The Diagnostic Accuracy of the Hyperextension–Internal Rotation (HERI) Test in Assessing Anterior Glenohumeral Instability

Abstract

Background: To assess the diagnostic accuracy of the hyperextension–internal rotation (HERI) test to assess anterior glenohumeral instability. Methods: This study evaluated a total of 100 patients. The HERI test was performed on both the stable and unstable shoulder of each patient. The range of extension (°) was measured by a goniometer on both shoulders. Statistical analysis was performed to assess the diagnostic accuracy of the HERI test. Results: A total of 89% of patients showed evidence of an increased extension angle on the abnormal side by a mean of 84.19° compared with the normal contralateral side with a mean of 71.23°. The mean difference in extension angles between the normal and abnormal sides was 12.96°. In diagnosing anterior glenohumeral instability, the HERI test had a sensitivity of 88.8% (95% CI, 79.6–98%), a specificity of 87.3% (95% CI, 80.7–93.9%), and an accuracy of 88%. The positive predictive value was 87% (95% CI, 83.4–92.5%); the negative predictive value was 89% (95% CI, 84.6–93.3%); the positive likelihood ratio was 6.97 (95% CI, 3.4–10.4); the negative likelihood ratio was 0.13 (95% CI, 0.08–0.17); and the diagnostic odds ratio was 54.15 (95% CI, 47.2–61.0). Conclusions: The results of this diagnostic study indicate that the HERI test is a highly sensitive and specific test in assessing anterior glenohumeral instability without causing pain or producing a sense of instability.

1. Introduction

The glenohumeral joint is the most mobile and most commonly dislocated joint; 95–97% of these dislocations are in an anterior direction [1,2]. Nearly half occurs in people aged 15–29 years, and young males are at a higher risk of shoulder dislocation and recurrence [3,4]. The risk of recurrence is 75% in males aged younger than 19 years, predominately in the first two years following the dislocation [4]. Shoulder joint stabilizers are divided into dynamic and static stabilizers; dynamic stabilizers include periscapular muscles and rotator interval and rotator cuff muscles. The rotator cuff muscle is located near the center of the joint rotation, thereby decreasing the sheer force and helping centralize the humeral head on the glenoid [5]. Static shoulder stabilizers include the labrum, articular congruity, negative intraarticular pressure, and glenohumeral ligaments. Multiple shoulder biomechanical and cadaveric studies have proven the role of glenohumeral ligaments, specifically the inferior glenohumeral ligament (IGHL) and capsule, in maintaining the functional stability of the shoulder [6,7,8,9,10,11,12]. The IGHL complex is a hammock-like structure composed of three characteristic parts: the anterior band, axillary pouch, and posterior band [13].

Various shoulder clinical tests have been developed by shoulder surgeons to differentiate between normal laxity and pathological shoulder instability [14], with the apprehension test and the relocation test being the most common [15,16]. These tests depend on either the reproducibility of symptoms (pain or a sense of instability) or relieving them to be considered positive tests, which is difficult to achieve for all patients due to a contracted muscle being secondary to pain, a fear of instability, or a combination of the two [17]. Some authors have considered pain as a diagnostic criterion with the clinical tests of apprehension, relocation, and surprise [18,19]. Although pain triggered during shoulder examination is an indication for ongoing pathology in the glenohumeral joint, it is considered an obstacle to achieving patient cooperation during the physical examination, especially to detecting subtle instability [20]. To avoid pain and apprehension, many other tests have been proposed in the literature; for example, Gagey et al. described the hyperabduction test for inferior glenohumeral ligament assessment by examining the abduction range of motion passively [21]. Moreover, Lafosse et al. [10] developed the hyperextension–internal rotation (HERI) (the hyperextension rotation interne in French) test in an attempt to demonstrate that patients with chronic anterior instability could have an increase in the range of the glenohumeral extension angle of the unstable joint compared to the contralateral normal shoulder due to an injury of the IGHL anterior band and inferior capsule during shoulder dislocation.

The purpose of this study was to assess the sensitivity, specificity, and accuracy of the HERI test in the diagnosis of chronic anterior shoulder instability as compared with magnetic resonance imaging (MRI) findings and to examine the use of the HERI test as a clinical diagnostic test to accurately diagnose injuries of the anterior band of the IGHL and inferior capsule. We hypothesized that the HERI test would be a high-quality diagnostic tool.

2. Materials and Methods

This was a prospective cohort study conducted on patients who presented with a history of chronic shoulder instability, which was defined as at least two episodes of shoulder dislocations that occurred on different occasions. We conducted this study in our academic institution from January 2018 through December 2019. Our inclusion criteria were any patient with unilateral chronic anterior shoulder instability with a stable contralateral shoulder; we excluded any patient with a previous history of shoulder surgery on either side or any type of shoulder instability other than anterior instability. Our study protocol was approved by our institutional review board and all participants signed an informed consent form before enrollment.

All patients underwent a thorough history, physical examination (including apprehension and HERI tests), and standard radiographs. As described by Lafosse et al. [10], the HERI test was performed on both shoulders of each patient starting with the normal side, then the abnormal side, by two fellowship-trained orthopedic surgeons separately, then the median of extension angles was calculated. The test was performed as follows: the examiner stood behind the patient to test the right shoulder, and stabilized the left upper limb by putting his left elbow over the left scapula of the patient with maximum elevation of the left upper limb to prevent bending forward, and maintained an upright position of the patient’s spine. The examiner grasped the patient’s right forearm in pronation and fully extended the elbow, then pulled it backward until reaching maximum resistance. In this position, the right glenohumeral joint was in hyperextension and internal rotation, where maximum tension was directed against the anterior band of the IGHL and inferior capsule without inducing dislocation or apprehension.

The HERI test was repeated on the contralateral shoulder and the calculated extension angles from both sides were compared (Figure 1). Extension angles were measured by a goniometer and read by either a sports medicine fellow or a senior orthopedic resident. During the apprehension test and the HERI test, we documented if patients felt pain, apprehension, or neither.

MRI was used as the gold standard to confirm the status of the labrum. All MRI scans were reviewed by the same musculoskeletal specialist radiologist. The MRI protocol was fixed for our patients, and involved imaging in the abduction–external rotation (ABER) position to put the anterior band of the inferior glenohumeral ligament (IGHL) and antero-inferior capsule under maximum tension.

Data were analyzed using Statistical Package for Social Studies (SPSS 22; IBM Corp., New York, NY, USA). Continuous variables were expressed as mean ± standard deviation. Categorical variables were expressed as percentages. An AUROC was calculated using a ROC curve. Sensitivity, specificity, the area under the curve (AUC), the short distance method, and the Youden index method were used to determine the cut-off value for the affected side. A p-value < 0.05 was considered statistically significant.

3. Results

Overall, 100 patients met our inclusion criteria; 90 patients were male (90%) and 10 were female (10%), with a mean age of 25.87 years (range, 17–33 years). The mean number of dislocation episodes was 4.3, with a standard deviation of 1.6. In regard to the laterality of dislocation, 90 patients suffered from right shoulder instability, while only 10 had left shoulder instability, as shown in

Table 1. Characteristics of participants.

		Number	%
Gender	Male	90.00	90.00
	Female	10.00	10.00
Side of Instability	Right	90.00	90.00
	Left	10.00	10.00
Anterior Apprehension Test	Positive	98.00	98.00
	Negative	2.00	2.00
X-ray Results	Normal	97.00	97.00
	Hill–Sachs	3.00	3.00
MRI	Normal	4.00	4.084
	Branket	94.00	95.92
Level of Activity	Athletic	19.00	19.19
	Non Athletic	80.00	80.81
		Mean	± SD
Age		25.87	7.90
Number of Times Dislocated		4.32	1.68
Duration of Instability		2.00	1.62
Extension Angle of Affected Side		84.19	5.78
Extension Angle of Normal Side		71.23	5.39
Difference Between Affected and Normal Extension Angles		12.96	5.89

.

The mean extension angle of the unstable shoulders was 84.19° (78.41–89.97°), and the normal shoulders’ mean extension angle was 71.23° (65.84–76.62°). The mean difference of extension angles between the normal and abnormal sides was 12.96°. The distribution of our patients based on the difference in extension angles is shown in

Table 2. Distribution of patients based on the difference in extension angles.

Extension Angle	Number of Patients
No Difference	11
5–10°	5
10–20°	82
>20°	2

.

In our study, using the receiver operating characteristic curve, we calculated the Yuden index, which showed an optimal cutoff of 75.5, as shown in

Table 3. Sensitivity and specificity of the HERI test.

Optimal Cutoff	Sensitivity	Specificity	AUC	Short Distance	Yuden Index
75.5	0.888	0.873	0.923	0.029	0.760
Area Under the Curve
Area		p Value	Asymptomatic 95% Confidence Interval
			Lower Bound	Upper Bound
0.923		<0.001	0.884	0.962

, and the area under the receiver operating characteristic curve (AUROC) was 92.3% (Figure 2). Moreover, the HERI test was sensitive in 88.8% (95% CI, 79.6–98%) and specific in 87.3% (95% CI, 80.7–93.9%) of cases, as shown in Table 3. Furthermore, the positive predictive value of the HERI test was 87% (95% CI, 83.4–92.5%) and the negative predictive value was 89% (95% CI, 84.6–93.3%); the positive likelihood ratio was 6.97 (95% CI, 3.4–10.4), while the negative likelihood ratio was 0.13 (95% CI, 0.08–0.17) and the diagnostic odds ratio was 54.15 (95% CI, 47.2–61.0).

None of our patients experienced pain or apprehension when the HERI test was performed on the abnormal side, while 73 patients were apprehensive, 14 patients felt pain, and 11 experienced both during the anterior apprehension test, and the remaining two had negative apprehension tests.

All patients underwent basic radiological imaging, including anteroposterior, lateral, and axillary views, upon their initial clinic visit. Of these, 97 patients had normal X-rays, 3 patients only showed a Hill–Sachs lesion, and 1 patient showed an intra-articular osseous fragment. The mean defect sizes were 4.5 ± 3.0 mm in depth and 9.5 ± 4.0 mm in width. MRI was performed for all patients involved in this study, with results as follows: a Bankart lesion was seen in 94 patients (94%), of whom 6 patients showed an osseous Bankart lesion < 25% of the glenoid diameter and only 1 patient showed an osseous Bankart lesion > 25% of the glenoid diameter, which was evident in plain X-rays. No patient was reported to have a rotator cuff tear in our study group. We followed patients who were operated on during our study; 75 patients were taken for diagnostic and therapeutic arthroscopy and all of them showed evidence of a torn labrum and an anterior band of the IGHL.

4. Discussion

The glenohumeral joint is the most mobile joint in the human body; this wide range of motion puts the joint at risk of instability and dislocation. The glenohumeral joint has two groups of stabilizers: statics and dynamic stabilizers. The static stabilizers of the glenohumeral joints are the joint’s bony anatomy, labrum, negative intraarticular pressure, and glenohumeral ligaments [22]. The IGHL is a critical structure for shoulder stabilization: it prevents anterior shoulder instability by resisting excessive translation in an abducted and externally rotated shoulder, which is the most common position in which anterior shoulder dislocation occurs [3].

The shoulder apprehension test is one of the well-known and universally used examinations for anterior shoulder instability; it is positive if the patient experiences pain, discomfort, or a sense of instability during examination. This test has high specificity, ranging from 71% to 98%; unfortunately, its sensitivity varied widely in the literature, ranging from 52% to 98% [23]. Furthermore, relocation and release tests are usually performed in the same setting of an apprehension examination; the relocation test is positive if the patient’s sense of apprehension is relived after applying anterior-to-posterior pressure directly on the humeral head, while a release or surprise test is when the examiner suddenly removes their hand and the patient experiences apprehension again. Relocation and release tests had reported sensitivities of 45.83% and 63.89%, respectively [11]. Moreover, the load-shift test is also a widely used test in examining anterior shoulder instability, with 100% specificity. However, it can miss half of the patients, as its sensitivity is only 50% [14] compared to the most important finding of the present study: the HERI test showed 88.8% sensitivity and 87.3% specificity. Furthermore, none of our patients or Lafosse et al.’s [10] patients experienced pain or apprehension.

In our study group, the mean difference of extension angles between the normal and abnormal sides was 12.96° compared to 14.5°, which was the reported mean difference of the extension angles of Lafosse’s clinical study group [10]. We had 11 patients who had similar extension angles on both shoulders. Although there is no clear clinical explanation, we have noticed those patients had only two or three episodes of instability over a 6-month period. Thus, we assume that with each instability episode the capsule and ligaments were being stretched out. The difference in extension angles was bigger in unstable shoulders with a higher number of dislocation episodes compared to unstable shoulders with fewer episodes of dislocation. Interestingly, due to the internal rotation of the humerus during the test, which does not allow the engagement of the humeral head over the rim of the glenoid, the presence or absence of bony loss in either the glenoid or humeral head have no clinical effect on HERI test results.

This study was initiated to critically compare the HERI test with MRI findings that confirm the presence of a Bankart lesion in young patients with recurrent anterior instability of the shoulder.

The HERI test has limited clinical value when assessing patients with a previous history of shoulder surgeries, hyperlaxity, multidirectional instability, a recent instability event, or bilateral anterior shoulder instability. Additionally, the HERI test cannot be performed by a single examiner, as an assistant is required to measure the extension angle.

There were some limitations in this study. First, our study included only patients who were symptomatic and sought medical care due to reduced functionality. So, HERI test accuracy in asymptomatic and first-time shoulder dislocators is unknown. Second, MRI was our radiological reference, but not MR arthrography. However, we think conventional MRI scans are utilized to characterize lesions developed in chronic glenohumeral instability, which is a clinical diagnosis. Further studies need to be performed to validate the clinical accuracy of the HERI test coupled with a group of other common physical examination tests used to clinically assess anterior glenohumeral instability.

5. Conclusions

In conclusion, the results of this diagnostic study support the hypothesis that the HERI test is a highly sensitive and specific diagnostic tool for assessing the status of anterior shoulder instability. Moreover, it is a pain-free, simple physical examination that does not produce a sense of instability in patients.