Modernising tactile acuity assessment; clinimetrics of semi-automated tests and effects of age, sex and anthropometry on performance

Abstract

Background: Reduced tactile acuity has been observed in several chronic pain conditions and has been proposed as a clinical indicator of somatosensory impairments related to the condition. As some interventions targeting these impairments have resulted in pain reduction, assessing tactile acuity may have significant clinical potential. While two-point discrimination threshold (TPDT) is a popular method of assessing tactile acuity, large measurement error has been observed (impeding responsiveness) and its validity has been questioned. The recently developed semi-automated 'imprint Tactile Acuity Device' (iTAD) may improve tactile acuity assessment, but clinimetric properties of its scores (accuracy score, response time and rate correct score) need further examination.

Aims: Experiment 1: To determine inter-rater reliability and measurement error of TPDT and iTAD assessments. Experiment 2: To determine internal consistencies and floor or ceiling effects of iTAD scores, and investigate effects of age, sex, and anthropometry on performance.

Methods: Experiment 1: To assess inter-rater reliability (ICC_(2,1)) and measurement error (coefficient of variation (CoV)), three assessors each performed TPDT and iTAD assessments at the neck in forty healthy participants. Experiment 2: To assess internal consistency (ICC_(2,k)) and floor or ceiling effects (skewness z-scores), one hundred healthy participants performed the iTAD's localisation and orientation tests. Balanced for sex, participants were equally divided over five age brackets (18-30, 31-40, 41-50, 51-60 and 61-70). Age, sex, body mass index (BMI) and neck surface area were assessed to examine their direct (using multiple linear regression analysis) and indirect (using sequential mediation analysis) relationship with iTAD scores.

Results: Mean ICC_(2,1) was moderate for TPDT (0.70) and moderate-to-good for the various iTAD scores (0.65-0.86). The CoV was 25.3% for TPDT and ranged from 6.1% to 16.5% for iTAD scores. Internal consistency was high for both iTAD accuracy scores (ICC_(2,6) = 0.84; ICC_(2,4) = 0.86). No overt floor or ceiling effects were detected (all skewness z-scores < 3.29). Accuracy scores were only directly related to age (decreasing with increasing age) and sex (higher for men).

Discussion: Although reliability was similar, iTAD scores demonstrated less measurement error than TPDT indicating a potential for better responsiveness to treatment effects. Further, unlike previously reported for TPDT, iTAD scores appeared independent of anthropometry, which simplifies interpretation. Additionally, the iTAD assesses multiple aspects of tactile processing which may provide a more comprehensive evaluation of tactile acuity. Taken together, the iTAD shows promise in measuring tactile acuity, but patient studies are needed to verify clinical relevance.

Keywords: Chronic pain; Locognosia; Neck; Neurologic examination; Reliability; Tactile acuity; Touch perception.

Figure 1. TPDT assessment procedure.

Example of the assessment of a hypothetical two-point discrimination threshold (TPDT). Assessment is based on a forced-choice response (one or two points), alternating four runs with either increasing or decreasing caliper distances. Steps taken are either in five mm. (first run) or two mm. (other runs). Three consecutive reports of either one or two points indicates a reversal. Mean of the four reversals is calculated for the TPDT score.

Figure 2. The imprint Tactile Acuity Device (iTAD), containing twelve build-in vibrotactile stimulators (top), and wirelessly connected tablet.

The iTAD performs two tactile acuity tests: (1) the localisation test (bottom left) where the perceived location of the tactile stimulus is selected and (2) the orientation test (bottom right) where the perceived location of a second tactile stimulus, relative to a first, is selected. For both tests, as well as the overall score (mean of both tests), the accuracy score (i.e., percentage correct), the average response time and the rate correct score (number of correct responses per minute of response activity) is calculated.

Figure 3. Results sequential mediation analyses.

Relationships between demographics (sex and age), anthropometrics (body mass index (BMI) and neck surface area (NSA)) and iTAD accuracy scores for the localisation test (A) and orientation test (B). Relationships are expressed in semi-partial correlations (sr) and unstandardized regression coefficients (b), including their level of significance (p). Coding for sex: female =0 and male =1.

Figure 4. Scatter plots of localisation and orientation accuracy scores.

Scatter plots of iTAD accuracy scores as a function of age and sex. Scores are displayed for the localisation (A) and orientation (B) test. Lines represent the least squares regressions.