Erasmus MC Modification of the Revised Nottingham Sensory Assessment

1. Abstract

The Erasmus MC Modification of the Revised Nottingham Sensory Assessment (EmNSA) is a standardized, clinically refined neurorehabilitation instrument developed to evaluate somatosensory function in individuals recovering from acute, subacute, and chronic stroke or other central nervous system pathologies. Developed as an evolution of Lincoln et al.’s original Nottingham Sensory Assessment (NSA) and the subsequent Revised Nottingham Sensory Assessment (rNSA), the EmNSA optimizes psychometric robustness, intra- and inter-rater reproducibility, and clinical feasibility by streamlining sensory testing protocols. The instrument systematically evaluates three critical sensory modalities: light touch sensation (tactile perception mediated via the dorsal column-medial lemniscal system), sharp-blunt discrimination (nociceptive and thermal-adjacent discriminative sensation mediated through the spinothalamic tract), and proprioception (kinesthetic and articular position sense). The scale tests these modalities across four defined anatomical segments of the upper extremity (fingers, hand, forearm, upper arm / shoulder) and four segments of the lower extremity (toes, foot, lower leg, upper leg / hip), yielding 24 discrete clinical items. Each item is scored on an authentic 3-point ordinal scale (0 = absent, 1 = impaired, 2 = normal), providing subscale scores of 0 to 16 for each sensory modality, regional extremity subscores of 0 to 24, and a composite total score ranging from 0 to 48. Psychometric investigations have demonstrated superior inter-rater and test-retest reliability compared to earlier iterations, with weighted kappa coefficients typically exceeding 0.70 for light touch and sharp-blunt discrimination and up to 0.90 for joint movement appreciation. The EmNSA exhibits robust construct, convergent, and discriminant validity when evaluated alongside motor impairment indices such as the Fugl-Meyer Assessment and functional outcome measures like the Barthel Index. Administered in approximately 10 to 15 minutes for an extremity, the EmNSA remains an essential diagnostic and evaluative tool in clinical neurology, neurorehabilitation research, and physical therapy.

2. Keywords

Erasmus MC Modification of the Revised Nottingham Sensory Assessment, EmNSA, Somatosensory Impairment, Stroke Rehabilitation, Light Touch, Sharp-Blunt Discrimination, Proprioception, Neurorehabilitation, Psychometrics, Kinesthesia

3. Authors

The Erasmus MC Modification of the Revised Nottingham Sensory Assessment was developed and psychometrically validated by an interdisciplinary clinical research team in the Department of Rehabilitation Medicine at Erasmus University Medical Center (Erasmus MC) in Rotterdam, The Netherlands:

• Florike Stolk-Hornsveld, PT, MSc: Physical Therapist and Clinical Researcher, Department of Rehabilitation Medicine, Erasmus MC, Rotterdam, The Netherlands.
• Judy L. Crow, PhD: Neurorehabilitation Researcher and Physiotherapist, Erasmus MC and International Stroke Recovery Research Consortia.
• Eline P. Hendriks, PT: Neurorehabilitation Clinician, Department of Physical Therapy and Rehabilitation, Erasmus MC.
• Rob van der Baan, PT: Department of Rehabilitation Medicine, Erasmus MC.
• Bianca C. Harmeling-van der Wel, MD, PhD: Physiatrist and Clinical Epidemiologist, Department of Rehabilitation Medicine, Erasmus MC, Rotterdam, The Netherlands.

4. Purpose

Somatosensory dysfunction is observed in approximately 50% to 80% of individuals following a cerebrovascular accident (stroke). Deficits in tactile perception, two-point discrimination, nociceptive discrimination, and proprioceptive kinesthesia negatively impact fine motor coordination, balance, postural equilibrium, ambulation, and the performance of activities of daily living (ADLs). Despite its profound clinical consequence, somatosensation has historically been measured in clinical practice using non-standardized bedside routines (e.g., erratic cotton-swab touches, uncalibrated safety pins) characterized by high subjective variability and poor psychometric reliability.

The primary clinical and scientific purpose of the EmNSA is to deliver a standardized, repeatable, and rapid clinical instrument capable of:

• Accurately diagnosing specific somatosensory impairments across dermatomal and articular regions of both the paretic and non-paretic limbs.
• Differentiating between superficial exteroceptive tactile deficits, lateralized spinothalamic discriminative deficits, and deep kinesthetic proprioceptive dysfunctions.
• Informing targeted neurorehabilitation interventions, such as task-specific sensory training, somatosensory retraining protocols, sensory electrical stimulation, and constraint-induced movement therapy.
• Monitoring longitudinal neurological recovery and neuroplastic reorganization over time in clinical trials and clinical practice.
• Assessing fall risk and hand functional capacity by establishing sensory baselines that underpin motor output and feedforward motor planning.

The original Nottingham Sensory Assessment, while foundational, was burdened by lengthy administration times (frequently exceeding 45 minutes) and problematic inter-tester reliability on complex stereognosis and two-point discrimination subtests. The EmNSA was deliberately engineered to resolve these psychometric inefficiencies. By eliminating modalities with poor inter-rater concordance and standardizing anatomical contact zones, patient positioning, and testing rules (utilizing a clear “three-trial” evaluation format per item), the EmNSA provides a feasible, 10- to 15-minute diagnostic protocol suitable for routine acute, subacute, and long-term neurorehabilitation settings.

5. Psychological and Neurological Construct

The construct assessed by the EmNSA is human somatosensation, specifically operationalized across three functional neuroanatomical dimensions: tactile exteroception (light touch), discriminative nociception (sharp-blunt discrimination), and deep mechanoreception (proprioception). These modalities represent distinct ascending sensory tracts and cortical processing pathways within the central nervous system:

5.1. Light Touch (Tactile Exteroception)

Light touch sensation evaluates the threshold and integrity of cutaneous low-threshold mechanoreceptors (such as Meissner corpuscles, Merkel disks, and hair follicle receptors). These peripheral inputs ascend via heavily myelinated A-beta afferents through the dorsal column-medial lemniscal pathway to the gracile or cuneate nuclei in the medulla, cross in the medial lemniscus, synapse in the ventral posterolateral (VPL) nucleus of the thalamus, and project to the primary somatosensory cortex (Brodmann areas 3b and 1) in the postcentral gyrus. Clinical failure to recognize cotton-wool stimulation reflects interrupted ascending peripheral conduction or cortical/subcortical infarction, directly impeding tactile exploration and spontaneous limb usage.

5.2. Sharp-Blunt Discrimination (Protective and Cutaneous Discrimination)

Sharp-blunt discrimination tests the neurosensory system’s ability to distinguish between punctate, high-threshold nociceptive inputs and blunt mechanoreceptive pressure. This modality engages small-diameter thinly myelinated A-delta fibers and unmyelinated C fibers projecting through the dorsal horn of the spinal cord, crossing via the anterior white commissure, and ascending through the spinothalamic tract to the VPL and posterior thalamic nuclei before terminating in S1 and the secondary somatosensory area (S2). Impairment indicates compromised protective sensation, increasing vulnerability to peripheral injury, decubitus formation, and unconscious microtrauma.

5.3. Proprioception (Kinesthetic Articular Sense)

Proprioception measures articular joint movement awareness and directional appreciation. Mediated by complex mechanoreceptors including muscle spindles (Ia and II afferents), Golgi tendon organs (Ib afferents), and Ruffini endings within the joint capsules, proprioceptive afferents ascend through the dorsal columns to the thalamus and higher-order parietal processing centers (Brodmann areas 3a and 2, with dense projections to the posterior parietal cortex and primary motor cortex). Preserved proprioception is essential for predictive motor control, kinesthetic body schema mapping, and adjusting grip force or stride length without visual feedback.

6. Theoretical Framework

The EmNSA is grounded in neurocomputational sensorimotor control theory and sensory feedback processing models (e.g., Wolpert, Ghahramani, and Flanagan’s internal models of motor control). Under this theoretical paradigm, the motor system relies on continuous internal forward models that predict the sensory consequences of movement commands. Efference copies of descending motor commands are compared against incoming somatosensory reafference in the parietal and cerebellar cortices. When somatosensory afference is distorted, interrupted, or abolished following cortical, subcortical, or brainstem stroke:

• Sensory prediction error calculations are disrupted, resulting in motor dysmetria, limb unsteadiness, ataxia, and excessive grip forces.
• The phenomenon of learned non-use (as formulated by Taub and colleagues) rapidly emerges: patients consciously and unconsciously avoid utilizing a paretic limb that yields no intelligible somatosensory feedback, even when motor efferent pathways have partially recovered.
• Sensorimotor integration deficits prevent neuroplastic synaptic reorganization, retarding clinical recovery.

Furthermore, the EmNSA aligns with modern psychometric Classical Test Theory (CTT) and Item Response Theory (IRT) frameworks for neurological assessment. The original Nottingham scale contained redundant, poorly standardized items exhibiting excessive measurement noise. By systematically eliminating high-error items (such as stereognosis, which requires functional grasping ability and thus confounds motor weakness with sensory failure), the EmNSA isolates sensory detection from motor execution, yielding clean, construct-pure measurements of afferent processing.

7. Validity

Empirical investigations have established strong measurement validity across multiple clinical cohorts:

7.1. Construct and Content Validity

Content validity was confirmed through formal consensus panels of rehabilitation physicians, physical therapists, and neuroscientists who evaluated the operational definitions, tactile stimulation methods, and joint movement protocols. The scale demonstrates hierarchical construct validity: sensory deficits are most pronounced distally (fingers and toes) and attenuate proximally (upper arm and hip), reflecting classical somatotopic representations along the sensory homunculus.

7.2. Convergent and Criterion Validity

The EmNSA demonstrates statistically significant convergent validity when evaluated against comparable sensory and sensorimotor instruments:

• Strong correlations with the sensory subscale of the Fugl-Meyer Assessment (FMA-Sensory), with Spearman’s rho coefficients typically ranging from $r_s = 0.68$ to $0.84$ ($p < 0.001$).
• Moderate-to-high correlations with functional motor performance of the upper limb (e.g., Action Research Arm Test [ARAT], $r_s = 0.52 – 0.65$), supporting the theoretical link between sensory acuity and functional dexterity.
• Significant positive association with the Barthel Index and Functional Independence Measure (FIM), demonstrating that preserved EmNSA scores correspond to superior daily functional autonomy.

7.3. Discriminant and Known-Groups Validity

The EmNSA successfully discriminates between healthy age-matched control participants (who consistently achieve ceiling scores of 48) and individuals with stroke presenting varying anatomical lesion locations (e.g., cortical versus subcortical infarcts). Studies show that patients with pure cortical middle cerebral artery (MCA) strokes exhibit pronounced deficits in sharp-blunt discrimination and proprioception, whereas small lacunar capsular strokes frequently spare sharp-blunt discrimination while impairing tactile sensation.

8. Reliability

The clinical utility of any sensory evaluation tool hinges on its stability across raters and measurement sessions. The original Nottingham Sensory Assessment displayed poor inter-rater reliability for specific items (Cohen’s kappa falling below 0.40). The Erasmus MC modification resolved these discrepancies through precise mechanical rules (three trials per site) and strict scoring criteria:

8.1. Inter-Rater Reliability

In the seminal psychometric validation study by Stolk-Hornsveld et al. (2006) involving 70 stroke survivors, inter-rater reliability between independent physical therapists was evaluated across all segments:

• Light Touch: Upper extremity Cohen’s kappa ($kappa$) values ranged from $0.62$ to $0.88$; lower extremity values ranged from $0.58$ to $0.82$, reflecting substantial to almost perfect inter-rater concordance.
• Sharp-Blunt Discrimination: Kappa values ranged between $0.65$ and $0.86$ across all tested segments.
• Proprioception: Inter-rater agreement for joint movement appreciation and directional identification ranged from $kappa = 0.68$ (fingers and toes) to $kappa = 0.91$ (shoulder and hip).

8.2. Intra-Rater and Test-Retest Reliability

Intra-rater test-retest reliability evaluated across repeated administrations within stable intervals (24 to 48 hours) yielded intraclass correlation coefficients (ICC) exceeding $0.85$ for extremity subscale totals and $0.92$ for the EmNSA total composite score, demonstrating exceptional stability in non-recovering phases.

9. Factor Analysis and Dimensionality

Exploratory factor analysis (EFA) and subsequent confirmatory psychometric evaluations (including Mokken scale analysis and Rasch measurement models) have demonstrated a clear multidimensional yet internally consistent structural architecture:

9.1. Factor Structure

Factor analytic models of the 24 EmNSA items consistently identify a robust three-factor solution corresponding to the three tested sensory modalities:

• Factor 1: Tactile Exteroception (Light Touch): Items 1–4 and 13–16 load heavily onto this factor (factor loadings ranging from $0.72$ to $0.89$), accounting for a substantial proportion of shared variance.
• Factor 2: Nociceptive Discrimination (Sharp/Blunt): Items 5–8 and 17–20 display distinct factor loadings (ranging from $0.69$ to $0.85$) onto a discrete discriminative dimension, confirming its neuroanatomical segregation via the spinothalamic pathway.
• Factor 3: Articular Proprioception: Items 9–12 and 21–24 load onto the deep kinesthetic factor (loadings from $0.74$ to $0.92$), demonstrating independent variance from cutaneous superficial sensation.

9.2. Rasch Analysis and Hierarchical Ordering

Rasch item-difficulty analyses confirm a distinct hierarchical gradient across anatomical segments. Distal items (fingers, toes) possess the highest item difficulty (i.e., require higher neurosensory integrity to register a normal score), whereas proximal segments (upper arm, shoulder, upper leg, hip) exhibit lower item difficulty. Differential item functioning (DIF) across age and biological sex has been demonstrated to be negligible, confirming measurement invariance across diverse clinical demographic groups.

10. Instrument / Measurement Tool

• Test Type: Clinician-administered standardized neurosensory observation performance test.
• Target Population: Adults and older adults with neurological disorders (primarily stroke, traumatic brain injury, spinal cord pathology, or polyneuropathy).
• Administration Time: Approximately 10 to 15 minutes for the upper extremity; 10 to 15 minutes for the lower extremity (20–30 minutes for the complete bilateral evaluation).
• Required Materials:
  • Standardized piece of dry cotton wool (for light touch).
  • Neurological safety pin or Neurotip (with standardized sharp and blunt ends for sharp/blunt discrimination).
  • Visual occlusion blindfold or cardboard screening barrier.
  • Standardized EmNSA scoring record sheet.
• Testing Procedure Rules:
  • Testing begins distally and proceeds proximally (e.g., fingers $\rightarrow$ hand $\rightarrow$ forearm $\rightarrow$ upper arm).
  • Patient’s vision must be occluded (using a blindfold, screen, or having the patient turn their head and close eyes).
  • Each specific anatomical point receives exactly three applications or passive joint movements.
  • For light touch and sharp/blunt: cotton wool or pin is applied perpendicularly without scraping or dragging.
  • For proprioception: the examiner grasps the joint laterally along the bony prominences (avoiding muscle belly pressure that could provide cutaneous pressure cues) and passively displaces the joint through small ranges of motion (approximately 10 degrees).
• Item Count: 24 total clinical items (12 Upper Extremity, 12 Lower Extremity).
  • 8 Light Touch items (4 upper, 4 lower).
  • 8 Sharp/Blunt Discrimination items (4 upper, 4 lower).
  • 8 Proprioception items (4 upper, 4 lower).
• Response Format and Authentic Scoring Scale:
  • Light touch & Sharp/blunt discrimination:
    • 0 = Absent: No sensations felt / no sharp or blunt distinction.
    • 1 = Impaired: Sensations felt, but not all three correctly or unable to reliably distinguish.
    • 2 = Normal: All three touches felt correctly / correctly distinguishes sharp and blunt 3 out of 3 times.
  • Proprioception:
    • 0 = Absent: No movement appreciation.
    • 1 = Impaired: Direction of movement not perceived correctly or movement perceived without directional sense.
    • 2 = Normal: Movement appreciated and direction accurately identified 3 out of 3 times.
• Scoring Structure:
  • Upper Extremity Subtotal: Range 0 to 24.
  • Lower Extremity Subtotal: Range 0 to 24.
  • Modality Subtotals: Light Touch (0–16), Sharp/Blunt (0–16), Proprioception (0–16).
  • Total EmNSA Composite Score: Range 0 to 48 (higher scores indicate intact somatosensation; 48 reflects completely intact sensory functioning).

11. Permissions, Fee, and Test Year

The Erasmus MC Modification of the Revised Nottingham Sensory Assessment was published in 2006 by Florike Stolk-Hornsveld and colleagues at the Erasmus University Medical Center. The tool was developed for open clinical and scientific use within physical therapy, occupational therapy, and neurorehabilitation clinics. The instrument is considered non-commercial and in the public domain for academic research and standard clinical practice, provided appropriate citation of the foundational validation article is maintained. No licensing fees, commercial royalties, or administrative purchase costs are required to utilize the test form or protocol.

12. References

• Connell, L. A., Lincoln, N. B., & Radford, K. A. (2008). Somatosensory impairment after stroke: Frequency of different deficits and their recovery. Clinical Rehabilitation, 22(8), 758–767. https://doi.org/10.1177/0269215508090674
• Lincoln, N. B., Jackson, J. M., & Adams, S. A. (1998). Reliability and evaluative properties of the Nottingham Sensory Assessment for stroke patients. Clinical Rehabilitation, 12(1), 23–28. https://doi.org/10.1191/026921598672927231
• Meyer, S., De Baets, L., Krumlinde-Sundholm, L., & Peeters, A. (2016). Assessment of somatosensory functioning in stroke survivors: Systematic review of psychometric properties. Neurorehabilitation and Neural Repair, 30(8), 731–744. https://doi.org/10.1177/1545968315624784
• Stolk-Hornsveld, F., Crow, J. L., Hendriks, E. P., van der Baan, R., & Harmeling-van der Wel, B. C. (2006). The Erasmus MC modifications to the (revised) Nottingham Sensory Assessment: A reliable somatosensory assessment measure for patients with intracranial disorders. Clinical Rehabilitation, 20(2), 160–172. https://doi.org/10.1191/0269215506cr932oa
• Tyson, S. F., Crow, J. L., Connell, L., Winward, C., & Hillier, S. (2013). Sensory impairments of the lower limb after stroke: A pooled analysis of individual patient data. Topics in Stroke Rehabilitation, 20(5), 441–449. https://doi.org/10.1310/tsr2005-441

13. Items of the Scale (Questionnaire)