falls Prediction Model for Parkinson’s Disease: development and preliminary validation. Movement Disorders, 28(7), 934-941.
• KNGF-richtlijn Ziekte van Parkinson (2016). Nederlandse versie. Koninklijk Nederlands Genootschap voor Fysiotherapie.

Items of the Three-Step Falls Prediction Model

IMPORTANT: The following scale items must be preserved in their original language and must not be changed in any way.

The source material indicates that the instrument is composed of three questions/components derived from the Vragenlijst valgeschiedenis, the New Freezing of Gait questionnaire (NFOG-Q), and the Tien meter looptest (10MWT), resulting in a score range of 0-11. The specific text of these three scored items was not provided in the source content.

The three steps used to derive the score are:

1. Assessment of Fall History (Vragenlijst valgeschiedenis).
2. Assessment via the New Freezing of Gait questionnaire (NFOG-Q).
3. Assessment via the 10-Meter Walk Test (10MWT).