Despite improvement in surgical techniques,1 the failure rate is high after cuff repair.2,3 Risk factors for failure to heal include fatty infiltration of the muscle, retraction of the musculotendinous unit, and probably inappropriate postoperative rehabilitation protocols.2,4,5 Preventing gap formation between a repaired tendon and bone as a result of active or passive loading of the construct is considered essential for successful healing. Although passive mobilization may often be well tolerated,6–9 convincing data exist that the arm position at rest has an important effect on the strain on the repair of the rotator cuff,10 particularly in the early post-operative weeks.11 A biomechanical analysis concluded that the best arm position to protect rotator cuff repairs from excessive tension would be between 58° and 109°, depending on the type of tear pattern.12,13 Intraoperative measurements in patients and animal models10,14,15 have shown decreased strain in the supraspinatus with increasing abduction, and this has been confirmed in computational modeling.
On the basis of such data, the current authors attempted to protect all transtendinous supraspinatus repairs in an abduction brace for 6 weeks with early passive mobilization above the abduction angle of the brace. Theoretically, the abduction brace unloads and protects the supraspinatus repair. For this to function, however, the rotator cuff muscles must not be under active tension. The authors hypothesized that the patients' ability to relax the shoulder muscles depends on a variety of factors, such as the arm position in space, the size of the cuff repair, and pain, and that the loading of the abduction brace can be used as a surrogate for the degree of relaxation. The purpose of this study was to measure the load on a conventional abduction brace in different positions during the early postoperative period to determine a most effective position for protection of rotator cuff repair.
Design of the Abduction Brace and Measurement Positions
A conventional shoulder abduction brace (abduction brace “Comfort”; Pompa AG, Hausen, Switzerland) was modified to allow for measurement of load transfer of the arm onto the brace by measuring the bending moment generated by the weight of the arm. This was achieved with a torque-sensitive hinge construction (Figure 1) located in the axilla of the brace. The brace allowed different positions in space, either in the true frontal or the scapular plane, using abduction angles of 30°, 50°, 70°, and 90°. The rotation of the arm was always 90° to the scapular plane with 90° of flexion in the elbow.
Sensor Design and User Interface
Two perpendicular general-purpose strain gauges in stacked rosette configuration (C2A-06-062WW-120, Micro Measurements; Vishay Precision Group Inc, Malvern, Pennsylvania) were glued on the center to enable torsion measurement. Lead wires were twisted to minimize noise and connected to a Ni9235 (National Instruments, Austin, Texas) quarter-bridge strain gauge module for dynamic strain measurement while using 2 independent channels (Figure 1). A LabView (National Instruments) runtime engine–based graphical user interface was programmed to allow quick data acquisition during patient recordings. For initial adjustment and scaling, a 10-N standard weight was used with a 15-cm lever arm. Measured voltage was calculated in a moment (Nm). Standard weights were used to test the accuracy of the sensor (±0.037 Nm). The safety, sensitivity, and reproducibility of the complete setup were tested and verified at different time points on volunteers prior to use on patients.
Calculation of Muscular Abduction Moment
With decreasing shoulder abduction angle, the resulting moment and recorded load on the abduction brace is reduced as geometrically explained in Figure 2. Therefore, the absolute recorded values of the shoulder brace (MSAB) were corrected by formula 1, below, to calculate the true amount of unloading of the shoulder muscles (MM) by subtracting them from the angle-dependent arm weight (MG_arm):
Calculation of muscular abduction moment. MSAB is the measured moment from the shoulder abduction brace. MM is the moment generated by shoulder muscles. MG_arm represents the moment generated by the patient's arm weight (FG_arm). Moment MG_arm is dependent on the lever arm x (distance to the glenohumeral joint and the arm's center of mass) and the arm mass. In a 90° configuration, it can be assumed that the lever arm x has reached its maximum length (xmax). By means of trigonometry, a theoretical moment MG_arm can therefore be calculated for other angles beta (30°, 50°, and 70°). By rearranging the sum of the moment equation, the theoretical moment generated by the muscles MM can therefore be calculated.
Prior to the patient assessments, the experimental brace was tested on each of 8 healthy volunteers—staff members of the hospital (residents, consultants, and scientists) without shoulder pain or previous shoulder surgeries—to establish the feasibility and reproducibility of the method.
Once institutional review board approval was obtained, the authors prospectively enrolled 22 patients who underwent arthroscopic repair of a rotator cuff tear involving the supraspinatus tendon and consented to participate in this trial. Eleven of the 22 supraspinatus tears were combined with partial subscapularis repairs, 8 with partial infraspinatus repairs, and 2 shoulders underwent 3-tendon repairs. Patients with supraspinatus tendon-tear repair that made an abduction brace necessary during the postoperative course, a fully paralyzed shoulder by the interscalene catheter, and complete measurements at days 1 and 2 were included.
For ethical reasons and to simulate the situation on the second postoperative day when the arm was neurologically recovered, volunteers were tested without regional anesthesia. The protocol included torque measurements in the true frontal and the scapular planes in 30°, 50°, 70°, and 90° of abduction for 5 seconds after the patient indicated comfort in each new position (Figures 3–4).
Arm position in the shoulder abduction brace: true frontal plane (A) and scapular plane (B).
Arm position in different abduction angles in the shoulder abduction brace: 30° (A), 50° (B), 70° (C), and 90° (D).
After this, the volunteers were asked which position they judged to be most comfortable.
According to the standardized protocol for treatment of postoperative pain, all operated on patients received an indwelling, interscalene catheter with continuous analgesia causing paresis of the relevant shoulder muscles (C5/6) for the operative and first postoperative days. Paresis was considered complete if the patients experienced no pain, had no sensitivity to prick in the segments C5 and C6, were unable to contract the deltoid or external rotators, and had a negative biceps tendon reflex. To obtain measurements with complete paresis, the first series of measurements was performed on the first postoperative day. To obtain measurements with regained activity of the shoulder musculature, the authors repeated the same measurements on the second postoperative day after full clinical recovery of sensory and motor innervation.
The measurement protocol included torque measurements in the true frontal and the scapular planes in 30°, 50°, 70°, and 90° each, for 5 seconds in neutral rotation, after the patient indicated comfort in each new position (Figures 3–4). Two patients reported pain at 90° of abduction, Therefore, the measurements were performed at 30°, 50°, and 70° for them. Subsequent to the measurements on the second postoperative day, the patients were asked which position they considered most comfortable. On the basis of convincing biomechanical12,13 and experimental14 data, the authors decided not to include an abduction position lower than 30° in the tests.
Evaluation of the Data
The differences in torque values from day 1 to day 2 were calculated relative to the starting position of 90° of abduction of each day to compensate for different absolute arm weights.
It was further analyzed whether the amount of relaxation and weight transfer to the brace correlated with the size of the tear, the most comfortable position of the brace, and whether a bony acromioplasty was performed.
Statistical analysis was conducted with SPSS Statistics for Windows, version 24.0, software (IBM Corp, Armonk, New York). Because visual inspection of the data revealed no deviation from normal distribution, parametric statistical tests were employed. To test a difference in brace torque between day 1 and day 2, a paired samples t test was performed for each tested abduction angle and over both arm positions. Statistical significance was set at P<.05.
To test the test-retest reliability of the device and the measurement procedure in the healthy volunteer group, an inter-class correlation coefficient was collected. This was achieved on absolute agreement based on a two-way mixed effects model and a single measures approach was used by repeated measurements.
Torque values of the control group at 90° of abduction in the true frontal plane were 3.3±1.0 Nm for day 1 and 3.7±1.0 Nm for day 2. For the scapular plane, they were 3.0±0.8 Nm and 3.1±0.9 Nm on day 1 and day 2, respectively (Figure 5). All of the volunteers preferred the scapular plane position at 30° of abduction. The repeatability of the tests was very high (interclass correlation coefficient, 0.967).
Load on the brace in the true frontal and the scapular planes on day 1 and day 2. The green line indicates patients' load, and the blue line indicates the load on the brace of the healthy control group. Positive brace loads indicate more weight onto the brace, while negative values indicate involvement of abduction muscles.
The torque values at 90° of abduction in the true frontal plane were 3.4±1.1 Nm for day 1 and 3.2±0.8 Nm for day 2. For the scapular plane, they were 3.5±0.8 Nm and 3.3±0.7 Nm, respectively.
The differences in the lower abduction angles relative to the 90° starting position for the true frontal and for the scapular planes are provided in Figure 5. Interestingly, with full relaxation of day 1, a decrease of the total load transmitted to the brace was noted for the lower positions in both planes, also after correction as indicated in formula 1.
On day 2, however, after recovery of the sensory and motor innervation, a higher proportion of the arm weight was transmitted to the brace over all abduction angles for the frontal and the scapular planes (difference: 0.24±0.70 Nm, P=.002, and 0.28±0.50 Nm, P<.001, respectively). There was no significant correlation with tear size, tear location (anterior or posterior), fatty infiltration (Goutallier classification), or performance of an acromioplasty with maximal brace loading for any position or plane on both days.
The majority of the patients preferred a low position (10 patients, 30°; 9 patients, 50°). One patient preferred 70° of abduction, and 2 patients were undecided. This was in line with the overall better load transfer to the brace with a low arm position. Subjectively, 15 patients preferred the scapular and 5 preferred the true frontal plane position of the brace. This was not related to the tear location (anterior vs posterior). Because the arm was anesthetized on day 1, evaluation of the preferred position was only recorded on day 2 after the arm was fully awake.
In addition to protecting a detached and repaired deltoid in open rotator cuff repair, the use of an abduction brace has a biomechanical rationale and therefore also a potential benefit after arthroscopic supraspinatus repairs. First, the arm rests in a defined position. As opposed to the use of a sling, there is little possibility of the patient making involuntary, potentially harmful movements with the arm in a position that is already associated with maximal tension in the supraspinatus unit.16 Second, the imposed abduction position approximates the insertion of the repaired rotator cuff to the musculature and thus reduces passive tension and stress on the repair.14 Third, the brace is an unambiguous visual signal in the social environment of the patient, making it clear that the arm cannot be used and must be protected.
There is general consensus11 that, particularly in the first 6 weeks postoperatively, passive or active overtensioning of the repair should be avoided to protect the repair. A biomechanical model has suggested that the higher the abduction angle, the better the protection of the repair should be.12,13 However, the current authors have observed that in the immediate postoperative period, some patients tend to actively lift the brace through muscular contraction rather than deposit the arm on the brace. This observation would represent a situation that is counterproductive. Therefore, it was a goal of this study to determine a position of the abduction brace that allows patients to relax their arm, thereby presumably protecting a supraspinatus repair.
Both of the authors' hypotheses—that a lower arm position (30° to 50°) is more effective for relaxing the shoulder and is subjectively more comfortable—were confirmed. Unexpected, however, was that these were particularly true with the arm fully awake. With the arm anesthetized, less weight was transmitted in lower abduction angles.
A possible explanation for this finding may be an initially increased tissue stiffness of the shoulder abductors as a result of postoperative swelling and edema caused by arthroscopic fluid and surgical trauma, which normalizes rapidly on the second postoperative day. Because the authors did not anticipate this possibility when planning the study, swelling was not quantified in these patients. It should be considered in further evaluation of this topic. Another explanation would be that muscle retraction and atrophy with a long-standing tendon tear will lead to a passive resistance against shoulder adduction, which can reach considerable loads.14 This passive tension was shown to rapidly decrease to 0 within the first 24 to 48 hours in an in vivo sheep model.15 The current authors think that, during rehabilitation, there is greater therapeutic impact in avoiding active contraction of a muscle that is already pretensioned than optimizing the passive muscle tension on the repair, so that an optimal relaxation with optimal weight deposition on the brace is sought.
In this study, there was no correlation between preoperative tendon retraction and optimally relaxing arm position. However, all repairs could be completed in a watertight fashion because none of the involved rotator cuff muscles exceeded Goutallier stage 2 (18 patients had subscapularis Goutallier 0 or 1). Therefore, muscle elasticity was not grossly pathological, so a correlation cannot be expected. The numerical findings regarding the degree of abduction were similar in the frontal and the scapular planes. The only difference was the greater comfort provided by the scapular plane fitting, a finding that was confirmed by the volunteers. The authors will therefore standardize the brace to the most comfortable position, which has been found to be an abduction between 30° and 50° in the scapular plane (Figure 6). In addition to subjective comfort, this position has the advantages of the hand being in front of the body and able to hold light objects and the arm being in a less conflicting position.
The most preferred arm position in the brace in the scapular plane in 30° of abduction.
A limitation of this experimental setup was that the torque sensor was located in the axilla and therefore not exactly in the center of rotation. Thus, the measured torque and the results from formula 1 are only an approximation of the true values (Figure 7). Further, the measured torque values will depend on arm size and weight and on the degree of flexion of the elbow. The exact center of gravity of the arm is unknown. To overcome these limitations, the authors standardized the values as relative data compared with the starting position at 90° of abduction of each day. The resulting weight and torque on the brace was presumed to be a combination of active and passive forces. However, with recovery from the plexus block on day 2, the loading on the brace significantly increased at lower abduction angles. A relevant active component appears unlikely in these positions. Further limitations were the small number of patients and that the results had no correlation with the clinical outcome.
Position of the torque sensor in the abduction brace (A). Center of the rotation of the glenohumeral joint (B).
There is consensus that protection of a surgical rotator cuff repair is beneficial, particularly in the first 6 weeks postoperatively. Abduction braces are one potential, albeit controversial, way to provide such protection. The use of an abduction brace appears to be efficient. This study documented that an abduction brace fitted at an abduction angle between 30° and 50° in the scapular plane resulted in the largest load transmission from the arm to the brace and was subjectively most comfortable. Therefore, if an abduction brace is considered, this seems to be the most rational currently known position for the arm.
- Aurora A, McCarron J, Iannotti JP, Derwin K. Commercially available extracellular matrix materials for rotator cuff repairs: state of the art and future trends. J Shoulder Elbow Surg. 2007;16(5)(suppl):S171–S178. doi:10.1016/j.jse.2007.03.008 [CrossRef] PMID:17560804
- Meyer DC, Wieser K, Farshad M, Gerber C. Retraction of supraspinatus muscle and tendon as predictors of success of rotator cuff repair. Am J Sports Med. 2012;40(10):2242–2247. doi:10.1177/0363546512457587 [CrossRef] PMID:22926748
- Iannotti JP, Deutsch A, Green A, et al. Time to failure after rotator cuff repair: a prospective imaging study. J Bone Joint Surg Am. 2013;95(11):965–971. doi:10.2106/JBJS.L.00708 [CrossRef] PMID:23780533
- Gerber C, Schneeberger AG, Hoppeler H, Meyer DC. Correlation of atrophy and fatty infiltration on strength and integrity of rotator cuff repairs: a study in thirteen patients. J Shoulder Elbow Surg. 2007;16(6):691–696. doi:10.1016/j.jse.2007.02.122 [CrossRef] PMID:17931904
- Oh JH, Kim SH, Ji HM, Jo KH, Bin SW, Gong HS. Prognostic factors affecting anatomic outcome of rotator cuff repair and correlation with functional outcome. Arthroscopy. 2009;25(1):30–39. doi:10.1016/j.arthro.2008.08.010 [CrossRef] PMID:19111216
- Arndt J, Clavert P, Mielcarek P, Bouchaib J, Meyer N, Kempf JFFrench Society for Shoulder & Elbow (SOFEC). Immediate passive motion versus immobilization after endoscopic supraspinatus tendon repair: a prospective randomized study. Orthop Traumatol Surg Res. 2012;98(6) (suppl):S131–S138. doi:10.1016/j.otsr.2012.05.003 [CrossRef] PMID:22944392
- Keener JD, Galatz LM, Stobbs-Cucchi G, Patton R, Yamaguchi K. Rehabilitation following arthroscopic rotator cuff repair: a prospective randomized trial of immobilization compared with early motion. J Bone Joint Surg Am. 2014;96(1):11–19. doi:10.2106/JBJS.M.00034 [CrossRef] PMID:24382719
- Koh KH, Lim TK, Shon MS, Park YE, Lee SW, Yoo JC. Effect of immobilization without passive exercise after rotator cuff repair: randomized clinical trial comparing four and eight weeks of immobilization. J Bone Joint Surg Am. 2014;96(6):e44. doi:10.2106/JBJS.L.01741 [CrossRef] PMID:24647511
- Lee BG, Cho NS, Rhee YG. Effect of two rehabilitation protocols on range of motion and healing rates after arthroscopic rotator cuff repair: aggressive versus limited early passive exercises. Arthroscopy. 2012;28(1):34–42. doi:10.1016/j.arthro.2011.07.012 [CrossRef] PMID:22014477
- Hatakeyama Y, Itoi E, Pradhan RL, Urayama M, Sato K. Effect of arm elevation and rotation on the strain in the repaired rotator cuff tendon: a cadaveric study. Am J Sports Med. 2001;29(6):788–794. doi:10.1177/03635465010290061901 [CrossRef] PMID:11734494
- Thigpen CA, Shaffer MA, Gaunt BW, Leggin BG, Williams GR, Wilcox RB III, . The American Society of Shoulder and Elbow Therapists' consensus statement on rehabilitation following arthroscopic rotator cuff repair. J Shoulder Elbow Surg. 2016;25(4):521–535. doi:10.1016/j.jse.2015.12.018 [CrossRef] PMID:26995456
- Jackson M, Sylvestre É, Bleau J, Allard P, Begon M. Estimating optimal shoulder immobilization postures following surgical repair of massive rotator cuff tears. J Biomech. 2013;46(1):179–182. doi:10.1016/j.jbiomech.2012.09.026 [CrossRef] PMID:23116766
- Jackson M, Tétreault P, Allard P, Begon M. Optimal shoulder immobilization postures following surgical repair of rotator cuff tears: a simulation analysis. J Shoulder Elbow Surg. 2013;22(8):1011–1018. doi:10.1016/j.jse.2012.10.042 [CrossRef] PMID:23352183
- Hersche O, Gerber C. Passive tension in the supraspinatus musculotendinous unit after long-standing rupture of its tendon: a preliminary report. J Shoulder Elbow Surg. 1998;7(4):393–396. doi:10.1016/S1058-2746(98)90030-1 [CrossRef] PMID:9752650
- Meyer DC, Jacob HA, Nyffeler RW, Gerber C. In vivo tendon force measurement of 2-week duration in sheep. J Biomech. 2004;37(1):135–140. doi:10.1016/S0021-9290(03)00260-4 [CrossRef] PMID:14672577
- Lieber RL. Skeletal Muscle Structure, Function, and Plasticity: The Physiological Basis of Rehabilitation. 3rd ed. Baltimore, MD: Lippincott Williams and Wilkins; 2010.