Abstract
PURPOSE:
To investigate the test-retest reliability, side-to-side difference, responsiveness, and concurrent validity of skin blood flow (SkBF) measured by laser Doppler flowmetry (LDF) in individuals with acute or subacute lateral elbow tendinopathy (LET).
METHODS:
Eighteen individuals with acute or subacute LET were recruited for this study. SkBF was measured over the lateral epicondyle and common extensor origin on both ipsilateral and contralateral sides. The intraclass correlation coefficient (ICC3,1) was used to evaluate test-retest reliability. A paired t-test was used for comparing the side-to-side difference. Responsiveness was reported with a change score, paired t-test, effect size (ES), and standardized response mean (SRM). The concurrent validity of SkBF was investigated by correlating with a visual analog scale (VAS) pain intensity during resisted wrist extension isometric contraction using the Spearman correlation coefficient.
RESULTS:
Test-retest reliability was good at lateral epicondyle (ICC3,1 = 0.899), and at common extensor origin (ICC3,1 = 0.803). A side-to-side difference was found between the two sides (p < 0.001). For responsiveness, the change score for SkBF at the lateral epicondyle was – 8.04, and the common extensor origin was – 3.54. The ES and SRM ranged from – 0.71 to – 0.78. Concurrent validity was reported with a strong correlation with pain intensity (r = – 0.637).
DISCUSSION:
SkBF is a reliable and responsive variable for investigating the elbows with acute or subacute LET providing clinical information according to its inflammatory responses. However, the concurrent validity can be found only for SkBF at common extensor origin, which correlates with pain during resisted wrist extension isometric contraction.
Introduction
Lateral elbow tendinopathy (LET), commonly termed tennis elbow can be found in both men and women, and is most prevalent in the working-age population [1]. Lateral elbow pain is the most common characteristic, which is frequently aggravated by grasping during resisted wrist extension [2]. The causes of LET are related to the activities that involve repeated supination and pronation of the forearm with the elbow in extension, representing a degenerative process involving the origin of the extensor tendons at the lateral elbow [2, 3]. The repetitive stress and overuse cause microtrauma and partial tears that may progress to further damage of the extensor tendons [4, 5].
At the early stages, acute or subacute phase of LET, the inflammatory process is possibly evident [6], which is characterized by pain, swelling, heat, and redness, which result from microcirculatory reactions [7, 8]. The investigation of microcirculation is therefore an interesting option to detect the inflammatory process that occurred at the pathological site of the early stages of LET. However, the connection between the pathological site of LET and the change in microcirculation at the dermal level is still under researched. If the detection of the change of skin microcirculation is feasible, it should be beneficial since the investigation of the skin is non-invasive and can be used as an objective measurement for early detection in clinical investigation.
Laser Doppler flowmetry (LDF) can be used for measuring skin blood flow (SkBF) representing local microcirculation. The LDF is a non-invasive method for measuring microvascular blood perfusion at targeted tissues. Briefly, microvascular blood perfusion is estimated by the reflection of red blood cells per minute, the laser light is scattered by static and dynamic red blood cells. Scattering by a dynamic red blood cell results in a Doppler frequency shift, while light scattered by a static cell remains unshifted [9].
The invasive technique of LDF has been used to investigate the microcirculation of pathological rotator cuff tendon compared with normal rotator cuff tendon. The mean flux showed significantly higher in the pathologic tendon [10]. Also, the non-invasive technique of LDF has been reported as having good test-retest reliability for measuring skin blood flow over the lumbosacral area [11].
Considering SkBF as a clinical assessment variable for LET, the investigations regarding the reliability and validity of the non-invasive technique of LDF are interesting yet needed for a formal investigation. However, there is no study investigating test-retest reliability, the comparison between ipsilateral and contralateral sides or side-to-side difference, responsiveness, and concurrent validity. This study, therefore, aimed to investigate those aforementioned properties of SkBF obtained by LDF.
Materials and methods
Participants
Eighteen individuals with acute or subacute LET aged above 18 years were recruited into this study. The examination to identify LET consists of palpable local tenderness at the lateral epicondyle of the elbow, pain provocation at the lateral aspect of the elbow with resisted wrist, and middle finger extension during passive wrist flexion with ulnar deviation. The pain location was at the lateral aspect of the elbow during activities, the duration of symptoms was less than 3 months representing acute or subacute phases [12], and the intensity of pain at rest is at least 30 of 100 millimeters on the visual analog scale (VAS). The exclusion criteria were the use of corticosteroid injection at the affected lateral epicondyle, history of upper extremity surgery or fracture at the affected side, rheumatoid arthritis, carpal tunnel syndrome, cubital tunnel syndrome, and cervical radiculopathy. Written informed consent was obtained from each individual before participation. The data collection period was from December 2018 to October 2019. The study protocol and informed consent have been approved by Mahidol University Central Institutional Review Board (MU-CIRB), COA no. 2018/147.1807 with adherence to the Declaration ofHelsinki.
For sample size calculation, statulator.com, a web-based sample size calculation program was used, based on the data of SkBF at LET from our pilot study. The confidence interval was set at 95%, alpha level 5%, power 80%, mean difference 9.500, and standard deviation 9.264, then the calculated sample size of 15 subjects was required for a significant difference. We also recruited 20% more to cover all possible missing values that might happen, therefore, 18 subjects were recruited.
Demographic and baseline data of the participants
Besides age, sex, weight, height, and body mass index (BMI), all participants were interviewed about the duration of symptoms reported as total days since the onset. Also, the ipsilateral side and dominant hand, left or right, were identified. All participants in this study had unilateral LET. Pain intensity was measured with a horizontal 100-mm visual analog scale (VAS) at rest and during resisted wrist extension isometric contraction. The disability was assessed by the Thai-version abbreviated version of the original disabilities of the arm, shoulder, and hand questionnaire (QuickDASH) consisting of the disability and symptoms questionnaire and two additional modules, which each reported the percentage of disability ranging from 0 – 100, the higher score the greater disability, which good content validity and excellent internal consistency have been previously reported [13].
Skin blood flow (SkBF)
Laser Doppler flowmetry (LDF) (Moor Instruments Ltd, Devon, UK) was used for measuring SkBF representing local microcirculation. The data was calculated mathematically by using moorVMS-PC software (Moor Instruments Ltd, Devon, UK) reporting as SkBF expressed as perfusion unit (PU). For the starting position of SkBthe F measurement, participants sat and placed their pronated forearms on a plinth. The examiner placed skin probes fixed to the skin with a probe holder adhered to the skin over the lateral epicondyle and common extensor origin on both ipsilateral and contralateral sides, then measured absolute blood flow units. SkBF was measured continuously for 3 minutes, the average value was calculated to minimize its variability [14] and used for data analysis.
Pain intensity
Pain intensity was assessed with a 100-mm horizontal line VAS to indicate each participant’s intensity of pain during isometric contraction with resisted wrist extension as a provocative test. VAS pain intensity during resisted wrist extension isometric contraction was used as a reference measurement for concurrent validity.
Procedures
Baseline assessments included SkBF at lateral epicondyle and common extensor origin on both ipsilateral and contralateral sides and VAS during resisted wrist extension isometric contraction. The measurements of SkBF at both locations were done twice with a 15-minute rest interval for investigating test-retest reliability at the ipsilateral side. SkBF was also measured at the contralateral side at both locations for side-to-side comparison. Then, all participants were asked to perform an isometric contraction with resisted wrist extension to express the pain intensity at a painful area for analyzing concurrent validity [15]. For interventions, the participants underwent therapeutic exercises [16] and high-intensity laser therapy (HILT) for LET. Therapeutic exercises consist of the 30-second active exercise of the wrist from full extension to flexion and the 45-second isometric exercise of wrist extension. Fifteen repetitions were done for each exercise. HILT (MLS®, ASA Srl, Vicenza, Italy) was applied into 2 phases: scanning and trigger point phases. For the scanning phase, wrist extensor muscle origin and lateral epicondyle were treated with 300 Hz, 25% intensity for 4 minutes. For the trigger point phase, the laser probe was vertically and statically pointed upon the targeted areas with 100 Hz, 25% intensity for 20 seconds, total HILT treatment time was approximately 5 minutes and total energy was 320 joules. The interventions were given twice a week, for two weeks, for 4 sessions. At the end of the treatment program, the participants rested for 30 minutes and were assessed for post-test data including SkBF at lateral epicondyle and common extensor origin on the ipsilateral side, and VAS pain intensity during resisted wrist extension isometriccontraction.
Statistical analysis
Data were analyzed using SPSS software (IBM SPSS Statistics for Windows, Version 23, Armonk, NY, USA). Shapiro-Wilk test was used to test for the distribution of the data. The variables were normally distributed except baseline and post-test VAS pain intensity during resisted wrist extension isometric contraction. A p-value of less than 0.05 is indicated as being statistically significant. For test-retest reliability, the intraclass correlation coefficient of the two-way mixed model, single measures, and consistency (ICC3,1) was used. In the interpretation of ICC, a value less than 0.5 was considered poor, 0.5 to 075 was moderate, 0.75 to 0.9 was good, and greater than 0.9 was excellent reliability [17]. For side-to-side differences, the ipsilateral and contralateral sides were compared by using a paired t-test. Responsiveness was reported with change score, paired t-test, effect size (ES), and standardized response mean (SRM): change score is the difference between baseline and post-test derived by subtracting post-test value from the baseline value. ES was calculated by the mean change divided by the standard deviation of the baseline score [18]. For interpretation, ES less than 0.2 is considered small, 0.5 moderate, and 0.8 large from Cohen’s suggestion [19]. SRM is similar to ES, calculated by the mean change divided by the standard deviation of the change. The interpretation for SRM is as well as ES, or 0.2, 0.5, and 0.8 represent small, moderate, and large respectively [20].
Concurrent validity of SkBF at ipsilateral lateral epicondyle and common extensor origin was done by correlating with visual analog scale (VAS) pain intensity during resisted wrist extension isometric contraction. Due to the non-normal distribution in baseline and post-test VAS pain intensity during resisted wrist extension isometric contraction. Therefore, the Spearman correlation coefficient was used. The interpretation of the correlation coefficient is as follows; r <0.3 is weak, 0.3 to 0.5 moderate, and >0.5 strong correlation [21].
Results
The demographic and baseline data of the participants are shown in Table 1.
Demographic and baseline data of the participants (n = 18)
Demographic and baseline data of the participants (n = 18)
SD = standard deviation; BMI = body mass index; VAS = visual analog scale; mm = millimeters; QuickDASH = abbreviated version of the original disabilities of the arm, shoulder and hand (DASH) questionnaire
For test-retest reliability, it was reported as good reliability for SkBF at lateral epicondyle (ICC3,1 = 0.899, P < 0.001), and at common extensor origin (ICC3,1 = 0.803, P = 0.003) as shown in Table 2.
Test-retest reliability (ICC3,1) of SkBF at lateral epicondyle and common extensor origin (n = 9)
SkBF = skin blood flow; PU = perfusion unit; ICC3,1 = intraclass correlation coefficient of the two-way mixed model, consistency; CI = confidence interval.
For side-to-side difference, the comparisons between the ipsilateral and contralateral sides were done, they were significantly different between the two sides at lateral epicondyle (t = 6.037, P < 0.001), and at common extensor origin (t = 5.501, P < 0.001) as shown in Table 3.
Side-to-side difference between ipsilateral and contralateral sides (n = 18)
SD = standard deviation; SkBF = skin blood flow; PU = perfusion unit; P-value from paired t-test.
For responsiveness as shown in Table 4, the computations were done according to the baseline and post-test data. The change score for SkBF (PU) at lateral epicondyle was – 8.04, while at common extensor origin was – 3.54. The statistically significant differences between the baseline and post-test data were found at the lateral epicondyle (t = 3.234, P = 0.005) and common extensor origin (t = 3.212, P = 0.005). The ES and SRM were reported as – 0.78 and – 0.76 for SkBF at the lateral epicondyle, and – 0.71 and – 0.76 for SkBF at common extensor origin, respectively.
Responsiveness of SkBF at lateral epicondyle and common extensor origin at the ipsilateral side (n = 18)
SkBF = skin blood flow; PU = perfusion unit; SD = standard deviation; SRM = standardized response mean; P-value from paired t-test.
Concurrent validity, as presented in Table 5, reported a strong correlation between SkBF at common extensor origin and VAS pain intensity during resisted wrist extension isometric contraction (Spearman r = – 0.637, P = 0.005). While weak correlations were found between SkBF at lateral epicondyle and VAS pain intensity either at rest or during resisted wrist extension isometric contraction, also, SkBF at common extensor origin and VAS pain intensity at rest.
Concurrent validity of baseline SkBF at ipsilateral lateral epicondyle and common extensor origin with baseline visual analog scale (VAS) pain intensity at rest and during resisted wrist extension isometric contraction determined by Spearman correlation coefficient (n = 18)
SkBF = skin blood flow; PU = perfusion unit; VAS = visual analog scale; mm = millimeters.
This study aimed to investigate the test-retest reliability, side-to-side difference, responsiveness, and concurrent validity of SkBF measured by LDF in individuals with acute or subacute LET. Mean age of our participants was 44±15 years old, 10 women (55.6%) and 8 men (44.4%), which corresponded to the previous research stating that the occurrence of LET is most common in the 4th and 5th decades of life, regardless of sex [13]. The duration of symptoms was about 26 days on average, which belonged to acute or subacute stages of LET since the criteria for chronic LET is greater than 3 months [12].
Test-retest reliability, we did by averaging the value of SkBF during the period of 3-minute data collection to reduce the variability [14] since the data of SkBF has high variability because of its sensitivity to the change of local microcirculation of targeted tissues. With this technique, we found good test-retest reliability for both locations, lateral epicondyle (ICC3,1 = 0.899) and common extensor origin (ICC3,1 = 0.803). However, we did not study inter-tester reliability because there was a measurement researcher in this study. The future study can investigate inter-tester reliability if that study requires more than one measurement researcher to collect SkBF. The good test-retest reliability in this study corresponded with the previous study [11] reporting ICC3,5 of 0.89 for investigating mean tissue blood flow at the lumbosacral area for a measurement period of 5 minutes.
We detected significant side-to-side differences for both locations between the ipsilateral and contralateral sides of LET. The ipsilateral side had higher blood flow compared to the contralateral side. This finding demonstrated the increased local microcirculation on the targeted tissue with LET at the early stages. The increased microcirculation as the response to the inflammatory process includes the increase of vascular permeability and the rate of proliferation of blood vessels [22].
The responsiveness is meaningful for the rehabilitation process because it can demonstrate the patient’s recovery over time [15], representing the clinical significance in addition to the statistical significance [17, 23]. In this study, we found that SkBF at both locations was responsive as reported with the ES and SRM ranging from – 0.71 to – 0.78. The change score for lateral epicondyle was – 8.04 and for common extensor origin was – 3.54 with a statistically significant difference. The change score was higher at the lateral epicondyle than that of common extensor origin, which represented the greater extent of the variability of microcirculation on the skin over the lateral epicondyle rather than the common extensor origin. The difference might be due to the density of capillaries between the two areas and their reactivity to the inflammatory process, which needs further investigation.
We found a strong correlation only between SkBF at common extensor origin and the VAS pain intensity during resisted wrist extension isometric contraction at the post-test session. The post-test SkBF values were less than the baseline SkBF values, similar to the contralateral side at the baseline, which represented the decrease of microcirculation in the targeted areas. The reduced SkBF at the post-test session represented the change of the inflammatory process to the next phase, by which the reduction or cessation of tissue infiltration and the initiation of healing can be observed [7]. However, the correlations at baseline sessions between SkBF at both locations and VAS pain intensity during resisted wrist extension isometric contraction were low. This might be due to the high variability of blood perfusion during the beginning period of LET. Therefore, the use of SkBF as a clinical outcome should be considered together with other clinical outcomes such as pain, disability, and other associated variables for more understanding of the patient’s health status.
The limitation of this study, SkBF seems to be appropriate to be used for superficial structures, however, this technique is probably limited for deep structures, other techniques or instruments should be considered instead. Another consideration is about the duration of onset, the chronic state awaits further investigation. The sample size of this study was only 18, this was from sample size calculation. Future studies can recruit more subjects with wider aspects of clinical characteristics of LET such as different onset or different subgroups of the severity of tendinopathy [24].
In conclusion, the assessment of LET, SkBF can be considered since it was proven to be reliable and responsive for investigating the elbows with acute or subacute LET. However, the concurrent validity was limited, it was found only for SkBF at common extensor origin correlating with pain during resisted wrist extension isometric contraction. The objective assessment in the clinical setting for individuals with acute or subacute LET, and other clinical outcomes such as pain, disability, and related functional tests should be added for more understanding.
Footnotes
Conflict of interest
The authors have no conflict of interest to report.