Concussion produces multisystem functional disturbance that structural imaging routinely misses, and concussion medicine has no equivalent of the standardized multidomain disability scales — EDSS, NIHSS, mRS, FIM — that transformed the rest of neurology. The Short-Term Concussion Disability (ST-CD) Scale, derived from the Comprehensive Functional Neurological Evaluation (C-FNE™), aggregates objective findings from nine functional systems into a single ordinal score from 0 to 10 in 0.5-point increments, each step anchored to a defined proportion of waking time during which function is materially impacted. This manuscript sets out the conceptual foundation, measurement model, development history, scoring architecture, design decisions and rejected alternatives, and intended clinical use. It presents no psychometric validation data: the scale is explicitly an investigational instrument, is not diagnostic, does not establish causation or permanent impairment, and a staged validation program remains to be completed.
Manuscript status and scope. This manuscript is a descriptive scale-development and methodology report. It presents the conceptual foundation, design rationale, development history, scoring architecture, and intended clinical application of the Short-Term Concussion Disability (ST-CD) Scale. It does not present psychometric validation data. Reliability, validity, responsiveness, and predictive performance have not yet been formally established, and the scale should be regarded as an investigational clinical instrument pending the validation program outlined in Section 12.0.
Traumatic brain injury (TBI) is among the leading causes of acquired neurological disability worldwide, affecting millions of individuals annually through motor vehicle collisions, falls, sports-related injuries, occupational trauma, military service, and assaults.1–5 Although many traumatic brain injuries are classified as mild traumatic brain injuries (mTBI), commonly referred to as concussions, the designation mild describes the initial injury severity rather than the potential magnitude or duration of neurological dysfunction experienced by the patient. Many individuals continue to experience persistent deficits affecting cognition, balance, vision, vestibular function, executive performance, emotional regulation, sleep, and activities of daily living long after the initial injury.2–7
The societal burden associated with concussion extends well beyond the acute injury. Persistent neurological dysfunction contributes to reduced workplace productivity, prolonged disability, increased healthcare utilization, diminished quality of life, and substantial economic costs to patients, employers, healthcare systems, and society.3,4 While many individuals recover within several weeks, a significant proportion experience persistent symptoms requiring multidisciplinary evaluation and rehabilitation.2,6 Accurate identification and quantification of neurological dysfunction during this period are essential for directing treatment, monitoring recovery, facilitating return-to-work and return-to-play decisions, and supporting medicolegal determinations when appropriate.2,8
Despite substantial advances in concussion research over the past two decades, objectively quantifying the overall functional consequences of concussion remains a significant clinical challenge. Conventional structural imaging modalities, including computed tomography (CT) and magnetic resonance imaging (MRI), are invaluable for identifying hemorrhage, skull fractures, cerebral edema, and other structural pathology requiring urgent intervention.9 However, these modalities frequently appear normal in patients with concussion despite clinically meaningful neurological dysfunction.10–13 Advanced neuroimaging techniques, including diffusion tensor imaging (DTI), have improved scientific understanding of group-level white matter abnormalities following concussion but remain limited in routine clinical practice because of variability in acquisition protocols, lack of standardized normative databases, and uncertain interpretation at the individual patient level.12–15 Consequently, functional assessment rather than structural imaging continues to form the cornerstone of concussion evaluation and management.8,9
Current concussion assessment incorporates numerous clinical instruments, including symptom inventories, neurological examinations, computerized cognitive testing, vestibular and oculomotor assessments, balance testing, psychological screening instruments, and functional performance measures.8,16–21 Each contributes valuable information regarding specific aspects of neurological function; however, these assessments are frequently interpreted independently and often fail to provide a standardized measure of the patient’s overall functional neurological disability. As a result, clinicians may successfully identify individual impairments while lacking an objective method to quantify the cumulative impact of multisystem neurological dysfunction on daily function, treatment planning, and recovery monitoring.
This challenge is not unique to concussion medicine. Throughout neurology, standardized disability scales have fundamentally transformed patient care by providing objective methods for quantifying disease severity, monitoring progression, facilitating communication among healthcare professionals, standardizing research outcomes, and supporting clinical decision-making. The Expanded Disability Status Scale (EDSS) for multiple sclerosis, the National Institutes of Health Stroke Scale (NIHSS), the Modified Rankin Scale (mRS), and the Functional Independence Measure (FIM) have become integral components of neurological practice because they provide a common language for describing functional disability beyond diagnosis alone.22–26
Concussion medicine currently lacks a similarly comprehensive multidomain disability framework capable of objectively integrating neurological findings across the vestibular, oculomotor, cognitive, balance, motor, autonomic, cranial nerve, and psychological systems commonly affected following injury. Although symptom severity scales remain valuable for documenting the patient’s subjective experience, symptom burden alone does not consistently correlate with objectively measurable neurological dysfunction or functional capacity.17,19,20 Patients with similar symptom scores may demonstrate markedly different objective deficits, recovery trajectories, rehabilitation needs, and long-term outcomes. Conversely, clinically significant neurological dysfunction may persist despite partial symptom resolution, emphasizing the need for a standardized framework that extends beyond subjective symptom reporting alone.11,12,20
To address this unmet clinical need, the Comprehensive Functional Neurological Evaluation (C-FNE™) was developed as a standardized multidomain assessment integrating objective neurological measurements across the functional systems most commonly disrupted following concussion. Through standardized clinical examination techniques and instrumented assessments, the C-FNE provides clinicians with reproducible measures of neurological function that extend beyond structural imaging and symptom inventories alone.
Building upon this comprehensive examination framework, the Short-Term Concussion Disability (ST-CD) Scale was subsequently developed to provide a standardized measure of current functional neurological disability following concussion. Rather than functioning as a diagnostic instrument, the ST-CD Scale synthesizes objective findings generated during the C-FNE into a clinically meaningful representation of the patient’s overall neurological disability during the acute and subacute phases of recovery. The scale is intended to facilitate communication among healthcare providers, support individualized treatment planning, monitor recovery over time, assist return-to-function decision-making, and establish a common language for describing concussion-related disability across clinical practice, research, and medicolegal settings.
The ST-CD Scale has undergone continuous iterative refinement through approximately seven years of routine clinical application within a multidisciplinary concussion program and has been incorporated into more than 5,000 patient evaluations. This extensive clinical experience has informed the evolution of the examination framework, domain selection, scoring methodology, and interpretation guidelines presented in this manuscript. Although formal psychometric studies evaluating reliability, construct validity, criterion validity, responsiveness, and predictive performance remain an important future objective, the purpose of this manuscript is not to present definitive validation data. Rather, its purpose is to describe the conceptual foundation, scientific rationale, development process, methodology, and intended clinical application of the ST-CD Scale as an objective framework for measuring short-term functional neurological disability following concussion.
Terminology. Throughout this manuscript, the term concussion-type traumatic brain injury (cTBI) is used to describe the subset of traumatic brain injuries traditionally classified as concussion or uncomplicated mild traumatic brain injury (mTBI). The designation is intended solely as a concise descriptive term for this clinical phenotype and is not proposed as a replacement for existing diagnostic classifications. It is used throughout this manuscript for clarity and consistency. Section 1.1 retains the conventional terms where the equivalence between them is being established; the abbreviated form is used thereafter.
Accurate assessment of concussion remains one of the most complex challenges in contemporary neurological practice. Although substantial advances have been made in the diagnosis and management of mild traumatic brain injury (mTBI), clinicians continue to face significant difficulties in objectively characterizing the functional consequences of injury, monitoring recovery, and communicating the severity of neurological dysfunction using standardized methods.2,8,9
Unlike many neurological disorders characterized by discrete structural lesions or well-defined pathological processes, concussion primarily represents a disturbance of neurological function. Biomechanical forces transmitted to the brain during injury initiate a complex cascade of metabolic, electrophysiological, vascular, inflammatory, and neurochemical changes that may disrupt neuronal communication without producing abnormalities detectable by conventional structural imaging.9–13 As a result, many patients demonstrate significant neurological dysfunction despite normal radiographic findings, creating an inherent disconnect between structural assessment and clinical presentation.
Computed tomography (CT) remains the imaging modality of choice for the acute evaluation of head trauma because of its rapid acquisition, widespread availability, and high sensitivity for detecting life-threatening conditions such as intracranial hemorrhage, skull fractures, mass effect, and cerebral edema.9 However, CT was never intended to evaluate the subtle functional disturbances that characterize uncomplicated cTBI. Consequently, the majority of patients diagnosed with cTBI demonstrate normal CT findings despite clinically significant neurological symptoms and objective functional deficits.8,9
Magnetic resonance imaging (MRI) offers substantially greater soft tissue resolution than CT and is capable of identifying structural abnormalities including diffuse axonal injury, cerebral contusions, ischemic changes, and other parenchymal pathology not visible on CT. Nevertheless, conventional MRI also frequently appears normal in patients with cTBI, particularly during the acute and subacute stages following injury.10–12 The absence of structural abnormalities should therefore not be interpreted as evidence that neurological function remains intact.
Advanced neuroimaging techniques, including diffusion tensor imaging (DTI), susceptibility-weighted imaging (SWI), functional magnetic resonance imaging (fMRI), magnetic resonance spectroscopy (MRS), and positron emission tomography (PET), have significantly expanded scientific understanding of cTBI-related brain dysfunction. Numerous investigations have demonstrated alterations in white matter integrity, functional connectivity, cerebral metabolism, and network organization following cTBI.11–15 However, despite their value as research tools, these technologies remain limited in routine clinical practice due to variability in acquisition protocols, differences among imaging platforms, lack of universally accepted normative datasets, limited reproducibility across institutions, and uncertainty regarding interpretation at the individual patient level.12–15 Consequently, these modalities currently complement rather than replace comprehensive clinical assessment.
Importantly, structural imaging and functional neurological assessment should not be viewed as competing approaches but rather as complementary components of comprehensive concussion evaluation. Structural imaging identifies anatomical injury, whereas functional assessment evaluates the neurological consequences of injury on human performance.
Patient-reported symptom inventories have become central components of concussion management because they provide insight into the patient’s subjective experience and facilitate serial monitoring throughout recovery. Instruments such as the Post-Concussion Symptom Scale (PCSS), Rivermead Post-Concussion Symptoms Questionnaire (RPQ), and Sports Concussion Assessment Tool (SCAT) are widely used in both clinical practice and research.16–19
Despite their widespread acceptance, symptom inventories possess inherent limitations. Symptom reporting is influenced by numerous factors including mood disorders, sleep quality, anxiety, depression, chronic pain, medication effects, cultural differences, litigation status, patient expectations, and individual symptom tolerance.20–22 Furthermore, patients may intentionally or unintentionally underreport or overreport symptoms for a variety of reasons, introducing variability that is independent of neurological function.
Perhaps more importantly, symptom severity does not consistently correlate with objectively measurable neurological dysfunction. Individuals reporting relatively mild symptoms may demonstrate significant vestibular impairment, oculomotor dysfunction, balance deficits, or cognitive abnormalities, whereas others reporting severe symptom burden may demonstrate comparatively limited objective findings.17,19,23 Consequently, symptom inventories should be viewed as valuable measures of patient perception rather than direct measures of functional neurological disability.
Modern concussion management frequently incorporates multiple objective assessments targeting specific neurological systems, including vestibular evaluation, oculomotor testing, computerized cognitive assessment, instrumented balance analysis, dual-task performance, reaction time testing, cranial nerve examination, and psychological screening.16–25
Although each assessment contributes important information regarding a particular aspect of neurological function, these instruments are generally interpreted independently. Clinicians must mentally synthesize findings across numerous tests without the benefit of a standardized framework capable of integrating objective abnormalities into a single measure of overall neurological disability. This fragmented approach increases variability in clinical interpretation, complicates communication among providers, and limits the ability to objectively monitor overall neurological recovery over time.
Similar challenges historically existed in other neurological disorders before the development of standardized disability scales. In multiple sclerosis, stroke, spinal cord injury, and rehabilitation medicine, objective disability instruments transformed clinical care by providing common metrics for describing disease severity, monitoring progression, evaluating treatment response, and facilitating research.22–26 Concussion medicine has experienced substantial advances in individual assessment technologies; however, a comparable multidomain disability framework has not been widely adopted.
The challenge in concussion medicine is therefore not the absence of objective testing, but rather the absence of a standardized methodology for integrating objective neurological findings into a reproducible measure of current functional disability. Existing assessments successfully identify abnormalities within individual neurological systems, yet clinicians frequently lack a common language to quantify the cumulative functional impact of multisystem dysfunction.
An integrated framework should preserve the strengths of individual assessments while synthesizing objective findings into a standardized representation of overall neurological disability. Such a framework has the potential to improve interdisciplinary communication, support individualized treatment planning, facilitate serial monitoring of recovery, standardize research outcomes, and enhance the consistency of medicolegal and disability evaluations.
The Comprehensive Functional Neurological Evaluation (C-FNE) and the Short-Term Concussion Disability (ST-CD) Scale were developed to address this unmet need. Together, they provide a standardized approach for objectively evaluating multiple neurological systems and quantifying the patient’s current level of functional neurological disability during the recovery process following cTBI.
This manuscript has five objectives. First, to establish the conceptual and measurement basis on which a disability scale for concussion should rest — the international framework for functioning and disability, the lineage of established neurological disability instruments, the construct being measured, the measurement model adopted, and the justification for an ordinal structure (Sections 2.0 and 3.0). Second, to document the development history of the C-FNE and the ST-CD Scale, including the reasoning behind domain and instrument selection and the successive revisions that produced the current version (Section 4.0). Third, to state every substantive design decision with its rationale and its accepted cost, together with the alternative approaches considered and the grounds on which each was rejected (Section 5.0). Fourth, to describe the examination and the scale in sufficient technical detail that an independent clinician or investigator could reproduce the scoring procedure (Sections 6.0 and 7.0). Fifth, to define the appropriate uses, the boundaries of appropriate use, and the validation work that must be completed before the instrument can be recommended for general adoption (Sections 8.0 through 12.0).
The manuscript deliberately follows a sequential argument rather than a defense of a proprietary method. The problem is stated first; the inadequacy of existing approaches is examined; the meaning of disability is defined; the case for objective measurement is made; the examination that generates the objective data is described; the conversion of those data into a standardized score is specified; clinical use is explained; and the outstanding scientific obligations are stated plainly.
Before a new disability instrument can be justified, the construct it intends to measure must be defined precisely. Terms such as injury, impairment, disability, and handicap are frequently used interchangeably in clinical, insurance, and legal settings, yet they describe distinct phenomena that carry different clinical and administrative consequences. Conflation of these terms is one of the principal sources of inconsistency in concussion documentation.
The World Health Organization’s International Classification of Functioning, Disability and Health (ICF) provides the accepted international vocabulary for describing human functioning.27 Within that framework, impairment denotes a problem in body function or structure, such as reduced vestibulo-ocular gain, saccadic dysmetria, or diminished postural stability. Activity limitation denotes difficulty executing a task, such as reading a page of text, descending stairs, or sustaining attention through a meeting. Participation restriction denotes a problem experienced in life situations, such as an inability to work a full shift, drive, or return to study. Contextual factors — environmental and personal — modulate the relationship among these levels.27
Concussion is unusual in that impairment can be substantial while remaining invisible on structural imaging, and activity limitation can be profound while conventional bedside neurological examination appears unremarkable. A useful concussion disability scale must therefore be anchored in measurable impairment, but must express its output in terms that relate to activity and participation. The ST-CD Scale is constructed on precisely this logic: its inputs are objective impairment measures, and its output is expressed as a graded statement about functional capacity in daily life.
Definition used throughout this manuscript. Short-term concussion disability is the magnitude of functional limitation attributable to objectively demonstrable neurological dysfunction, measured at a defined point during the acute or subacute recovery period, before maximal medical improvement has been established. It is a state measure, not a permanent characteristic of the patient, and it is expected to change as recovery progresses or fails to progress.
Four instruments dominate the measurement of neurological disability and each contributes a distinct design lesson to the present work.
The Expanded Disability Status Scale (EDSS), introduced by Kurtzke in 1983 for multiple sclerosis, remains the most direct structural antecedent of the ST-CD Scale.22 Its defining feature is a two-level architecture: the examiner first grades several Functional Systems from the neurological examination, then converts those graded systems into a single ordinal score from 0 to 10 in half-point increments. The lower portion of the EDSS range is governed principally by examination findings, while the upper range is governed by ambulatory capacity. The EDSS demonstrates that an ordinal global score can be rendered reproducible when it is explicitly derived from a structured examination rather than from global clinical impression — the principle later formalized in standardized Neurostatus scoring and rater certification.22,28 It also illustrates the recognized weaknesses of ordinal disability scales: non-linear step intervals, uneven sensitivity across the range, and heavy weighting of a single domain at the upper end.28,29
The National Institutes of Health Stroke Scale (NIHSS) demonstrates the value of item-level operational definitions.23 Every item specifies exactly what the examiner does and exactly what score each observed behavior receives. This granularity, coupled with mandatory rater training, is the principal reason the NIHSS achieves acceptable inter-rater agreement across thousands of examiners in routine clinical settings. The lesson adopted here is that reproducibility is a function of scoring rules, not of examiner expertise alone.
The modified Rankin Scale (mRS) demonstrates the communicative power of a short, interpretable ordinal scale.24 Its six grades are immediately meaningful to clinicians, patients, payers, and courts. Its known weakness — substantial inter-rater variability when applied without a structured interview — reinforces the same lesson drawn from the NIHSS.24,30
The Functional Independence Measure (FIM) demonstrates the value of anchoring grades to observable performance of daily activities and to the amount of assistance required.25,26 Its influence on the ST-CD Scale lies in the decision to anchor each global step to a statement about real-world functional capacity rather than to an abstract severity adjective.
A fifth instrument, the Glasgow Outcome Scale–Extended (GOS-E), is widely used in traumatic brain injury outcome research.31 Its coarse eight-category structure is well suited to describing global outcome after moderate and severe TBI but is insufficiently sensitive to discriminate among patients with cTBI, the overwhelming majority of whom fall within its top one or two categories. This ceiling effect is a central reason a concussion-specific scale is required.
Table 1. Design characteristics of established neurological disability scales and the corresponding design decisions adopted for the ST-CD Scale.
| Instrument | Population | Range / structure | Principal strength | Design lesson adopted for ST-CD |
|---|---|---|---|---|
| EDSS22 | Multiple sclerosis | 0–10, 0.5 steps; 7–8 Functional Systems | Global score explicitly derived from a structured neurological examination | Two-level architecture: Functional System Scores aggregated into a global 0–10 ordinal score in 0.5 steps |
| NIHSS23 | Acute stroke | 0–42 summed items | Item-level operational definitions; strong inter-rater agreement with training | Every domain grade is tied to a stated measurement threshold, not to impression |
| mRS24 | Stroke / general neurology | 0–6 ordinal | Immediately interpretable across disciplines | Interpretive category bands layered over the numeric continuum |
| FIM25,26 | Rehabilitation | 18 items, 7 levels each | Anchored to observable performance and assistance required | Global steps anchored to real-world functional capacity, not severity adjectives |
| GOS-E31 | Moderate–severe TBI | 1–8 categories | Standard endpoint in TBI outcome trials | Demonstrates the ceiling effect that a concussion-specific scale must avoid |
Three arguments support anchoring a concussion disability scale in objective measurement.
The measurement argument. Symptom report and functional disability are related but distinct constructs. Correlation between symptom burden and objective performance after cTBI is consistently modest, and the direction of divergence is not predictable at the individual level.17,19,23 An instrument built primarily on symptom report therefore measures perception with reasonable fidelity and function with poor fidelity.
The clinical argument. Treatment after cTBI is domain-specific. Vestibular rehabilitation addresses vestibular and balance dysfunction; oculomotor therapy addresses convergence, saccadic, and pursuit deficits; cognitive rehabilitation addresses attention and processing speed; behavioral health addresses mood, anxiety, traumatic stress, and insomnia.32–36 A global symptom score cannot direct therapy because it does not identify which system is failing. A domain-resolved objective assessment can.
The evidentiary argument. Concussion is frequently evaluated in contexts where the findings will be scrutinized — disability determination, workers’ compensation, personal injury litigation, return-to-duty and return-to-play decisions. In these settings, self-report is routinely discounted. Instrument-derived measurements obtained under standardized conditions are not immune to challenge, but they shift the discussion from whether the patient is credible to whether the measurement is valid, which is a scientifically tractable question. This consideration must never distort clinical judgment, and Section 10.0 addresses the ethical boundaries directly.
A distinction that is frequently blurred in practice, and that the ST-CD Scale deliberately enforces, is the difference between short-term disability and permanent impairment. Permanent impairment rating — as codified in the American Medical Association Guides to the Evaluation of Permanent Impairment and analogous jurisdictional schedules — is a determination of permanent loss, appropriately performed only after the patient has reached maximal medical improvement (MMI), meaning that further material recovery is not anticipated.37 Applying an impairment rating before MMI systematically misrepresents the patient’s condition, in either direction.
The ST-CD Scale occupies the interval before MMI. It answers a different question: how disabled is this patient right now, and by which systems? Serial ST-CD scores across the recovery period generate a disability trajectory. When a subsequent permanent impairment evaluation becomes appropriate, that trajectory constitutes documented evidence of the course of recovery, but the ST-CD score is not itself an impairment rating and must not be converted into one. This boundary is restated in Sections 8.0, 8.0, and 9.0 because its violation is the most likely misuse of the instrument.
Table 2. Distinction between short-term disability measurement and permanent impairment rating.
| Dimension | ST-CD Scale (short-term disability) | Permanent impairment rating |
|---|---|---|
| Timing | Acute and subacute phase; any point before maximal medical improvement | Only after maximal medical improvement is established |
| Question answered | What is the patient’s current functional neurological disability, and which systems drive it? | What permanent loss of function remains after recovery has plateaued? |
| Expected behavior over time | Expected to change; serial measurement is the point | Expected to be stable; a single determination |
| Primary clinical use | Directing rehabilitation, monitoring recovery, graduated return to activity | Final disability apportionment and settlement of permanent loss |
| Relationship | Documents the recovery trajectory that precedes and informs impairment evaluation | Builds upon documented disability trajectories; does not replace them |
A clinical instrument should state what it claims to measure, on what theory of measurement it rests, and what class of statistical treatment its output will and will not support. Instruments that omit this are frequently found, years later, to have been measuring something other than what their users assumed. This section sets out the measurement framework of the ST-CD Scale before any description of its content, so that the design decisions in Section 5.0 can be evaluated against a stated position rather than inferred from the result.
The construct the ST-CD Scale is intended to measure is defined as follows.
Construct. Current functional neurological disability attributable to objectively demonstrable dysfunction across the neurological systems commonly disrupted by cTBI, measured at a defined point during the acute or subacute recovery period.
The construct includes: the magnitude and distribution of measurable dysfunction across vestibular, oculomotor, postural, cognitive, cranial nerve and sensory, motor, autonomic and headache, psychological, and participation domains, expressed as its consequence for daily function.
The construct excludes: diagnosis; causation; symptom burden as experienced; suffering; permanent impairment; prognosis; effort or credibility; and disability arising from non-neurological injury sustained in the same event.
Two features of this definition carry consequences throughout the manuscript. First, the construct is defined as a state, not a trait: it is expected to change, and an instrument measuring it must be capable of registering change. Second, the construct is defined as attributable to measurable dysfunction, which deliberately excludes disability that the patient experiences but that no measurement supports. This is a real limitation and is acknowledged as such in Section 11.0; it is accepted because the alternative — admitting unmeasured disability into a scale that claims objectivity — would defeat the purpose of building the instrument.
Three commitments define the measurement position adopted here.
Measurement precedes interpretation. A domain grade is produced by comparing a measurement against a stated threshold. The clinician’s judgment enters at defined points — anchor reconciliation and final assignment — and when it does, it is recorded. Judgment distributed invisibly throughout a scoring process is indistinguishable from unreliability.
Transparency is prior to accuracy. Between a more accurate procedure that cannot be written down and a less accurate procedure that can, the framework selects the latter. A procedure that is fully specified can be tested, criticized, and improved; an accurate but unspecified procedure cannot be, and its accuracy cannot be verified by anyone other than its author.
The instrument is falsifiable. Each proposition the framework advances — that the nine domains cohere as a measurement system, that the derivation rules order patients correctly, that the anchors mean what they say — is an empirical claim capable of being refuted by the studies in Section 12.0. The framework is written so that such refutation would be detectable rather than absorbed.
The distinction between reflective and formative measurement models is central to how the ST-CD Scale should be evaluated, and it is frequently overlooked in clinical scale development.38,39
In a reflective model, the items of a scale are interchangeable manifestations of a single underlying latent trait. Depression questionnaires are the standard example: low mood, anhedonia, and sleep disturbance are held to be expressions of an underlying depressive state, they are expected to correlate strongly with one another, and removing one item should not materially change what the scale measures. Reflective models justify internal-consistency statistics such as Cronbach’s alpha and are the natural target of common-factor analysis.
In a formative model, the components jointly constitute the construct rather than reflecting it. Socioeconomic status is the classic example: income, education, and occupation are not interchangeable expressions of an underlying quantity — they compose the quantity, they need not correlate, and deleting one changes the construct itself.39
The ST-CD Scale is a formative composite. A patient may have profound vestibular dysfunction with entirely normal cognition, or severe cognitive dysfunction with a normal vestibular examination. These domains are not expected to correlate, because they index anatomically and physiologically distinct systems that cTBI may disrupt independently. Removing the vestibular domain would not simply reduce the scale’s reliability; it would change what the scale measures.
Three consequences follow, and each is a commitment the authors are prepared to be held to.
The ST-CD Scale is ordinal by design rather than by concession, and the reasoning should be stated rather than assumed.
Why not a continuous score. A continuous scale would imply that the underlying quantity is measured on an interval or ratio metric — that the difference between 2.0 and 3.0 equals the difference between 8.0 and 9.0. No such claim is defensible for a composite of heterogeneous domain grades. Presenting a continuous score would encode a measurement claim the instrument cannot support, would invite parametric analysis that the data do not license, and would imply a precision that the underlying gradings do not possess.40,41
Why not a simple category system. A purely categorical instrument — mild, moderate, severe — would be honest about its measurement properties but would lack the resolution required to track recovery. A patient improving within the moderate band would register no change at all, which defeats the purpose of a state measure intended for serial use.
Why an ordinal scale with interpretive bands. The chosen structure preserves the resolution needed for serial measurement while retaining the interpretability of categories. The numeric score orders patients; the category band communicates. The two are reported together precisely so that neither the false precision of the number nor the coarseness of the category operates alone.
What follows from ordinality. ST-CD scores should be summarized by median and range rather than mean and standard deviation; group comparisons should use non-parametric methods; and change scores should not be computed arithmetically until a Rasch or equivalent analysis has established whether an interval transformation is supportable.40,41 These constraints are stated here so that subsequent investigators inherit them rather than rediscover them.
The two components of this framework are frequently conflated, and the distinction is worth stating precisely: the C-FNE is a measurement procedure; the ST-CD Scale is a scoring model applied to its output.
Table 3. Distinction between the C-FNE as a measurement procedure and the ST-CD Scale as a scoring model.
| Dimension | C-FNE | ST-CD Scale |
|---|---|---|
| What it is | A standardized multidomain examination procedure | An ordinal scoring model applied to the examination’s output |
| What it produces | Quantitative measurements across nine functional systems | Nine domain grades and a single anchored global score |
| What it answers | Which systems are dysfunctional, and by how much? | What is the overall magnitude of current functional disability? |
| Can it exist alone? | Yes — the examination has clinical value with no score attached | No — the scale has no meaning without a completed examination |
| Substitutability | Equivalent instrumentation may be substituted where the measured physiological quantity is preserved | The scoring model is fixed; substituting derivation rules produces a different instrument |
| Failure mode | Incomplete or technically invalid measurement | Correct measurement combined by an inappropriate rule |
The practical significance of the distinction is that the two components fail independently and must be validated independently. A perfectly executed C-FNE combined by a poor scoring rule yields a misleading score from sound data. A well-designed scoring rule applied to a poorly executed examination yields a misleading score from sound reasoning. The validation programme in Section 12.0 therefore evaluates measurement quality and scoring-model quality as separate questions.
A reasonable objection to any disability scale is that disability is a lived experience and that quantifying it is a category mistake. The objection deserves an answer rather than dismissal.
The framework does not claim to measure the experience of disability. It claims to measure something narrower and more tractable: the extent to which measurable neurological dysfunction restricts function. That quantity is measurable for the same reason visual acuity, ambulatory capacity, and pulmonary function are measurable — it varies systematically, it can be observed under standardized conditions, and it orders individuals in a way that predicts what they can do. The ICF framework rests on precisely this proposition, and every disability instrument in neurological practice is an application of it.27
What the framework accepts, in consequence, is that the score is a description of function and not a description of the patient. A patient with an ST-CD score of 6.5 is not “a 6.5”; the score states how much of that person’s day is currently constrained by dysfunction that was measured. This distinction is not merely rhetorical. It determines how the score should be written in a report, how it should be explained to a patient, and how it should be read by anyone else.
This section documents how the examination and the scale came into being. It is included deliberately and at length, because the credibility of a clinical instrument rests substantially on whether it emerged from sustained clinical use and iterative correction or was constructed as a theoretical exercise and applied afterward. The ST-CD Scale is a product of the former process.
The framework originated in 2021 within a multidisciplinary outpatient practice evaluating patients referred after motor vehicle collisions, falls, occupational injuries, assaults, and sport-related trauma. The referral pattern was consequential for design: a substantial proportion of patients presented in the subacute period, weeks to months after injury, rather than on the sideline or in the emergency department. Sideline and emergency-department instruments — which are optimized for rapid triage and removal-from-play decisions — were poorly matched to this population.19,32
Three recurring clinical failures motivated the work. First, patients arrived carrying normal CT or MRI reports that had been interpreted by prior providers, employers, and insurers as evidence that no injury existed, despite unambiguous functional deficits on examination. Second, when multiple objective tests were performed, each was reported separately and no document stated what the aggregate meant; referring providers received a folder of results rather than an answer. Third, when asked to characterize severity, clinicians within the practice produced inconsistent descriptions of comparable patients, because severity was being assigned by impression rather than by rule.
The first two problems required a standardized multidomain examination. The third required a scale.
Six principles were established at the outset and have governed every subsequent revision.
Nine functional systems were selected. Candidate domains were evaluated against four criteria: documented frequency of involvement after cTBI in the peer-reviewed literature; availability of an objective or semi-objective measure suitable for routine outpatient use; demonstrated responsiveness to targeted rehabilitation; and non-redundancy with the domains already selected.
Vestibular, oculomotor, and postural control domains were included first, reflecting the high prevalence and rehabilitative tractability of vestibulo-ocular and balance dysfunction after cTBI.16,20,21,32,33 Cognitive processing was included on the strength of the extensive literature documenting deficits in attention, processing speed, and working memory after cTBI.6,10,34 Cranial nerve and sensory function and motor and strength function were included because cTBI frequently coexists with cervical, peripheral, and cranial nerve injury in the same mechanism, and because these findings are readily quantified.35 Autonomic and headache function was included because post-traumatic headache, photophobia, phonophobia, and exercise intolerance are among the most persistent and functionally limiting sequelae.36,42 Psychological and behavioral function was included because anxiety, depression, traumatic stress, and insomnia both accompany and independently prolong concussion recovery, and because validated brief instruments exist for each.43–47 Finally, activities of daily living and functional participation was retained as a distinct domain, reflecting the ICF principle that impairment and participation are not the same construct and that a disability scale must ultimately speak to the latter.27
Two candidate domains were considered and excluded. Isolated musculoskeletal pain was excluded as a scored domain because it is frequently attributable to concurrent injury rather than to neurological dysfunction and would confound the neurological construct; it is documented in the report but does not generate an ST-CD domain grade. Symptom burden as such was excluded as a scored domain for the reasons given in Section 1.2.2, and is recorded alongside the score as context.
Instruments were selected for standardization, availability in an outpatient setting, tolerability by symptomatic patients, and generation of numeric output permitting threshold-based grading. Videonystagmography was adopted for vestibular and oculomotor assessment because it provides quantified recordings of spontaneous, gaze-evoked, and positional nystagmus, smooth pursuit gain, saccadic accuracy and latency, optokinetic response, and caloric responses.48,49 Instrumented posturography was adopted for postural control because force-platform measurement of center-of-pressure excursion is substantially more sensitive and more reproducible than observational balance scoring.20,21,50 Computerized cognitive testing was adopted for the cognitive domain to provide normatively referenced performance across attention, processing speed, working memory, and executive tasks.34,51 Quantitative dynamometry was adopted for motor and strength assessment.52 Validated brief screening instruments were adopted for the psychological domain, and a validated functional-impact instrument for the participation domain.43–47
It is a deliberate feature of the framework that the domain grading thresholds are specified in terms of the physiological quantity measured rather than in terms of a particular manufacturer’s device, so that equivalent instrumentation may be substituted. It is a corresponding limitation, addressed in Section 11.0, that equivalence across devices has not been formally demonstrated.
Development followed a sequence that is conventional in instrument construction but is rarely documented in clinical scale papers: identification of a clinical need, review of the relevant literature, definition of the construct, selection of domains and instruments, construction of a prototype scoring architecture, pilot implementation, iterative refinement under routine use, and standardization. Figure 2 sets out that sequence and marks the point the framework has currently reached.
Five phases are distinguishable in retrospect. The phase boundaries were not planned in advance; they are identified here by the substantive change that closed each one.
Phase I — Concept development
The initial phase established the clinical need described in Section 4.1, surveyed the existing concussion assessment literature and the established neurological disability instruments, and produced a first prototype consisting of a fixed examination battery with a global clinician severity impression appended. The prototype was used clinically from the outset; it was never a paper exercise.
Phase II — Pilot implementation and initial functional systems
The second phase introduced the concept of separately graded functional systems, replacing the single global impression with a set of domain grades. The domain set at this stage was narrower than the current nine and was assembled around vestibular, oculomotor, balance, and cognitive function. Scoring at this stage combined the domain grades by averaging.
Phase III — Abandonment of averaging
The third phase produced the most consequential single change in the development of the instrument. Averaging was abandoned after repeated observation that it produced scores contradicting the observed function of patients with severe single-system involvement. The EDSS-style derivation described in Section 7.3 — in which the highest domain grades and the number of involved domains jointly determine the global score — was adopted in its place, and the domain set was expanded toward its present composition.22
Phase IV — Anchoring and standardization
The fourth phase replaced adjectival severity descriptors with the time-and-capacity anchors described in Section 7.2, after referring providers, therapists, and non-clinical readers were found to be interpreting the same descriptor differently. This phase also introduced examiner standardization, the requirement that departures from the derived score be documented, and formal version control of the scoring rules.
Phase V — Not Tested rules and stabilization
The final phase addressed the systematic underestimation of disability in patients unable to complete the examination battery, introducing the formal Not Tested status and the associated global-scoring rules described in Section 7.7. The scoring rules were then frozen for the purpose of preparing this manuscript and the validation programme; no substantive change has been made since stabilization, and any subsequent change will be issued as a numbered version.
The scoring rules are maintained under formal version control. Each version carries a number, an effective date, a statement of what changed, and the rationale for the change. Scores are recorded against the version under which they were assigned, so that scores generated under different versions are not silently compared. This practice began during Phase IV; scores generated before that point are not version-identified and should not be pooled with later scores for research purposes.
The ST-CD Scale did not appear in its present form. The table below records the four substantive scoring changes, each driven by a specific failure of the preceding version.
Table 4. Development phases of the ST-CD Scale and the problem each phase addressed.
| Phase | Approach | Problem identified in routine use | Change made |
|---|---|---|---|
| I (initial) | Global clinician severity impression, 0–10 | Comparable patients received scores differing by two or more points depending on the examiner; severity could not be defended when questioned | Score required to be derived from documented examination findings rather than assigned globally |
| II | Unweighted average of domain findings | Averaging diluted severe single-system dysfunction: a patient with profound vestibular loss and otherwise normal systems scored as mildly disabled, contrary to observed function | Averaging abandoned; adopted an EDSS-style derivation in which the highest domain grades and the number of involved domains jointly determine the global score22 |
| III | EDSS-style derivation with descriptive severity words | Severity adjectives (“moderate,” “severe”) were interpreted differently by referring providers, therapists, and non-clinical readers | Every global step anchored to an explicit statement of the proportion of waking time functionally impacted and the corresponding functional capacity |
| IV (current) | Anchored derivation with defined handling of untested domains and mandatory clinician confirmation | Aborted or incomplete testing (symptom intolerance, calibration failure, deferred cognitive testing) produced spuriously low scores when missing domains were scored as normal | Formal “Not Tested” status introduced; rules specified for when a global score may and may not be reported (Section 7.6) |
Phase IV also incorporated a set of internal consistency rules derived from repeated observation, intended to reduce systematic under-detection. Examples include the requirement that any abnormal ocular-motor measurement anywhere in the battery be reflected in the oculomotor domain regardless of whether the patient reports visual symptoms; that objectively abnormal postural control be reflected in the balance domain regardless of self-reported steadiness; and that documented insomnia be evaluated together with fatigue and cognitive complaint, given their consistent co-occurrence. These rules exist because the alternative — permitting the absence of a self-reported symptom to override an objective abnormality — reintroduces precisely the subjectivity the framework was built to eliminate.
The framework has been applied in more than 5,000 patient evaluations across a multi-state outpatient practice over approximately seven years. This experience informed domain selection, threshold placement, the handling of incomplete testing, and the wording of the anchors. It should be understood for what it is and what it is not. It is a substantial body of applied clinical use that has exposed the framework to a wide range of presentations, ages, injury mechanisms, and comorbidities, and that has driven successive correction of the instrument. It is not a validation dataset. The evaluations were performed for clinical purposes under routine conditions, without a predefined research protocol, without independent duplicate scoring for reliability estimation, and without prospectively collected reference-standard outcomes. No inferential claims are made from this experience, and none should be drawn from it. The validation program required to convert clinical experience into psychometric evidence is specified in Section 12.0.
Statement on data. No patient-level data, outcome statistics, accuracy estimates, or reliability coefficients are reported in this manuscript. Where the volume of clinical application is described, it is offered solely as a statement of development context.
Every scale embodies a series of choices, and a scale whose choices are undocumented is difficult to evaluate and impossible to improve deliberately. This section states each substantive design decision, the reasoning behind it, and the cost accepted in making it. It then sets out the alternative approaches that were considered and the grounds on which each was rejected.
The decisions are presented with their trade-offs rather than as unqualified merits. Every one of them purchases something at a price, and a reviewer who identifies the price has not found a flaw the authors overlooked.
Table 5. Principal design decisions, rationale, and accepted cost.
| Decision | Rationale | Alternative rejected | Cost accepted |
|---|---|---|---|
| Nine functional systems | Covers the systems with documented involvement after cTBI, each with an available objective measure and a distinct rehabilitative pathway | Fewer domains (3–5); more domains (12+) | Examination burden of 90–150 minutes; incomplete batteries are common |
| 0–4 domain grading | Five levels distinguish measurable-but-silent from intermittent, consistent, and disabling involvement without exceeding the resolution the underlying measurements support | Binary normal/abnormal; 0–6 as in some EDSS systems | Coarser than the underlying continuous measurements |
| Global range 0–10 | Preserves structural comparability with the EDSS and with the 0–10 severity conventions already familiar to clinicians | 0–100; 0–6 categorical | The upper range is rarely used in concussion |
| 0.5 increments | Provides resolution sufficient to register clinically meaningful change on serial examination without implying precision the gradings cannot support | Whole-point steps; continuous scoring | Twenty-one levels invite over-interpretation of small differences |
| Composite rather than profile alone | A single value is required for communication, triage, trend display, and administrative use | Domain profile only | Aggregation obscures the profile unless reporting is constrained (Section 7.8) |
| Derivation by highest grade and number of domains | Reproduces the EDSS principle that focal severe and diffuse mild involvement can yield comparable disability by different routes | Unweighted average; weighted sum | Rules are consensus-derived rather than empirically weighted |
| Anchoring to proportion of time impacted | Time-and-capacity statements are answerable from documented activity tolerance; severity adjectives are not | Adjectival anchors; symptom-count anchors | Requires a functional history of adequate quality |
| Mandatory clinician assignment | Preserves accountability and permits documented departure where the derivation does not fit the patient | Fully algorithmic scoring | Introduces a source of variance that must be measured, not assumed away |
| Formal Not Tested status | Prevents the systematic underestimation of disability in patients too symptomatic to complete the battery | Imputation; scoring omissions as normal | Global score is unavailable in some examinations |
| Ordinal treatment | The composite cannot support interval claims without empirical demonstration | Continuous or interval scoring | Constrains statistical analysis to non-parametric methods |
The number nine was an outcome rather than a target. Domains were admitted if they satisfied four criteria simultaneously: documented frequency of involvement after cTBI; availability of an objective or semi-objective measure suitable for routine outpatient use; a distinct rehabilitative pathway, such that identifying the domain changes what is done for the patient; and non-redundancy with domains already admitted.
Fewer domains were considered. A three- to five-domain instrument covering vestibular, oculomotor, cognitive, and balance function would be substantially faster to administer and would capture the presentations most characteristic of concussion. It was rejected because it would systematically miss the patient whose disability is driven by post-traumatic headache with photophobia, or by insomnia and anxiety, and because those presentations are common and are among the most functionally limiting.
More domains were also considered. Candidate additions included autonomic cardiovascular function as a separate domain from headache, olfactory function, sleep architecture as distinct from insomnia severity, and vocational capacity as distinct from general participation. Each was rejected on grounds of measurement burden, absence of a routinely available objective measure, or redundancy with an admitted domain. Two candidates were rejected on construct grounds: isolated musculoskeletal pain, because it is frequently attributable to concurrent non-neurological injury and would confound the neurological construct; and symptom burden as such, for the reasons given in Section 1.2.2.
Why 0–4 rather than binary. A binary normal/abnormal classification would be maximally reproducible and minimally informative. It cannot distinguish the patient with a measurable caloric asymmetry who functions normally from the patient whose vestibular dysfunction has ended their ability to drive, and that distinction is the entire clinical question.
Why five levels rather than more. Grading resolution should not exceed the reliability of the underlying judgment. Five levels — absent, measurable but silent, intermittent, consistent, and disabling — correspond to distinctions that examiners can be trained to make consistently. Seven-level grading was piloted informally and abandoned when adjacent categories proved not to be reliably separable.
Why 0.5 global increments. Whole-point steps across a 0–10 range give eleven levels, which is too coarse to register the improvement typically observed between monthly reassessments; a patient may improve materially without crossing a whole-point boundary. Continuous scoring was rejected under Section 3.4. Half-point steps give twenty-one levels, which matches both the resolution of the derivation rules and the convention of the antecedent instrument.22 The accepted cost is the temptation to over-interpret single-step differences, which Section 8.3 addresses directly.
The strongest argument against a composite is that it discards information: nine grades contain more than one number does. That argument is correct, and the framework accepts it — which is why Section 7.8 requires that the profile always accompany the score.
The argument for the composite is practical rather than theoretical. Communication, triage, trend display, care coordination, and administrative decision-making all require a single value; where clinicians do not supply one, others construct one informally and without rules. The choice is not between a composite and no composite. It is between a composite whose derivation is documented and a composite assembled ad hoc by whoever needs a number.
A fully algorithmic score would be perfectly reproducible, and reproducibility is not the same as validity. Algorithms fail predictably at the boundaries of their design: the patient with documented pre-existing vestibular disease, the patient whose participation is limited by a concurrent orthopaedic injury, the patient for whom normative reference data do not apply. In each case a rule-derived score would be defensible and wrong.
The framework retains the clinician as the final authority and requires that any departure from the derived value be recorded with its reason. This is the same structure adopted by standardized neurological scoring systems that combine defined examination rules with certified rater judgment.28 It converts clinician variance from an invisible property of the instrument into a measurable and auditable one — and the inter-rater study specified in Section 12.2 is designed to measure exactly that.
Severity adjectives do not survive transfer between readers. “Moderate” is interpreted differently by a neurologist, a physical therapist, a claims adjuster, and a patient, and the differences are systematic rather than random. Anchoring each global step to the proportion of waking time during which function is materially impacted replaces an adjective with a question that can be answered from documented activity tolerance, work status, driving status, and examination findings.
The approach borrows from the FIM’s use of observable performance and required assistance as grade anchors rather than severity descriptors.25,26 Its cost is a dependence on the quality of the functional history: where that history is thin, the anchor cannot discriminate, and the derivation must carry the score alone. This is a genuine weakness and is stated in Section 11.0.
The following approaches were evaluated during development. Each is a reasonable option, several are in widespread use, and the grounds for rejection are specific to the construct defined in Section 3.1 rather than to any deficiency in the instruments themselves.
Table 6. Alternative approaches considered during development and grounds for rejection.
| Alternative | What it would have provided | Grounds for rejection |
|---|---|---|
| Symptom inventories alone (PCSS, RPQ, SCAT)16,53 | Rapid, validated, low-burden serial measurement of the patient’s experience | Measures perception rather than function; correlates only modestly with objective performance; vulnerable to contextual and incentive effects. Retained as context alongside the score, not as a scored domain |
| Structural or diffusion imaging | Objective, quantitative, widely accepted as evidence | Frequently normal in concussion; DTI lacks standardized acquisition, normative databases, and individual-level interpretability. Complements rather than measures function12–15 |
| Neuropsychological testing alone | Deep, normatively referenced characterization of cognition | Misses vestibular, oculomotor, postural, autonomic, and participation dysfunction entirely; a patient may be cognitively intact and severely disabled |
| A single validated screening tool (VOMS)16 | Brief, validated, sensitive to vestibulo-ocular provocation | Designed as a screening instrument for a specific system; provokes symptoms rather than quantifying function; not constructed to support a global disability statement |
| A validated handicap inventory (DHI)54 | Validated, patient-reported, directly addresses functional handicap | Self-report instrument confined to dizziness-related handicap; does not span the domain set and reintroduces the symptom-report limitations above |
| Machine-learning derived score | Empirically optimized weighting; potentially superior predictive accuracy | Requires a large labelled dataset with a reference standard that does not currently exist; produces weights that cannot be explained to a patient, a therapist, or a court. Deferred, not excluded — appropriate once the validation dataset exists |
| Weighted or unweighted average of domain grades | Simple, transparent, arithmetically defensible | Averaging dilutes severe single-system dysfunction; a patient with profound vestibular loss and eight normal domains scores as mildly disabled, contrary to observed function. Empirically abandoned in Phase II (Section 4.6) |
| Continuous 0–100 scoring | Fine resolution; supports parametric analysis | Encodes an interval-measurement claim the composite cannot support; implies precision the domain grades do not possess41,55 |
| Adoption of an existing scale (GOS-E, mRS, FIM)24–26,31 | Established, validated, familiar, comparable across studies | Ceiling effects place nearly all concussion patients in the top one or two categories; none resolves the domain profile required to direct treatment |
Two of these decisions are explicitly provisional. Machine-learning derivation of domain weights is deferred rather than rejected, and becomes appropriate once the predictive validity cohort described in Section 12.1 provides labelled outcome data. Interval transformation of the ordinal structure is likewise deferred pending the Rasch analysis described in Section 12.3. In both cases the current design is the defensible position given present evidence, not a permanent commitment.
The framework is not proposed as a replacement for the instruments in current use. The relationship is one of scope: existing tools answer narrower questions well, and the ST-CD Scale answers a question none of them was built to address.
Table 7. Scope of instruments in current concussion practice compared with the C-FNE and the ST-CD Scale.
| Instrument | Primary question | Data type | Domains covered | Multidomain composite | Directs treatment | Validated |
|---|---|---|---|---|---|---|
| SCAT653,56 | Is this an acute concussion? | Mixed self-report and brief performance | Symptoms, orientation, brief cognition, balance screen | No | Limited | Yes |
| PCSS / RPQ16–18 | How does the patient feel? | Self-report | Symptom burden | No (symptom sum only) | No | Yes |
| VOMS16 | Does vestibulo-ocular provocation reproduce symptoms? | Provoked symptom report | Vestibulo-ocular | No | Partly | Yes |
| DHI54 | How handicapping is the dizziness? | Self-report | Dizziness handicap | No | Partly | Yes |
| CT / MRI9–11 | Is there structural injury? | Imaging | Structure only | No | Rarely | Yes |
| DTI12–15 | Is white matter microstructure altered? | Imaging | Structure only | No | No | Research use |
| Posturography20,21,50 | How stable is postural control? | Instrumented | Balance | No | Yes, within domain | Yes |
| Computerized cognition34,51 | How efficient is cognition? | Instrumented | Cognition | No | Yes, within domain | Yes |
| C-FNE | Which systems are dysfunctional, and how much? | Instrumented, multidomain | Nine functional systems | Provides the inputs | Yes | Not yet |
| ST-CD Scale | What is the overall magnitude of current disability? | Derived composite | Nine functional systems | Yes | Yes, via the profile | Not yet |
The final column is the one that matters most in evaluating this manuscript. Every established instrument in the table carries published validation evidence; the two instruments proposed here do not. That asymmetry is the reason the validation programme in Section 12.0 is specified in the detail that it is, and the reason the framework is described throughout as investigational.
The ST-CD Scale cannot be understood independently of the examination that generates its inputs. A scale is only as good as the measurements it aggregates, and a reader evaluating the ST-CD Scale is entitled to know, in detail, where its numbers come from. This chapter describes the C-FNE: its origin, its design philosophy, its relationship to the conventional neurological examination, its content, its standardization, and the ways in which it differs from routine cTBI evaluation.
The C-FNE is a standardized, sequenced, multidomain neurological evaluation performed in a single outpatient session, typically requiring 90 to 150 minutes depending on patient tolerance.
During development of the examination framework, the authors introduced the term Comprehensive Functional Neurological Evaluation (C-FNE) to describe the standardized multidomain examination protocol developed for the objective assessment of patients with cTBI. The designation distinguishes the examination from conventional neurological examinations by emphasizing its standardized integration of vestibular, oculomotor, balance, cognitive, cranial nerve, motor, autonomic, psychological, and functional participation assessments into a single examination framework.
The components of the designation were selected to describe the examination accurately rather than to differentiate it commercially, and each carries a specific meaning within this framework.
Table 8. Components of the C-FNE designation and what each denotes.
| Term | What it denotes |
|---|---|
| Comprehensive | The multidomain nature of the examination: nine functional systems assessed within a single session under a single protocol, rather than a battery assembled ad hoc from separately ordered tests |
| Functional | That the examination evaluates neurological function rather than structural pathology — the distinction developed in Sections 1.2.1 and 2.1, and the reason the examination has diagnostic value where imaging is normal |
| Neurological | That the assessment is of the nervous system, distinguishing it from orthopaedic, musculoskeletal, and symptom-inventory-based examinations that may be performed on the same patient for other purposes |
| Evaluation | The integration of multiple objective examination procedures into a unified assessment yielding a single interpretive output, rather than a set of independently reported test results |
The C-FNE was constructed to answer one question with as little ambiguity as possible: which neurological systems are dysfunctional in this patient, and to what degree? Every design feature follows from that question.
It is diagnostic of dysfunction, not of injury. The examination characterizes the functional state of the nervous system. It does not establish that a concussive event occurred, and it does not attribute the findings to any particular event; both remain clinical determinations made on the whole record (Sections 8.1 and 9.6).
It is designed for the subacute presentation. Sideline and emergency-department instruments are optimized for rapid triage and removal-from-play decisions in the minutes to hours after injury. The C-FNE is optimized for the patient presenting weeks to months later, in whom the question is no longer whether to remove from activity but which systems remain impaired and what should be done about them.
It is designed to be performed identically by different examiners. Test order, instructions, setup, stopping rules, and documentation are specified. Where a choice existed between a more sensitive procedure that varied by examiner and a more reproducible procedure that did not, the reproducible procedure was selected (Section 4.2).
It is designed to produce numbers, not impressions. Each domain yields a quantitative or threshold-referenced output. This is what permits the examination to feed a scoring model at all; an examination reported in narrative alone cannot support one.
It is designed to survive scrutiny. Calibration status, protocol adherence, deviations, and incomplete measurements are recorded on the face of the report rather than omitted. An examination whose limitations are documented is more defensible than one that appears seamless.
The C-FNE does not replace the conventional neurological examination and is not an alternative to it. It presupposes it. A standard history, general neurological examination, and appropriate imaging remain the basis for identifying the pathology that must not be missed — haemorrhage, mass lesion, cervical spine injury, focal deficit indicating a localizable lesion. Where such findings are present, the C-FNE is not the relevant next step.
What the C-FNE adds is resolution in the region where the conventional examination is least informative. The bedside neurological examination is a highly efficient instrument for detecting focal structural pathology and a comparatively insensitive one for detecting the distributed functional dysfunction that characterizes cTBI. A patient may have entirely normal strength, reflexes, sensation, gait, and cranial nerve examination and still demonstrate a substantial caloric asymmetry, reduced pursuit gain, posturographic instability under degraded sensory conditions, and processing speed two standard deviations below normative expectation. The conventional examination is not wrong in such a patient; it is being asked a question it was not built to answer.
Most patients with cTBI receive some combination of symptom inventory, brief cognitive screening, and vestibulo-ocular screening, frequently across several appointments and several providers. The differences between that pattern and the C-FNE are procedural rather than conceptual, and they are the differences that determine whether the output can support a scoring model.
Table 9. Procedural differences between routine cTBI evaluation and the C-FNE.
| Dimension | Routine evaluation | C-FNE |
|---|---|---|
| Composition | Assembled from separately ordered tests, often across providers and visits | Fixed protocol performed in a single session under one supervising clinician |
| Test order | Determined by scheduling and availability | Standardized to control fatigue effects and to place provocative testing after measures it would contaminate |
| Domain coverage | Driven by presenting complaint; asymptomatic domains frequently untested | All nine domains assessed regardless of presenting complaint, so silent dysfunction is detected |
| Output | Separate reports interpreted independently | Integrated profile with cross-domain consistency examined explicitly (Section 6.8) |
| Missing data | Typically unrecorded; an untested domain is indistinguishable from a normal one | Formally recorded as Not Tested with classified reason (Section 7.7) |
| Quality control | Variable and often undocumented | Calibration status, protocol adherence, and deviations documented on the report |
| Interpretive endpoint | Clinician synthesis, unwritten | Domain grades and a derived global score with the derivation stated (Section 7.0) |
None of these differences implies that routine evaluation is performed poorly. They reflect a different purpose: routine evaluation is organized around answering the patient’s presenting complaint, and the C-FNE is organized around producing a complete and comparable measurement set.
Test order is standardized for two reasons: to control for fatigue effects on performance-based measures, and to ensure that provocative testing is performed after measures that would be contaminated by symptom exacerbation. The general sequence proceeds from history and symptom documentation, to cranial nerve and sensorimotor examination, to computerized cognitive testing, to instrumented posturography, to the videonystagmographic battery with positional and caloric testing performed last, to quantitative strength testing, and finally to the standardized psychological and functional-impact instruments.
Because provocative vestibular testing may exacerbate symptoms, it is placed after cognitive testing rather than before it. Any deviation from the standard sequence, and any test not completed, is recorded on the face of the report, because both affect the interpretation of the resulting score.
Each of the nine functional systems is graded from 0 to 4 on the basis of the measurements obtained. Grade 0 denotes performance within normative limits. Grade 1 denotes a measurable abnormality without demonstrable functional consequence. Grade 2 denotes abnormality producing intermittent limitation of function. Grade 3 denotes abnormality producing consistent limitation requiring modification of activity. Grade 4 denotes abnormality producing loss of the function or dependence on assistance. A tenth possibility, NT (Not Tested), is recorded when the measurement could not be obtained; NT is never scored as 0.
Table 10. The nine ST-CD functional systems, their principal objective measures, and the domain construct graded.
| FS | System | Principal objective measures | Construct graded 0–4 |
|---|---|---|---|
| FS1 | Vestibular function | Videonystagmography: spontaneous and gaze-evoked nystagmus, Dix–Hallpike and positional testing, caloric responses and unilateral weakness calculation | Degree of peripheral and/or central vestibular dysfunction and its provoked functional consequence |
| FS2 | Oculomotor function | Recorded smooth pursuit gain, saccadic accuracy and latency, optokinetic response, gaze stability, convergence and accommodation assessment | Degree of impairment of visual tracking, fixation, and near-work tolerance |
| FS3 | Postural control and balance | Instrumented force-platform posturography across standardized stance conditions; composite stability score | Degree of postural instability and fall risk under progressively reduced sensory input |
| FS4 | Cognitive processing | Computerized testing of attention, processing speed, working memory, verbal and visuospatial memory, and executive tasks, normatively referenced | Degree of impairment of cognitive efficiency relative to normative expectation |
| FS5 | Cranial nerve and sensory function | Structured cranial nerve examination; visual field, pupillary, extraocular, facial sensory and motor, auditory and vestibulocochlear screening; dermatomal sensory testing | Degree of cranial nerve and sensory dysfunction |
| FS6 | Motor and strength function | Quantitative grip dynamometry with side-to-side comparison; segmental strength testing; coordination and fine motor performance | Degree of motor weakness, asymmetry, or incoordination |
| FS7 | Autonomic and headache function | Headache frequency, duration and provocation profile; photophobia and phonophobia documentation; orthostatic and exertional tolerance assessment | Degree to which headache, sensory sensitivity, and autonomic intolerance restrict sustained activity |
| FS8 | Psychological and behavioral function | Validated brief instruments for anxiety, depression, traumatic stress, and insomnia severity, scored to published cut-points43–46 | Degree of psychological and sleep dysfunction affecting neurological recovery and function |
| FS9 | Activities of daily living and participation | Validated functional-impact instrument covering work, social, and domestic functioning; documented work, school, and driving status | Degree of restriction in real-world participation |
Because the ST-CD Scale is a formative composite (Section 3.3), the defensibility of the instrument rests substantially on whether the domain set is complete and non-redundant. Each domain is therefore argued individually below: what it measures, why cTBI disrupts it, why it is separable from the others, and what its inclusion changes about the patient’s care.
The peripheral vestibular apparatus and its central connections through the vestibular nuclei, cerebellum, and vestibulo-ocular and vestibulospinal pathways are among the structures most vulnerable to the rotational and linear accelerations that produce concussion. Otoconial displacement generates post-traumatic benign paroxysmal positional vertigo; shear and inflammatory injury to central vestibular pathways produces gaze instability, motion intolerance, and impaired sensory reweighting.49,57,58 Vestibular dysfunction is quantifiable by caloric and positional testing, is among the most common objective findings after cTBI, and is directly rehabilitable through canalith repositioning, gaze stabilization, and habituation.33,48,59 It is separable from FS3 in that a patient may have a normal vestibular examination with impaired postural control from proprioceptive or central integrative causes, and the therapeutic implication differs accordingly.
Ocular motility depends on widely distributed circuitry — brainstem burst generators, cerebellar vermis, frontal and parietal eye fields, and the dorsolateral prefrontal cortex — which makes recorded eye movement an unusually sensitive index of diffuse network disruption.60 Convergence insufficiency, accommodative dysfunction, saccadic dysmetria, and reduced pursuit gain are documented in a substantial proportion of patients after cTBI, particularly adolescents, and account for much of the near-work and screen intolerance that limits return to school and to office-based work.60,61 The domain is separable from FS4 because oculomotor dysfunction impairs the delivery of visual information to cognitive systems that may themselves be intact, and separable from FS1 because central oculomotor abnormality occurs with entirely normal peripheral vestibular function.
Postural stability requires continuous integration of vestibular, visual, and somatosensory input with cerebellar and brainstem motor output, and the reweighting of those inputs when one becomes unreliable.20,21 Concussion impairs this integrative process rather than any single input, which is why instrumented posturography under progressively degraded sensory conditions detects dysfunction that observational balance testing does not.50,62,63 The domain carries direct safety implications — fall risk, driving, work at height, return to contact activity — and is separable from FS1 and FS6 because the deficit lies in integration rather than in any contributing channel.
The neurometabolic cascade following cTBI, together with disruption of long white-matter association tracts and of the frontoparietal and salience networks, produces a characteristic pattern of reduced processing speed, impaired sustained and divided attention, and reduced working-memory capacity.64–66 These deficits are frequently the limiting factor for occupational and academic function even when the patient is physically well, and they respond to structured cognitive rehabilitation and graded cognitive loading.34,67 Normatively referenced computerized testing makes the domain quantifiable at the individual level.51
The mechanism that produces cTBI frequently produces concurrent injury to cranial nerves at their skull-base exits and to cervical roots and peripheral nerves — olfactory disruption, trigeminal sensory change, facial weakness, auditory and vestibulocochlear involvement, and cervical radicular findings.35,68,69 These findings are separable from the diffuse network dysfunction indexed by FS2 and FS4, are localizable, and materially change the differential diagnosis and the referral pathway. Their exclusion would leave a category of objectively demonstrable neurological injury unrepresented in a scale that claims to measure neurological disability.
Motor findings after cTBI arise from several sources: corticospinal and cerebellar dysfunction, cervical and peripheral nerve injury sustained in the same event, and disuse. Quantitative dynamometry with side-to-side comparison, together with coordination and fine motor testing, provides an objective and reproducible measure with established normative reference values.52 The domain is retained despite the relative infrequency of severe motor findings in uncomplicated cTBI, because when such findings are present they are functionally decisive and because their objectivity is difficult to dispute.
Post-traumatic headache is the most prevalent persistent symptom after cTBI and has a distinct pathophysiology involving trigeminovascular activation and central sensitization.36 It travels with photophobia, phonophobia, and exertional and orthostatic intolerance, the last reflecting disturbance of autonomic cardiovascular regulation and of cerebral autoregulation.70,71 This domain frequently determines how long a patient can sustain any activity at all, and it responds to specific intervention — headache management and graded subthreshold aerobic reconditioning — that differs entirely from vestibular or cognitive rehabilitation.42 A scale omitting it would misclassify the patient whose objective testing is otherwise near-normal but who cannot tolerate more than an hour of activity.
Anxiety, depression, traumatic stress, and insomnia are not merely reactions to concussion; they are among the most consistently identified predictors of prolonged recovery, and insomnia in particular exerts a measurable effect on cognitive and vestibular performance.72–75 Each is measurable with a brief validated instrument scored against published cut-points, which makes the domain gradable by rule rather than by impression.43–46 Its inclusion is also a safeguard: without it, a patient whose disability is predominantly psychological and sleep-related would receive a low score from an instrument that had never looked.
The preceding eight domains measure impairment. This domain measures participation, and the ICF framework is explicit that the two are distinct and that neither predicts the other reliably.27 Its inclusion serves two functions within the scale. Substantively, it ensures that the construct defined in Section 3.1 — disability, not laboratory abnormality — is directly represented. Procedurally, it provides the independent check used in anchor reconciliation (Section 7.4), where participation is compared against the impairment-derived score and material disagreement between them is treated as a finding rather than smoothed away.
Domain grading thresholds are specified in the C-FNE scoring manual in terms of the measured quantity. Two illustrative conventions indicate the level of specificity intended. In the postural control domain, a composite posturographic stability score below the normative threshold constitutes an abnormal finding, with the magnitude of the deficit and the pattern of failure across sensory conditions determining whether the grade is 2, 3, or 4. In the psychological domain, grades follow the published severity cut-points of the constituent instruments rather than clinician impression, so that a moderate score on a validated anxiety instrument produces the same domain grade irrespective of examiner.
Reproducibility depends on procedures that are frequently left implicit. The following are specified and documented for every examination.
These procedures are stated here because they are the substance of what makes an examination standardized, and because a reviewer or opposing expert evaluating the framework will ask about each of them. They are specified in full in the examiner manual (Supplement B), which has not yet been published; until it is, the claim of standardization rests on internal documentation rather than on material an independent party can inspect.
A methodological feature of the C-FNE, and one of the principal arguments for a multidomain examination over any single test, is that the component assessments overlap physiologically and can therefore be cross-checked against one another. Every individual neurological test carries limitations and, standing alone, a degree of diagnostic uncertainty. Vestibular dysfunction identified on caloric testing predicts a characteristic pattern of posturographic failure; oculomotor abnormality on recorded pursuit predicts reduced performance on visually mediated cognitive tasks; cranial nerve findings should correspond to the sensory distribution reported.
When findings converge, confidence in the profile increases. When they diverge — for example, when severe self-reported dysfunction accompanies uniformly normal instrumented measurement, or when a single isolated abnormality appears without any physiologically related finding — the divergence is documented explicitly and the examination is interpreted with corresponding caution. Divergence is informative rather than embarrassing, and recording it is part of the method.
Incomplete examinations are common in this population and are a genuine methodological problem. Patients may be unable to tolerate caloric irrigation, may terminate positional testing because of provoked migraine, may fail oculomotor calibration, or may have cognitive testing deferred to a subsequent visit. Early versions of the framework treated missing measurements as normal, which systematically underestimated disability in precisely the most symptomatic patients — those too symptomatic to complete the battery.
The current framework requires that any measurement not obtained be recorded as Not Tested, with the reason stated. Not Tested is not a grade of 0, is not imputed, and is displayed on the report. The consequences for global scoring are specified in Section 7.7.
The C-FNE and the ST-CD Scale are distinct components with distinct failure modes, as set out in Section 3.5. The relationship is worth restating at the close of this chapter because it is the point most frequently lost: the C-FNE is the examination; the ST-CD Scale is what is done with its output. The examination has clinical value with no score attached — it identifies which systems are dysfunctional and directs treatment accordingly, and a clinician who performs it and never computes a global score has still learned what they needed to know. The scale, by contrast, has no meaning without a completed examination, cannot be estimated from records, and cannot be assigned by anyone who did not measure the patient (Section 9.6).
This asymmetry has a practical implication for how the framework should be evaluated. A reader who is unpersuaded by the scoring model may still find the examination protocol useful, and the two claims should be assessed separately.
The ST-CD Scale is a two-level ordinal instrument, structurally modeled on the EDSS.22 At the first level, nine Functional System Scores are graded 0 to 4 from the objective measurements described in Section 6.6. At the second level, those grades are converted, by explicit rule, into a single global score ranging from 0 to 10 in 0.5-point increments. Six interpretive category bands are layered over the numeric continuum to support communication with non-specialist readers.
Three features distinguish the ST-CD Scale from its antecedent. First, its Functional System Scores are derived from instrumented measurement rather than from bedside examination alone. Second, its global steps are anchored to a stated proportion of waking time during which function is materially impacted, rather than to ambulatory capacity, because ambulation is rarely the limiting factor after cTBI. Third, it is explicitly a short-term state measure rather than a longitudinal disease-course measure.
Each global step is defined by two statements: the proportion of the patient’s waking time during which neurological dysfunction materially affects function, and the corresponding capacity for work, study, driving, and independent daily activity. The time-based anchor exists to eliminate the interpretive drift that severity adjectives invariably produce. Two clinicians may disagree about whether a presentation is “moderate”; they are far less likely to disagree about whether a patient is functionally impacted for roughly a quarter or roughly half of the waking day, because that question is answerable from documented activity tolerance, work status, and examination findings.
Table 11. ST-CD global scale: steps, anchors, and category bands. Percentages refer to the proportion of waking time during which neurological dysfunction materially affects function.
| Score | Category band | Time impacted | Functional anchor |
|---|---|---|---|
| 0 | No disability | 0% | All nine functional systems within normative limits. No activity restriction. Full participation in work, study, driving, and physical activity. |
| 0.5 | Slight | <5% | Isolated measurable abnormality in one system without demonstrable functional consequence. Full participation maintained. |
| 1.0 | Slight | ≈5–10% | Measurable abnormality in one or more systems producing brief, infrequent limitation. Full duties maintained without accommodation. |
| 1.5 | Slight | ≈10% | Abnormality in more than one system with brief limitation. Occasional self-directed pacing; no formal accommodation required. |
| 2.0 | Mild | ≈15% | Consistent measurable dysfunction producing intermittent limitation. Full duties maintained but with increased effort or fatigue. |
| 2.5 | Mild | ≈15–20% | Intermittent limitation in more than one domain. Informal accommodation (breaks, reduced screen time) begins. |
| 3.0 | Mild | ≈20% | Intermittent limitation requiring documented accommodation for a portion of the day. Full-time duties maintained. |
| 3.5 | Mild | ≈25% | Limitation for approximately one quarter of the waking day. Duties maintained with restrictions; some activities deferred. |
| 4.0 | Moderate | ≈30% | Consistent dysfunction requiring modification of most daily activities. Reduced hours or restricted duties typically required. |
| 4.5 | Moderate | ≈35–40% | Modification required across multiple domains. Driving may be restricted to short familiar routes; sustained screen or reading work limited. |
| 5.0 | Moderate | ≈45% | Function materially impacted for approximately half of the waking day. Part-time or substantially modified duty; supervision of complex tasks. |
| 5.5 | Moderate | ≈50% | Half the waking day materially impacted. Independent in self-care; unable to sustain full occupational or academic demand. |
| 6.0 | Severe | ≈55–60% | Majority of the waking day impacted. Occupational or academic activity generally not sustainable; assistance required for complex tasks. |
| 6.5 | Severe | ≈65% | Multiple systems severely involved. Driving generally not advisable. Household and community activity substantially restricted. |
| 7.0 | Severe | ≈70% | Most of the waking day impacted. Independent in basic self-care but requires assistance with instrumental activities of daily living. |
| 7.5 | Severe | ≈75% | Severe multisystem dysfunction. Regular assistance required; activity tolerance limited to short intervals. |
| 8.0 | Severe | ≈80% | Function preserved only for brief intervals. Assistance required for most instrumental and some basic activities. |
| 8.5 | Extreme | ≈85% | Severe involvement across the majority of functional systems. Daily assistance required; independent community activity not possible. |
| 9.0 | Extreme | ≈90% | Near-continuous functional impact. Dependent for most activities of daily living; supervision required. |
| 9.5 | Extreme | ≈95% | Continuous functional impact across nearly all systems. Dependent for essentially all activities. |
| 10 | Extreme | 100% | Total functional dependence attributable to neurological dysfunction across all assessed systems. |
The upper range of the scale is included for completeness of the ordinal continuum and to preserve structural comparability with the EDSS. In practice, scores at the extreme end of the range are uncommon in cTBI and should prompt reconsideration of the diagnosis, evaluation for complicating pathology, or reassessment of examination validity.
The provisional global score is derived from the profile of Functional System Scores according to Table 12. The derivation is driven jointly by the highest domain grade recorded and by the number of domains involved, reproducing the EDSS principle that severe dysfunction in a single system and mild dysfunction across many systems can produce comparable overall disability, but by different routes.
Table 12. Derivation of the provisional ST-CD global score from the Functional System Score profile.
| Functional System Score profile | Provisional global score | Governing principle |
|---|---|---|
| All nine systems graded 0 | 0 | No measurable dysfunction |
| One system graded 1; all others 0 | 0.5 | Isolated subclinical abnormality |
| Two or more systems graded 1; none higher | 1.0–1.5 | Diffuse subclinical abnormality |
| One system graded 2; others 0–1 | 1.5–2.0 | Single-system functional involvement |
| Two systems graded 2 | 2.5–3.0 | Two-system functional involvement |
| Three or more systems graded 2, or one system graded 3 | 3.5–4.0 | Diffuse mild involvement or focal moderate involvement |
| Two systems graded 3 | 4.5–5.0 | Two-system moderate involvement |
| Three systems graded 3, or one system graded 4 | 5.5–6.0 | Diffuse moderate involvement or focal severe involvement |
| Four systems graded 3, or two systems graded 4 | 6.5–7.0 | Extensive moderate or dual severe involvement |
| Three systems graded 4 | 7.5–8.0 | Multisystem severe involvement |
| Four or more systems graded 4 | 8.5–9.0 | Widespread severe involvement |
| Six or more systems graded 4 with dependence for basic activities of daily living (FS9 = 4) | 9.5–10 | Global functional dependence |
Where a profile maps to a range of two adjacent half-steps, the higher value is selected when the participation domain (FS9) exceeds the median grade of the remaining domains, and the lower value is selected when it falls below. This rule operationalizes the ICF principle that the global score is a statement about function, not a summary of laboratory abnormality.27
The provisional score derived from Table 12 is then compared against the time-and-capacity anchor in Table 11. Where the derived score and the anchor agree, the derived score stands. Where they disagree by half a step, the anchor governs. Where they disagree by a full step or more, the discrepancy is treated as a signal that either the domain grades or the functional history requires re-examination, and the reason for the final score is documented explicitly in the report.
This reconciliation step exists because the two levels of the instrument measure the construct from different directions — upward from measured impairment and downward from observed participation — and material disagreement between them is clinically meaningful information rather than a defect to be smoothed over.
The derived and reconciled value is a recommendation. The final ST-CD score is assigned by the licensed clinician performing or supervising the evaluation. When the assigned score differs from the derived value, the clinician records both values and the reason for the departure. Acceptable reasons include documented pre-existing dysfunction affecting one or more domains, concurrent non-neurological injury independently limiting participation, medication effects, examination validity concerns, and developmental or age-related factors affecting the applicability of normative reference data.
Requiring the departure to be written down is the point. It converts what would otherwise be unexplained variability into an auditable clinical judgment, and it makes systematic examiner drift detectable.
Severity adjectives were removed from the global scale in Phase IV for the reasons given in Section 4.6, and the same reasoning applies to domain grading. Each of the five domain grades is therefore defined operationally — by what must be demonstrable for the grade to be assigned — rather than by an adjective the examiner is left to interpret.
Table 13. Operational definitions of the five domain grades. These definitions apply across all nine functional systems; the measurement thresholds that trigger each grade are domain-specific and are stated in the scoring manual.
| Grade | Label | What must be demonstrable | What must NOT be present |
|---|---|---|---|
| 0 | Within normative limits | All measurements in the domain fall within the applicable normative reference range | — |
| 1 | Measurable, functionally silent | At least one measurement falls outside normative limits | No demonstrable consequence for any activity; no accommodation of any kind; no activity avoided or modified |
| 2 | Intermittent limitation | Abnormal measurement plus documented limitation occurring some of the time, under specific conditions, or on provocation | Function is not limited under ordinary conditions; no sustained modification of activity is required |
| 3 | Consistent limitation | Abnormal measurement plus limitation present under ordinary conditions, requiring documented modification of activity, accommodation, or restriction | The function is not lost; assistance from another person is not required |
| 4 | Function lost or assistance required | Abnormal measurement plus loss of the function, or dependence on another person or on an assistive device to perform it | — |
| NT | Not Tested | The measurement was attempted or indicated but not obtained; the reason is recorded | NT is never assigned where a valid measurement exists; NT is not a grade of 0 and is never imputed |
Two features of these definitions are deliberate. First, grades 2 through 4 each require both an abnormal measurement and a demonstrable functional consequence; an abnormal measurement alone can never exceed grade 1. This prevents the accumulation of physiologically minor abnormalities into a high score. Second, the “what must not be present” column is included because upward drift across adjacent grades is the most common form of scoring error observed during development, and exclusion criteria constrain it more effectively than inclusion criteria alone.
Not Tested is a single status but not a single situation, and the reasons differ in what they imply about the patient. The reason is therefore classified and recorded, because a measurement missing for a technical reason carries no clinical information while a measurement missing because the patient could not tolerate it carries a great deal.
Table 14. Classification of reasons a measurement may be unobtainable, and the handling rule for each.
| Class | Examples | Handling | Bearing on other domains |
|---|---|---|---|
| Technical failure | Equipment malfunction, calibration failure, software error, recording artifact | NT recorded with cause; repeat measurement scheduled where clinically indicated | None. Carries no information about the patient |
| Patient intolerance | Provoked migraine terminating positional testing; refusal of caloric irrigation after symptom exacerbation; inability to keep eyes open | NT recorded with cause and with the point in the battery at which testing stopped | Bears on FS7 and FS9 and is documented there; does not generate a grade in the untested domain |
| Competing impairment | Pain limiting strength testing; visual impairment limiting oculomotor testing; wheelchair use limiting posturography | NT recorded with the specific limiting factor named | The limiting impairment is documented; attribution to the index injury is a separate determination (Section 9.3) |
| Deferral | Cognitive testing deferred to a subsequent visit; examination performed within 24–48 hours of injury with testing intentionally postponed | NT recorded with the scheduled reassessment date | None, provided the reassessment occurs and is reported |
| Communication barrier | Language, hearing, or literacy barrier affecting instructed tasks or self-report instruments | NT recorded; where an interpreter or translated instrument is available and validated, testing is repeated with it | Normative applicability is questionable even where a score is obtained; documented as a validity caveat |
| Developmental or pre-morbid factor | Learning disability, attention-deficit disorder, or established cognitive impairment affecting the applicability of normative data | A grade may be assigned where the pre-morbid baseline is documented; otherwise NT with reason | Documented as a confounder affecting the specific domain (Section 9.3) |
| Effort or validity concern | Performance inconsistent with observed function; internally contradictory findings; measurements inconsistent with physiological expectation | The measurement is recorded and flagged as of uncertain validity; a domain grade is not assigned on data flagged invalid | Requires formal performance validity assessment; see Sections 9.4 and 9.6 |
| Medication effect | Sedating medication, vestibular suppressants, or stimulants active at the time of examination | Measurement recorded with the medication and timing documented; grade assigned with the caveat stated | Documented as a confounder; repeat testing considered if the effect is likely material |
The general principle behind these rules is that missing data are recorded, classified, and made visible rather than resolved. Imputation would produce a complete-looking score from an incomplete examination, and the reader of the report would have no way of knowing which had occurred.
The following rules govern global scoring when domains carry Not Tested status.
A composite score aggregates information, and aggregation carries a characteristic risk: the composite acquires an apparent authority that its components did not individually possess, and the components are subsequently read through the composite rather than on their own terms. The ST-CD Scale is explicitly constructed to resist this, and the following principle governs its interpretation.
The component findings retain their individual clinical meaning and are reported independently of the global score. An abnormal instrument-derived result is not rendered more reliable by the fact that it contributes to the ST-CD Scale, and a normal result is not diminished by the fact that other domains are abnormal. Each component must be interpreted according to its own measurement characteristics, published evidence base, normative reference population, technical limitations, and relevance to the individual examined. A caloric asymmetry means what the vestibular literature says it means, whether the patient’s global score is 2.0 or 8.0. A normal cognitive battery means what the normative data say it means, whether or not five other domains are abnormal.
Two consequences follow. First, the global score adds information about pattern and magnitude; it subtracts nothing and overrides nothing. Where a component finding and the global score appear to point in different directions, the component finding is not the error to be corrected — the divergence is itself the finding, and Section 6.8 requires that it be documented.
Second, and stated as a requirement rather than a recommendation: the ST-CD global score is not reported in isolation. Every report of an ST-CD score includes the complete Functional System profile and the measurements from which each domain grade was derived. A score transmitted without its underlying profile — in a referral note, a summary letter, a claims record, or an expert report — has been stripped of the information that gives it meaning and should be treated by the recipient as incomplete rather than as a finding.
Reporting rule. An ST-CD global score is reported only together with (a) the nine Functional System grades, (b) the objective measurements supporting each grade, (c) all domains carrying Not Tested status and the reason for each, and (d) any departure of the assigned score from the derived score with the clinician’s stated reason. A global score presented without these elements is not a complete report of the instrument’s output.
The following two examples are constructed illustrations created for this manuscript to demonstrate the mechanics of derivation. They do not represent actual patients and contain no patient data.
Example A — Focal vestibular presentation
An adult presents six weeks after a rear-impact collision. Videonystagmography demonstrates positional nystagmus consistent with posterior canal involvement and a substantial unilateral caloric weakness; the patient reports provoked vertigo on positional change several times daily and has stopped driving (FS1 = 3). Posturography demonstrates a composite stability score well below the normative threshold with failure concentrated in conditions requiring vestibular input (FS3 = 3). Recorded oculomotor testing is normal (FS2 = 0). Cognitive testing is within normative limits (FS4 = 0). Cranial nerve, sensory, and motor examinations are normal (FS5 = 0, FS6 = 0). Headache is present but infrequent and does not limit activity (FS7 = 1). Psychological screening instruments are below clinical cut-points (FS8 = 0). The patient has returned to work at full hours but has ceased driving and avoids overhead and rapid positional activity (FS9 = 2).
Two systems are graded 3, which maps to a provisional global score of 4.5 to 5.0. FS9 (grade 2) falls below the median grade of the remaining involved domains, selecting the lower value: 4.5. The anchor for 4.5 — modification required across multiple domains, driving restricted — corresponds to the documented history, so no reconciliation adjustment is required. Final score: ST-CD 4.5, Moderate disability, driven by FS1 and FS3.
Example B — Diffuse multisystem presentation
An adult presents three months after a side-impact collision. Recorded oculomotor testing demonstrates reduced pursuit gain and saccadic dysmetria with near-work intolerance (FS2 = 3). Cognitive testing demonstrates impairment of processing speed and working memory relative to normative expectation, with corresponding difficulty completing occupational tasks (FS4 = 3). Posturography is abnormal with moderate deficit (FS3 = 2). Vestibular testing demonstrates positional findings without caloric asymmetry (FS1 = 2). Post-traumatic headache occurs most days with marked photophobia limiting screen tolerance to short intervals (FS7 = 3). Psychological instruments demonstrate moderate anxiety, moderate depression, and severe insomnia (FS8 = 3). Cranial nerve and motor examinations are normal (FS5 = 0, FS6 = 0). The patient is working reduced hours with accommodations and requires assistance with complex household management (FS9 = 3).
Four systems are graded 3 (FS2, FS4, FS7, FS8), with an additional two graded 2, mapping to a provisional global score of 6.5 to 7.0. FS9 is graded 3, equal to the median of the involved domains, and the anchor comparison is therefore decisive: the documented capacity — reduced hours sustained, independent in basic self-care, assistance required for instrumental activities — corresponds to the 6.5 anchor rather than the 7.0 anchor. Final score: ST-CD 6.5, Severe disability, driven by FS2, FS4, FS7, and FS8.
The two examples arrive at different scores by different routes and, importantly, generate different treatment priorities: Example A directs canalith repositioning and vestibular rehabilitation, whereas Example B directs oculomotor therapy, cognitive rehabilitation, headache management, and behavioral health intervention. This is the practical argument for reporting the Functional System profile alongside the global score rather than the global score alone.
The relationship among the components of this framework is hierarchical, and stating it explicitly prevents the most common category error — treating a score as a diagnosis.
| Level | Statement |
|---|---|
| Diagnosis | Concussion (mild traumatic brain injury) is the clinical diagnosis, established by accepted diagnostic criteria on the basis of mechanism, acute clinical features, and clinical course.76 |
| Characterization | The C-FNE is the comprehensive objective assessment that characterizes which neurological systems are dysfunctional and to what degree. |
| Quantification | The ST-CD Scale is the standardized quantitative expression of the patient’s current functional neurological disability, derived from the C-FNE. |
| Intervention | Treatment planning is directed by the specific domain deficits identified by the C-FNE and summarized by the ST-CD score. |
| Long-term assessment | Permanent impairment assessment, when appropriate after maximal medical improvement, builds upon the documented disability trajectory rather than replacing it. |
The ST-CD Scale does not diagnose concussion, does not establish causation, and does not substitute for clinical diagnosis. A patient may carry a high ST-CD score for reasons unrelated to the index injury; the score quantifies disability, and attribution of that disability remains a separate clinical determination requiring history, temporal sequence, and consideration of alternative explanations.
The principal clinical value of the framework lies not in the global score but in the Functional System profile that produces it. Contemporary concussion management is domain-specific and increasingly active rather than rest-based.32,77,78 A profile in which the vestibular and postural domains carry the highest grades directs canalith repositioning, gaze stabilization, habituation, and balance retraining. A profile dominated by the oculomotor domain directs vergence and oculomotor rehabilitation and near-work pacing. A profile dominated by the cognitive domain directs cognitive rehabilitation, graded cognitive loading, and academic or occupational accommodation.34 A profile dominated by the autonomic and headache domain directs headache management and graded subthreshold aerobic reconditioning.42 A profile dominated by the psychological domain directs behavioral health intervention and sleep management, which are not adjunctive considerations but determinants of the recovery trajectory.43–46,72
Two patients with an identical global score of 5.0 may therefore require entirely different treatment programs. Reporting the global score without the domain profile discards the information that determines care.
The ST-CD Scale is designed for repeated administration. Serial scores across the recovery period generate a trajectory that is more informative than any single measurement, because it distinguishes improving, plateaued, and deteriorating courses. A patient improving from 6.5 to 4.5 to 2.5 across successive evaluations is following an expected recovery course; a patient stable at 5.5 across three evaluations separated by two months each is not, and that observation should prompt reconsideration of the treatment plan, evaluation for unrecognized contributing pathology, and assessment of the psychological and sleep domains, which are among the more consistently identified modifiers of prolonged recovery.72
An important caution applies. The magnitude of change that constitutes real clinical improvement rather than measurement variability — the minimal clinically important difference — has not been established for this scale. Until it is, changes of a single half-step should not be interpreted as meaningful, and clinical decisions should rest on the direction and consistency of the trajectory together with the underlying domain data rather than on small numeric differences.
Return-to-activity decisions are among the most consequential and least standardized determinations in concussion care. The ST-CD framework contributes to these decisions in two ways. The global score and its anchor provide a defensible statement of overall capacity that corresponds directly to the language of work restriction — full duty, modified duty, reduced hours, off work. The domain profile identifies which specific demands are problematic, allowing restrictions to be written precisely rather than generically: a patient whose limitation is concentrated in the oculomotor and cognitive domains may safely perform physical work while being unable to sustain screen-based tasks, and a generic restriction would serve that patient poorly in both directions.
The framework informs but does not replace return-to-play or return-to-duty protocols, which incorporate graduated exertional progression and sport-specific or duty-specific criteria beyond the scope of a disability scale.32
A single documented score with a defined anchor allows a physician, physical therapist, occupational therapist, neuropsychologist, case manager, employer, and patient to hold the same conversation. This is the practical benefit that the EDSS, NIHSS, and mRS delivered in their respective fields, and it is achieved by the existence of a shared vocabulary rather than by any property of the numbers themselves.22–24
Reports generated from the framework state the global score, the category band, the nine Functional System grades, the measurements underlying each grade, all Not Tested domains with reasons, the derived score where it differs from the assigned score, and the clinician’s reason for any departure. The requirement that the global score never travel without this supporting profile is stated in Section 7.8 and applies to every context in which the score is communicated, including referral notes and summary correspondence where the temptation to transmit the number alone is greatest.
A substantial proportion of concussion evaluations occur in contexts where the findings will be reviewed by parties with adverse interests. Addressing this openly is preferable to pretending otherwise, and the framework was designed with these settings in mind. That design intent also creates specific obligations, which this section states.
Concussion presentations are frequently characterized as subjective. Where an evaluation rests on symptom report alone, that characterization has force. Where a structured objective examination has been performed, it does not. The relevant sequence is that subjective symptoms prompt a medically appropriate evaluation, and objective instrumented measurement then independently characterizes the neurological dysfunction. Recorded eye-movement data, caloric responses, force-platform center-of-pressure measurements, normatively referenced computerized cognitive performance, and quantitative dynamometry are measurements obtained by calibrated equipment, not self-report.
This does not render such measurements beyond challenge. It relocates the challenge to questions that can be examined scientifically: whether the equipment was calibrated, whether the protocol was followed, whether normative reference data are applicable to this patient, whether effort was adequate, and whether the findings are internally consistent. Those are the correct questions, and a well-documented evaluation should be able to answer all of them.
As stated in Section 2.4, the ST-CD score is a short-term disability measure and is not a permanent impairment rating. Reports should state this explicitly, because a numeric score presented without that qualification invites conversion into a permanency conclusion by readers unfamiliar with the distinction. Where a permanent impairment determination is subsequently required, it must be performed after maximal medical improvement using the applicable impairment schedule; the serial ST-CD record then serves as documentation of the course of recovery preceding that determination.37
The ST-CD Scale measures current disability without regard to cause. Where pre-existing dysfunction is documented — prior vestibular disorder, prior brain injury, attention-deficit disorder, pre-existing anxiety, depression, or insomnia, degenerative cervical disease, prior stroke, or age-related decline — the affected domains must be identified and the pre-existing contribution described. Where possible, the score should be reported alongside an explicit statement of which domain grades are affected by pre-existing conditions and whether the clinician considers the post-injury presentation to represent new dysfunction, exacerbation of prior dysfunction, or unchanged prior dysfunction.
This is a clinical judgment, not a computation, and the framework does not generate it. Presenting an apportioned score as though the instrument produced it would misrepresent the method.
Performance-based measurement is not immune to suboptimal effort, and the incentive structures present in disability and litigation contexts are well documented in the neuropsychological literature.79,80 Several features of the framework provide partial protection: physiological measures such as caloric response and positional nystagmus are not under voluntary control; cross-domain triangulation (Section 6.8) makes internally inconsistent profiles detectable; and divergence between self-report and instrumented measurement is documented rather than reconciled silently.
Partial protection is not complete protection. The framework does not currently incorporate formal performance validity or symptom validity testing, and it should not be represented as containing an effort measure. Where validity is a material question, formal performance validity testing administered and interpreted by a qualified examiner is the appropriate instrument, and this limitation is restated in Section 11.0.79,80
Reports intended for review outside the treating relationship should contain the elements listed in Table 15. The list is deliberately specific: each element corresponds to a question that a competent reviewer will ask, and a report that omits an element invites the inference that the answer would have been unfavorable. Where an element cannot be supplied, the report should say so and state why.
Table 15. Minimum reporting dataset for an ST-CD evaluation intended for review outside the treating relationship.
| Element | Content required | Question it answers |
|---|---|---|
| Examination context | Date of examination; interval from date of injury; referral source; examination setting | When in the recovery course does this measurement sit? |
| Examiner | Identity, credentials, and training of the individual performing and of the individual interpreting the examination | Was the examination performed and interpreted by someone qualified to do so? |
| Instrumentation | Devices used, software versions, calibration status and date, and the normative reference dataset applied to each measure | Are the measurements technically sound and appropriately referenced? |
| Protocol adherence | Test sequence used; any deviation from the standard sequence with reason | Were the conditions of measurement standardized? |
| Functional System profile | All nine domain grades with the specific measurement supporting each | What was actually found, domain by domain? |
| Not Tested domains | Every domain not assessed, with the reason and whether the omission would plausibly have raised the score | What is missing, and does its absence bias the result? |
| Score derivation | Derived score, assigned score, and the clinician’s stated reason for any departure between them | How was the number reached, and who is accountable for it? |
| Internal consistency | Convergent and divergent findings across physiologically related domains (Section 6.8) | Do the findings cohere, and where they do not, was that noticed? |
| Confounders | Documented pre-existing conditions, concurrent injuries, medications, and comorbidities, with the domains each affects | What else could explain these findings? |
| Scope statement | Explicit statement that the score reflects current short-term disability, is not a permanent impairment rating, and does not establish causation | What does this number not mean? |
The ST-CD Scale may be encountered in clinical, administrative, insurance, disability, and legal settings, because cTBI-related functional limitations frequently affect employment, education, driving, independent activity, and claims arising from traumatic injury. Its use in such settings requires careful distinction between the underlying clinical evidence and the investigational status of the composite scale. That distinction is not a formality: the individual measurements rest on established literature and validated instruments, whereas the method of combining them does not yet rest on published psychometric evidence, and the two should not be presented as though they carried equivalent evidentiary weight.
The ST-CD global score does not replace the patient’s history, physical and neurological examination, medical records, differential diagnosis, or the individual measurements from which the Functional System Scores are derived. It is a structured method of organizing and communicating those findings. The clinical significance of an ST-CD score must therefore be evaluated by examining the complete evidentiary foundation, including the instruments used, testing conditions, normative comparisons, quality-control procedures, identified confounders, internal consistency of the findings, and the clinician’s documented reasoning — which is precisely the material specified in Table 15.
The scale has not undergone sufficient independent psychometric evaluation to establish a known error rate, minimal clinically important difference, diagnostic sensitivity or specificity, or predictive accuracy for long-term outcome, and it has never been evaluated as an instrument for determining causation. It must not be represented as having established performance in any of those areas; the full statement of what is not known appears in Section 11.0. Until such evidence is available, the scale is most appropriately used as a transparent, reproducible framework for describing the pattern and estimated magnitude of current functional neurological disability and for tracking change across serial examinations.
Qualification to administer and interpret. Administration and interpretation require a clinician trained in the constituent assessments and in the scoring procedure, working within the scope of an appropriate professional license. The scale is not intended for independent interpretation by non-clinical reviewers, claims personnel, or attorneys, and a score cannot be validly assigned, revised, or rebutted by a reviewer who has not examined the patient and who is working from records alone. Structured neurological scoring instruments have consistently required rater training to achieve acceptable agreement, and there is no reason to expect this one to be an exception.23,28
Retrospective assignment. The ST-CD Scale is a measurement instrument, not a chart-review rubric. A score may be assigned only from a contemporaneously performed C-FNE. It may not be reconstructed retrospectively from records, imputed from symptom documentation, estimated from a partial dataset, or applied to a historical presentation the examiner did not measure. A number generated by any of those routes is not an ST-CD score and should not be described as one.
Causation. In an individual case, the ST-CD score does not independently establish that a particular event caused the measured abnormalities. Causation remains a separate clinical determination requiring consideration of the temporal relationship, injury mechanism, contemporaneous symptoms, prior neurological and psychological history, intervening events, alternative diagnoses, consistency across records, examination findings, and the overall weight of the medical evidence.37,76 The instrument measures the magnitude and pattern of disability present at the time of examination; the reasoning that connects that disability to a specific event is the clinician’s, is made on the record, and is subject to the same scrutiny as any other clinical opinion.
The following statements define the limits of appropriate application and are intended to be read as constraints rather than as caveats.
This is the principal limitation and it is not a minor one. No inter-rater reliability, test–retest reliability, internal structure, convergent or discriminant validity, responsiveness, minimal clinically important difference, or predictive validity data are available for the ST-CD Scale. The framework is internally coherent and derived from established scale-design principles, but internal coherence is not evidence of measurement validity. Until the studies described in Section 12.0 are completed, the scale should be regarded as a structured documentation framework with plausible measurement properties rather than as a validated outcome measure.
The framework was developed within a single multidisciplinary practice, by a small group of clinicians, in a population weighted toward motor vehicle collision injury presenting in the subacute period. Thresholds, anchors, and internal consistency rules reflect that population and that clinical culture. Generalizability to sport-related concussion presenting acutely, to military populations, to emergency department populations, and to pediatric populations is unestablished. Multi-site application by independent examiners is a prerequisite for any general claim.
The ST-CD Scale is ordinal. The interval between 2.0 and 2.5 is not necessarily equivalent to the interval between 7.0 and 7.5, and the numbers should not be treated as continuous measurements for arithmetic purposes. This limitation is shared by the EDSS and by ordinal disability scales generally, and it has been the subject of sustained methodological criticism.29,40 Parametric analysis of ST-CD scores without prior demonstration of interval properties, for example through Rasch analysis, would be inappropriate.40
The derivation rules in Table 12 determine how domain grades combine into a global score, and those rules were established by clinical consensus informed by the EDSS structure, not by empirical weighting derived from outcome data. It is entirely possible that empirical analysis would demonstrate that certain domains warrant greater weight in predicting functional outcome than the current rules assign. Establishing empirically derived weights is a specific objective of the validation program.
Domain grades depend on instrumented measurements whose normative reference values are device-specific and, in several domains, derived from populations that may not match the patient being examined. Equivalence across manufacturers and platforms has not been demonstrated. Where normative data are absent or inapplicable — as is frequently the case in pediatric vestibular and posturographic testing — the resulting grade carries greater uncertainty than the numeric output implies, and this should be documented.81
As stated in Section 9.4, the framework does not incorporate performance validity or symptom validity measures. In populations where secondary gain is a material consideration, this is a substantive gap, and integration of established validity indicators is a priority for subsequent versions.79,80
Cervical injury, peripheral vestibular disease unrelated to trauma, pre-existing psychiatric illness, chronic pain, medication effects, obstructive sleep apnea, and primary headache disorders all influence domain grades. The framework documents these but does not disentangle them, and no algorithmic adjustment for comorbidity exists.
This section states the studies required to establish whether the ST-CD Scale measures what it is intended to measure. It is written in the detail that it is because the credibility of an instrument-development paper rests less on what its authors claim than on whether they have specified, in advance and in terms that permit failure, what would have to be shown.
The programme follows accepted standards for the development and evaluation of clinical outcome assessments.38,39,41,82–86 Within that terminology the ST-CD Scale is a clinician-reported outcome measure (ClinRO): its output is generated by a trained clinician applying defined criteria to observed and instrumented findings, rather than reported directly by the patient. That classification determines which measurement properties are relevant — inter-rater reliability is central, and the internal-consistency statistics appropriate to reflective patient-reported instruments are not applicable, for the reasons set out in Section 3.3.87,88
Thirteen studies are specified across six stages. Each is stated with its design and its acceptance criterion, so that failure to meet a criterion is as informative as success and cannot be absorbed by post hoc reinterpretation.
Table 16. Specified validation studies, designs, and pre-stated acceptance criteria.
| # | Study | Design | Acceptance criterion | Stage |
|---|---|---|---|---|
| 1 | Inter-rater reliability | ≥100 consecutive patients scored independently by two blinded examiners from identical recorded examination data | ICC ≥ 0.80 (global); weighted κ ≥ 0.60 (each domain)89,90 | 1 |
| 2 | Test–retest reliability | Repeat examination at 7–14 days in clinically stable patients; interval chosen to limit both practice effects and genuine recovery | ICC ≥ 0.80; SEM reported to permit calculation of minimal detectable change | 1 |
| 3 | Component structure | Principal component analysis of domain grades to test non-redundancy of the domain set | No domain loading redundantly; no domain contributing negligibly across the sample | 2 |
| 4 | Rasch analysis | Rating-scale model applied to the ordinal structure; step ordering, fit statistics, and person–item targeting examined55,91 | Ordered thresholds across all steps; infit/outfit within conventional limits; no disordered categories | 2 |
| 5 | Differential item functioning | DIF analysis across age band, sex, injury mechanism, and litigation status | No substantial DIF on any domain across any grouping variable | 2 |
| 6 | Construct validity | A priori hypotheses stated in advance regarding direction and magnitude of correlation with GOS-E, mRS, and functional status31,92,93 | ≥75% of pre-stated hypotheses confirmed | 3 |
| 7 | Convergent validity | Correlation with instruments measuring related constructs within each domain | Moderate-to-strong correlation with domain-matched measures | 3 |
| 8 | Discriminant validity | Correlation with symptom-report instruments, hypothesized in advance to be moderate rather than strong | Correlation with symptom inventories materially lower than with function-based measures; a high correlation would falsify the construct claim in Section 3.1 | 3 |
| 9 | Known-groups validity | Comparison across groups expected to differ: full duty vs. off work; within vs. beyond expected recovery window | Statistically and clinically meaningful separation between groups | 3 |
| 10 | Responsiveness | Longitudinal measurement across recovery; effect size and standardized response mean calculated | SRM ≥ 0.80 in patients independently rated as improved | 4 |
| 11 | Minimal clinically important difference | Anchor-based estimation using patient- and clinician-rated global change, supported by distribution-based estimates39,82 | A defensible MCID established with confidence interval reported | 4 |
| 12 | Predictive validity | Prospective cohort; baseline score and profile tested against return to work or study, symptom persistence beyond three months, and function at 6 and 12 months, adjusted for established prognostic factors72,74,75 | Independent predictive contribution beyond established prognostic variables | 5 |
| 13 | Multicentre external validation and independent replication | Application across independent practices, examiners, instrumentation platforms, and referral populations, including non-litigated and sport-related cohorts, led by investigators with no involvement in development | Reliability and validity estimates within the confidence intervals of the development-site findings | 6 |
Inter-rater reliability is the first and most consequential study, because the framework deliberately retains clinician judgment at two points (Sections 7.4 and 5.5). That decision was defended in Section 7.5 on the grounds that it converts clinician variance into something measurable; the obligation created by that defence is to measure it. Scoring will be performed independently by two examiners blinded to one another and to the treating clinician’s assignment, working from identical recorded examination data so that measurement variance and scoring variance are separated. Agreement will be reported at the level of the global score and separately for each of the nine domains, following established reporting guidance for reliability studies.89,90
Test–retest reliability presents a design tension specific to a state measure: the instrument is intended to detect change, so any interval long enough to eliminate practice effects is long enough to permit genuine recovery. The 7- to 14-day interval in clinically stable patients is a compromise, and the resulting estimate will therefore be a lower bound on stability rather than a pure reliability coefficient. Stating this in advance is preferable to reporting an attenuated coefficient later as though it were unexpected.
Structural analysis of a formative composite asks different questions from those asked of a reflective scale, and Section 4.3 sets out why. The relevant questions are whether each domain contributes non-redundantly, whether the ordinal steps of the global scale are ordered as intended and are used across their range, and whether the instrument behaves equivalently across subgroups. Rasch analysis addresses the second and, if thresholds prove ordered and fit acceptable, may support an interval transformation that would relieve the analytic constraints stated in Section 3.4.55,91 If thresholds prove disordered, the correct response is to collapse or redefine the affected steps, and that possibility is accepted in advance.
The discriminant validity study deserves particular emphasis because it is constructed to be capable of falsifying the central claim of this manuscript. Section 1.2.2 and Section 2.3 assert that functional disability and symptom burden are distinct constructs. If ST-CD scores were to correlate strongly with symptom inventory totals, that assertion would be wrong, and the instrument would be an expensive route to information obtainable from a questionnaire. The hypothesis is stated in advance, in that direction, for that reason.
Convergent and known-groups hypotheses will likewise be registered before data collection, with the expected direction and approximate magnitude of each correlation specified. Post hoc hypothesis generation from an observed correlation matrix is not validation.
Until the minimal clinically important difference is established, the interpretive caution in Section 8.3 stands: single half-step changes should not be treated as meaningful, and clinical decisions should rest on trajectory and on the underlying domain data. The MCID study is placed at Stage 4 rather than earlier because it depends on a stable and reliable instrument; estimating a meaningful difference before reliability is known would produce a figure without a defensible interpretation.
The final stage is the one the authors are least able to conduct and the one that matters most. The framework was developed within a single practice by a small group of clinicians, in a population weighted toward motor vehicle collision injury presenting subacutely (Section 11.2). Every estimate produced by Stages 1 through 5 at that site will carry that limitation. Independent replication by investigators with no involvement in development, using different instrumentation and different referral populations, is the only evidence that would establish the instrument as generalizable, and the authors regard it as a condition of general adoption rather than an optional extension.
To that end, the scoring manual, examiner manual, and derivation rules are available to independent investigators on request, and the authors undertake to publish replication findings irrespective of direction.
The present manuscript is the first in a planned sequence. The sequence deliberately mirrors the pattern by which established clinical instruments were introduced and subsequently validated: description first, comparison and validation afterward.
| Paper | Subject | Purpose |
|---|---|---|
| 1 | Development of the C-FNE and the ST-CD Scale (this manuscript) | Establishes the construct, measurement model, development history, design decisions, architecture, and intended application |
| 2 | The C-FNE compared with structural and diffusion imaging | Examines the complementary roles of functional assessment and neuroimaging in characterizing concussion |
| 3 | Reliability and structural evaluation of the ST-CD Scale | Reports Stages 1 and 2 of the roadmap |
| 4 | Validity, responsiveness, and predictive performance | Reports Stages 3 through 5 |
| 5 | Clinical outcomes of ST-CD-directed treatment | Evaluates whether domain-directed treatment guided by the framework improves outcomes |
Concussion medicine possesses an extensive array of objective assessment technologies and no standardized method for integrating their output into a statement of how disabled the patient actually is. The consequence is that clinicians identify impairments accurately and characterize disability inconsistently, and that the patient’s functional status is described differently by every provider who encounters it.
The ST-CD Scale was developed to address that gap. It applies an architecture with a long record in neurology — objectively graded functional systems aggregated into an anchored global ordinal score — to the specific pattern of multisystem dysfunction that follows concussion. Its inputs are instrumented measurements rather than symptom report; its outputs are anchored to real-world functional capacity rather than to severity adjectives; its derivation is written down rather than held as clinical impression; and its scope is explicitly bounded to the short-term recovery period before maximal medical improvement.
What the scale offers today is a structured, reproducible, and auditable method for describing short-term concussion disability, developed through sustained clinical application and successive correction. What it does not yet offer is psychometric evidence that it measures what it is intended to measure, with acceptable reliability, adequate responsiveness, and useful predictive value. That evidence is obtainable, the studies required to obtain it are specified in Section 12.0, and the appropriate description of the instrument until those studies are complete is investigational.
The broader argument, however, does not depend on this particular scale. Concussion disability is measurable, it should be measured objectively and multidimensionally, and the field would benefit from a common language for describing it. If the ST-CD Scale proves to be the wrong instrument, the case for building a better one stands.
Ethics approval. This manuscript describes the conceptual development, architecture, and intended application of a clinical assessment framework. It reports no patient-level data, no human subjects research, and no analysis of clinical outcomes. The illustrative scoring examples in Section 7.7 are constructed for demonstration and do not represent identifiable individuals. Institutional review board approval was therefore not required. All studies proposed in Section 12.0 will require prospective ethics approval and informed consent.
Consent for publication. Not applicable; no identifiable patient information is reported.
Data availability. No datasets were generated or analyzed for this manuscript. The C-FNE scoring manual and the ST-CD derivation rules are available from the corresponding author for research and non-commercial educational use.
Funding. No external funding was received for the preparation of this manuscript.
Competing interests. The authors are clinicians and officers of the practices within which the C-FNE and the ST-CD Scale were developed and are used clinically. See the trademark notice below. The authors therefore have a professional and commercial interest in the framework described. This disclosure is made in full, and readers should weigh the descriptive claims in this manuscript accordingly. The absence of independent validation, disclosed throughout, is the appropriate context in which to evaluate any claim made here.
Trademark notice. C-FNE™ (Comprehensive Functional Neurological Evaluation) is a trademark of NeuroLyteDx and is used in this manuscript solely to identify the standardized examination framework described herein. No claim is made to any term other than the designation itself, and the abbreviated form C-FNE is used without the symbol throughout the remainder of the manuscript in accordance with scientific publishing convention.
Author contributions. W.C. conceived the framework, developed the examination and scoring architecture, and drafted the manuscript. I.D. contributed to clinical development, examination protocol design, and manuscript revision. A.G. and C.R. contributed clinical and medical oversight and critical revision. All authors reviewed and approved the final manuscript and accept responsibility for its content.
Statement on scale status. The ST-CD Scale is an investigational clinical instrument. It has not been cleared or approved by any regulatory authority, is not a substitute for clinical judgment, and should not be used as the sole basis for diagnostic, treatment, employment, disability, or legal determinations.
The material required to reproduce an ST-CD score exceeds what a journal article can accommodate, and compressing it into the main text would displace the scientific argument without making the instrument reproducible. It is therefore issued as supplementary material. The controlling question a reviewer will ask of any new scale is whether an independent examiner could produce the same score from the same patient; the supplements exist to make the answer to that question verifiable rather than asserted.
Table 17. Supplementary materials accompanying this manuscript.
| Ref. | Supplement | Contents | Status |
|---|---|---|---|
| A | Scoring Manual | Every domain threshold with its normative source; grade assignment rules for all nine systems; decision tree for each domain; derivation table; anchor reconciliation procedure; tie-breaking rules; override criteria and documentation requirements; missing-data rules; artifact and protocol-deviation rules; version history | Under compilation from internal scoring documentation |
| B | Examiner Manual | Examination sequence; setup and calibration procedures for each instrument; patient instruction scripts; tolerance and stopping rules; documentation standards | Under compilation |
| C | Training and Certification Protocol | Training curriculum; supervised scoring requirement; certification standard; periodic re-certification and drift monitoring | To be developed alongside the inter-rater study (Section 12.2) |
| D | Example Reports | De-identified specimen reports demonstrating the minimum reporting dataset (Table 15), including a complete examination, an examination with two domains Not Tested, and an examination for which no global score is issued | To be prepared under institutional review with appropriate consent |
| E | Worked Scoring Cases | Constructed cases spanning the scale range with full derivation shown, for examiner training and for reader verification of the derivation rules | Two cases appear in Section 7.9; the full set is in preparation |
| F | Normative Reference Tables | Normative reference values applied in each domain, with source, reference population, and applicability caveats stated for each | Requires compilation and, where age-specific data are unavailable, explicit acknowledgment (Section 11.5) |
| G | Version History | Numbered version record of the scoring rules with effective dates, changes, and rationale (Section 4.7) | Maintained internally; to be published with the instrument |
| H | Reviewer FAQ | Anticipated questions regarding measurement model, ordinality, clinician judgment, missing data, and appropriate statistical treatment | Drafted |
Statement on supplementary status. Supplements A, B, and F are the materials on which reproducibility depends, and they are not yet complete. Until they are published in full, an independent examiner cannot verify that the scoring procedure is reproducible, and the framework should be evaluated accordingly. The authors regard publication of the complete Scoring Manual as a precondition of the reliability study specified in Section 12.2, not a subsequent step.
Numbered in order of first citation. Formatted in APA 7th edition style.
| Abbreviation | Definition |
|---|---|
| ADL | Activities of daily living |
| C-FNE | Comprehensive Functional Neurological Evaluation |
| C-FNE | Comprehensive Functional Neurological Evaluation |
| ClinRO | Clinician-reported outcome measure |
| cTBI | Concussion-type traumatic brain injury |
| CT | Computed tomography |
| DTI | Diffusion tensor imaging |
| EDSS | Expanded Disability Status Scale |
| FIM | Functional Independence Measure |
| fMRI | Functional magnetic resonance imaging |
| FS | Functional System (FS1 through FS9 within the ST-CD framework) |
| GOS-E | Glasgow Outcome Scale – Extended |
| ICC | Intraclass correlation coefficient |
| ICF | International Classification of Functioning, Disability and Health |
| MCID | Minimal clinically important difference |
| MMI | Maximal medical improvement |
| MRI | Magnetic resonance imaging |
| MRS | Magnetic resonance spectroscopy |
| mRS | Modified Rankin Scale |
| mTBI | Mild traumatic brain injury |
| NIHSS | National Institutes of Health Stroke Scale |
| NT | Not Tested (a formal status within the C-FNE, distinct from a grade of 0) |
| PCSS | Post-Concussion Symptom Scale |
| PET | Positron emission tomography |
| RPQ | Rivermead Post-Concussion Symptoms Questionnaire |
| SCAT | Sport Concussion Assessment Tool |
| ST-CD | Short-Term Concussion Disability (Scale) |
| SWI | Susceptibility-weighted imaging |
| TBI | Traumatic brain injury |
| VNG | Videonystagmography |
| VOMS | Vestibular/Ocular Motor Screening |