Video Analysis of Primary Shoulder Dislocations in Rugby Tackles.

BACKGROUND: Characteristics of rugby tackles that lead to primary anterior shoulder dislocation remain unclear.
PURPOSE: To clarify the characteristics of tackling that lead to shoulder dislocation and to assess the correlation between the mechanism of injury and morphological damage of the glenoid. STUDY
DESIGN: Case series; Level of evidence, 4.
METHODS: Eleven elite rugby players who sustained primary anterior shoulder dislocation due to one-on-one tackling between 2001 and 2014 were included. Using an assessment system, the tackler's movement, posture, and shoulder and head position were evaluated in each phase of tackling. Based on 3-dimensional computed tomography, the glenoid of the affected shoulder was classified into 3 types: intact, erosion, and bone defect. Orientation of the glenoid defect and presence of Hill-Sachs lesion were also evaluated.
RESULTS: Eleven tackles that led to primary shoulder dislocation were divided into hand, arm, and shoulder tackle types based on the site at which the tackler contacted the ball carrier initially. In hand and arm tackles, the tackler's shoulder joint was forcibly moved to horizontal abduction by the impact of his upper limb, which appeared to result from an inappropriate approach to the ball carrier. In shoulder tackles, the tackler's head was lowered and was in front of the ball carrier at impact. There was no significant correlation between tackle types and the characteristics of bony lesions of the shoulder.
CONCLUSION: Although the precise mechanism of primary anterior shoulder dislocation could not be estimated from this single-view analysis, failure of individual tackling leading to injury is not uniform and can be caused by 2 main factors: failure of approach followed by an extended arm position or inappropriate posture of the tackler at impact, such as a lowered head in front of the opponent. These findings indicate that injury mechanisms should be assessed for each type of tackle, as it is unknown whether external force to the glenoid is different in each mechanism during shoulder dislocation.

The incidence and severity of shoulder dislocation among rugby players are very high,[4,10] and the injuries occur as a result of various mechanisms such as tackling, breakdown, falling, and scoring during competition.[1] Our previous study[7] demonstrated that 67% of primary anterior shoulder dislocation in rugby players occurs during tackling. Depending on the situation, there are various types of tackles in rugby[3,9]: arm tackle (the tackler stops the ball carrier by using his or her arms), jersey tackle (the tackler stops the ball carrier by pulling his or her jersey), shoulder tackle (the tackler initially contacts the ball carrier with his or her shoulder then takes hold of the opponent’s body), and tap tackle (the tackler dives to touch the ankle or foot of the ball carrier).

Several studies that analyzed tackling in rugby players using video records[2,8,12] have described that an anterior shoulder dislocation may occur when the tackler’s elbow is positioned straight at the time of impact with the ball carrier and the arm is forcibly moved in the posterior direction. In this situation, during contact, the affected shoulder is subjected to a combination of horizontal abduction or abduction and external rotation. However, the characteristics of tackling that lead to primary anterior shoulder dislocation in rugby have not been well investigated, and several factors may be involved.

The aims of this study were to use video analysis to investigate the characteristics of tackles that lead to shoulder dislocation when a player performs a one-on-one tackle, as well as to show the relevance of the mechanisms of injury and the morphological damage of the glenoid based on image findings. We hypothesized that these rugby tackles could be characterized and various factors might be related, and that there may be a relationship between the mechanisms of injury and the morphological damage of the glenoid.

Methods

Eleven elite rugby players who sustained primary anterior shoulder dislocation or subluxation due to one-on-one tackling between 2001 and 2014, and diagnosed in our hospital, were included in the present study (Table 1). The inclusion criteria were a definitive diagnosis of primary anterior shoulder dislocation or subluxation with objective imaging studies in our institution and no history of previous surgery on the affected shoulder. Athletes were excluded from the study if the injury occurred in a tackle that involved more than 2 players, if the videos could not be analyzed due to insufficient image quality, or if the players had previous history of anterior shoulder dislocation to the affected side of the shoulder. The study was approved by the institutional review board of our hospital. Video recordings were obtained either from the players themselves or from the staff of their team. Two authors (N.M., T.K.), 1 a senior orthopaedic surgeon and the other a physical therapist, who have covered a professional rugby team together for more than a decade, reviewed all the injuries recorded in the video in blinded fashion. When the judgment was not consistent, the final decision was achieved through a discussion among 3 reviewers (N.M., T.K., T.M.).

TABLE 1

Participant Demographics (N = 11)

Age, y, mean ± SD	25.3 ± 5.9
Height, cm, mean ± SD	180.7 ± 9.4
Weight, kg, mean ± SD	93.8 ± 13.6
Body mass index, kg/m2, mean ± SD	28.6 ± 2.5
Grade (collegiate/professional), n/N	5/6
Side of the affected shoulder (right/left), n/N	4/7
Position (forwards/backs), n/N	4/7

Because we believe that assessing the mechanisms of the injury is essential, each tackle was evaluated by our original assessment system modified from the past reports,[3,6,9] which was designed to categorize the tackle into 1 of 3 types based on the initial impact site—hand tackle, arm tackle, or shoulder tackle (Table 2). Each tackle was also divided into 3 phases—approach, impact, and postimpact. Then, the distance between the players in the approach phase, the shoulder and the head position in the impact phase, and the direction of shoulder motion in the angle of abduction, horizontal abduction, and external rotation in the postimpact phase were analyzed.

TABLE 2

Description of Tackle Characteristics

Term		Explanation
Approach phase
Orientation of tackler in relation to ball carrier  (Hendricks et al 2014[6])	In front	Tackler and ball carrier moving head-on toward each other
	Side	Tackler moving in from the ball carrier’s side
	Oblique	Tackler moving into ball carrier at an angle
	Behind	Tackler chasing ball carrier toward own try-line
Leading leg	Same	Ipsilateral side of the shoulder contact
	Opposite	Contralateral side of the shoulder contact
Distance between the leading leg of tackler and target  (Hendricks et al 2014,[6] modified)	Near	Within a half body length of ball carrier
	Moderate	Within one body length of ball carrier
	Distant	Greater than one body length from ball carrier
Head position (Hendricks et al 2014,[6] modified)	Up and forward	Toward the ball carrier
	Down	Toward the ground
Impact phase
Type of tackle	Hand tackle	First contact point of tackler to ball carrier is between the hand and forearm
	Arm tackle	First contact point of tackler to ball carrier is between the forearm and upper arm
	Shoulder tackle	First contact point of tackler to ball carrier is between the shoulder and neck
Head placement (Hendricks et al 2014[6])	Above	Tackler’s head higher than ball carrier’s body during contact
	Beside	Tackler’s head next to ball carrier’s body during contact
	In front	Tackler’s head in front of ball carrier’s body during contact
	Behind	Tackler’s head at the back of ball carrier’s body during contact
Position of shoulder joint at impacta	Abd <90°	Shoulder joint of tackler is located >90° at the moment of impact
	Abd = 90°	Shoulder joint of tackler is located ≈90° at the moment of impact
	Abd >90°	Shoulder joint of tackler is located <90° degree at the moment of impact
	ER	Shoulder joint of tackler is located in external rotation at the moment of impact
	Ntrl	Shoulder joint of tackler is not located in external or internal rotation at the moment of impact
	IR	Shoulder joint of tackler is located in internal rotation at the moment of impact
Postimpact phase
Main direction of shoulder motion (Longo et al 2011[8])	Abd	The upper arm of tackler moved abduction after impact
	H.abd	The upper arm of tackler moved horizontal abduction after impact
	ER	The upper arm of tackler moved external rotation after impact
Leg drive (McIntosh et al 2010[9])	Active	With leg drive and forward momentum
	Passive	Without leg drive and forward momentum
Tackle result (Hendricks et al 2014[6])	Unsuccessful	Identified when ball carrier is able to offload ball, or break an attempted tackle
	Successful	Tackler prevents ball carrier from progressing toward his try-line

Shoulder angle and movement is shown at angle of the upper arm to the trunk. Abd, abduction; ER, external rotation; H.Abd, horizontal abduction; IR, internal rotation; Ntrl, neutral.

The shape of the glenoid on the affected side was classified into 3 types—intact, erosion, and bone defect—based on the reconstructed clockface image from computed tomography (CT), and the humeral head was examined by one of the authors (T.K.), who was blinded to the patient data, to determine whether a Hill-Sachs lesion was present. According to a past report,[11] the orientation of the glenoid defect was also evaluated in each shoulder and was defined as a line passing through the center of the glenoid and perpendicular to the defect margin, expressed as degrees from the long axis (Figure 1). For statistical analyses, a Fisher exact test was applied to examine any correlation between the tackle types and the degree of glenoid bone loss. A Kruskal-Wallis test was also applied to determine the significant differences in the orientation of the glenoid defect for each type of tackle. All analyses were processed by using SPSS for Mac software program (version 21; IBM Corp). P < .05 was considered statistically significant.

Figure 1.

The orientation of the glenoid defect (angle a).

Results

A total of 11 tackles that led to primary anterior shoulder dislocation were evaluated by our assessment system (see Table A1 in the Appendix for details). Interrater reliability of the assessment between the 2 authors was relatively high (κ = 0.85-1.0) (see Table A2 in the Appendix). Of these, the numbers of hand, arm, and shoulder tackles were 4, 4, and 3, respectively. The characteristic of hand tackle injuries (n = 4) was that most of the ball carriers changed their running course just before the approach phase of the tackle (Figure 2). Thus, tacklers failed to close the distant ball carrier in the approach phase and caught the ball carrier with full arm extension in the impact phase. These situations resulted in the shoulder oriented at a humerothoracic angle of >90° of abduction; after the impact of the tackle, the shoulder was horizontally abducted in all the cases.

TABLE A1

Tackle Characteristics

	Approach Phase				Impact Phase			Postimpact Phase
	Orientation of Tackler Relative to Ball Carrier	Leading Leg	Distance Between the Leading Leg of Tackler and Target	Head Position	Type of Tackle	Head Placement	Position of Shoulder Joint at Impact	Main Direction of Shoulder Motion	Leg Drive	Tackle Result
1	In front	Same	Moderate	Up and forward	Hand	Beside	Abd>90°, ER	H. Abd	Passive	Unsuccessful
2	In front	Same	Moderate	Down	Hand	Beside	Abd>90°, ER	H. Abd	Passive	Unsuccessful
3	In front	Same	Distant	Up and forward	Hand	Beside	Abd>90°, ER	H. Abd	Passive	Unsuccessful
4	Behind	Same	Distant	Up and forward	Hand	Behind	Abd>90°, ER	H. Abd	Passive	Unsuccessful
5	In front	Same	Near	Down	Arm	Beside	Abd=90°, ER	H. Abd	Passive	Successful
6	Oblique	Same	Near	Down	Arm	Beside	Abd=90°, ER	H. Abd	Passive	Successful
7	Side	Same	Near	Up and forward	Arm	Behind	Abd>90°, ER	H. Abd	Passive	Unsuccessful
8	In front	Opposite	Moderate	Down	Arm	Beside	Abd>90°, ER	H. Abd	Active	Unsuccessful
9	In front	Opposite	Near	Down	Shoulder	In front	Abd<90°, Ntr	H. Abd	Active	Successful
10	In front	Opposite	Near	Down	Shoulder	In front	Abd<90°, Ntr	H. Abd	Active	Successful
11	Oblique	Same	Near	Down	Shoulder	In front	Abd<90°, Ntr	H. Abd	Active	Unsuccessful

Abd, abduction; H. Abd, horizontal abduction; ER, external rotation; Ntr, neutral.

TABLE A2

Interrater Reliability of the Tackle Assessment

Description	κ Value	Concordance Rate (%)
Orientation of tackler in relation to ball carrier	1	100
Leading leg	1	100
Distance between the leading leg of tackler and target	0.85	90.9
Type of tackle	0.86	90.9
Head position	1	100
Position of shoulder joint at impact	1	100
Movement of upper arm after contact	1	100
Leg drive	1	100
Tackle result	1	100

Figure 2.

Case of the hand tackle. (A) Cognition: Tackler (TC; asterisk) stands face to face with ball carrier (BC) who has changed his running course just before the frame. (B) Approach: TC attempts to catch the distant BC with full arm extension. (C) Impact: TC touches the BC with his hand. (D) Outcome: TC’s shoulder orientation is forced to full abduction and horizontal abduction by the BC’s moment.

In arm tackle injuries (n = 4), the characteristics of the tackles were similar to those of hand tackles (Figure 3). The tacklers could not make contact with the ball carrier at an appropriate site of the shoulder because they could not get close enough; they came into contact with the ball carrier between the forearm and upper arm and eventually could not stop the ball carrier’s movement completely, although the tackler sometimes brought the ball carrier to the ground. At the time of impact (and injury), the humerothoracic angle was approximately 90° of abduction with external rotation, followed by horizontal abduction after the impact.

Figure 3.

Case of the arm tackle. (A) Cognition: Tackler (TC; asterisk) stands face to face with ball carrier (BC). (B) Approach: TC attempts to catch the distant BC by diving. (C) Impact: TC impacts the BC by his arm. (D) Outcome: TC’s shoulder orientation is forced to horizontal abduction by the BC’s moment.

In contrast to the 2 tackles mentioned above, several different points were noted in the shoulder tackle injuries (n = 3) (Figure 4). The tacklers successfully approached, and their shoulder came into contact with the midline of the torso of the ball carrier. In all cases, however, it was noted that the tacklers placed their lowered head in front of the ball carrier. At the time of the injury, the humerothoracic angle was approximately 90° of abduction with neutral rotational position; the shoulder was subsequently horizontally abducted in all cases.

Figure 4.

Cases of the shoulder tackle. (A) Cognition: Tackler (TC; asterisk) stands face to face with ball carrier (BC). (B) Approach: TC successfully closes in front of the BC. TC’s neck flexed position is emphasized in all the cases (arrowhead). (C) Impact: TC impacts the BC by his shoulder. At this moment, TC’s head in front of the running course of the BC is also noticed. (D) Outcome: TC completely stops the BC movement in 2 of the 3 cases shown.

The assessment of the shape of the glenoid on the injured side revealed bone defect in 4 cases and erosion in 4 cases, while Hill-Sachs lesions in the humeral head were observed in 9 cases (81.8%). No relationship was observed between the type of tackle and the shape of the glenoid on the affected shoulder (P = .55). There was also no significant difference in the orientation of the glenoid defect among the 3 types of tackles (P = .67) (Table 3).

TABLE 3

Summary of the Tackle Types and Computed Tomography Findings

Tackle Type, n	Glenoid Rim Morphologya			Mean Degrees of Glenoidb Lesion	Hill-Sachs Lesion, n (% Row)
	Intact	Erosion	Bone Defect
Hand, 4	1	2	1	94.3	3 (75.0)
Arm, 4	2	0	2	93.2	4 (100.0)
Shoulder, 3	0	2	1	92.3	2 (66.7)
Total, 11	3	4	4	—	9/11 (81.8)

No significant difference in the glenoid rim morphology between the 3 types (Fisher exact test P = .55).

No significant difference in the orientation of the glenoid defect between the 3 types (Kruskal-Wallis test P = .67).

Discussion

Regarding rugby game defense strategy, Hendricks and Lambert[5] and Hendricks et al[6] reported that the arm tackle and the jersey tackle had lower success rates than the shoulder tackle. They recommended that coaches should encourage players to improve their footwork during tackling and to focus on maintaining a square body position when making shoulder contact in order to decrease the risk of injury and to perform a successful tackle. Other reports[2,8,12] have mentioned that anterior shoulder dislocation occurs due to the levering force caused by a forcible backward movement from abduction of the shoulder by tackling in professional-level rugby players. Therefore, it has been proposed that tacklers should contact the opposing player with their shoulder to minimize the risk of anterior shoulder dislocation.[2]

In the present study, the characteristics of the tackles were evaluated from 11 videos that captured primary anterior shoulder dislocation. We categorized the tackles based on the initial site of impact with the ball carrier, and several characteristics could be pointed out in each tackle type. In cases of hand and arm tackles (n = 8), the tackler failed to align his body to the ball carrier in the approach phase, which resulted in an incomplete tackle with unfavorable shoulder orientation in the subsequent contact phase. In these situations, the levering force that occurred during forcible horizontal abduction in the glenohumeral joint appears to result in anterior shoulder dislocation.

In case of the shoulder tackle (n = 3), the tacklers successfully kept their body position square to the ball carrier in the approach phase, and the subsequent contact occurred firmly with their shoulder. It was noted, however, that the tacklers placed their lowered head in front of the ball carrier in all 3 cases. These characteristics were apparently different from the other 2 tackle types. In this type of injury, we assumed that tackler’s inappropriate posture with lowered and contralaterally rotated head influences the glenohumeral joint via scapula orientation, which seems to be a primary factor in the cause of anterior shoulder dislocation. It remains unknown whether the mechanism of the shoulder tackle involves levering force or direct arm contusion to the glenoid.

In the present study, the tackles that led to primary shoulder dislocation were categorized into 1 of 3 types for the purpose of addressing these mechanisms in detail, as described above. To verify whether the mechanisms of each tackle type were consistent, we examined bone injury of the affected shoulder by using reconstructed CT images. Regarding the CT image examinations, even though 72.7% of the athletes had a glenoid lesion, no significant difference in the degrees and orientations were found between tackle types. It seems that a part of these results was due to small sample size, which is a major limitation of this study, although the high-quality images of primary shoulder dislocation during tackling are difficult to take.[2,8,12] Thus, a future study with a larger population might enable indication of the force on the glenoid during tackling.

In addition to the major limitation described above, another limitation is that the present study was based on the collected videos that were recorded from a single video camera, which meant that we could not accurately measure the angle of the shoulder. Thus, to determine the mechanism of primary anterior shoulder dislocation, we would need to perform an investigation using multiangle video recordings or a different experimental approach, such as a 3-dimensional motion analysis.

Conclusion

Although the precise mechanism of primary anterior shoulder dislocation could not be estimated from this single-view analysis, our study demonstrated that failure of individual tackling that leads to injury is not uniform and can be caused by 2 main factors: failure of approach followed by extended arm position or inappropriate posture of the tackler at the impact, with lowered head in front of the opponent. These findings indicate that injury mechanisms should be assessed in each type of tackling because it remains unknown whether external force to the glenoid is different in each mechanism during shoulder dislocation.