Is the YES/NO classification accurate in screening scapular dyskinesis in asymptomatic individuals? – A novel validation study utilizing surface electromyography as a surrogate measure in identifying movement asymmetries

ABSTRACT

Background

Scapular dyskinesis is a known risk factor for shoulder pain, making it important to screen for prevention. Physical therapists screen scapular dyskinesis by visually comparing asymmetries in scapular movement during overhead reach using the Scapular Dyskinesis Test Yes/No classification (Y/N). Although scapular kinematics has been used to quantify scapular dyskinesis, current measurement techniques are inaccurate. Optimal scapular muscle activity is crucial for normal shoulder function and is measured using surface electromyography (sEMG). Research suggests that impaired scapular muscles can lead to scapular dyskinesis. Despite kinematics being a poor reference standard, there is currently no validated method to identify movement asymmetries using muscle activity as an alternative. We utilized sEMG to establish Y/N’s validity. We hypothesized that Y/N is a valid tool using sEMG as a viable surrogate measure for identifying scapular dyskinesis.

Methods

We employed a known-groups (symmetrical vs. asymmetrical shoulders) validity design following the Standards for Reporting Diagnostic Accuracy Studies. Seventy-two asymptomatic subjects were evaluated using Y/N as the index test and sEMG as the reference standard. We created a criterion to assign the sEMG as the reference standard to establish the known groups. We calculated the sensitivity (Sn), specificity (Sp), positive and negative predictive values (PPV, NPV), likelihood ratios (LR+, LR-), and diagnostic odds ratio (DOR) using a 2 × 2 table analysis.

Results

The diagnostic accuracy values were Sn = 0.56 (0.37–0.74), Sp = 0.36 (0.08–0.65), PPV = 0.68 (0.49–0.88), NPV = 0.25 (0.04–0.46), LR+ = 0.87 (0.50–1.53), and LR- = 1.22 (0.50–2.97).

Conclusion

The Y/N’s diagnostic accuracy was poor against the sEMG, suggesting clinicians should rely less on Y/N to screen scapular dyskinesis in the asymptomatic population. Our study demonstrated that sEMG might be a suitable alternative as a reference standard in validating methods designed to screen movement asymmetries.

Introduction

Scapular dyskinesis is defined as altered scapular position or movement patterns during functional activities [1,2]. It is commonly associated with shoulder pain [3–5], but it may also be present in asymptomatic individuals [6–8]. More recent evidence suggests that scapular dyskinesis is a risk factor for shoulder pain [9] suggesting the need for screening as a preventive measure.

Physical therapists screen for scapular dyskinesis by visually comparing scapular movement asymmetries in overhead reach using the Scapular Dyskinesis Test [2]. The patient performs repeated shoulder elevation and lowering with weights on both hands, while the therapist observes scapular motion behind. Dyskinesis is categorized into three types: Type 1: Prominence of the inferior angle, Type 2: Prominence of the medial border, and Type 3: Dysrhythmia or excessive or premature movement of the scapula, each observed on a single plane of motion [10]. However, the number of possible abnormal movement patterns and combinations may make it difficult for therapists to agree on a final label. An alternative method known as the Yes/No classification (Y/N) simplifies this process by merely identifying the presence or absence of asymmetry between the shoulders. This approach is more inclusive and does not require therapists to observe multiple planes of motion, making it more reliable [11].

Considering a movement impairment, scapular kinematics have been used to validate [11,12] and quantify the presence of scapular dyskinesis, particularly in shoulder pathologies [13–16]. However, all the current scapular kinematic measurement techniques are inaccurate [17,18]. The thick layer of soft tissue covering the scapula leads to movement artifacts of at least 5º on shoulder angles between 90–120º and more on greater angles during kinematic measurements [19,20].

Optimal muscle activity of the upper (UT), lower (LT), and middle (MT) trapezii and serratus anterior (SA), collectively known as scapular muscles, is responsible for the scapular motion required to maintain normal shoulder function [21]. Muscle activity is accurately measured as amplitude by surface electromyography (sEMG) [22]. Surface electromyographic studies have significantly increased our understanding of shoulder control of the scapular muscles [23–25]. Data suggest impaired scapular muscles contribute to scapular dyskinesis [3,15,16,24,26–28]. Although kinematics is known to be a poor reference standard, no method, including the Y/N, has been validated using muscle activity as an alternative reference standard to identify movement asymmetries.

Therefore, the study’s aim was to establish Y/N’s construct validity amongst asymptomatic individuals by utilizing sEMG. We hypothesized that Y/N is a valid tool in identifying scapular dyskinesis using sEMG to quantify muscle activity as a viable surrogate measure for identifying movement asymmetries.

Methods

Study design

A known-groups validity design was employed, involving a prospective blind comparison of the Y/N (index test) and the sEMG (reference standard) in a consecutive series of subjects from a relevant clinical population (asymptomatic).

We selected sEMG as the reference standard due to the inadequacy of scapular kinematic measurements in demonstrating reliability as a gold standard [17,18]. In contrast, electromyography shows promise as it continues to enhance our understanding of the control exerted by the scapular muscles on the shoulder [23–25].

The study was designed in accordance with the 2015 Standards for Reporting Diagnostic Accuracy Studies (STARD) [29].

Subjects

Asymptomatic adults between 18 and 35 years old were included using word of mouth and referrals in the State of Georgia, United States of America in 2020. Table 1 summarizes the exclusion criteria. Approval was obtained from the Institutional Review Board (IRB) of Augusta University, and all subjects read and signed a consent form before participating in our study. A screening tool was used to identify potential subjects, including existing medical problems, medications, and pain ratings. Potential subjects were later invited to undergo evaluation via the Y/N for eligibility, described next.

Table 1.

Exclusion criteria.

Any of the following:
Shoulder pain with activity of 2/10 or greater on the numeric pain rating scale
History of shoulder pain within the past year
Adhesive capsulitis, defined as a loss of greater than 50% in passive shoulder range of motion in shoulder external rotation and one other plane of motion
Previous shoulder surgery within the past year
History of shoulder fracture
Systemic musculoskeletal disease (rheumatoid arthritis, fibromyalgia, etc.)
Shoulder pain that was reproduced with active/passive cervical spine motion.

Procedures and instrumentation

Protocol for the scapular dyskinesis test Yes/No classification

To identify eligible subjects, the Y/N was performed and video recorded for later evaluation for the presence or absence of scapular dyskinesis (Figure 1). Male participants were asked to remove their shirts, while women wore sports bras to expose both scapulae. Using a metronome at a rate of 60 bpm, participants performed five consecutive nonstop repetitions of bilateral, active, and weighted 120º of shoulder flexion using dumbbells based on their body weight: 1.4 kg (3 lb) for those weighing <68.1 kg (150 lb) and 2.3 kg (5 lb) for those >68.1 kg (150 lb) according to the Scapular Dyskinesis Test protocol by McClure at al [12,30].

Figure 1.

Scapular dyskinesis test Yes/No classification and video recording set-up.

A guide was placed in front of the subjects (2 ft. from the subject’s toes) to standardize the shoulder flexion and assure accuracy in all five repetitions. The guide was a pole made of an eight-foot PVC pipe on a wooden base. A spring clamp with handles wrapped with bright neon orange tape was clamped on the pole for easy visibility. Subjects’ shoulders were passively elevated to align with a goniometer (fixed at 120º), and they were asked to hold their shoulders in that position. The clamp was moved roughly at the level of the subjects’ middle fingers or the level they would remember to raise their arms during the test. To establish reliability between repetitions, after determining the clamp’s ideal height on the pole, subjects were asked to put their arms down to their sides, raise back up again at the clamp level, and hold. The fixed goniometer was placed back at the shoulders one at a time to check if the shoulders were aligned with the goniometer. If not, this process was repeated until the subjects accurately elevated both arms to align with the goniometer.

To record the movement, a high-definition digital camera on a tripod equipped with lighting was set up behind the participant 1 m away at the level of the seventh thoracic spinous process (between the inferior angles of the scapulae). Each video was saved on an MP4 format labeled by their respective unidentified subject number assigned earlier during the consent process. All videos were held on a secure Box folder (server) provided by the IRB. Videos were later evaluated using Y/N to determine the presence (Yes, asymmetrical shoulders) or absence (No, symmetrical shoulders) of scapular dyskinesis.

The definition of operational terms is as follows:

Yes – Scapular dyskinesis is present (asymmetrical shoulders). Either or both of the following motion abnormalities may be present on either shoulder: 1) dysrhythmia – the scapula demonstrates premature or excessive elevation or protraction, non-smooth or stuttering motion during arm elevation or lowering, or rapid downward rotation during arm lowering; or 2) winging – the scapula’s medial border or inferior angle is posteriorly displaced away from the posterior thorax.

NO – Scapular dyskinesis is not present (symmetrical shoulders). Both scapulae are stable with minimal motion during the initial 30º to 60º of shoulder elevation. Smoothly and continuously rotates upward during elevation, and smoothly and continuously rotates downward during humeral lowering. No evidence of winging is present.

Surface electromyography protocol

While subjects performed the Y/N and video recorded, muscle activities were simultaneously collected from the eight scapular muscles of both shoulders of each subject using the Ultium® wireless sEMG system (Noraxon U.S.A. Inc., Scottsdale, Arizona). All eight sensors simultaneously collected and transmitted muscle activity in real-time, sampled at 2000 hz. The system’s receiver received the signal with a 24-bit internal sampling resolution, common-mode rejection ratio < −100 dB, input impedance exceeding 1,000 MΩ, and sEMG resolution between 0.3 and 1.1 μV. Dual disposable wet-gel self-adhesive Ag/AgCl snap electrodes (Noraxon U.S.A. Inc., Scottsdale, Arizona), with an interelectrode distance of approximately 2 cm, were donned on the eight muscles as recommended by SENIAM [31] and on the SA, as suggested by Boettcher [32]. Table 2 shows the landmarks for each muscle’s electrode placement location. The sEMG sensors and matching electrodes were attached using the provided wire leads (Figure 2).

Table 2.

Surface electromyography electrode placement location landmarks.

Muscle	Electrode Location
UT	Midway between the acromion C7 spinous process
MT	Midway between the medial border of the scapula and the spine, at the level of T3
LT	2/3 on the line from the root of the scapula to the T8 spinous process
SA	Over the 7th rib in the anterior axillary line

UT = Upper trapezius; MT = Middle trapezius; LT = Lower trapezius; SA = Serratus anterior.

Figure 2.

Electrode and sensor placement on the subject.

Raw sEMG data were processed and analyzed in real-time with myoRESEARCH® software (MR3) and myoMUSCLE® software module (Noraxon U.S.A. Inc., Scottsdale, Arizona). The sEMG baseline check feature was activated before data collection. While the subject remained still, the levels were visually checked. If any single or a combination of the following were detected: impedance above 100 kΩ (ideally <10 kΩ), root-mean-square (RMS) >5 μV, and a median frequency signal between 50 and 60 hz, it may suggest that skin prep and electrode placement may not be sufficient to produce a clean signal. If so, the process was repeated until the baseline levels were below the cutoff values.

For normalization, subjects performed two maximum voluntary isometric contractions (MVIC) to reference each muscle’s electrical activity [33,34]. Muscles were tested as described in previous studies [35,36]. Using a metronome at 60 bpm, subjects generated a maximum force over 3 s, followed by an additional 5 s as this force was maintained [37]. A 30-s rest period was given between trials to minimize fatigue. The highest ½ second of activity represented 100% activation.

For data reduction, the software’s online processing was activated. The following were applied to the raw sEMG signal in real time during data collection: 1) full-wave rectification; 2) bandpass filter between 20 and 500 hz; 3) RMS smoothing over a 300-ms window for dynamic contractions [38]; 4) and ECG artifact elimination using the software’s algorithm on a minimum of five clean heartbeats. For each subject, the mean EMG activity from the three middle trials was ensemble-averaged and expressed as 100% MVIC (% MVIC) to represent the mean muscular activity of the shoulder muscles during the test.

Establishment of sEMG symmetrical and asymmetrical groups (known groups)

The sEMG was assigned as the reference standard in identifying symmetrical and asymmetrical known groups. The basic principle was to ‘hand over’ the expertise of a human (expert PTs) to the machine (sEMG). This was accomplished by creating a sEMG data criterion to determine what constitutes symmetrical versus asymmetrical shoulders of asymptomatic subjects. The criterion was later used to label the subjects into these known groups. There were three steps involved in establishing this using a decision tree (Figure 3):

Figure 3.

Decision tree for the establishment of sEMG symmetrical and asymmetrical groups (known groups).

Note. SYM = symmetrical; ASYM = asymmetrical; ROC = receiver operating curve analysis; LR+ = positive likelihood ratio; LR- = negative likelihood ratio

Step 1 – The Expertise: Twenty (10 symmetrical and 10 asymmetrical) of the eligible subjects were labeled by two musculoskeletal physical therapy (PT) experts (κ = 0.85) from our previously published study [39].

Step 2 – Creating the Criterion (the ‘hand-over’): Using the sEMG data (mean difference) from the symmetrical and asymmetrical subjects grouped by the expert PTs in Step 1, a criterion was created by establishing cutoff levels from the % MVIC difference of each muscle pair groups. We employed the Receiver Operating Curve analyses (ROC) to establish this criterion. The values with the best balance between sensitivity (Sn) and specificity (Sp), highest positive likelihood ratio (LR+), and lowest negative likelihood ratio (LR-) were chosen as the cutoff level for each muscle. Receiver operating curves were calculated with SPSS version 27 (IBM Corp., Armonk, NY). The cutoff values established by ROCs for each muscle were used to decide whether individual muscles were symmetrical or asymmetrical. Table 3 summarizes the cutoff values established by ROCs for each muscle used to decide individual muscles if it was symmetrical or asymmetrical.

Table 3.

Receiver operating curve analyses cut-off levels.

Muscle	Cut-off
UT	7.10
MT	2.80
LT	6.75
SA	20.84

UT = upper trapezius; MT = middle trapezius; LT = lower trapezius; SA = serratus anterior. Values are in % MVIC.

Any sEMG value from the subjects’ muscles greater or equal to its respective cutoff level was considered asymmetrical for that muscle. Conversely, any sEMG value less than its respective cutoff level was considered symmetrical. A majority rule was adopted to determine every subject’s final label – greater than 50%. This meant that a minimum of three out of four muscles must be labeled symmetrical or asymmetrical for every subject to get a final label of either symmetrical or asymmetrical according to the sEMG. Naturally, equivalent labels of the resulting two muscle pairs – two symmetrical and two asymmetrical (e.g. UT and MT = symmetrical; LT and SA = asymmetrical) were possible. To address subjects exhibiting a 50–50 split between a symmetrical and asymmetrical set of muscles, posttest probabilities of positive (LR+) and negative likelihood ratios (LR-) for every possible muscle pair were used. Muscle pairs with the highest posttest probabilities for LR+ and lowest posttest probabilities for LR- were used as the benchmarks to decide on the final label of such test subjects. Posttest probabilities were calculated based on a pretest probability of 50% using Excel version 16.46 (Microsoft, Corp.). Table 4 summarizes all possible muscle pairs’ posttest probabilities.

Table 4.

Possible muscle pairs’ posttest probabilities.

	Post-test Probability
LR	UT+MT	UT+LT	UT+SA	MT+LT	MT+SA	LT+SA
LR+	96.55%	94.12%	94.12%	96.55%	96.55%	94.12%
LR-	7.62%	5.88%	5.88%	7.62%	7.62%	5.88%

UT = upper trapezius; MT = middle trapezius; LT = lower trapezius; SA = serratus anterior; LR = likelihood ratio; LR+ = positive likelihood ratio; LR- = negative likelihood ratio. Posttest probabilities were based on 50% pretest probability.

*Percentage value in bold are highest or lowest posttest LR+ or LR- probabilities, respectively.

To apply these benchmarks, the following rules were adopted: 1) Those whose LR+ posttest probability of any of the muscle pairs were higher than the other pair, the final overall label for that subject was asymmetrical; or 2) those whose LR- posttest probability of any of the muscle pair were lower than the other pair, the final overall label for that subject was symmetrical; or 3) those whose both muscle pairs were still opposing after applying the criteria, the subject was removed from the final validation.

Step 3 – Labeling: The criterion was later applied to the remaining eligible subjects for final validation – labeling to establish symmetrical from asymmetrical known groups according to the sEMG.

Establishment of Y/N symmetrical and asymmetrical groups

The expert PTs rated the remaining eligible subjects’ videos. The ratings were tallied for final labeling. Subjects with disagreements between raters were resolved by a backup musculoskeletal PT rater with similar experience. All raters were blinded to each other’s ratings and the sEMG results.

Sample size estimation

To establish the construct validity of Y/N using sEMG, the sample size was estimated through previous studies [11,12] and PASS software (NCSS, Utah, USA) sample size estimation for sensitivity (Sn) and specificity (Sp) analyses [40]. Based on the prevalence of scapular dyskinesis between 60% and 70% [6–8] with a significance level of 0.05 and power of 80%, the minimum sample size required to test the null hypothesis of Sn/Sp = 0.50 versus the alternative hypothesis Sn/Sp = 0.80 was 30 subjects (20 asymmetrical with scapular dyskinesis, 10 symmetrical without scapular dyskinesis).

Statistical methods

To validate the Y/N (index test) against the sEMG (reference standard), a standard two-by-two (2 × 2) contingency table analysis was utilized, wherein symmetrical and asymmetrical groups from the sEMG were columns, while symmetrical and asymmetrical groups of the Y/N were rows. Diagnostic accuracy metrics, including Sn, Sp, LR+, LR-, positive predictive value (PPV), and negative predictive value (NPV), were calculated in Excel version 16.46 (Microsoft, Corp.).

Results

Seventy-two out of 100 consecutive asymptomatic participants were screened and consented for eligibility. The demographic characteristics of the subjects were as follows: The mean age was 24 (±3) years, 54% (39) were women, 64% (46) were right-handed, and 51% (37) had a history of overhead sports. Figure 4 summarizes the flow of subjects through the study. Table 5 is the 2 × 2 contingency table used for computing diagnostic accuracy metrics. Table 6 summarizes the diagnostic accuracy metrics derived from the 2 × 2 table analysis.

Figure 4.

Participants’ flow through the study.

Note. SYM = symmetrical; ASYM = asymmetrical; Y/N = scapular dyskinesis test yes-no classification; sEMG = surface electromyography

Table 5.

2×2 contingency table.

		sEMG
		+ (ASYM)		- (SYM)	Total
Y/N	+ (ASYM)	15		7	22
	- (SYM)	12		4	16
Total		27		11	38

Y/N = scapular dyskinesis test yes-no classification; sEMG = surface electromyography; SYM = symmetrical; ASYM = asymmetrical.

Table 6.

Values of diagnostic accuracy metrics.

Diagnostic Accuracy Metric	Value	CI
Sn	0.56	0.37–0.74
Sp	0.36	0.08–0.65
PPV	0.68	0.49–0.88
NPV	0.25	0.04–0.46
LR+	0.87	0.50–1.53
LR-	1.22	0.50–2.97

Sp = specificity; SN = sensitivity; PPV = positive predictive value; NPV = negative predictive value; LT+ = positive likelihood ratio; LR- = negative likelihood ratio; CI = confidence interval.

Discussion

Our objective for this study was to establish Y/N’s construct validity among asymptomatic individuals by utilizing sEMG to quantify muscle activity as a viable surrogate measure for identifying movement asymmetries.

The results revealed that the Sp and NPV of Y/N are only 36% and 25%, respectively, indicating a higher risk of false-negative findings by failing to identify individuals who genuinely have symmetrical scapular motions (no dyskinesis) [41]. Conversely, the Y/N may be better in identifying those who genuinely have asymmetrical shoulders (with dyskinesis) by decreasing the risk of false-negative findings due to higher Sn (56%) and PPV (68%) [41]. These findings were similar to the validation study results by Uhl et al., even though the values of our study were much lower. Ul et al. concluded that the Y/N might be a good screening tool when used in the asymptomatic population due to high Sn (78%) and PPV (76%) despite low Sp (38%) and NPV (40%) [11]. Though the results were comparable, it is worth noting that the study by Uhl et al. used kinematics as the reference standard, including the study by Tate et al. [12] that initially validated the Y/N. With the lack of reliability of scapular kinematic measurements [17,18], it may not be appropriate to make a similar conclusion that the Y/N could be a valid tool. For this same reason and for the increasing contribution of electromyography in understanding the scapular muscles’ control of the shoulder, we chose the sEMG as the reference standard to validate the Y/N in our study.

Moreover, the results of the likelihood ratios further contradict the validity of the Y/N. Likelihood ratios combine Sn and Sp information into a fraction used to quantify shifts in probability [42]. According to Jaeschke et al., recommended ratios are LR+ >10 and LR- <0.1 for large and conclusive shifts in probability, while LR+ between 1 and 2 and LR- between 0.5 and 1 has a small and rarely significant change in probability shifts [43]. Based on an LR+ of 0.87 and LR- of 1.22, the Y/N cannot shift probabilities in either direction. LRs of Y/N should not be considered in the context of probability shifts. In their systematic review, Wright et al. drew a similar conclusion that none of the scapular physical examination tests, including the Y/N, demonstrated the ability to alter posttest probability to enable the diagnostic process [44]. However, all the validation studies included in the review by Wright et al. used symptomatic subjects and specific pathologies as reference standards instead of the deliberate use of asymptomatic subjects for the specific purpose of our study.

Although Y/N has excellent reliability, as demonstrated in our previous study [39], it lacks diagnostic accuracy in identifying shoulder asymmetries against the sEMG. Asymmetries detected by the Y/N as an observed movement do not appear to correlate with the expected scapular muscle activity. Huang et al. (2017, 2015b) found similar findings in their investigation [26,45]. They concluded that a change in muscle activation might not correspond to scapular movement during arm elevation – a plausible explanation for why studies that specifically addressed scapular kinematics to treat shoulder pain conflicted. According to systematic reviews by Bury et al. and Reijneveld et al., employing scapula-focused treatments led to improvements in pain [46,47]. However, the effect on kinematics was inconsistent, with some studies showing no changes [48] and others reporting worsened kinematics [49,50]. The expected actions of scapular muscles, such as upward and downward rotators or internal and external rotators, might not accurately reflect the actual movement of the scapula during overhead activities. It is conceivable that scapular dyskinesis and scapular muscles may not be as closely related as conventionally understood. This raises consequential questions about whether movement asymmetries, even in healthy individuals, need to be addressed. If so, are the exercises for the muscles thought to be responsible appropriate? Scapular dyskinesis is just one example of impaired movement, along with other issues such as dynamic knee valgus, excessive lumbar rotation, and forward head posture during arm elevation. Perhaps what is often labeled as movement impairments may fall within the range of normal variability or could be interpreted from an alternative theoretical standpoint, such as those outlined in motor control theories.

Despite the unexpected results, our study demonstrated that the sEMG might be a capable alternative for identifying movement asymmetries by quantifying muscle activity. It may be more reflective of the true behavior of the muscles relative to the movement they produce, including those known as movement asymmetries, such as scapular dyskinesis.

Study limitations

The SDT is very subjective. It may be possible that all observers committed an expectation bias due to a known expected outcome. This may have influenced the scapular dyskinesis labeling because raters ‘see what they want to see.’ In this case, the presence of scapular dyskinesis.

Most of the experiments took place during the SARS-CoV-2 pandemic [51]. At its height, the rating period lasted 10 weeks, which may have introduced history and timing biases ranging from subject recruitment to rater performance.

Since the study used convenience sampling (sampling bias), this may contribute to the weak generalizability of the results. It may be possible that the sample was non-representative of the general population due to the nature of subject enrollment which is all volunteers (response bias).

Despite the strict protocols, random errors might have occurred when using sEMG. It may have been insensitive to measuring a particular outcome or unable to perform equally in different subjects due to body size, percentage of adipose tissue, etc. (spectrum bias). A crosstalk phenomenon [52] may have occurred when using sEMG. It is also possible that electrocardiographic artifacts were recorded alongside the signals intended only for the scapular muscles due to their proximity to the heart [53]. Nonetheless, the proprietary software algorithm used in our study has addressed this issue. In addition, strict sEMG protocol based on expert recommendations was followed, including motor point location, sensor placement parallel to the muscle, and vigorous skin preparation [31,32,36,54].

In the process of establishing the sEMG symmetrical and asymmetrical known groups, Step 2 (Creating the Criterion), also known as the hand-over process, included subjects (10 symmetrical, 10 asymmetrical) who were labeled by the expert PTs using the Y/N itself (index test). Although the criterion sources are recognized experts, the sEMG, which became the reference standard, still depended on the experts’ clinical judgment that utilized the same test in question (Y/N). This could be a potential source for review bias (reference judged by the examiner not blinded to the test result) and incorporation bias (a reference not independent of the test) [42]. To avoid this, the 20 subjects used in the handover process were not included in the final contingency table analysis.

Conclusion

The Y/N does not appear to be a valid test with clinical value when tested against sEMG. Therefore, the Y/N may not be a valid tool to screen scapular dyskinesis in asymptomatic individuals.

Although results show that Y/N may not be accurate in screening for scapular dyskinesis, it may be premature to suggest abandoning the test altogether. Since the study was conducted in a laboratory setting, it does not necessarily reflect how Y/N is typically performed in clinical practice. Strict laboratory protocols do not represent real-world scenarios. For this reason, it is not recommended to make a drastic change in clinical practice due to a single controversial result, as shown in our study. It may be more appropriate for clinicians to rely less on the Y/N when suspecting scapular dyskinesis in asymptomatic individuals.

Furthermore, our study may be the first to demonstrate that sEMG could be a suitable substitute as a reference standard in validating methods designed to screen movement asymmetries.

Biographies

Lawrence S. Ramiscal PT, DPT, Ph.D., is an Associate Professor and Director of Research and Faculty Development at Lincoln Memorial University’s Physical Therapy Program in Knoxville, Tennessee. A Board-Certified Orthopaedic Clinical Specialist since 2007 and an AAOMPT Fellow since 2009, he has over 25 years of experience in outpatient orthopedics, including over a decade as a private practice owner. Dr. Ramiscal’s research focuses on understanding orthopedic movement impairments through motor control theories, manual therapy effectiveness, and the reliability and validity of diagnostic tests. He is a founding board member and past vice president of the Philippine Association of Orthopaedic Manual Physical Therapists and serves as Editor-in-Chief of the Asian Journal of Physical Therapy.

Lori A. Bolgla PT, Ph.D., MAcc, ATC, is a Professor in the Department of Physical Therapy and holds the Kellett Chair in Allied Health Sciences at Augusta University in Augusta, Georgia. A licensed physical therapist and certified athletic trainer, her research centers on kinesiological applications of surface electromyography to improve neuromuscular function, injury prevention, and rehabilitation outcomes. Widely published and a sought-after speaker, Dr. Bolgla is also a dedicated educator and mentor, advancing evidence-based practices in physical therapy and allied health sciences.

Chad E. Cook PT, Ph.D., MBA, FAPTA is the Director of Clinical Facilitation Research and the Duke Center of Excellence in Manual and Manipulative Therapy in the Department of Orthopaedic Surgery at Duke University in Durham, North Carolina, where he is also a Professor and core faculty member. An AAOMPT Fellow for over 25 years, his research focuses on improving patient examination processes and validating clinical tools in physical therapy. Dr. Cook has authored three textbooks, published over 350 peer-reviewed articles, and delivered lectures worldwide on orthopedic examination and treatment. Renowned for his contributions to evidence-based practice, he is a leading figure in musculoskeletal care and manual therapy research.

John S. Magel PT, Ph.D., D.Sc., is a Research Associate Professor in the Department of Physical Therapy and Athletic Training at the University of Utah in Salt Lake City, Utah, where he obtained his Ph.D. in Rehabilitation Sciences in 2015. He completed his manual therapy fellowship and Doctor of Science in Physical Therapy from the US Army-Baylor University program in 2001. An AAOMPT Fellow, he received the AAOMPT Mennell Service Award in 2018. His research focuses on managing musculoskeletal conditions, and he received the Academy of Orthopaedic Physical Therapy’s Stephen J. Rose Excellence in Research Award. Currently, he leads an NIH-funded clinical trial on integrating mindfulness into physical therapy for chronic musculoskeletal pain and long-term opioid use.

Stephen A. Parada M.D., is a distinguished Orthopedic Surgeon and Professor in the Department of Orthopaedic Surgery at the Medical College of Georgia, Augusta University in Agusta, Georgia. He serves as Vice Chair for Orthopaedic Research and Chief of Orthopaedic Shoulder Surgery, specializing in shoulder surgery and sports medicine. Dr. Parada received the GME Exemplary Teaching Award from the Medical College of Georgia in 2020 and 2021 for his contributions as an educator and mentor to medical students and residents. He has also published extensively on shoulder arthroplasty outcomes and educational media for orthopedic conditions.

Raymond Chong Ph.D., is a Professor and Chair of the Department of Nutrition and Dietetics at Augusta University in Augusta, Georgia. He previously chaired the Department of Interdisciplinary Health Sciences from 2018 to 2023, overseeing the graduate Dietetics program and the Applied Health Sciences PhD and Master of Public Health programs. Dr. Chong is also a visiting professor at Université Paris-Sud 11 in France. He earned his BS in Physical Education in 1991, MS in 1993, and Ph.D. in Exercise and Movement Science in 1997 from the University of Oregon. His research focuses on human movement, aging, and neurological diseases like Parkinson’s. He has published over 60% of his peer-reviewed papers as either the first or senior author and serves as a reviewer for the U.S. Veterans Affairs Office of Research & Development.

Funding Statement

The author(s) reported that there is no funding associated with the work featured in this article.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/10669817.2024.2436402