Abstract
Two-point discrimination (TPD) testing and two-point estimation (TPE) methods are often used to determine tactile acuity, but their reliability is uncertain. Our aim in this study was to assess the reliability of TPD and TPE measurements in the same lumbar spine region of healthy young adults. Participants were 37 young adults (16 females and 21 males; age range: 20–27 years – M age = 21.65, SD = 1.9), who received two tests (TPD and TPE) to determine their tactile acuity. The tests were performed in the lumbar spine area, at the level of the L3 segment, bilaterally, with a 10-minute interval between the measurements. The first test session assessed the reliability of measurements performed by two examiners (inter-rater reliability); and, after 7 days, another examiner repeated the measurements (intra-rater reliability). The reliability of both tests was evaluated using intra-class correlation coefficients (ICC), and results revealed high intra-rater, and inter-rater repeatability for immediate administration of TPD and TPE tests (i.e, after 10 minutes) and moderate repeatability when they were performed at a 7-day interval. Thus, we confirmed high reliability of TPD and TPE assessments of tactile acuity repeated at a short time interval and moderate reliability after 7 days. There was slightly higher reliability for the TPE method.
Keywords
Introduction
Tactile acuity refers to the precision with which individuals sense touch. It is largely affected by the pain experienced in the tested area, the patient’s age, previous injuries to the nervous system, problems in the somatosensory system and other neurological disorders (Dinse et al., 2006; Sathian et al., 1997). Assessment of tactile sensation is useful in (i) the examination of a patient’s neurological status (Serrada et al., 2019), with impaired tactile perception a sign of increasing disease symptoms and improved touch sensation a sign of recovery; and (ii) the training of sensory perception in blind patients (Noh et al., 2015). The somatosensory system, responsible for the ability to distinguish information through the sense of touch, allows people to discriminate between two simultaneous tactile stimuli (Kandel et al., 2012). There are a few types of tactile discrimination. One of the best known, most common and most studied methods is the two-point discrimination test (TPD) (i.e., the ability to differentiate between two different tactile stimuli applied relatively close to one another) (Catley et al., 2013). However, several authors have questioned whether the TPD test is reliable, due to variations in force when applying the stimulus. Too strong a stimulus may cause pain and result in higher test values, while too weak a stimulus may not be perceived by the patient (Craig & Johnson, 2000; Adamczyk, Saulicz, et al., 2018).
Adamczyk, Luedtke, et al. (2018) were pioneers of using estimation in tactile perception in physiotherapeutic assessment. An estimator is a formula for calculating or estimating a value based on observed data. These authors created a test for evaluating tactile acuity based on the patient’s estimate of the distance between two stimuli (i.e., the two-point estimation task – TPE). In the clinical setting, the shorter duration of the TPE, compared to the TPD, is an advantage (Adamczyk, Sługocka, et al., 2019). The TPE involves simultaneously touching two points on the body of the examinee; the patient is then asked to determine the distance between the perceived points of touch. During the development of this method, the form in which the patient might provide feedback appeared problematic. Initially, the two feedback options considered were verbal and manual. In the manual form, the patient adjusted a calliper to illustrate the perceived distance between the two points of touch, while, in the verbal form, the patient verbally described the perceived distance between the two points in centimeters. Studies revealed that the manual assessment provided more reliable and repeatable measurements (Adamczyk, Sługocka, et al., 2019).
The assessment of reliability of both the TPD and TPE methods in the lumbar area was performed in patients with chronic pain symptoms by comparing the precision of the tactile sensation on the painful and pain-free sides of the body (Adamczyk, Sługocka, et al., 2019; Wang et al., 2020). This reliability assessment was conducted at different lumbar levels (L3 segment vs. L5 segment) and at different time intervals (7 days vs. 24 hours). The common factor was the distance of five cm from the midline in the lumbar area (Adamczyk, Sługocka, et al., 2019). Due to the place of measurement and the methodology of TPE (calliper range of 120 mm), the site for the TPE was necessarily more peripheral, compared to the TPD test. These various methodological difficulties led us, in this study, to re-assess the reliability of these tactile acuity tests by performing them in the same area of the lumbar spine in healthy individuals on both sides of the body.
Method
Participants
Participant Characteristics.
CV - Coefficient of variation SD – Standard deviation.
Test Procedure
The tests were performed by a physician with four years of professional experience (Examiner A – OS) and a physiotherapist with two years of professional experience (Examiner B – JK). Both examiners were previously trained in performing TPD and TPE tests; and, at baseline testing, they had already conducted such tests in over 200 healthy individuals and patients in a clinical setting. For the measurements during the tests, we used two identical callipers (Precision Measuring, digital calliper 150 mm). Both instruments were equipped with an electronic display that offered readings with a precision of up to 0.01 mm. The tests were conducted on a Meden-in-med therapeutic table with a face hole in the head section, supporting the participant’s face during all measurements. The order of the tests performed (TPD or TPE) and the side of the body (right or left) for which the test was begun were selected randomly by someone who did not perform the measurements directly (MS). This same person read the measurement results from the electronic displays and recorded them on a dedicated result form. The randomly selected order of the tests and the body side from which the measurements began were then continued for that participant for both stages of the study (i.e, during session I and session II, conducted 7 days later). Measurements during both sessions were always performed in the same room, at the same approximate room temperature (18–20°C). Participants were in a prone position with their backs uncovered during the entire test. After each touch, each participant was instructed to declare the number of points in which they felt this touch. If they felt the touch at one point, they were to say “one,” if at two points, they were to say “two,” and, if they were not certain, they were to say, “I don’t know.” The test began with the location of the processus spinosus of the third vertebra of the lumbar spine. From this site, we measured 15 cm on both sides, perpendicular to the longitudinal axis of the lumbar spine, and we marked the points with a black felt tip pen. These points were then touched at the beginning of the TPD test, and they were also the central points for the application of the calliper tips during the TPE test.
The TPD measurements were conducted following the test protocol described by Adamczyk et al. (2016). One TPD measurement comprised two cycles, always performed in the same order. At the start of the test, the pre-determined point was touched by the joined calliper arms, and the arms were gradually moved apart, until the participant reported perception of touch in two places. For certainty, the participant’s body surface was touched with the calliper again, first with one arm of the calliper, then with both arms of the calliper; and, when normal perception of touch was confirmed (at one point and then at two points), the assistant (MS) was shown the digital display on the calliper. Without informing anyone, the assistant then recorded the value in the result tables. At the beginning of the descending run, the calliper arms were extended 10 mm over the recorded value of tactile perception of two points, and the space between the arms was gradually reduced until the participant felt touch at one point. Having confirmed this measurement with the participant by touching again, the assistant recorded the value. The obtained results were averaged, and they were considered one measurement of TPD. The entire procedure was repeated twice on the same side, and analogous measurements were performed on the other side of the spine.
Two callipers were used for the TPE tests. One was used by the examiner for the measurements, and the other one, with the arms joined and the display facing the opposite direction, was held by the participant. The calliper arms were set at 120 mm apart (Adamczyk et al., 2018b, 2019a; Wand et al., 2010). Before the measurement, the participants were instructed to move the arms of the calliper as far apart as the distance they felt between the points of touch on their back, and to lift the calliper up, so that the assistant could record the displayed value. When the assistant said “done,” the participant squeezed the calliper arms together. Each such measurement was repeated three times on each side of the back.
The first test, comprising a set of TPD and TPE measurements on both sides of the body, was performed by Examiner A, who then left the room where the study was conducted. Next, the assistant went to the room where Examiner B was waiting and invited participants to the testing room. Examiner B was informed which test to perform and on which side to start. 10 minutes after the end of the test conducted by Examiner A, Examiner B started to perform measurements in the analogous order and number. The second measurement session was performed only by Examiner A 7 days later. For a given participant, the measurements were conducted in the same order (regarding the type of test and side of the body) as during the first session. Having performed all measurements, Examiner A left the room where the tests were conducted and returned after 10 minutes to repeat them, maintaining the same order and quantity of measurements.
Statistical Analysis
Statistical analyses were conducted using the Statistica 10.3 software. Intra-rater reliability was assessed based on the calculations of the results obtained in the consecutive testing sessions. The inter-rater reliability between Examiner A and Examiner B was established, as well as the intra-rater reliability for the measurements conducted by Examiner A on the first day and 7 days later. Three correlations were verified in total. For the calculations, we used the ICC (inter-class correlation coefficient statistical model from McGraw & Wong, 1996): “Two-way mixed effects, consistency, multiple raters/measurements.” ICC is a statistic used for measurements on organized groups of data to verify if the data in the group resemble each other (if they correlate) (Koch, 1982). We used a scale proposed by Portney and Watkins (2000) to interpret the results as follows: • 0.75–1.0 was considered a high degree of correlation (high reliability of results); • 0.50–0.75 was considered a moderate degree of correlation (moderate reliability of results); • up to 0.50 was considered a low degree of correlation (low reliability of results).
The standard error of measurement and the smallest detectable difference were derived from the following formulae: • SEM = SD × √1−ICC • SDD = 1.96 × SEM × √2
Results
Means (Standard Deviations) and 95% Confidence Intervals of TPD and TPE Measurements Obtained by One Examiner and by Two Examiners.
Two-Point Discrimination (TPD) and Two-Point Estimation (TPE) Tests (One Examiner).
ICC3.1 – one measurement result.
ICC3.2 -two measurement result.
ICC3.3 -three measurement result.
SEM - standard error of measurement.
95% CI – 95% confidence interval.
Two-Point Discrimination (TPD) and Two-Point Estimation (TPE) Tests (Two Examiners).
ICC3.1 – one measurement result.
ICC3.2 -two measurement result.
ICC3.3 -three measurement result.
SEM - standard error of measurement.
95% CI – 95% confidence interval.
Reliability of Two-point Discrimination (TPD) and Two-point Estimation (TPE) Tests – One Examiner
The TPD test conducted at 15 cm laterally from the spine for one examiner revealed a high degree of correlation between the tests with a 10-minute interval, and a moderate degree of correlation for the tests with a 7-day interval. The lowest ICC value for the tests with a 10-minute interval was 0.86 for ICC3.1 on the right side, and the highest value was 0.94 for ICC3.3 on the right side. The results of the test after 7 days demonstrated a moderate degree of correlation. The lowest value was 0.53 for ICC3.1 on the right side, and the highest value was 0.61 for ICC3.1 on the left side. In the TPE test for one examiner, the assessment of the precision of the measurements conducted with a 10-minute interval revealed a high reliability (the lowest value was 0.91 for ICC3.1 on the right side and the highest was 0.93 for ICC3.2 on the right side and IC3.3 for the left side). The results of the test after 7 days shows us moderate to high degree of correlation with lowest value being 0,64 for ICC3.1 on right side and highest 0,81 for ICC3.3 on the left side.
Reliability of Two-Point Discrimination (TPD) and Two-Point Estimation (TPE) Tests – Two Examiners
Regarding two examiners, the TPD test demonstrated a high degree of correlation between examiners, with the lowest value of 0.75 for ICC3.1 on the left side and the highest value of 0.81 for ICC3.3 on the left side. The TPE test revealed a similar precision of measurements, as demonstrated by the ICC values presenting a high degree of correlation, where the lowest ICC value was 0.78 for ICC3.1 on the right side, and the highest value was 0.88 for ICC3.3 on the left side.
Discussion
Our data demonstrated a high degree of intra-rater correlation (reliability) between two measurements by the same examiner on the TPD and TPE assessments for immediate re-testing (within 10 minutes), a moderately high degree of inter-rater correlation between two examiners, and a moderately high intra-rater correlation between two measurements by the same examiner for temporally distant retesting (after 7 days). Previously, four research teams assessed the inter-rater and intra-rater reliability of the TPD test. Adamczyk et al. (2016) conducted such assessment at the level of the L3 segment in a group of 21 healthy individuals, using two examiners. The intra-rater reliability was ICC = 0.72 and the inter-rater reliability was ICC = 0.56. Ehrenbrusthoff et al. (2016) demonstrated at the L5 segment level that intra-rater reliability was ICC = 0.8 and inter-rater reliability was ICC = 0.53. Catley et al. (2013) used 28 examiners and 28 healthy patients and analyzed the reliability of the TPD test in the lumbar section at the L3 segment level, finding an intra-rater reliability of ICC = 0.81 and an inter-rater reliability of ICC = 0.66. Finally, Wang et al. (2020) verified the reliability of the TPD test on the painful and pain-free side in patients with chronic lumbar back pain (CLBP), dividing participants into young patients (Y-CLBP) and elderly patients (O-CLBP) and found intra-rater ICC values for the groups to range from 0.74 to 0.85, and inter-rater ICC values to range between 0.66 and 0.75. Our study provided the following results: ICC = 0.86–0.94 for a single examiner and 0.75 – 0.81 for two examiners. Compared to previous studies, our results are moderately higher than those observed by other researchers. The difference may be due to the site where the stimulus was delivered, the applied pressure, the intervals between tests, our participant population, and the fact that in some prior studies the reliability of the TPD test was evaluated in patients with back pain.
Our findings were most consistent with those of Adamczyk et al. (2016), as our measurements were performed at the same level (L3 segment) and on healthy participants. The differences between both studies included the size of the study populations (37 participants in our study vs. 21 in Adamczyk et al., 2016), the place of measurement (15 cm from the spinal axis in our study and a distance of 5 cm and only on the left side in Adamczyk et al. (2016), and the participants’ sex (both sexes in our study and males only in Adamczyk et al. (2016). The “better” results of the TPD precision test which are found in Adamczyk’s works could also be affected by the fact that the examiners had extensive experience in this type of measurements, and they had previously participated in a few projects associated with the assessment of superficial sensation (Adamczyk et al., 2016, 2018a, 2018b, 2019b). This can be demonstrated indirectly by the average TPD values on the left side of the body. In the study by Adamczyk et al. (2016), these mean values, as well as the difference between the two series of measurements (50.56–54.6 mm) for Examiner A were higher compared to Examiner A in our study (41.3–42.2 mm). However, it is possible that these differences were due to the different measurement sites (more peripheral from the spinal axis) and the high percentage of females (43.2% of the study group) participating in our study.
The reliability of the TPE test was previously evaluated in two scientific studies. Luedtke et al. (2018) verified the reliability of both the manual and verbal versions of the TPE test for the painful and pain-free side of the cervical spine. They found the inter-rater reliability based on the manual version of the test to be higher than for the verbal version (manual: ICC = 0.75–0.91; verbal: ICC = 0.53–0.88). The intra-rater reliability with a 2-day interval between tests was also higher in the manual than verbal version (manual: ICC = 0.75–0.91; verbal: ICC = 0.67–0.84). Considering radically different levels at which measurements were performed and the fact that the later measurements were conducted after 2 days in these studies, it is difficult to compare these studies directly with ours. Of noted, however, inter-rater reliability was high and similar across studies. More consistent with our findings were the results from Wang et al. (2020), although Wang et al.’s (2020) measurements were performed at a different level (L5 segment) and site (5 cm). In their study on the manual version of the TPE test, Wang et al. (2020) reported an ICC of 0.82–0.85 for one examiner and an ICC of 0.70–0.85 for two examiners. Our values were similarly high (one examiner: ICC= 0.91–0.93; two examiners: ICC = 0.78–0.88). Again, these minor differences, favoring our study, are probably due to differences between the measurement sites and the fact that we used only young, healthy participants, while Wang et al. (2020) used symptomatic patients (with CLBP) and participants in two age groups: 18–35 years old and 36–65 years old.
Limitations and Directions for Further Research
As our study involved only healthy participants, its limitations may include failure to control for the menstrual cycle phase in female participants. As water is retained in the body during the second phase of the cycle, including in the subcutaneous tissue, tactile acuity may then be affected. Many women experience transient tactile hyperesthesia in the second phase of their menstrual cycle. As there may have been adverse effects on TPD and TPE for some female participants in the second phase of their menstrual cycle during the later test session (after 7 days), future assessments of the precision of temporally distant measurements (performed within a range of several days) should be designed so that tests are conducted in females in the same phase of their menstrual cycle, perhaps after first estimating the degree to which menstrual cycle phase affects women’s tactile acuity.
Conclusion
No prior investigators have assessed the reliability of the TPD and TPE tests in healthy young females and males with a 7-day interval between test sessions. Our results (TPD: ICC = 0.53–0.61; TPE: ICC = 0.64–0.81) demonstrate moderate or moderately low TPD reliability and moderate/moderately high TPE reliability. This variability could be due to environmental factors (e.g., ambient temperature, time of day, changes in the atmospheric pressure), or changes in the participant’s psychophysiological status (emotions, agitation, fatigue, etc.). Based on these results, the TPD and TPE tests can be considered reliable methods of assessing tactile acuity in the lumbar spine area (at the level of the L3 segment) in healthy individuals. The reliability of two measurements is sufficient for the determination of the patient’s tactile acuity. Only the test with a 7-day interval demonstrates a moderate reliability, and it requires at least three measurements.
Both the TPD test and the TPE test are reliable methods of assessing tactile acuity in the lumbar spine area (at the level of the L3 segment) in healthy individuals. In the measurements performed within a short time interval, both one examiner (intra-examiner reliability) and two examiner (inter-examiner reliability) measurements reflected highly reliable results (ICC ±0.75) on both sides of the body. For measurements conducted 1 week apart, the TPE test showed higher reliability than the TPD test. Moderate reliability for the right side of the body and high reliability for the left side was achieved with only a single measurement. The 7-day reliability of the TPD test was moderate, and increasing the number of measurements did not enhance it.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
