Shoulder & Elbow

The dilemma for the treating surgeon is deciding which patients to treat with RCR and who to treat with rTSA. Predicting which surgical procedure will provide a better functional result is difficult and controversial.⁷ The RCR method is a bone-conserving procedure with relatively low surgical risk and allows the option for rTSA to be performed as a salvage surgery should repair fail. It also may be less costly in the appropriate population.⁸ However, large rotator cuff tears in elderly patients have low healing potential, and the prospect of participating in a lengthy rehabilitation after an operation that may not prove successful can be deterring.^9,10 In the elderly population, rTSA may be a reliable option, as tendon healing of the cuff is not necessary to restore function. However, rTSA does not conserve bone, provides a non-anatomic solution, and has had historically high complication rates.^4,5

In an effort to aid in the decision-making process when considering these 2 surgical options, we compared RCR and rTSA performed at a single institution for MCT in patients >65 years. Our aim was to identify preoperative patient variables that influence a surgeon’s decision to proceed with 1 of the 2 procedures. Moreover, we evaluated clinical outcomes in these 2 patient populations. We hypothesized that (1) patients selected for rTSA would have worse preoperative function, less range of motion, more comorbidities, more evidence of radiographic subluxation, and a higher likelihood of having undergone previous RCR than those selected for RCR, and (2) both RCR and rTSA would be successful and result in improved clinical outcomes with high patient satisfaction.

MATERIALS AND METHODS

PATIENT SELECTION

We performed a retrospective chart review using our practice database of all patients undergoing arthroscopic RCR and rTSA for any indication by the senior author (M.A.F.) between January 2004 and April 2015. A total of 1503 RCRs and 1973 rTSAs were conducted during the study period. Patient medical records were reviewed, and those meeting the following criteria were included in the study: >65 years at the time of surgery, MCT, no preoperative glenohumeral arthritis, minimum follow-up of 12 months, functional deltoid muscle on physical examination, and no prior shoulder surgery except for RCR or diagnostic arthroscopy. A total of 92 patients who underwent arthroscopic RCR and 89 patients who underwent rTSA met the inclusion criteria. For patients with bilateral shoulder surgery, we measured each shoulder independently. Three patients underwent bilateral rTSA, and 3 patients underwent bilateral RCR, leaving 95 shoulders in the RCR group and 92 in the rTSA group. The Western Institutional Review Board determined this study to be exempt from review.

RADIOGRAPHIC EVALUATION

All patient charts included a radiology report and documented interpretation of the images by the treating surgeon prior to surgery. Radiographs were assessed to assure the absence of preoperative glenohumeral osteoarthritis. The images were also graded based on the Hamada classification.¹¹ Stage 1 is associated with minimal radiographic change with an acromiohumeral interval (AHI) >6 mm; stage 2 is characterized by narrowing of the AHI <6 mm; and Stage 3 is defined by narrowing of the AHI with radiographic changes of the acromion. Stages 4 and higher include arthritic changes to the glenohumeral joint, and they were not included in the study population. The AHI measurements and the presence or absence of glenohumeral subluxation were documented.

Continue to: MASSIVE CUFF TEAR DETERMINATION...

We defined MCT on the basis of previously described criteria of tears involving ≥2 tendons or tears measuring ≥5 cm in greatest dimension.^12,13 Patient charts were screened, and those whose clinical notes or radiology reports indicated an absence of MCT were excluded. Preoperative imaging of the remaining patients was then evaluated by 3 fellowship-trained shoulder surgeons to confirm MCT in all patients with a clinically documented MCT, as well as to assess those who had insufficient documentation of tear size in the notes.

Advanced imaging was evaluated for fatty atrophy of the rotator cuff musculature, and Goutallier classification was assigned.^14,15 Length of retraction was measured from the tendon end to the medial aspect of the footprint on coronal imaging, and the subscapularis and teres minor were assessed and documented as torn or intact.^16,17

DATA COLLECTION

We reviewed clinical charts and patient questionnaire forms from both the preoperative and follow-up visits. Clinical data collected included gender, age at surgery, active range of motion (forward elevation, abduction, external rotation, and internal rotation), comorbidities, smoking status, BMI, history of shoulder surgery, and any postoperative complications or need for secondary surgery. All patients completed patient-centered questionnaires regarding shoulder pain and dysfunction at each visit or via telephone communication with clinic staff. Outcome measurements used for analysis included American Shoulder and Elbow Surgeons (ASES) Score, simple shoulder test (SST), visual analog score (VAS) pain scale, and patient-reported satisfaction (Graded 1-10; 1 = poor outcome; 4 = satisfactory outcome; 7 = good outcome; 10 = excellent outcome).

DATA ANALYSIS AND STATISTICAL METHODS

Statistical tests were selected based on the result of Shapiro–Wilk test for normality. Continuous variables were evaluated with either independent t test or Mann–Whitney U test. Dependent t test was used to evaluate outcome variables. For categorical variables, either Pearson’s χ²or Fisher’s exact test was performed depending on the sample size. Alpha was set at P =.05.

Continue to: RESULTS...

PREOPERATIVE CHARACTERISTICS

Of the 187 shoulders in the study group, 95 had RCR and 92 had rTSA. Demographic information and preoperative variables for both groups are summarized in Table 1 and Table 2. Patients in the RCR group had greater preoperative forward elevation, abduction, and external rotation and higher preoperative functional scores than those in the rTSA group. Patients in the rTSA group were older and more likely to be female than those in the RCR group. More patients in the rTSA group had undergone prior RCR compared with those in the RCR group. Each of these differences was statistically significant. Subjective pain scores, BMI, and comorbidities were similar between the 2 groups.

Table 1. Patient demographics

Abbreviations: BMI, body mass index; RCR, rotator cuff repair; rTSA, reverse total shoulder arthroplasty.

Table 2. Preoperative variables

Abbreviations: AB, abduction; ASES, American Shoulder and Elbow Society score; ER, external rotation; FE, forward elevation; IR, internal rotation; preop, preoperative; SST, simple shoulder test.

Radiographically, patients selected to undergo rTSA had a smaller AHI (4.8 vs 8.7, P < .0001) and more evidence of superior subluxation (50.6% vs 14.1%, P < .0001) than those in the RCR group. Average Hamada grade was 1.4 ± 0.7 and 2.2 ± 0.7 for the RCR and rTSA groups, respectively (P < .0001). Average Goutallier grade was similar between the groups (2.2 ± 0.6 for RCR vs 2.2 ± 0.8 for rTSA, P =.227), and 25.5% of the RCR group had Grade 3 or 4 atrophy compared with 28.1% of the rTSA group.

POSTOPERATIVE OUTCOMES

The average follow-up time was 44 months for RCR and 47 months for rTSA. Patients in the RCR and rTSA groups were highly satisfied with the surgery (8.5 ± 2.6 vs 8.2 ± 2.6, P = .461) and had significantly increased range of motion in all planes and improved functional scores (Table 3). The rTSA group had greater net improvement in forward elevation, abduction, and external rotation than the RCR group. Both groups demonstrated similar improvement in ASES, SST, and VAS pain scores.

Table 3. Postoperative outcomes

Abbreviations: AB, abduction; ASES, American Shoulder and Elbow Society score; ER, external rotation; FE, forward elevation; IR, internal rotation; SST, simple shoulder test; VAS – visual analog score.

In the RCR group, 5 patients (5.3%) required reoperation: 3 patients underwent conversion to rTSA, 1 patient underwent biceps tenotomy with subacromial decompression, and 1 patient underwent arthroscopic irrigation and debridement for a postoperative Propionibacterium acnes infection. In the rTSA group, 2 patients (2.2%) required reoperation: 1 patient underwent open reduction internal fixation for a scapula fracture that failed conservative management, and 1 patient had an open irrigation and debridement with polyethylene exchange for an acute postoperative infection of unknown source.

DISCUSSION

Massive, retracted rotator cuff tears are a common and difficult problem.¹ The treatment options are numerous and depend on a variety of preoperative factors including patient-specific characteristics and factors specific to the tear. For certain patients, nonoperative management may be a reasonable first step, as an MCT does not necessarily preclude painless, functional shoulder motion. Elderly, lower demand individuals have been shown to do well with physical rehabilitation.¹⁸ Similarly, for the same category of elderly patients who do not respond to conservative measures, arthroscopic tendon debridement with or without subacromial decompression and/or biceps tenotomy may be effective.^1,19 This technique has been described as “limited goals surgery;” despite some mixed results in the literature, multiple studies have reported symptomatic and functional improvement after simple debridement.^2,19–21The consensus among several authors has been that this procedure continues to play a role for elderly, low-demand patients whose functional goals are limited and whose primary complaint is pain.^1,2,20

For the majority of patients with MCT who desire pain relief and a restoration of shoulder function, RCR remains the gold standard of treatment and should be the primary aim if feasible. Complete RCR has consistently outperformed both partial repair and debridement in multiple studies in terms of pain relief and functional improvement.^10,21,22However, elderly patients with chronic, massive tears, particularly in the setting of muscle atrophy, are at high risk of failure with attempted cuff repair.^9,23 Novel techniques such as superior capsular reconstruction and subacromial balloon spacer implantation may offer a minimally invasive method of re-centering the humeral head and stabilizing the glenohumeral joint; however, these new treatment options lack any long-term data in the literature to support their widespread use.^24–26 Alternatively, rTSA has been shown to be a reliable option to restore shoulder function in the setting of a massive irreparable rotator cuff tear, even in the absence of arthritis.^5,27-31

Continue to: The decision-making process...

The decision-making process for selecting RCR or rTSA in the setting of MCT without arthritis in the older population (age >65 years) remains challenging. We attempted to quantify the data of a high-volume surgeon and identify the differences and similarities between those patients selected for either procedure. At our institution, we generally performed rTSA on patients with low preoperative range of motion, poor function based on SST and ASES scores, small AHI, and strong evidence of superior subluxation. We were also more likely to perform rTSA if the patient had a history of rotator cuff surgery. There was a predilection for older age and female gender in those who underwent rTSA.

For our study, we elected to focus on patients >65 years. In our experience, the choice of which surgical procedure to perform is generally easier in younger patients. Most surgeons appropriately opt for an arthroscopic procedure or tendon transfer to preserve bone and maintain the option of rTSA as a salvage procedure if necessary in the future. Studies have reported that <60 years is a predictor of poor outcome with rTSA, and patients <65 years who undergo rTSA have been shown to have high complication rates.^30-32 Furthermore, the longevity of the implant in young patients is a significant concern, and revision surgery after rTSA is technically demanding and known to result in poor functional outcomes.^32,33

Although the indications for rTSA are expanding, attempts at RCR in the setting of MCT remain largely appropriate. Preserved preoperative anterior elevation >90° has been associated with loss of motion after rTSA and poor satisfaction, and one should exercise caution when considering rTSA in this setting.³ The current study confirmed that even older patients with MCT may be very satisfied with arthroscopic RCR (Figure 1). Both range of motion and function significantly improved, and patients were largely satisfied with the procedure with an average self-reported outcome of good to excellent. At the time of final follow-up for this study, only 3 shoulders in the RCR group had undergone conversion to rTSA. This number may be expected to rise with long follow-up periods, and we feel that prolonging the time before arthroplasty is generally in the best interest of the patient.

Our results were consistent with several reported studies in which RCR has been shown to be successful in the setting of MCT.^34–37 Henry and colleagues³⁶ performed a systematic review that evaluated 954 patients who underwent partial or complete anatomic RCR for MCT. Although the average age was 63 years (range, 37–87), functional outcome scores, VAS pain score, and overall range of motion consistently and significantly improved.

rTSA may be a “more reliable” option than RCR in treating MCT in the older population because it does not rely on tendon healing. However, the relationship between tendon healing and clinical outcomes after RCR is unclear. The aforementioned systematic review reported re-tear rates to be as high as 79%, but several studies have reported high satisfaction even in the setting of retear.³⁶ Yoo and colleagues³⁸ and Chung and colleagues⁹ reported re-tear rates of 45.5% and 39.8%, respectively, but both studies noted that there was no difference in outcome measures between those patients with and without re-tears. In particular, for patients who have had no prior rotator cuff surgery, an attempt at arthroscopic repair may be a prudent option with relatively low risk.

Although certain patients may clinically improve despite suffering a re-tear (or inability to heal in the first place), others continue to experience pain and dysfunction that negatively affect their quality of life.^39–41 These patients are more often appropriate candidates for rTSA. Indeed, several studies have demonstrated a higher re-tear rate in patients with a history of surgery than in those without.^23,31,38,42 Shamsudin and colleagues⁴³ found revision arthroscopic RCR, even in a younger age group with tears of all sizes, to be twice as likely to re-tear. Notably, re-tear after revision repair may be more likely to be symptomatic, as these re-tears are routinely associated with pain, stiffness, and loss of function. Even in the hands of experienced surgeons in a younger population, revision repair has only been able to reverse pseudoparalysis in 43% of patients, leading to only 39% return to sport or full activity.⁴⁴ In examining our data, we were much less likely to perform an RCR in patients who had a history of cuff repair surgery than in those without this history.

Continue to: Overall, those patients selected for rTSA...

Overall, those patients selected for rTSA in our study population performed well postoperatively (Figure 2 and Figure 3). Vast improvements were noted in range of motion, function, and pain scores at final follow up. Moreover, no patients in the study group required revision arthroplasty during the follow-up period. Although the average follow-up period was only 47 months, these results suggested that elderly patients with MCT without arthritis may be particularly ideal candidates for rTSA with regard to implant survival and anticipated revision rate when chosen appropriately.

Several weaknesses were noted within this paper. First, the study was retrospective, precluding randomization of treatment groups and standardization of data collection and follow-up. The outcomes of RCR and rTSA could not be compared directly due to the inherent selection bias. The groups clearly differed in many respects, and these preoperative factors likely played a role in postoperative outcomes. However, the primary goal of this study was not to compare outcomes of the treatment groups but to analyze the patterns of patient selection by an experienced treating surgeon and contribute to published data that each surgery can be successful in this patient population when chosen appropriately.

Second, our data were based on a single surgeon’s decisions, and results may not be generalizable. Furthermore, the senior author has had a longstanding interest in reverse shoulder arthroplasty and has published data illustrating successful outcomes for rTSA in patients with MCT. For this reason, one could presume that there may have been some bias toward treating patients with rTSA. However, we feel that the senior author’s unique and longstanding experience in treating MCT allows for a thorough evaluation and comparison of preoperative variables and outcomes declared within this study. Indeed, many patients included in this study were referred from outside institutions specifically for rTSA but instead were deemed more appropriate candidates for RCR and underwent successful arthroscopic repair, a common scenario which served as an impetus for this study.

CONCLUSION

RCR and rTSA are both viable options for patients >65 years with MCT without arthritis. Treatment must be individualized for each patient with careful consideration of a number of preoperative variables and patient characteristics. At our institution, patients with previous RCR, decreased range of motion, poor function, and strong radiographic evidence of subluxation are more likely to undergo rTSA. When chosen appropriately, both RCR and rTSA can result in improved range of motion, function, and high patient satisfaction in this patient population.

The dilemma for the treating surgeon is deciding which patients to treat with RCR and who to treat with rTSA. Predicting which surgical procedure will provide a better functional result is difficult and controversial.⁷ The RCR method is a bone-conserving procedure with relatively low surgical risk and allows the option for rTSA to be performed as a salvage surgery should repair fail. It also may be less costly in the appropriate population.⁸ However, large rotator cuff tears in elderly patients have low healing potential, and the prospect of participating in a lengthy rehabilitation after an operation that may not prove successful can be deterring.^9,10 In the elderly population, rTSA may be a reliable option, as tendon healing of the cuff is not necessary to restore function. However, rTSA does not conserve bone, provides a non-anatomic solution, and has had historically high complication rates.^4,5

In an effort to aid in the decision-making process when considering these 2 surgical options, we compared RCR and rTSA performed at a single institution for MCT in patients >65 years. Our aim was to identify preoperative patient variables that influence a surgeon’s decision to proceed with 1 of the 2 procedures. Moreover, we evaluated clinical outcomes in these 2 patient populations. We hypothesized that (1) patients selected for rTSA would have worse preoperative function, less range of motion, more comorbidities, more evidence of radiographic subluxation, and a higher likelihood of having undergone previous RCR than those selected for RCR, and (2) both RCR and rTSA would be successful and result in improved clinical outcomes with high patient satisfaction.

MATERIALS AND METHODS

PATIENT SELECTION

We performed a retrospective chart review using our practice database of all patients undergoing arthroscopic RCR and rTSA for any indication by the senior author (M.A.F.) between January 2004 and April 2015. A total of 1503 RCRs and 1973 rTSAs were conducted during the study period. Patient medical records were reviewed, and those meeting the following criteria were included in the study: >65 years at the time of surgery, MCT, no preoperative glenohumeral arthritis, minimum follow-up of 12 months, functional deltoid muscle on physical examination, and no prior shoulder surgery except for RCR or diagnostic arthroscopy. A total of 92 patients who underwent arthroscopic RCR and 89 patients who underwent rTSA met the inclusion criteria. For patients with bilateral shoulder surgery, we measured each shoulder independently. Three patients underwent bilateral rTSA, and 3 patients underwent bilateral RCR, leaving 95 shoulders in the RCR group and 92 in the rTSA group. The Western Institutional Review Board determined this study to be exempt from review.

RADIOGRAPHIC EVALUATION

All patient charts included a radiology report and documented interpretation of the images by the treating surgeon prior to surgery. Radiographs were assessed to assure the absence of preoperative glenohumeral osteoarthritis. The images were also graded based on the Hamada classification.¹¹ Stage 1 is associated with minimal radiographic change with an acromiohumeral interval (AHI) >6 mm; stage 2 is characterized by narrowing of the AHI <6 mm; and Stage 3 is defined by narrowing of the AHI with radiographic changes of the acromion. Stages 4 and higher include arthritic changes to the glenohumeral joint, and they were not included in the study population. The AHI measurements and the presence or absence of glenohumeral subluxation were documented.

Continue to: MASSIVE CUFF TEAR DETERMINATION...

We defined MCT on the basis of previously described criteria of tears involving ≥2 tendons or tears measuring ≥5 cm in greatest dimension.^12,13 Patient charts were screened, and those whose clinical notes or radiology reports indicated an absence of MCT were excluded. Preoperative imaging of the remaining patients was then evaluated by 3 fellowship-trained shoulder surgeons to confirm MCT in all patients with a clinically documented MCT, as well as to assess those who had insufficient documentation of tear size in the notes.

Advanced imaging was evaluated for fatty atrophy of the rotator cuff musculature, and Goutallier classification was assigned.^14,15 Length of retraction was measured from the tendon end to the medial aspect of the footprint on coronal imaging, and the subscapularis and teres minor were assessed and documented as torn or intact.^16,17

DATA COLLECTION

We reviewed clinical charts and patient questionnaire forms from both the preoperative and follow-up visits. Clinical data collected included gender, age at surgery, active range of motion (forward elevation, abduction, external rotation, and internal rotation), comorbidities, smoking status, BMI, history of shoulder surgery, and any postoperative complications or need for secondary surgery. All patients completed patient-centered questionnaires regarding shoulder pain and dysfunction at each visit or via telephone communication with clinic staff. Outcome measurements used for analysis included American Shoulder and Elbow Surgeons (ASES) Score, simple shoulder test (SST), visual analog score (VAS) pain scale, and patient-reported satisfaction (Graded 1-10; 1 = poor outcome; 4 = satisfactory outcome; 7 = good outcome; 10 = excellent outcome).

DATA ANALYSIS AND STATISTICAL METHODS

Statistical tests were selected based on the result of Shapiro–Wilk test for normality. Continuous variables were evaluated with either independent t test or Mann–Whitney U test. Dependent t test was used to evaluate outcome variables. For categorical variables, either Pearson’s χ²or Fisher’s exact test was performed depending on the sample size. Alpha was set at P =.05.

Continue to: RESULTS...

PREOPERATIVE CHARACTERISTICS

Of the 187 shoulders in the study group, 95 had RCR and 92 had rTSA. Demographic information and preoperative variables for both groups are summarized in Table 1 and Table 2. Patients in the RCR group had greater preoperative forward elevation, abduction, and external rotation and higher preoperative functional scores than those in the rTSA group. Patients in the rTSA group were older and more likely to be female than those in the RCR group. More patients in the rTSA group had undergone prior RCR compared with those in the RCR group. Each of these differences was statistically significant. Subjective pain scores, BMI, and comorbidities were similar between the 2 groups.

Table 1. Patient demographics

Abbreviations: BMI, body mass index; RCR, rotator cuff repair; rTSA, reverse total shoulder arthroplasty.

Table 2. Preoperative variables

Abbreviations: AB, abduction; ASES, American Shoulder and Elbow Society score; ER, external rotation; FE, forward elevation; IR, internal rotation; preop, preoperative; SST, simple shoulder test.

Radiographically, patients selected to undergo rTSA had a smaller AHI (4.8 vs 8.7, P < .0001) and more evidence of superior subluxation (50.6% vs 14.1%, P < .0001) than those in the RCR group. Average Hamada grade was 1.4 ± 0.7 and 2.2 ± 0.7 for the RCR and rTSA groups, respectively (P < .0001). Average Goutallier grade was similar between the groups (2.2 ± 0.6 for RCR vs 2.2 ± 0.8 for rTSA, P =.227), and 25.5% of the RCR group had Grade 3 or 4 atrophy compared with 28.1% of the rTSA group.

POSTOPERATIVE OUTCOMES

The average follow-up time was 44 months for RCR and 47 months for rTSA. Patients in the RCR and rTSA groups were highly satisfied with the surgery (8.5 ± 2.6 vs 8.2 ± 2.6, P = .461) and had significantly increased range of motion in all planes and improved functional scores (Table 3). The rTSA group had greater net improvement in forward elevation, abduction, and external rotation than the RCR group. Both groups demonstrated similar improvement in ASES, SST, and VAS pain scores.

Table 3. Postoperative outcomes

Abbreviations: AB, abduction; ASES, American Shoulder and Elbow Society score; ER, external rotation; FE, forward elevation; IR, internal rotation; SST, simple shoulder test; VAS – visual analog score.

In the RCR group, 5 patients (5.3%) required reoperation: 3 patients underwent conversion to rTSA, 1 patient underwent biceps tenotomy with subacromial decompression, and 1 patient underwent arthroscopic irrigation and debridement for a postoperative Propionibacterium acnes infection. In the rTSA group, 2 patients (2.2%) required reoperation: 1 patient underwent open reduction internal fixation for a scapula fracture that failed conservative management, and 1 patient had an open irrigation and debridement with polyethylene exchange for an acute postoperative infection of unknown source.

DISCUSSION

Massive, retracted rotator cuff tears are a common and difficult problem.¹ The treatment options are numerous and depend on a variety of preoperative factors including patient-specific characteristics and factors specific to the tear. For certain patients, nonoperative management may be a reasonable first step, as an MCT does not necessarily preclude painless, functional shoulder motion. Elderly, lower demand individuals have been shown to do well with physical rehabilitation.¹⁸ Similarly, for the same category of elderly patients who do not respond to conservative measures, arthroscopic tendon debridement with or without subacromial decompression and/or biceps tenotomy may be effective.^1,19 This technique has been described as “limited goals surgery;” despite some mixed results in the literature, multiple studies have reported symptomatic and functional improvement after simple debridement.^2,19–21The consensus among several authors has been that this procedure continues to play a role for elderly, low-demand patients whose functional goals are limited and whose primary complaint is pain.^1,2,20

For the majority of patients with MCT who desire pain relief and a restoration of shoulder function, RCR remains the gold standard of treatment and should be the primary aim if feasible. Complete RCR has consistently outperformed both partial repair and debridement in multiple studies in terms of pain relief and functional improvement.^10,21,22However, elderly patients with chronic, massive tears, particularly in the setting of muscle atrophy, are at high risk of failure with attempted cuff repair.^9,23 Novel techniques such as superior capsular reconstruction and subacromial balloon spacer implantation may offer a minimally invasive method of re-centering the humeral head and stabilizing the glenohumeral joint; however, these new treatment options lack any long-term data in the literature to support their widespread use.^24–26 Alternatively, rTSA has been shown to be a reliable option to restore shoulder function in the setting of a massive irreparable rotator cuff tear, even in the absence of arthritis.^5,27-31

Continue to: The decision-making process...

The decision-making process for selecting RCR or rTSA in the setting of MCT without arthritis in the older population (age >65 years) remains challenging. We attempted to quantify the data of a high-volume surgeon and identify the differences and similarities between those patients selected for either procedure. At our institution, we generally performed rTSA on patients with low preoperative range of motion, poor function based on SST and ASES scores, small AHI, and strong evidence of superior subluxation. We were also more likely to perform rTSA if the patient had a history of rotator cuff surgery. There was a predilection for older age and female gender in those who underwent rTSA.

For our study, we elected to focus on patients >65 years. In our experience, the choice of which surgical procedure to perform is generally easier in younger patients. Most surgeons appropriately opt for an arthroscopic procedure or tendon transfer to preserve bone and maintain the option of rTSA as a salvage procedure if necessary in the future. Studies have reported that <60 years is a predictor of poor outcome with rTSA, and patients <65 years who undergo rTSA have been shown to have high complication rates.^30-32 Furthermore, the longevity of the implant in young patients is a significant concern, and revision surgery after rTSA is technically demanding and known to result in poor functional outcomes.^32,33

Although the indications for rTSA are expanding, attempts at RCR in the setting of MCT remain largely appropriate. Preserved preoperative anterior elevation >90° has been associated with loss of motion after rTSA and poor satisfaction, and one should exercise caution when considering rTSA in this setting.³ The current study confirmed that even older patients with MCT may be very satisfied with arthroscopic RCR (Figure 1). Both range of motion and function significantly improved, and patients were largely satisfied with the procedure with an average self-reported outcome of good to excellent. At the time of final follow-up for this study, only 3 shoulders in the RCR group had undergone conversion to rTSA. This number may be expected to rise with long follow-up periods, and we feel that prolonging the time before arthroplasty is generally in the best interest of the patient.

Our results were consistent with several reported studies in which RCR has been shown to be successful in the setting of MCT.^34–37 Henry and colleagues³⁶ performed a systematic review that evaluated 954 patients who underwent partial or complete anatomic RCR for MCT. Although the average age was 63 years (range, 37–87), functional outcome scores, VAS pain score, and overall range of motion consistently and significantly improved.

rTSA may be a “more reliable” option than RCR in treating MCT in the older population because it does not rely on tendon healing. However, the relationship between tendon healing and clinical outcomes after RCR is unclear. The aforementioned systematic review reported re-tear rates to be as high as 79%, but several studies have reported high satisfaction even in the setting of retear.³⁶ Yoo and colleagues³⁸ and Chung and colleagues⁹ reported re-tear rates of 45.5% and 39.8%, respectively, but both studies noted that there was no difference in outcome measures between those patients with and without re-tears. In particular, for patients who have had no prior rotator cuff surgery, an attempt at arthroscopic repair may be a prudent option with relatively low risk.

Although certain patients may clinically improve despite suffering a re-tear (or inability to heal in the first place), others continue to experience pain and dysfunction that negatively affect their quality of life.^39–41 These patients are more often appropriate candidates for rTSA. Indeed, several studies have demonstrated a higher re-tear rate in patients with a history of surgery than in those without.^23,31,38,42 Shamsudin and colleagues⁴³ found revision arthroscopic RCR, even in a younger age group with tears of all sizes, to be twice as likely to re-tear. Notably, re-tear after revision repair may be more likely to be symptomatic, as these re-tears are routinely associated with pain, stiffness, and loss of function. Even in the hands of experienced surgeons in a younger population, revision repair has only been able to reverse pseudoparalysis in 43% of patients, leading to only 39% return to sport or full activity.⁴⁴ In examining our data, we were much less likely to perform an RCR in patients who had a history of cuff repair surgery than in those without this history.

Continue to: Overall, those patients selected for rTSA...

Overall, those patients selected for rTSA in our study population performed well postoperatively (Figure 2 and Figure 3). Vast improvements were noted in range of motion, function, and pain scores at final follow up. Moreover, no patients in the study group required revision arthroplasty during the follow-up period. Although the average follow-up period was only 47 months, these results suggested that elderly patients with MCT without arthritis may be particularly ideal candidates for rTSA with regard to implant survival and anticipated revision rate when chosen appropriately.

Several weaknesses were noted within this paper. First, the study was retrospective, precluding randomization of treatment groups and standardization of data collection and follow-up. The outcomes of RCR and rTSA could not be compared directly due to the inherent selection bias. The groups clearly differed in many respects, and these preoperative factors likely played a role in postoperative outcomes. However, the primary goal of this study was not to compare outcomes of the treatment groups but to analyze the patterns of patient selection by an experienced treating surgeon and contribute to published data that each surgery can be successful in this patient population when chosen appropriately.

Second, our data were based on a single surgeon’s decisions, and results may not be generalizable. Furthermore, the senior author has had a longstanding interest in reverse shoulder arthroplasty and has published data illustrating successful outcomes for rTSA in patients with MCT. For this reason, one could presume that there may have been some bias toward treating patients with rTSA. However, we feel that the senior author’s unique and longstanding experience in treating MCT allows for a thorough evaluation and comparison of preoperative variables and outcomes declared within this study. Indeed, many patients included in this study were referred from outside institutions specifically for rTSA but instead were deemed more appropriate candidates for RCR and underwent successful arthroscopic repair, a common scenario which served as an impetus for this study.

CONCLUSION

RCR and rTSA are both viable options for patients >65 years with MCT without arthritis. Treatment must be individualized for each patient with careful consideration of a number of preoperative variables and patient characteristics. At our institution, patients with previous RCR, decreased range of motion, poor function, and strong radiographic evidence of subluxation are more likely to undergo rTSA. When chosen appropriately, both RCR and rTSA can result in improved range of motion, function, and high patient satisfaction in this patient population.

The dilemma for the treating surgeon is deciding which patients to treat with RCR and who to treat with rTSA. Predicting which surgical procedure will provide a better functional result is difficult and controversial.⁷ The RCR method is a bone-conserving procedure with relatively low surgical risk and allows the option for rTSA to be performed as a salvage surgery should repair fail. It also may be less costly in the appropriate population.⁸ However, large rotator cuff tears in elderly patients have low healing potential, and the prospect of participating in a lengthy rehabilitation after an operation that may not prove successful can be deterring.^9,10 In the elderly population, rTSA may be a reliable option, as tendon healing of the cuff is not necessary to restore function. However, rTSA does not conserve bone, provides a non-anatomic solution, and has had historically high complication rates.^4,5

In an effort to aid in the decision-making process when considering these 2 surgical options, we compared RCR and rTSA performed at a single institution for MCT in patients >65 years. Our aim was to identify preoperative patient variables that influence a surgeon’s decision to proceed with 1 of the 2 procedures. Moreover, we evaluated clinical outcomes in these 2 patient populations. We hypothesized that (1) patients selected for rTSA would have worse preoperative function, less range of motion, more comorbidities, more evidence of radiographic subluxation, and a higher likelihood of having undergone previous RCR than those selected for RCR, and (2) both RCR and rTSA would be successful and result in improved clinical outcomes with high patient satisfaction.

MATERIALS AND METHODS

PATIENT SELECTION

We performed a retrospective chart review using our practice database of all patients undergoing arthroscopic RCR and rTSA for any indication by the senior author (M.A.F.) between January 2004 and April 2015. A total of 1503 RCRs and 1973 rTSAs were conducted during the study period. Patient medical records were reviewed, and those meeting the following criteria were included in the study: >65 years at the time of surgery, MCT, no preoperative glenohumeral arthritis, minimum follow-up of 12 months, functional deltoid muscle on physical examination, and no prior shoulder surgery except for RCR or diagnostic arthroscopy. A total of 92 patients who underwent arthroscopic RCR and 89 patients who underwent rTSA met the inclusion criteria. For patients with bilateral shoulder surgery, we measured each shoulder independently. Three patients underwent bilateral rTSA, and 3 patients underwent bilateral RCR, leaving 95 shoulders in the RCR group and 92 in the rTSA group. The Western Institutional Review Board determined this study to be exempt from review.

RADIOGRAPHIC EVALUATION

All patient charts included a radiology report and documented interpretation of the images by the treating surgeon prior to surgery. Radiographs were assessed to assure the absence of preoperative glenohumeral osteoarthritis. The images were also graded based on the Hamada classification.¹¹ Stage 1 is associated with minimal radiographic change with an acromiohumeral interval (AHI) >6 mm; stage 2 is characterized by narrowing of the AHI <6 mm; and Stage 3 is defined by narrowing of the AHI with radiographic changes of the acromion. Stages 4 and higher include arthritic changes to the glenohumeral joint, and they were not included in the study population. The AHI measurements and the presence or absence of glenohumeral subluxation were documented.

Continue to: MASSIVE CUFF TEAR DETERMINATION...

We defined MCT on the basis of previously described criteria of tears involving ≥2 tendons or tears measuring ≥5 cm in greatest dimension.^12,13 Patient charts were screened, and those whose clinical notes or radiology reports indicated an absence of MCT were excluded. Preoperative imaging of the remaining patients was then evaluated by 3 fellowship-trained shoulder surgeons to confirm MCT in all patients with a clinically documented MCT, as well as to assess those who had insufficient documentation of tear size in the notes.

Advanced imaging was evaluated for fatty atrophy of the rotator cuff musculature, and Goutallier classification was assigned.^14,15 Length of retraction was measured from the tendon end to the medial aspect of the footprint on coronal imaging, and the subscapularis and teres minor were assessed and documented as torn or intact.^16,17

DATA COLLECTION

We reviewed clinical charts and patient questionnaire forms from both the preoperative and follow-up visits. Clinical data collected included gender, age at surgery, active range of motion (forward elevation, abduction, external rotation, and internal rotation), comorbidities, smoking status, BMI, history of shoulder surgery, and any postoperative complications or need for secondary surgery. All patients completed patient-centered questionnaires regarding shoulder pain and dysfunction at each visit or via telephone communication with clinic staff. Outcome measurements used for analysis included American Shoulder and Elbow Surgeons (ASES) Score, simple shoulder test (SST), visual analog score (VAS) pain scale, and patient-reported satisfaction (Graded 1-10; 1 = poor outcome; 4 = satisfactory outcome; 7 = good outcome; 10 = excellent outcome).

DATA ANALYSIS AND STATISTICAL METHODS

Statistical tests were selected based on the result of Shapiro–Wilk test for normality. Continuous variables were evaluated with either independent t test or Mann–Whitney U test. Dependent t test was used to evaluate outcome variables. For categorical variables, either Pearson’s χ²or Fisher’s exact test was performed depending on the sample size. Alpha was set at P =.05.

Continue to: RESULTS...

PREOPERATIVE CHARACTERISTICS

Of the 187 shoulders in the study group, 95 had RCR and 92 had rTSA. Demographic information and preoperative variables for both groups are summarized in Table 1 and Table 2. Patients in the RCR group had greater preoperative forward elevation, abduction, and external rotation and higher preoperative functional scores than those in the rTSA group. Patients in the rTSA group were older and more likely to be female than those in the RCR group. More patients in the rTSA group had undergone prior RCR compared with those in the RCR group. Each of these differences was statistically significant. Subjective pain scores, BMI, and comorbidities were similar between the 2 groups.

Table 1. Patient demographics

Abbreviations: BMI, body mass index; RCR, rotator cuff repair; rTSA, reverse total shoulder arthroplasty.

Table 2. Preoperative variables

Abbreviations: AB, abduction; ASES, American Shoulder and Elbow Society score; ER, external rotation; FE, forward elevation; IR, internal rotation; preop, preoperative; SST, simple shoulder test.

Radiographically, patients selected to undergo rTSA had a smaller AHI (4.8 vs 8.7, P < .0001) and more evidence of superior subluxation (50.6% vs 14.1%, P < .0001) than those in the RCR group. Average Hamada grade was 1.4 ± 0.7 and 2.2 ± 0.7 for the RCR and rTSA groups, respectively (P < .0001). Average Goutallier grade was similar between the groups (2.2 ± 0.6 for RCR vs 2.2 ± 0.8 for rTSA, P =.227), and 25.5% of the RCR group had Grade 3 or 4 atrophy compared with 28.1% of the rTSA group.

POSTOPERATIVE OUTCOMES

The average follow-up time was 44 months for RCR and 47 months for rTSA. Patients in the RCR and rTSA groups were highly satisfied with the surgery (8.5 ± 2.6 vs 8.2 ± 2.6, P = .461) and had significantly increased range of motion in all planes and improved functional scores (Table 3). The rTSA group had greater net improvement in forward elevation, abduction, and external rotation than the RCR group. Both groups demonstrated similar improvement in ASES, SST, and VAS pain scores.

Table 3. Postoperative outcomes

Abbreviations: AB, abduction; ASES, American Shoulder and Elbow Society score; ER, external rotation; FE, forward elevation; IR, internal rotation; SST, simple shoulder test; VAS – visual analog score.

In the RCR group, 5 patients (5.3%) required reoperation: 3 patients underwent conversion to rTSA, 1 patient underwent biceps tenotomy with subacromial decompression, and 1 patient underwent arthroscopic irrigation and debridement for a postoperative Propionibacterium acnes infection. In the rTSA group, 2 patients (2.2%) required reoperation: 1 patient underwent open reduction internal fixation for a scapula fracture that failed conservative management, and 1 patient had an open irrigation and debridement with polyethylene exchange for an acute postoperative infection of unknown source.

DISCUSSION

Massive, retracted rotator cuff tears are a common and difficult problem.¹ The treatment options are numerous and depend on a variety of preoperative factors including patient-specific characteristics and factors specific to the tear. For certain patients, nonoperative management may be a reasonable first step, as an MCT does not necessarily preclude painless, functional shoulder motion. Elderly, lower demand individuals have been shown to do well with physical rehabilitation.¹⁸ Similarly, for the same category of elderly patients who do not respond to conservative measures, arthroscopic tendon debridement with or without subacromial decompression and/or biceps tenotomy may be effective.^1,19 This technique has been described as “limited goals surgery;” despite some mixed results in the literature, multiple studies have reported symptomatic and functional improvement after simple debridement.^2,19–21The consensus among several authors has been that this procedure continues to play a role for elderly, low-demand patients whose functional goals are limited and whose primary complaint is pain.^1,2,20

For the majority of patients with MCT who desire pain relief and a restoration of shoulder function, RCR remains the gold standard of treatment and should be the primary aim if feasible. Complete RCR has consistently outperformed both partial repair and debridement in multiple studies in terms of pain relief and functional improvement.^10,21,22However, elderly patients with chronic, massive tears, particularly in the setting of muscle atrophy, are at high risk of failure with attempted cuff repair.^9,23 Novel techniques such as superior capsular reconstruction and subacromial balloon spacer implantation may offer a minimally invasive method of re-centering the humeral head and stabilizing the glenohumeral joint; however, these new treatment options lack any long-term data in the literature to support their widespread use.^24–26 Alternatively, rTSA has been shown to be a reliable option to restore shoulder function in the setting of a massive irreparable rotator cuff tear, even in the absence of arthritis.^5,27-31

Continue to: The decision-making process...

The decision-making process for selecting RCR or rTSA in the setting of MCT without arthritis in the older population (age >65 years) remains challenging. We attempted to quantify the data of a high-volume surgeon and identify the differences and similarities between those patients selected for either procedure. At our institution, we generally performed rTSA on patients with low preoperative range of motion, poor function based on SST and ASES scores, small AHI, and strong evidence of superior subluxation. We were also more likely to perform rTSA if the patient had a history of rotator cuff surgery. There was a predilection for older age and female gender in those who underwent rTSA.

For our study, we elected to focus on patients >65 years. In our experience, the choice of which surgical procedure to perform is generally easier in younger patients. Most surgeons appropriately opt for an arthroscopic procedure or tendon transfer to preserve bone and maintain the option of rTSA as a salvage procedure if necessary in the future. Studies have reported that <60 years is a predictor of poor outcome with rTSA, and patients <65 years who undergo rTSA have been shown to have high complication rates.^30-32 Furthermore, the longevity of the implant in young patients is a significant concern, and revision surgery after rTSA is technically demanding and known to result in poor functional outcomes.^32,33

Although the indications for rTSA are expanding, attempts at RCR in the setting of MCT remain largely appropriate. Preserved preoperative anterior elevation >90° has been associated with loss of motion after rTSA and poor satisfaction, and one should exercise caution when considering rTSA in this setting.³ The current study confirmed that even older patients with MCT may be very satisfied with arthroscopic RCR (Figure 1). Both range of motion and function significantly improved, and patients were largely satisfied with the procedure with an average self-reported outcome of good to excellent. At the time of final follow-up for this study, only 3 shoulders in the RCR group had undergone conversion to rTSA. This number may be expected to rise with long follow-up periods, and we feel that prolonging the time before arthroplasty is generally in the best interest of the patient.

Our results were consistent with several reported studies in which RCR has been shown to be successful in the setting of MCT.^34–37 Henry and colleagues³⁶ performed a systematic review that evaluated 954 patients who underwent partial or complete anatomic RCR for MCT. Although the average age was 63 years (range, 37–87), functional outcome scores, VAS pain score, and overall range of motion consistently and significantly improved.

rTSA may be a “more reliable” option than RCR in treating MCT in the older population because it does not rely on tendon healing. However, the relationship between tendon healing and clinical outcomes after RCR is unclear. The aforementioned systematic review reported re-tear rates to be as high as 79%, but several studies have reported high satisfaction even in the setting of retear.³⁶ Yoo and colleagues³⁸ and Chung and colleagues⁹ reported re-tear rates of 45.5% and 39.8%, respectively, but both studies noted that there was no difference in outcome measures between those patients with and without re-tears. In particular, for patients who have had no prior rotator cuff surgery, an attempt at arthroscopic repair may be a prudent option with relatively low risk.

Although certain patients may clinically improve despite suffering a re-tear (or inability to heal in the first place), others continue to experience pain and dysfunction that negatively affect their quality of life.^39–41 These patients are more often appropriate candidates for rTSA. Indeed, several studies have demonstrated a higher re-tear rate in patients with a history of surgery than in those without.^23,31,38,42 Shamsudin and colleagues⁴³ found revision arthroscopic RCR, even in a younger age group with tears of all sizes, to be twice as likely to re-tear. Notably, re-tear after revision repair may be more likely to be symptomatic, as these re-tears are routinely associated with pain, stiffness, and loss of function. Even in the hands of experienced surgeons in a younger population, revision repair has only been able to reverse pseudoparalysis in 43% of patients, leading to only 39% return to sport or full activity.⁴⁴ In examining our data, we were much less likely to perform an RCR in patients who had a history of cuff repair surgery than in those without this history.

Continue to: Overall, those patients selected for rTSA...

Overall, those patients selected for rTSA in our study population performed well postoperatively (Figure 2 and Figure 3). Vast improvements were noted in range of motion, function, and pain scores at final follow up. Moreover, no patients in the study group required revision arthroplasty during the follow-up period. Although the average follow-up period was only 47 months, these results suggested that elderly patients with MCT without arthritis may be particularly ideal candidates for rTSA with regard to implant survival and anticipated revision rate when chosen appropriately.

Several weaknesses were noted within this paper. First, the study was retrospective, precluding randomization of treatment groups and standardization of data collection and follow-up. The outcomes of RCR and rTSA could not be compared directly due to the inherent selection bias. The groups clearly differed in many respects, and these preoperative factors likely played a role in postoperative outcomes. However, the primary goal of this study was not to compare outcomes of the treatment groups but to analyze the patterns of patient selection by an experienced treating surgeon and contribute to published data that each surgery can be successful in this patient population when chosen appropriately.

Second, our data were based on a single surgeon’s decisions, and results may not be generalizable. Furthermore, the senior author has had a longstanding interest in reverse shoulder arthroplasty and has published data illustrating successful outcomes for rTSA in patients with MCT. For this reason, one could presume that there may have been some bias toward treating patients with rTSA. However, we feel that the senior author’s unique and longstanding experience in treating MCT allows for a thorough evaluation and comparison of preoperative variables and outcomes declared within this study. Indeed, many patients included in this study were referred from outside institutions specifically for rTSA but instead were deemed more appropriate candidates for RCR and underwent successful arthroscopic repair, a common scenario which served as an impetus for this study.

CONCLUSION

RCR and rTSA are both viable options for patients >65 years with MCT without arthritis. Treatment must be individualized for each patient with careful consideration of a number of preoperative variables and patient characteristics. At our institution, patients with previous RCR, decreased range of motion, poor function, and strong radiographic evidence of subluxation are more likely to undergo rTSA. When chosen appropriately, both RCR and rTSA can result in improved range of motion, function, and high patient satisfaction in this patient population.

Unsurprisingly, upper extremity injuries are reported to be up to 5 times more common in goalkeepers than in outfield players,^1,2 reaching a high rate of up to 18% of all injuries among professional goalkeepers. The usage of upper extremities to stop the ball and repeated reaching to the ball and landing on the ground with changing upper extremity positions are some of the contributors to the increased upper extremity injury risk in goalkeepers.

Following 57 male professional European soccer teams from 16 countries between the years 2001 and 2011, Ekstrand and colleagues¹ showed that 90% of upper extremity injuries are traumatic, and only 10% are related to overuse. They also reported that the most common upper extremity injury location is the shoulder/clavicle (56%), followed by the hand/finger/thumb (24%), elbow (10%), wrist (5%), forearm (4%), and upper arm (1%). Specifically, the 6 most common injuries are acromioclavicular joint (ACJ) sprain (13%), shoulder dislocation (12%), hand metacarpal fracture (8%), shoulder rotator cuff tendinopathy (6%), hand phalanx fracture (6%), and shoulder ACJ dislocation (5%). 

This article will discuss common upper extremity injuries observed in soccer players, focusing on proper diagnosis and optimal management.

Continue to: THE SHOULDER...

THE SHOULDER

The majority of upper extremity injuries in professional soccer players are shoulder injuries.^1,2,4 Almost a third of these injuries (28%) are considered severe, preventing participation in training and matches for 28 days or more.⁶Ekstrand and colleagues¹ reported that shoulder dislocation represents the most severe upper extremity injury with a mean of 41 days of absence from soccer. When considering the position of the player, they further demonstrated that absence from full training and matches is twice as long for goalkeepers as for outfield players, which reflects the importance of shoulder function for goalkeepers.

In terms of the mechanism of shoulder instability injuries in soccer players, more than half (56%) of these injuries occur with a high-energy mechanism in the recognized position of combined humeral abduction and external rotation against a force of external rotation and horizontal extension.³ However, almost a quarter (24%) occur with a mechanism of varied upper extremity position and low-energy trauma, and 20% of injuries are either a low energy injury with little or no contact or gradual onset. These unique characteristics of shoulder instability injuries in soccer players should be accounted for during training and may imply that current training programs are suboptimal for the prevention of upper extremity injuries and shoulder injuries. Ejnisman and colleagues² reported on the development of a Fédération Internationale de Football Association (FIFA) 11+ shoulder injury prevention program for soccer goalkeepers as one of the ways to promote training programs that address the risk of shoulder injuries.

Reporting on the management of severe shoulder injuries requiring surgery in 25 professional soccer players in England, between 2007 and 2011, Hart and Funk³ found that the majority of subjects (88%) reported a dislocation as a feature of their presentation. Twenty-one (84%) subjects were diagnosed with labral injuries, of which 7 had an associated Hill-Sachs lesion. Two (8%) subjects were diagnosed with rotator cuff tears requiring repair, and 2 (8%) subjects had a combination of rotator cuff and labral injury repair. All patients underwent arthroscopic repair, except for 5 who had a Latarjet coracoid transfer. Post-surgery, all players were able to return to unrestricted participation in soccer at a mean of 11.4 weeks, with no significant difference between goalkeepers and outfield players and no recurrences at a mean of 91 weeks’ follow-up. 

Up to one-third of shoulder instability injuries in soccer players are reported to be recurrences,^1,3 which emphasizes the need to carefully assess soccer players before clearing them to return to play. These data raise the controversy over the treatment of first time shoulder dislocators and may support early surgical intervention.^7-9 In terms of the preferred surgical intervention in these cases, Balg and Boileau¹⁰ suggested a simple scoring system based on factors derived from a preoperative questionnaire, physical examination, and anteroposterior radiographs to help distinguish between patients who will benefit from an arthroscopic anterior stabilization using suture anchors and those who will require a bony procedure (open or arthroscopic). Cerciello and colleagues¹¹ reported excellent results for bony stabilization (modified Latarjet) in a population of 26 soccer players (28 shoulders) affected by chronic anterior instability. Only 1 player did not return to soccer, and 18 players (20 shoulders, 71%) returned to the same level. One re-dislocation was noted in a goalkeeper 74 months after surgery.

An injury to the ACJ has been previously reported to be the most prevalent type of shoulder injury in contact sports.¹²In soccer, injury to the ACJ is responsible for 18% of upper extremity injuries, and the majority (72%) are sprains.¹Interestingly, but unsurprisingly, implications of such an injury differ significantly between goalkeepers and outfield players with up to 3 times longer required absence periods for goalkeepers vs outfield players sustaining the same injury.

ACJ injury is commonly the result of a direct fall on the shoulder with the arm adducted or extended. Six grades of ACJ injuries have been described and distinguished by the injured anatomical structure (acromioclavicular ligaments and coracoclavicular ligaments) and the direction and magnitude of clavicular dislocation.^13,14 Presentation will usually include anterior shoulder pain, a noticeable swelling or change in morphology of the lateral end of the clavicle (mainly in dislocation types), and sharp pain provoked by palpation of the ACJ. Radiographic imaging will confirm the diagnosis and help with identifying the specific grade/type of injury.

Decision making and management of acute ACJ injury should be based on the type/grade of injury. Nonoperative treatment is recommended for types I and II, and most athletes have a successful outcome with a full return to play.¹²Types IV, V, and VI are treated early with operative intervention, mostly due to the morbidity associated with prolonged dislocation of the joint and subsequent soft tissue damage.¹² Treatment of type III injury remains controversial. Pereira-Graterol and colleagues¹⁵ reported the effectiveness of clavicular hook plate (DePuy Synthes) in the surgical stabilization of grade III ACJ dislocation in 11 professional soccer players. At a mean follow-up of 4 years, they showed excellent functional results with full shoulder range of motion at 5 weeks and latest return to soccer at 6 months. The hook plate was removed after 16 weeks in 10 patients in whom no apparent complication was observed.

Continue to: THE ELBOW...

THE ELBOW

Ekstrand and colleagues¹ reported that 10% of all upper extremity injuries in professional soccer players are elbow injuries, of which only 19% are considered severe injuries that require more than 28 days of absence from playing soccer. The most common elbow injuries in their cohort were elbow medial collateral ligament (MCL) sprain and olecranon bursitis.

Elbow MCL is the primary constraint of the elbow joint to valgus stress, and MCL sprain occurs when the elbow is subjected to a valgus, or laterally directed force, which distracts the medial side of the elbow, exceeding the tensile properties of the MCL.¹⁶ A thorough physical examination that includes valgus stress tests through the arc of elbow flexion and extension to elicit a possible subjective feeling of apprehension, instability, or localized pain is essential for optimal evaluation and treatment.^16,17 Imaging studies (X-ray and stress X-rays, dynamic ultrasound, computed tomography [CT], magnetic resonance imaging [MRI], and MR arthrography) have a role in further establishing the diagnosis and identifying possible additional associated injuries.¹⁶ The treatment plan should be specifically tailored to the individual athlete, depending on his demands and the degree of MCL injury. In soccer, which is a non-throwing shoulder sport, nonoperative treatment should be the preferred initial treatment in most cases. Ekstrand and colleagues¹ showed that this injury requires a mean of 4 days of absence from soccer for outfield players and a mean of 21 days of absence from soccer for goalkeepers, thereby indicating more severe sprains and cautious return to soccer in goalkeepers. Athletes who fail nonoperative treatment are candidates for MCL reconstruction.¹⁶

The olecranon bursa is a synovium-lined sac that facilitates gliding between the olecranon and overlying skin. Olecranon bursitis is characterized by accumulation of fluid in the bursa with or without inflammation. The fluid can be serous, sanguineous, or purulent depending on the etiology.¹⁸ In soccer, traumatic etiology is common, but infection secondary to cuts or scratches of the skin around the elbow or previous therapeutic injections around the elbow should always be ruled out. Local pain, swelling, warmth, and redness are usually the presenting symptoms. Aseptic olecranon bursitis may be managed non-surgically with ‘‘benign neglect’’ and avoidance of pressure to the area, non-steroidal anti-inflammatory drugs, needle aspiration, corticosteroid injection, compression dressings, and/or padded splinting; whereas acute septic bursitis requires needle aspiration for diagnosis, appropriate oral or intravenous antibiotics directed toward the offending organism, and, when clinically indicated, surgical evacuation/excision of the bursa.¹⁸ When treating this condition with cortisone injection, possible complications, such as skin atrophy, secondary infection, and chronic local pain, have been reported and should be considered.¹⁹

Severe elbow injuries in professional athletes in general,^20-22 and soccer players specifically, are elbow subluxations/dislocations and elbow fracture. The mechanism of injury is usually contact injury with an opponent player or a fall on the palm with the arm extended. Posterolateral is the most common type of elbow dislocation. Elbow dislocations are further classified into simple (no associated fractures) and complex (associated with fractures) categories.²² Simple dislocations are usually treated with early mobilization after closed reduction; it is associated with a low risk for re-dislocation and with generally good results. The complex type of elbow fracture dislocation is more difficult to treat, has higher complication and re-dislocation rates, and requires operative treatment in most cases compared with simple dislocation.²² The “terrible triad” of the elbow (posterior elbow dislocation, radial head fracture, and coronoid fracture) represents a specific complex elbow dislocation scenario that is difficult to manage because of conflicting aims of ensuring elbow stability while maintaining early range of motion.²²

Isolated fracture around the elbow should be treated based on known principles of fracture management: mechanism of injury, fracture patterns, fracture displacement, intra-articular involvement, soft tissue condition, and associated injuries.

Continue to: THE WRIST...

THE WRIST

Ekstrand and colleagues¹ reported that 5% of all upper extremity injuries in their cohort of professional soccer players are wrist injuries, of which, only 2% are considered severe injuries that require >28 days of absence from playing soccer. The more common wrist injuries in soccer, which is considered a high-impact sport, are fractures (distal radius, scaphoid, capitate), and less reported injuries are dislocations (lunate, perilunate) and ligamentous injuries or tears (scapholunate ligament).²³

Distal radius fractures in high-impact sports, like soccer, usually occur as a result of a fall on an out-stretched hand and will usually be more comminuted, displaced, and intra-articular compared with low-impact sports.²³ All these aforementioned characteristics usually indicate surgical management of open reduction and internal fixation, which will allow for rapid start of rehabilitation and return to play.

Scaphoid fracture is the most common carpal bone fracture and presents unique challenges in terms of diagnosis and optimal treatment²⁴ in professional athletes. A typical injury scenario would be a player falling on an outstretched hand and sustaining a scaphoid fracture during a match or training session. The player may acknowledge some wrist pain but will often continue to play with minimal or no limitation. As wrist pain and swelling become more evident after the match/training session, the player will seek medical evaluation.²⁴ A complete wrist and upper extremity examination should be performed in addition to a specific assessment, which includes palpation of the distal scaphoid pole at the distal wrist flexion crease, palpation of the scaphoid waist through the wrist snuff box, and palpation dorsally just distal to the Lister tubercle at the scapholunate joint. Any wrist injury that results in decreased range of motion, snuff box swelling, and scaphoid tenderness should be further evaluated with imaging. Plain radiographs with special scaphoid views are the initial preferred imaging studies, but occult fracture will require an additional study such as a bone scan, CT, or MRI. Several studies have validated the benefit of MRI and the fact that it may outweigh the costs associated with lost productivity from unnecessary cast immobilization, especially in elite athletes.^23-25Casting the patient with a nondisplaced scaphoid waist fracture has been the traditional treatment; however, stiffness, weakness, and deconditioning that can occur with long-term casting required for scaphoid fractures are significant impairments for the professional athlete and usually end the player’s season. Surgical treatment, which was traditionally indicated for displaced or proximal pole fractures, is currently also considered for non-displaced scaphoid waist fractures in professional athletes. This treatment allows for a rapid return to the rehabilitation of the extremity and possible early return to professional sport. In view of the known complications (eg, malunion, nonunion, and avascular necrosis), return to soccer can be considered when imaging confirms advanced healing, which some consider as at least 50% of bone fracture bridging on CT scan, no pain, excellent motion, and at least 80% of normal grip strength.²⁴ Outfield players can return to play with a protective cast or brace until full healing is observed on imaging.

Continue to: THE HAND/FINGERS/THUMB...

THE HAND/FINGERS/THUMB

Almost a quarter of upper extremity injuries in professional soccer players were reported to involve the hand, fingers, and thumb. A quarter of them were classified as severe injuries requiring >28 days of absence from playing soccer.¹Specifically, hand metacarpal and phalanx fractures are the most common reported injuries in sports in general,²⁶ and in soccer,¹ and account for 14% of all upper extremity injuries¹ in professional soccer players. Goalkeepers require a functional hand to play, whereas an outfielder can play with protection on the injured area; thus, the period of absence from soccer in these injuries is significantly different between goalkeepers and outfielders with more than 4 times longer absence from soccer for a goalkeeper compared with an outfielder. The fifth ray has been shown to be the most frequently fractured ray in the hand in soccer with 51.7% of all hand fractures reported.²⁶ The common mechanism is a full hit on the hand, and a direct hit from the ball is another possible mechanism in goalkeepers.

In general, the diagnosis of hand injuries requires evaluation of the mechanism of injury and injury symptoms, proper and comprehensive physical examination of the whole extremity, and prompt imaging. In most cases, plain radiographs in several projections will suffice for the diagnosis of obvious fractures, but CT scan is an additional modality that allows for improved appreciation of occult or complex and comminuted fracture patterns. MRI or ultrasound can be used additionally whenever associated soft tissue injury is suspected. Optimal management of the hand is based on the specific characteristics of the fractures, which include location, direction of the fracture line, presence of comminution, displacement, articular involvement, and associated soft tissue injury. Nondisplaced extra-articular fractures often can be treated with buddy taping or splinting, whereas intra-articular fractures often require surgical treatment. Displaced fractures of the hand have a tendency to angulate volarly because of attachments of the interosseous muscles. Marginal fractures or avulsion fractures involving the metacarpals or phalanges can be sentinels of serious associated soft tissue injuries.²⁷

Phalangeal fractures can potentially affect the function of the entire hand; therefore, no malrotation is acceptable for phalangeal fractures because they can lead to overlap and malalignment of the digit. Displaced or malrotated fractures should be reduced either by closed or open techniques. Acceptable reduction is <6 mm of shortening, <15° of angulation, and no rotational deformity.^27,28 Nondisplaced phalangeal fractures can be treated nonoperatively with buddy taping and splinting with good results.²⁷ Interphalangeal (IP) dislocations can be reduced on the sidelines and then taped or splinted. Any injury with a force significant enough to cause joint dislocation indicates further evaluation for associated fractures and ligamentous injury or tear. The proximal interphalangeal (PIP) joint is the most common IP joint dislocation and is usually a dorsal dislocation. Reduction is often achieved by traction and flexion of the middle phalanx,²⁷ followed by splinting of the finger with the PIP in 30° of flexion or an extension block splint.²⁹ Successful reduction with no associated intra-articular fractures involving more than a third of the joint can be further managed nonoperatively with the splint, allowing 2 to 4 weeks for the volar plate, joint capsule, and collateral ligaments to heal. Additional 2 to 4 weeks of splinting with buddy taping to the adjacent finger is usually recommended.²⁹

The “Mallet finger” injury can be observed in goalkeepers and is caused by a flexion force on the tip of the finger while the distal interphalangeal (DIP) joint is extended. This force results in tearing of the extensor tendon or an avulsion fracture at the tendinous attachment on the dorsal lip of the distal phalangeal base. The classic mechanism of injury is an extended finger struck on the tip by a ball. Physical examination will indicate loss of DIP joint active extension, and the joint rests in an abnormally flexed position. Treatment typically consists of splinting the DIP joint in extension for 6 to 8 weeks. Operative treatment is reserved for severe injuries or fractures involving greater than one-third of the articular surface of the DIP joint or with failed nonoperative treatment.²⁷ 

Metacarpal fractures can be subdivided into distal, metacarpal neck, metacarpal shaft, and metacarpal base fractures. Metacarpal shaft fractures raise a specific concern regarding rotation, because even a small degree of rotation can create a substantial degree of deformity at the fingertip. This concern must be addressed during evaluation of the player. Fractures of the metacarpal base most commonly involve the fourth and fifth metacarpals and are often reduced easily but have a tendency to re-subluxate, which may indicate operative treatment. Most fractures of the metacarpals are low energy and result in simple fracture patterns that can be treated nonoperatively. Open reduction is reserved for high-energy trauma, fractures with excessive angulation, or multiple fractures.²⁷

Continue to: An important subgroup of metacarpal injuries...

An important subgroup of metacarpal injuries involves the base of the thumb. These injuries result from an axial load applied to the thumb. The most common injury is the “Bennett fracture,” which is an intra-articular fracture or dislocation involving the base of the first metacarpal. Bennett fractures are unstable fractures; unless properly recognized and treated, this intra-articular fracture subluxation may result in an unstable arthritic first carpometacarpal joint. These fractures are most commonly treated with closed or open reduction combined with internal fixation.²⁷ “Rolando fractures” are similar in location and etiology but are comminuted and usually require operative treatment.^{27, 29}

Another common hand injury found in soccer goalkeepers and involving the base of the thumb is disruption of the ulnar collateral ligament (UCL) of the first metacarpophalangeal (MCP) joint as a result of an acute radial or valgus stress on the thumb. Known as “gamekeeper’s thumb” or “skier’s thumb,” this injury can occur in the form of an avulsion fracture, an isolated ligament tear, or combined fracture and ligament rupture. On examination, swelling and tenderness over the thumb UCL are observed. A MCP joint stress test should be performed by gently applying a radially directed force to the thumb while stabilizing the metacarpal bone at both 0° and 30° at the MCP joint. Increased laxity, a soft or nonexistent end point, and gaping of the joint, as compared with the contralateral side, will indicate this injury.²⁹ Radiographs may show a small avulsion fracture fragment at the ulnar aspect of the base of the first metacarpal and at the attachment of the UCL. A Stener lesion is an abnormality that occurs when the thumb adductor muscle aponeurosis interposes between the 2 ends of the ruptured UCL, preventing UCL healing by immobilization alone. Ultrasound and MRI are additional imaging modalities that can assist with the diagnosis of a Stener lesion. The presence of a Stener lesion is a prime indication for surgical intervention. A nondisplaced fracture or isolated ligament injury with no evidence of a Stener lesion can be treated nonoperatively with splinting of the thumb and may lead to healing and restoration of stability. However, in professional players, surgical repair is often times preferred.²⁷

CONCLUSION

Upper extremity injuries are less common injuries among soccer players, but their prevalence is on the rise in recent years. Modern playing tactics and the increase in participation in soccer in younger age groups may be 2 contributing factors to this rise. Given the characteristics of their unique playing role and specific demands, the risk for upper extremity injuries among goalkeepers is significantly higher than that in outfielders and will usually result in a long absence period from soccer before they return to play. A thorough understanding of the mechanism of injury, players’ complaints and presentation, osseous and soft tissue involvement based on a systematic physical examination, imaging features, and treatment options is important for the optimal care of the players. Prompt and accurate diagnosis and appropriate management are essential for improved outcomes and timely return to play.

Unsurprisingly, upper extremity injuries are reported to be up to 5 times more common in goalkeepers than in outfield players,^1,2 reaching a high rate of up to 18% of all injuries among professional goalkeepers. The usage of upper extremities to stop the ball and repeated reaching to the ball and landing on the ground with changing upper extremity positions are some of the contributors to the increased upper extremity injury risk in goalkeepers.

Following 57 male professional European soccer teams from 16 countries between the years 2001 and 2011, Ekstrand and colleagues¹ showed that 90% of upper extremity injuries are traumatic, and only 10% are related to overuse. They also reported that the most common upper extremity injury location is the shoulder/clavicle (56%), followed by the hand/finger/thumb (24%), elbow (10%), wrist (5%), forearm (4%), and upper arm (1%). Specifically, the 6 most common injuries are acromioclavicular joint (ACJ) sprain (13%), shoulder dislocation (12%), hand metacarpal fracture (8%), shoulder rotator cuff tendinopathy (6%), hand phalanx fracture (6%), and shoulder ACJ dislocation (5%). 

This article will discuss common upper extremity injuries observed in soccer players, focusing on proper diagnosis and optimal management.

Continue to: THE SHOULDER...

THE SHOULDER

The majority of upper extremity injuries in professional soccer players are shoulder injuries.^1,2,4 Almost a third of these injuries (28%) are considered severe, preventing participation in training and matches for 28 days or more.⁶Ekstrand and colleagues¹ reported that shoulder dislocation represents the most severe upper extremity injury with a mean of 41 days of absence from soccer. When considering the position of the player, they further demonstrated that absence from full training and matches is twice as long for goalkeepers as for outfield players, which reflects the importance of shoulder function for goalkeepers.

In terms of the mechanism of shoulder instability injuries in soccer players, more than half (56%) of these injuries occur with a high-energy mechanism in the recognized position of combined humeral abduction and external rotation against a force of external rotation and horizontal extension.³ However, almost a quarter (24%) occur with a mechanism of varied upper extremity position and low-energy trauma, and 20% of injuries are either a low energy injury with little or no contact or gradual onset. These unique characteristics of shoulder instability injuries in soccer players should be accounted for during training and may imply that current training programs are suboptimal for the prevention of upper extremity injuries and shoulder injuries. Ejnisman and colleagues² reported on the development of a Fédération Internationale de Football Association (FIFA) 11+ shoulder injury prevention program for soccer goalkeepers as one of the ways to promote training programs that address the risk of shoulder injuries.

Reporting on the management of severe shoulder injuries requiring surgery in 25 professional soccer players in England, between 2007 and 2011, Hart and Funk³ found that the majority of subjects (88%) reported a dislocation as a feature of their presentation. Twenty-one (84%) subjects were diagnosed with labral injuries, of which 7 had an associated Hill-Sachs lesion. Two (8%) subjects were diagnosed with rotator cuff tears requiring repair, and 2 (8%) subjects had a combination of rotator cuff and labral injury repair. All patients underwent arthroscopic repair, except for 5 who had a Latarjet coracoid transfer. Post-surgery, all players were able to return to unrestricted participation in soccer at a mean of 11.4 weeks, with no significant difference between goalkeepers and outfield players and no recurrences at a mean of 91 weeks’ follow-up. 

Up to one-third of shoulder instability injuries in soccer players are reported to be recurrences,^1,3 which emphasizes the need to carefully assess soccer players before clearing them to return to play. These data raise the controversy over the treatment of first time shoulder dislocators and may support early surgical intervention.^7-9 In terms of the preferred surgical intervention in these cases, Balg and Boileau¹⁰ suggested a simple scoring system based on factors derived from a preoperative questionnaire, physical examination, and anteroposterior radiographs to help distinguish between patients who will benefit from an arthroscopic anterior stabilization using suture anchors and those who will require a bony procedure (open or arthroscopic). Cerciello and colleagues¹¹ reported excellent results for bony stabilization (modified Latarjet) in a population of 26 soccer players (28 shoulders) affected by chronic anterior instability. Only 1 player did not return to soccer, and 18 players (20 shoulders, 71%) returned to the same level. One re-dislocation was noted in a goalkeeper 74 months after surgery.

An injury to the ACJ has been previously reported to be the most prevalent type of shoulder injury in contact sports.¹²In soccer, injury to the ACJ is responsible for 18% of upper extremity injuries, and the majority (72%) are sprains.¹Interestingly, but unsurprisingly, implications of such an injury differ significantly between goalkeepers and outfield players with up to 3 times longer required absence periods for goalkeepers vs outfield players sustaining the same injury.

ACJ injury is commonly the result of a direct fall on the shoulder with the arm adducted or extended. Six grades of ACJ injuries have been described and distinguished by the injured anatomical structure (acromioclavicular ligaments and coracoclavicular ligaments) and the direction and magnitude of clavicular dislocation.^13,14 Presentation will usually include anterior shoulder pain, a noticeable swelling or change in morphology of the lateral end of the clavicle (mainly in dislocation types), and sharp pain provoked by palpation of the ACJ. Radiographic imaging will confirm the diagnosis and help with identifying the specific grade/type of injury.

Decision making and management of acute ACJ injury should be based on the type/grade of injury. Nonoperative treatment is recommended for types I and II, and most athletes have a successful outcome with a full return to play.¹²Types IV, V, and VI are treated early with operative intervention, mostly due to the morbidity associated with prolonged dislocation of the joint and subsequent soft tissue damage.¹² Treatment of type III injury remains controversial. Pereira-Graterol and colleagues¹⁵ reported the effectiveness of clavicular hook plate (DePuy Synthes) in the surgical stabilization of grade III ACJ dislocation in 11 professional soccer players. At a mean follow-up of 4 years, they showed excellent functional results with full shoulder range of motion at 5 weeks and latest return to soccer at 6 months. The hook plate was removed after 16 weeks in 10 patients in whom no apparent complication was observed.

Continue to: THE ELBOW...

THE ELBOW

Ekstrand and colleagues¹ reported that 10% of all upper extremity injuries in professional soccer players are elbow injuries, of which only 19% are considered severe injuries that require more than 28 days of absence from playing soccer. The most common elbow injuries in their cohort were elbow medial collateral ligament (MCL) sprain and olecranon bursitis.

Elbow MCL is the primary constraint of the elbow joint to valgus stress, and MCL sprain occurs when the elbow is subjected to a valgus, or laterally directed force, which distracts the medial side of the elbow, exceeding the tensile properties of the MCL.¹⁶ A thorough physical examination that includes valgus stress tests through the arc of elbow flexion and extension to elicit a possible subjective feeling of apprehension, instability, or localized pain is essential for optimal evaluation and treatment.^16,17 Imaging studies (X-ray and stress X-rays, dynamic ultrasound, computed tomography [CT], magnetic resonance imaging [MRI], and MR arthrography) have a role in further establishing the diagnosis and identifying possible additional associated injuries.¹⁶ The treatment plan should be specifically tailored to the individual athlete, depending on his demands and the degree of MCL injury. In soccer, which is a non-throwing shoulder sport, nonoperative treatment should be the preferred initial treatment in most cases. Ekstrand and colleagues¹ showed that this injury requires a mean of 4 days of absence from soccer for outfield players and a mean of 21 days of absence from soccer for goalkeepers, thereby indicating more severe sprains and cautious return to soccer in goalkeepers. Athletes who fail nonoperative treatment are candidates for MCL reconstruction.¹⁶

The olecranon bursa is a synovium-lined sac that facilitates gliding between the olecranon and overlying skin. Olecranon bursitis is characterized by accumulation of fluid in the bursa with or without inflammation. The fluid can be serous, sanguineous, or purulent depending on the etiology.¹⁸ In soccer, traumatic etiology is common, but infection secondary to cuts or scratches of the skin around the elbow or previous therapeutic injections around the elbow should always be ruled out. Local pain, swelling, warmth, and redness are usually the presenting symptoms. Aseptic olecranon bursitis may be managed non-surgically with ‘‘benign neglect’’ and avoidance of pressure to the area, non-steroidal anti-inflammatory drugs, needle aspiration, corticosteroid injection, compression dressings, and/or padded splinting; whereas acute septic bursitis requires needle aspiration for diagnosis, appropriate oral or intravenous antibiotics directed toward the offending organism, and, when clinically indicated, surgical evacuation/excision of the bursa.¹⁸ When treating this condition with cortisone injection, possible complications, such as skin atrophy, secondary infection, and chronic local pain, have been reported and should be considered.¹⁹

Severe elbow injuries in professional athletes in general,^20-22 and soccer players specifically, are elbow subluxations/dislocations and elbow fracture. The mechanism of injury is usually contact injury with an opponent player or a fall on the palm with the arm extended. Posterolateral is the most common type of elbow dislocation. Elbow dislocations are further classified into simple (no associated fractures) and complex (associated with fractures) categories.²² Simple dislocations are usually treated with early mobilization after closed reduction; it is associated with a low risk for re-dislocation and with generally good results. The complex type of elbow fracture dislocation is more difficult to treat, has higher complication and re-dislocation rates, and requires operative treatment in most cases compared with simple dislocation.²² The “terrible triad” of the elbow (posterior elbow dislocation, radial head fracture, and coronoid fracture) represents a specific complex elbow dislocation scenario that is difficult to manage because of conflicting aims of ensuring elbow stability while maintaining early range of motion.²²

Isolated fracture around the elbow should be treated based on known principles of fracture management: mechanism of injury, fracture patterns, fracture displacement, intra-articular involvement, soft tissue condition, and associated injuries.

Continue to: THE WRIST...

THE WRIST

Ekstrand and colleagues¹ reported that 5% of all upper extremity injuries in their cohort of professional soccer players are wrist injuries, of which, only 2% are considered severe injuries that require >28 days of absence from playing soccer. The more common wrist injuries in soccer, which is considered a high-impact sport, are fractures (distal radius, scaphoid, capitate), and less reported injuries are dislocations (lunate, perilunate) and ligamentous injuries or tears (scapholunate ligament).²³

Distal radius fractures in high-impact sports, like soccer, usually occur as a result of a fall on an out-stretched hand and will usually be more comminuted, displaced, and intra-articular compared with low-impact sports.²³ All these aforementioned characteristics usually indicate surgical management of open reduction and internal fixation, which will allow for rapid start of rehabilitation and return to play.

Scaphoid fracture is the most common carpal bone fracture and presents unique challenges in terms of diagnosis and optimal treatment²⁴ in professional athletes. A typical injury scenario would be a player falling on an outstretched hand and sustaining a scaphoid fracture during a match or training session. The player may acknowledge some wrist pain but will often continue to play with minimal or no limitation. As wrist pain and swelling become more evident after the match/training session, the player will seek medical evaluation.²⁴ A complete wrist and upper extremity examination should be performed in addition to a specific assessment, which includes palpation of the distal scaphoid pole at the distal wrist flexion crease, palpation of the scaphoid waist through the wrist snuff box, and palpation dorsally just distal to the Lister tubercle at the scapholunate joint. Any wrist injury that results in decreased range of motion, snuff box swelling, and scaphoid tenderness should be further evaluated with imaging. Plain radiographs with special scaphoid views are the initial preferred imaging studies, but occult fracture will require an additional study such as a bone scan, CT, or MRI. Several studies have validated the benefit of MRI and the fact that it may outweigh the costs associated with lost productivity from unnecessary cast immobilization, especially in elite athletes.^23-25Casting the patient with a nondisplaced scaphoid waist fracture has been the traditional treatment; however, stiffness, weakness, and deconditioning that can occur with long-term casting required for scaphoid fractures are significant impairments for the professional athlete and usually end the player’s season. Surgical treatment, which was traditionally indicated for displaced or proximal pole fractures, is currently also considered for non-displaced scaphoid waist fractures in professional athletes. This treatment allows for a rapid return to the rehabilitation of the extremity and possible early return to professional sport. In view of the known complications (eg, malunion, nonunion, and avascular necrosis), return to soccer can be considered when imaging confirms advanced healing, which some consider as at least 50% of bone fracture bridging on CT scan, no pain, excellent motion, and at least 80% of normal grip strength.²⁴ Outfield players can return to play with a protective cast or brace until full healing is observed on imaging.

Continue to: THE HAND/FINGERS/THUMB...

THE HAND/FINGERS/THUMB

Almost a quarter of upper extremity injuries in professional soccer players were reported to involve the hand, fingers, and thumb. A quarter of them were classified as severe injuries requiring >28 days of absence from playing soccer.¹Specifically, hand metacarpal and phalanx fractures are the most common reported injuries in sports in general,²⁶ and in soccer,¹ and account for 14% of all upper extremity injuries¹ in professional soccer players. Goalkeepers require a functional hand to play, whereas an outfielder can play with protection on the injured area; thus, the period of absence from soccer in these injuries is significantly different between goalkeepers and outfielders with more than 4 times longer absence from soccer for a goalkeeper compared with an outfielder. The fifth ray has been shown to be the most frequently fractured ray in the hand in soccer with 51.7% of all hand fractures reported.²⁶ The common mechanism is a full hit on the hand, and a direct hit from the ball is another possible mechanism in goalkeepers.

In general, the diagnosis of hand injuries requires evaluation of the mechanism of injury and injury symptoms, proper and comprehensive physical examination of the whole extremity, and prompt imaging. In most cases, plain radiographs in several projections will suffice for the diagnosis of obvious fractures, but CT scan is an additional modality that allows for improved appreciation of occult or complex and comminuted fracture patterns. MRI or ultrasound can be used additionally whenever associated soft tissue injury is suspected. Optimal management of the hand is based on the specific characteristics of the fractures, which include location, direction of the fracture line, presence of comminution, displacement, articular involvement, and associated soft tissue injury. Nondisplaced extra-articular fractures often can be treated with buddy taping or splinting, whereas intra-articular fractures often require surgical treatment. Displaced fractures of the hand have a tendency to angulate volarly because of attachments of the interosseous muscles. Marginal fractures or avulsion fractures involving the metacarpals or phalanges can be sentinels of serious associated soft tissue injuries.²⁷

Phalangeal fractures can potentially affect the function of the entire hand; therefore, no malrotation is acceptable for phalangeal fractures because they can lead to overlap and malalignment of the digit. Displaced or malrotated fractures should be reduced either by closed or open techniques. Acceptable reduction is <6 mm of shortening, <15° of angulation, and no rotational deformity.^27,28 Nondisplaced phalangeal fractures can be treated nonoperatively with buddy taping and splinting with good results.²⁷ Interphalangeal (IP) dislocations can be reduced on the sidelines and then taped or splinted. Any injury with a force significant enough to cause joint dislocation indicates further evaluation for associated fractures and ligamentous injury or tear. The proximal interphalangeal (PIP) joint is the most common IP joint dislocation and is usually a dorsal dislocation. Reduction is often achieved by traction and flexion of the middle phalanx,²⁷ followed by splinting of the finger with the PIP in 30° of flexion or an extension block splint.²⁹ Successful reduction with no associated intra-articular fractures involving more than a third of the joint can be further managed nonoperatively with the splint, allowing 2 to 4 weeks for the volar plate, joint capsule, and collateral ligaments to heal. Additional 2 to 4 weeks of splinting with buddy taping to the adjacent finger is usually recommended.²⁹

The “Mallet finger” injury can be observed in goalkeepers and is caused by a flexion force on the tip of the finger while the distal interphalangeal (DIP) joint is extended. This force results in tearing of the extensor tendon or an avulsion fracture at the tendinous attachment on the dorsal lip of the distal phalangeal base. The classic mechanism of injury is an extended finger struck on the tip by a ball. Physical examination will indicate loss of DIP joint active extension, and the joint rests in an abnormally flexed position. Treatment typically consists of splinting the DIP joint in extension for 6 to 8 weeks. Operative treatment is reserved for severe injuries or fractures involving greater than one-third of the articular surface of the DIP joint or with failed nonoperative treatment.²⁷ 

Metacarpal fractures can be subdivided into distal, metacarpal neck, metacarpal shaft, and metacarpal base fractures. Metacarpal shaft fractures raise a specific concern regarding rotation, because even a small degree of rotation can create a substantial degree of deformity at the fingertip. This concern must be addressed during evaluation of the player. Fractures of the metacarpal base most commonly involve the fourth and fifth metacarpals and are often reduced easily but have a tendency to re-subluxate, which may indicate operative treatment. Most fractures of the metacarpals are low energy and result in simple fracture patterns that can be treated nonoperatively. Open reduction is reserved for high-energy trauma, fractures with excessive angulation, or multiple fractures.²⁷

Continue to: An important subgroup of metacarpal injuries...

An important subgroup of metacarpal injuries involves the base of the thumb. These injuries result from an axial load applied to the thumb. The most common injury is the “Bennett fracture,” which is an intra-articular fracture or dislocation involving the base of the first metacarpal. Bennett fractures are unstable fractures; unless properly recognized and treated, this intra-articular fracture subluxation may result in an unstable arthritic first carpometacarpal joint. These fractures are most commonly treated with closed or open reduction combined with internal fixation.²⁷ “Rolando fractures” are similar in location and etiology but are comminuted and usually require operative treatment.^{27, 29}

Another common hand injury found in soccer goalkeepers and involving the base of the thumb is disruption of the ulnar collateral ligament (UCL) of the first metacarpophalangeal (MCP) joint as a result of an acute radial or valgus stress on the thumb. Known as “gamekeeper’s thumb” or “skier’s thumb,” this injury can occur in the form of an avulsion fracture, an isolated ligament tear, or combined fracture and ligament rupture. On examination, swelling and tenderness over the thumb UCL are observed. A MCP joint stress test should be performed by gently applying a radially directed force to the thumb while stabilizing the metacarpal bone at both 0° and 30° at the MCP joint. Increased laxity, a soft or nonexistent end point, and gaping of the joint, as compared with the contralateral side, will indicate this injury.²⁹ Radiographs may show a small avulsion fracture fragment at the ulnar aspect of the base of the first metacarpal and at the attachment of the UCL. A Stener lesion is an abnormality that occurs when the thumb adductor muscle aponeurosis interposes between the 2 ends of the ruptured UCL, preventing UCL healing by immobilization alone. Ultrasound and MRI are additional imaging modalities that can assist with the diagnosis of a Stener lesion. The presence of a Stener lesion is a prime indication for surgical intervention. A nondisplaced fracture or isolated ligament injury with no evidence of a Stener lesion can be treated nonoperatively with splinting of the thumb and may lead to healing and restoration of stability. However, in professional players, surgical repair is often times preferred.²⁷

CONCLUSION

Upper extremity injuries are less common injuries among soccer players, but their prevalence is on the rise in recent years. Modern playing tactics and the increase in participation in soccer in younger age groups may be 2 contributing factors to this rise. Given the characteristics of their unique playing role and specific demands, the risk for upper extremity injuries among goalkeepers is significantly higher than that in outfielders and will usually result in a long absence period from soccer before they return to play. A thorough understanding of the mechanism of injury, players’ complaints and presentation, osseous and soft tissue involvement based on a systematic physical examination, imaging features, and treatment options is important for the optimal care of the players. Prompt and accurate diagnosis and appropriate management are essential for improved outcomes and timely return to play.

Unsurprisingly, upper extremity injuries are reported to be up to 5 times more common in goalkeepers than in outfield players,^1,2 reaching a high rate of up to 18% of all injuries among professional goalkeepers. The usage of upper extremities to stop the ball and repeated reaching to the ball and landing on the ground with changing upper extremity positions are some of the contributors to the increased upper extremity injury risk in goalkeepers.

Following 57 male professional European soccer teams from 16 countries between the years 2001 and 2011, Ekstrand and colleagues¹ showed that 90% of upper extremity injuries are traumatic, and only 10% are related to overuse. They also reported that the most common upper extremity injury location is the shoulder/clavicle (56%), followed by the hand/finger/thumb (24%), elbow (10%), wrist (5%), forearm (4%), and upper arm (1%). Specifically, the 6 most common injuries are acromioclavicular joint (ACJ) sprain (13%), shoulder dislocation (12%), hand metacarpal fracture (8%), shoulder rotator cuff tendinopathy (6%), hand phalanx fracture (6%), and shoulder ACJ dislocation (5%). 

This article will discuss common upper extremity injuries observed in soccer players, focusing on proper diagnosis and optimal management.

Continue to: THE SHOULDER...

THE SHOULDER

The majority of upper extremity injuries in professional soccer players are shoulder injuries.^1,2,4 Almost a third of these injuries (28%) are considered severe, preventing participation in training and matches for 28 days or more.⁶Ekstrand and colleagues¹ reported that shoulder dislocation represents the most severe upper extremity injury with a mean of 41 days of absence from soccer. When considering the position of the player, they further demonstrated that absence from full training and matches is twice as long for goalkeepers as for outfield players, which reflects the importance of shoulder function for goalkeepers.

In terms of the mechanism of shoulder instability injuries in soccer players, more than half (56%) of these injuries occur with a high-energy mechanism in the recognized position of combined humeral abduction and external rotation against a force of external rotation and horizontal extension.³ However, almost a quarter (24%) occur with a mechanism of varied upper extremity position and low-energy trauma, and 20% of injuries are either a low energy injury with little or no contact or gradual onset. These unique characteristics of shoulder instability injuries in soccer players should be accounted for during training and may imply that current training programs are suboptimal for the prevention of upper extremity injuries and shoulder injuries. Ejnisman and colleagues² reported on the development of a Fédération Internationale de Football Association (FIFA) 11+ shoulder injury prevention program for soccer goalkeepers as one of the ways to promote training programs that address the risk of shoulder injuries.

Reporting on the management of severe shoulder injuries requiring surgery in 25 professional soccer players in England, between 2007 and 2011, Hart and Funk³ found that the majority of subjects (88%) reported a dislocation as a feature of their presentation. Twenty-one (84%) subjects were diagnosed with labral injuries, of which 7 had an associated Hill-Sachs lesion. Two (8%) subjects were diagnosed with rotator cuff tears requiring repair, and 2 (8%) subjects had a combination of rotator cuff and labral injury repair. All patients underwent arthroscopic repair, except for 5 who had a Latarjet coracoid transfer. Post-surgery, all players were able to return to unrestricted participation in soccer at a mean of 11.4 weeks, with no significant difference between goalkeepers and outfield players and no recurrences at a mean of 91 weeks’ follow-up. 

Up to one-third of shoulder instability injuries in soccer players are reported to be recurrences,^1,3 which emphasizes the need to carefully assess soccer players before clearing them to return to play. These data raise the controversy over the treatment of first time shoulder dislocators and may support early surgical intervention.^7-9 In terms of the preferred surgical intervention in these cases, Balg and Boileau¹⁰ suggested a simple scoring system based on factors derived from a preoperative questionnaire, physical examination, and anteroposterior radiographs to help distinguish between patients who will benefit from an arthroscopic anterior stabilization using suture anchors and those who will require a bony procedure (open or arthroscopic). Cerciello and colleagues¹¹ reported excellent results for bony stabilization (modified Latarjet) in a population of 26 soccer players (28 shoulders) affected by chronic anterior instability. Only 1 player did not return to soccer, and 18 players (20 shoulders, 71%) returned to the same level. One re-dislocation was noted in a goalkeeper 74 months after surgery.

An injury to the ACJ has been previously reported to be the most prevalent type of shoulder injury in contact sports.¹²In soccer, injury to the ACJ is responsible for 18% of upper extremity injuries, and the majority (72%) are sprains.¹Interestingly, but unsurprisingly, implications of such an injury differ significantly between goalkeepers and outfield players with up to 3 times longer required absence periods for goalkeepers vs outfield players sustaining the same injury.

ACJ injury is commonly the result of a direct fall on the shoulder with the arm adducted or extended. Six grades of ACJ injuries have been described and distinguished by the injured anatomical structure (acromioclavicular ligaments and coracoclavicular ligaments) and the direction and magnitude of clavicular dislocation.^13,14 Presentation will usually include anterior shoulder pain, a noticeable swelling or change in morphology of the lateral end of the clavicle (mainly in dislocation types), and sharp pain provoked by palpation of the ACJ. Radiographic imaging will confirm the diagnosis and help with identifying the specific grade/type of injury.

Decision making and management of acute ACJ injury should be based on the type/grade of injury. Nonoperative treatment is recommended for types I and II, and most athletes have a successful outcome with a full return to play.¹²Types IV, V, and VI are treated early with operative intervention, mostly due to the morbidity associated with prolonged dislocation of the joint and subsequent soft tissue damage.¹² Treatment of type III injury remains controversial. Pereira-Graterol and colleagues¹⁵ reported the effectiveness of clavicular hook plate (DePuy Synthes) in the surgical stabilization of grade III ACJ dislocation in 11 professional soccer players. At a mean follow-up of 4 years, they showed excellent functional results with full shoulder range of motion at 5 weeks and latest return to soccer at 6 months. The hook plate was removed after 16 weeks in 10 patients in whom no apparent complication was observed.

Continue to: THE ELBOW...

THE ELBOW

Ekstrand and colleagues¹ reported that 10% of all upper extremity injuries in professional soccer players are elbow injuries, of which only 19% are considered severe injuries that require more than 28 days of absence from playing soccer. The most common elbow injuries in their cohort were elbow medial collateral ligament (MCL) sprain and olecranon bursitis.

Elbow MCL is the primary constraint of the elbow joint to valgus stress, and MCL sprain occurs when the elbow is subjected to a valgus, or laterally directed force, which distracts the medial side of the elbow, exceeding the tensile properties of the MCL.¹⁶ A thorough physical examination that includes valgus stress tests through the arc of elbow flexion and extension to elicit a possible subjective feeling of apprehension, instability, or localized pain is essential for optimal evaluation and treatment.^16,17 Imaging studies (X-ray and stress X-rays, dynamic ultrasound, computed tomography [CT], magnetic resonance imaging [MRI], and MR arthrography) have a role in further establishing the diagnosis and identifying possible additional associated injuries.¹⁶ The treatment plan should be specifically tailored to the individual athlete, depending on his demands and the degree of MCL injury. In soccer, which is a non-throwing shoulder sport, nonoperative treatment should be the preferred initial treatment in most cases. Ekstrand and colleagues¹ showed that this injury requires a mean of 4 days of absence from soccer for outfield players and a mean of 21 days of absence from soccer for goalkeepers, thereby indicating more severe sprains and cautious return to soccer in goalkeepers. Athletes who fail nonoperative treatment are candidates for MCL reconstruction.¹⁶

The olecranon bursa is a synovium-lined sac that facilitates gliding between the olecranon and overlying skin. Olecranon bursitis is characterized by accumulation of fluid in the bursa with or without inflammation. The fluid can be serous, sanguineous, or purulent depending on the etiology.¹⁸ In soccer, traumatic etiology is common, but infection secondary to cuts or scratches of the skin around the elbow or previous therapeutic injections around the elbow should always be ruled out. Local pain, swelling, warmth, and redness are usually the presenting symptoms. Aseptic olecranon bursitis may be managed non-surgically with ‘‘benign neglect’’ and avoidance of pressure to the area, non-steroidal anti-inflammatory drugs, needle aspiration, corticosteroid injection, compression dressings, and/or padded splinting; whereas acute septic bursitis requires needle aspiration for diagnosis, appropriate oral or intravenous antibiotics directed toward the offending organism, and, when clinically indicated, surgical evacuation/excision of the bursa.¹⁸ When treating this condition with cortisone injection, possible complications, such as skin atrophy, secondary infection, and chronic local pain, have been reported and should be considered.¹⁹

Severe elbow injuries in professional athletes in general,^20-22 and soccer players specifically, are elbow subluxations/dislocations and elbow fracture. The mechanism of injury is usually contact injury with an opponent player or a fall on the palm with the arm extended. Posterolateral is the most common type of elbow dislocation. Elbow dislocations are further classified into simple (no associated fractures) and complex (associated with fractures) categories.²² Simple dislocations are usually treated with early mobilization after closed reduction; it is associated with a low risk for re-dislocation and with generally good results. The complex type of elbow fracture dislocation is more difficult to treat, has higher complication and re-dislocation rates, and requires operative treatment in most cases compared with simple dislocation.²² The “terrible triad” of the elbow (posterior elbow dislocation, radial head fracture, and coronoid fracture) represents a specific complex elbow dislocation scenario that is difficult to manage because of conflicting aims of ensuring elbow stability while maintaining early range of motion.²²

Isolated fracture around the elbow should be treated based on known principles of fracture management: mechanism of injury, fracture patterns, fracture displacement, intra-articular involvement, soft tissue condition, and associated injuries.

Continue to: THE WRIST...

THE WRIST

Ekstrand and colleagues¹ reported that 5% of all upper extremity injuries in their cohort of professional soccer players are wrist injuries, of which, only 2% are considered severe injuries that require >28 days of absence from playing soccer. The more common wrist injuries in soccer, which is considered a high-impact sport, are fractures (distal radius, scaphoid, capitate), and less reported injuries are dislocations (lunate, perilunate) and ligamentous injuries or tears (scapholunate ligament).²³

Distal radius fractures in high-impact sports, like soccer, usually occur as a result of a fall on an out-stretched hand and will usually be more comminuted, displaced, and intra-articular compared with low-impact sports.²³ All these aforementioned characteristics usually indicate surgical management of open reduction and internal fixation, which will allow for rapid start of rehabilitation and return to play.

Scaphoid fracture is the most common carpal bone fracture and presents unique challenges in terms of diagnosis and optimal treatment²⁴ in professional athletes. A typical injury scenario would be a player falling on an outstretched hand and sustaining a scaphoid fracture during a match or training session. The player may acknowledge some wrist pain but will often continue to play with minimal or no limitation. As wrist pain and swelling become more evident after the match/training session, the player will seek medical evaluation.²⁴ A complete wrist and upper extremity examination should be performed in addition to a specific assessment, which includes palpation of the distal scaphoid pole at the distal wrist flexion crease, palpation of the scaphoid waist through the wrist snuff box, and palpation dorsally just distal to the Lister tubercle at the scapholunate joint. Any wrist injury that results in decreased range of motion, snuff box swelling, and scaphoid tenderness should be further evaluated with imaging. Plain radiographs with special scaphoid views are the initial preferred imaging studies, but occult fracture will require an additional study such as a bone scan, CT, or MRI. Several studies have validated the benefit of MRI and the fact that it may outweigh the costs associated with lost productivity from unnecessary cast immobilization, especially in elite athletes.^23-25Casting the patient with a nondisplaced scaphoid waist fracture has been the traditional treatment; however, stiffness, weakness, and deconditioning that can occur with long-term casting required for scaphoid fractures are significant impairments for the professional athlete and usually end the player’s season. Surgical treatment, which was traditionally indicated for displaced or proximal pole fractures, is currently also considered for non-displaced scaphoid waist fractures in professional athletes. This treatment allows for a rapid return to the rehabilitation of the extremity and possible early return to professional sport. In view of the known complications (eg, malunion, nonunion, and avascular necrosis), return to soccer can be considered when imaging confirms advanced healing, which some consider as at least 50% of bone fracture bridging on CT scan, no pain, excellent motion, and at least 80% of normal grip strength.²⁴ Outfield players can return to play with a protective cast or brace until full healing is observed on imaging.

Continue to: THE HAND/FINGERS/THUMB...

THE HAND/FINGERS/THUMB

Almost a quarter of upper extremity injuries in professional soccer players were reported to involve the hand, fingers, and thumb. A quarter of them were classified as severe injuries requiring >28 days of absence from playing soccer.¹Specifically, hand metacarpal and phalanx fractures are the most common reported injuries in sports in general,²⁶ and in soccer,¹ and account for 14% of all upper extremity injuries¹ in professional soccer players. Goalkeepers require a functional hand to play, whereas an outfielder can play with protection on the injured area; thus, the period of absence from soccer in these injuries is significantly different between goalkeepers and outfielders with more than 4 times longer absence from soccer for a goalkeeper compared with an outfielder. The fifth ray has been shown to be the most frequently fractured ray in the hand in soccer with 51.7% of all hand fractures reported.²⁶ The common mechanism is a full hit on the hand, and a direct hit from the ball is another possible mechanism in goalkeepers.

In general, the diagnosis of hand injuries requires evaluation of the mechanism of injury and injury symptoms, proper and comprehensive physical examination of the whole extremity, and prompt imaging. In most cases, plain radiographs in several projections will suffice for the diagnosis of obvious fractures, but CT scan is an additional modality that allows for improved appreciation of occult or complex and comminuted fracture patterns. MRI or ultrasound can be used additionally whenever associated soft tissue injury is suspected. Optimal management of the hand is based on the specific characteristics of the fractures, which include location, direction of the fracture line, presence of comminution, displacement, articular involvement, and associated soft tissue injury. Nondisplaced extra-articular fractures often can be treated with buddy taping or splinting, whereas intra-articular fractures often require surgical treatment. Displaced fractures of the hand have a tendency to angulate volarly because of attachments of the interosseous muscles. Marginal fractures or avulsion fractures involving the metacarpals or phalanges can be sentinels of serious associated soft tissue injuries.²⁷

Phalangeal fractures can potentially affect the function of the entire hand; therefore, no malrotation is acceptable for phalangeal fractures because they can lead to overlap and malalignment of the digit. Displaced or malrotated fractures should be reduced either by closed or open techniques. Acceptable reduction is <6 mm of shortening, <15° of angulation, and no rotational deformity.^27,28 Nondisplaced phalangeal fractures can be treated nonoperatively with buddy taping and splinting with good results.²⁷ Interphalangeal (IP) dislocations can be reduced on the sidelines and then taped or splinted. Any injury with a force significant enough to cause joint dislocation indicates further evaluation for associated fractures and ligamentous injury or tear. The proximal interphalangeal (PIP) joint is the most common IP joint dislocation and is usually a dorsal dislocation. Reduction is often achieved by traction and flexion of the middle phalanx,²⁷ followed by splinting of the finger with the PIP in 30° of flexion or an extension block splint.²⁹ Successful reduction with no associated intra-articular fractures involving more than a third of the joint can be further managed nonoperatively with the splint, allowing 2 to 4 weeks for the volar plate, joint capsule, and collateral ligaments to heal. Additional 2 to 4 weeks of splinting with buddy taping to the adjacent finger is usually recommended.²⁹

The “Mallet finger” injury can be observed in goalkeepers and is caused by a flexion force on the tip of the finger while the distal interphalangeal (DIP) joint is extended. This force results in tearing of the extensor tendon or an avulsion fracture at the tendinous attachment on the dorsal lip of the distal phalangeal base. The classic mechanism of injury is an extended finger struck on the tip by a ball. Physical examination will indicate loss of DIP joint active extension, and the joint rests in an abnormally flexed position. Treatment typically consists of splinting the DIP joint in extension for 6 to 8 weeks. Operative treatment is reserved for severe injuries or fractures involving greater than one-third of the articular surface of the DIP joint or with failed nonoperative treatment.²⁷ 

Metacarpal fractures can be subdivided into distal, metacarpal neck, metacarpal shaft, and metacarpal base fractures. Metacarpal shaft fractures raise a specific concern regarding rotation, because even a small degree of rotation can create a substantial degree of deformity at the fingertip. This concern must be addressed during evaluation of the player. Fractures of the metacarpal base most commonly involve the fourth and fifth metacarpals and are often reduced easily but have a tendency to re-subluxate, which may indicate operative treatment. Most fractures of the metacarpals are low energy and result in simple fracture patterns that can be treated nonoperatively. Open reduction is reserved for high-energy trauma, fractures with excessive angulation, or multiple fractures.²⁷

Continue to: An important subgroup of metacarpal injuries...

An important subgroup of metacarpal injuries involves the base of the thumb. These injuries result from an axial load applied to the thumb. The most common injury is the “Bennett fracture,” which is an intra-articular fracture or dislocation involving the base of the first metacarpal. Bennett fractures are unstable fractures; unless properly recognized and treated, this intra-articular fracture subluxation may result in an unstable arthritic first carpometacarpal joint. These fractures are most commonly treated with closed or open reduction combined with internal fixation.²⁷ “Rolando fractures” are similar in location and etiology but are comminuted and usually require operative treatment.^{27, 29}

Another common hand injury found in soccer goalkeepers and involving the base of the thumb is disruption of the ulnar collateral ligament (UCL) of the first metacarpophalangeal (MCP) joint as a result of an acute radial or valgus stress on the thumb. Known as “gamekeeper’s thumb” or “skier’s thumb,” this injury can occur in the form of an avulsion fracture, an isolated ligament tear, or combined fracture and ligament rupture. On examination, swelling and tenderness over the thumb UCL are observed. A MCP joint stress test should be performed by gently applying a radially directed force to the thumb while stabilizing the metacarpal bone at both 0° and 30° at the MCP joint. Increased laxity, a soft or nonexistent end point, and gaping of the joint, as compared with the contralateral side, will indicate this injury.²⁹ Radiographs may show a small avulsion fracture fragment at the ulnar aspect of the base of the first metacarpal and at the attachment of the UCL. A Stener lesion is an abnormality that occurs when the thumb adductor muscle aponeurosis interposes between the 2 ends of the ruptured UCL, preventing UCL healing by immobilization alone. Ultrasound and MRI are additional imaging modalities that can assist with the diagnosis of a Stener lesion. The presence of a Stener lesion is a prime indication for surgical intervention. A nondisplaced fracture or isolated ligament injury with no evidence of a Stener lesion can be treated nonoperatively with splinting of the thumb and may lead to healing and restoration of stability. However, in professional players, surgical repair is often times preferred.²⁷

CONCLUSION

Upper extremity injuries are less common injuries among soccer players, but their prevalence is on the rise in recent years. Modern playing tactics and the increase in participation in soccer in younger age groups may be 2 contributing factors to this rise. Given the characteristics of their unique playing role and specific demands, the risk for upper extremity injuries among goalkeepers is significantly higher than that in outfielders and will usually result in a long absence period from soccer before they return to play. A thorough understanding of the mechanism of injury, players’ complaints and presentation, osseous and soft tissue involvement based on a systematic physical examination, imaging features, and treatment options is important for the optimal care of the players. Prompt and accurate diagnosis and appropriate management are essential for improved outcomes and timely return to play.

The purpose of this study is to detail the postoperative complications that occur following TSA using a large national database. Specifically, our goals are to determine the incidence and types of complications after shoulder arthroplasty, determine the patient factors that are associated with these complications, and evaluate the effects of these complications on postoperative in-hospital mortality and length of stay (LOS). Our hypothesis is that there would be a correlation between specific patient factors and complications and that these complications would adversely correlate to patient postoperative outcomes.

METHODS

DESIGN

We conducted a retrospective analysis of TSAs captured by the Nationwide Inpatient Sample (NIS) database between 2006 and 2010. The NIS is the largest all-payer inpatient database that is currently available to the public in the United States.⁷

The NIS is a part of the Healthcare Cost and Utilization Project funded by the Agency for Healthcare Research and Quality (AHRQ) and the US Department of Health and Human Services. The NIS database is designed to approximate a 20% sample of US hospitals and the patients they serve, including community, academic, general, and specialty-specific hospitals such as orthopedic hospitals.⁷ The 2010 update of the NIS database contains discharge data from 1051 hospitals across 45 states, with a representative sample of >39 million inpatient hospital stays.⁷ The NIS database and its data sources have been independently validated and assessed for quality each year since 1988.⁸Furthermore, comparative analysis of multiple database elements and distributions has been validated against standard norms, including the National Hospital Discharge Survey.⁹ The NIS database has been used in numerous published studies.^2,10,11

PATIENT SELECTION

The yearly NIS databases from 2006 to 2010 were compiled. Patients aged ≥40 years who underwent a TSA were identified using the International Classification of Diseases, 9th Revision (ICD-9), procedural code 81.80. Exclusion criteria were patients with a primary or a secondary diagnosis of humeral or scapular fracture, chronic osteomyelitis, rheumatologic diseases, or evidence of concurrent malignancy (Figure 1).

Native to NIS are patient demographics, including age, sex, and race. Patient comorbidities as described by Elixhauser and colleagues¹² are also included in the database.

Continue to: OUTCOMES...

OUTCOMES

The primary outcome of this study was a description of the type and frequency of postoperative complications of TSA. To conduct this analysis, we queried the TSA cohort for specific ICD-9 codes representing acute cardiac, central nervous system, infectious, gastrointestinal, genitourinary, postoperative shock, renal, respiratory, surgical, vascular, and wound complications. The ICD-9 codes used to identify complications were modeled according to previous literature on various surgical applications and were further parsed to reflect only acute postoperative diagnoses^13-15(see the Appendix for the comprehensive list of ICD-9 codes).

Two additional outcomes were analyzed, including postoperative mortality and LOS. Postoperative mortality was defined as death occurring prior to discharge. We calculated the average LOS among the complication and the noncomplication cohort.

STATISTICAL ANALYSIS

Patient demographics and target outcomes of the study were analyzed by frequency distribution. Where applicable, the chi-square and the Student’s t tests were used to confirm the statistical difference for dichotomous and continuous variables, respectively. Multivariate regressions were performed after controlling for possible clustering of the data using a generalized estimating equation following a previous analytical methodology.^16-20 The results are reported with odds ratios and 95% confidence intervals where applicable, all statistical tests with P ≤ 0.05 were considered to be significant, and all statistical tests were two-sided. We conducted all analyses using SAS, version 9.2 (SAS Institute).

RESULTS

From 2006 to 2010, a weighted sample of 141,973 patients was found to undergo a TSA. After applying our inclusion and exclusion criteria, our study cohort consisted of 125,766 patients (Figure 1).

Continue to: OVERALL TSA COHORT DEMOGRAPHICS...

OVERALL TSA COHORT DEMOGRAPHICS

The average age of the TSA cohort was 69.4 years (standard deviation [SD], 21.20), and 54.1% were females. The cohort had significant comorbidities, with 83.3% of them having at least 1 comorbidity at the time of surgery. Specifically, 31.3% of the patients had 1 comorbidity, 26.5% had 2 comorbidities, and 25.4% had ≥3 comorbidities. Hypertension was the most common comorbidity present in 66.2% of patients, and diabetes was the second most common comorbidity with a prevalence of 16.8%.

COMPLICATION COHORT DEMOGRAPHICS

An overall postoperative complication rate of 6.7% (weighted sample of 8457 patients) was noted in the overall TSA cohort. The TSA cohort was dichotomized into patients who suffered at least 1 complication (weighted, n = 8457) and patients undergoing routine TSAs (weighted, n = 117,308). The average age was significantly higher in the complication vs routine cohort (71.38 vs 69.27 years, P < 0.0001). Similarly, there were significantly more comorbidities (2.51 vs 1.71, P < 0.0001) in the complication cohort.

COMPLICATIONS

We noted a complication rate of 6.7% (weighted sample of 8457 patients). A single complication was noted in 5% of these patients, whereas 1.3% and 0.4% of the patients had 2 and ≥3 complications, respectively. Respiratory abnormalities (2.9%), acute renal failure (0.8%), and cardiac complications (0.8%) were the most prevalent complications after TSA. The list of complications is detailed in Figure 2. Logistic regression analysis of patient characteristics predicting complications showed that advanced age (odds ratio [OR], 2.1 in those aged ≥85 years) and increasing number of comorbidities (≥3; OR, 3.5) were most significant in predicting complications (all P < 0.0001) (Figure 3). Despite the ubiquity of hypertension in this patient population, it was not a significant predictor of complication (OR, 0.9); in contrast, pulmonary disorders (OR, 5.1) and fluid and electrolyte disorders (4.0) were most strongly associated with the development of a postoperative complication after surgery (Figure 4).

EFFECT OF COMPLICATIONS ON LOS

The average length of hospitalization was 2.3 days (95% confidence interval, 2.22-2.25) among the entire cohort. The average LOS was longer in the complication cohort (3.9 days) than in patients who did not have a complication (2.1 days, P < 0.0001). Of the specific complications noted, hemodynamic shock (11.8 days); infectious, most commonly pneumonia (7.6 days); and vascular complications (6.9 days) were associated with the longest hospitalizations. This result is summarized in Figure 5.

MORTALITY

An overall postoperative (in-house) mortality rate of 0.07% was noted (weighted, n = 88). Comparison between the patient cohort that died vs those who survived TSA resulted in significant differences in the rates of complications. Complications that were most significantly different between the cohorts included cardiac (60.47% vs 0.75%, P < 0.0001), postoperative shock (26.61% vs 0.04%, P < 0.0001), and respiratory complications (43.1% vs 2.8%, P < 0.0001). It is important to note that the overall rate of postoperative shock was exceedingly low in the TSA cohort, but it was highly prevalent in the mortality cohort, occurring in 26.61% of patients. A summary of the mortality statistics is presented in Figure 6.

Continue to: DISCUSSION...

DISCUSSION

TSA continues to be associated with high levels of satisfaction;¹ as a result, its incidence is increasing.² As our understanding and efficiency improves nationally, it is imperative that we determine the short-term and longer-term outcomes and complications. In addition, the factors that may affect prognosis must be elucidated to provide a more individualized and effective standard of care. To date, most of the outcome studies of TSA have evaluated long-term outcomes and specific implant-related complications.^1,5,6,21,22 Our intent was to evaluate the complications that occur in the postoperative period and their effect on unique “patient care” outcomes. With knowledge of these complications and the predisposing factors, we can better assess patients, risk-stratify, and provide appropriate guidelines.

We noted that complications occurring after TSA are not uncommon, with >6% of patients suffering a postoperative complication. In this study, the number of complications noted was associated with worse patient outcomes. In addition, we noted that patients undergoing a TSA have a significant burden of comorbidities; however, hematologic and fluid disorders (eg, iron deficiency anemia, pulmonary circulatory disorders, and fluid imbalances) were most important in predicting postoperative complications.

Increased LOS in the hospital after TSA was associated with the occurrence of complications. Of all noted complications, shock and infectious and vascular complications led to the longest hospitalizations. Hospital-acquired pneumonia was the most common infectious etiology, while pulmonary embolism and deep vein thrombosis were the most consistent vascular complications. Although seldom studied in the TSA population, a similar finding has been noted in patients after THA. O’Malley and colleagues,²³ using the American College of Surgeon’s National Surgical Quality Improvement Program database, identified independent factors that were associated with complications and average prolonged LOS. They noted that the occurrence of major complications was associated with a prolonged LOS. Some, but not all the major complications, included organ space infection, cardiac events, pneumonia, and venous thromboembolic events.²³ Therefore, attempts to limit the amount of time spent in hospitals and control the associated costs must focus on managing the incidence of complications.

Postoperative mortality after TSA was uncommon, occurring in 0.07% of the patients in this study. The low incidence of mortality noted in this study is probably related to the fact that our data represent mortality, whereas in the hospital and, unlike most mortality studies, it does not account for patient demise that may occur in the months after surgery. Other reports have noted that mortality occurs in <1.5% of these patients.^24-28 Singh and colleagues²⁵ observed in their evaluation of perioperative mortality after TSA a mortality rate of 0.8% with 90 days after 4380 shoulder replacements performed at their institution. Using multivariate analysis, they were able to identify associations between mortality and increasing American Society of Anesthesiology (ASA) class and Charlson Comorbidity Index. These results in relation to ours would indicate that the majority of patients who die after shoulder arthroplasty do so after initial discharge. Although we could not determine a causal relationship between mortality and patient comorbidities, we noted that certain complications strongly correlated with mortality. In patients who died, there was a relatively high incidence of cardiac (60.5%) and respiratory (43.1%) complications. Similarly, although postoperative shock was almost nonexistent in the patients who survived surgery (0.04%), it was much more common in the patients who suffered mortality (26.6%).

This study is not without limitations. Data were extracted from a national database, therefore precluding the inclusion of specific details of surgery and functional assessment. Inherent to ICD-9 coding, we were unable to assess the exact detail and severity of complications. For instance, we cannot be certain what criteria were used to define “acute renal failure” for each patient. This study is retrospective in nature and therefore adequate randomization and standardization of patients is not possible. Similarly, the nature of the database may not allow for exacting our inclusion and exclusion criteria. However, the large sample size of the patient population lessens the chance of potential biases and type 2 errors. Prior to October 2010, reverse shoulder arthroplasty was coded under the ICD-9procedural code 81.80 as TSA. Therefore, there is some overlap between TSA and reverse shoulder arthroplasty in our data. Reverse shoulder arthroplasty is now coded under ICD-9 procedural code 81.88. It is possible that results may differ if reverse shoulder arthroplasty were excluded from our patient cohort. This can be an area of future research.

CONCLUSION

Although much is known about the long-term hardware and functional complications after TSA, in this study, we have attempted to broaden the understanding of perioperative complications and the associated sequelae. Complications are common after TSA surgery and are related to adverse outcomes. In the setting of healthcare changes, the surgeon and the patient must understand the cause, types, incidence, and outcomes of medical and surgical complications after surgery. This allows for more accurate “standard of care” metrics. Further large-volume multicenter studies are needed to gain further insight into the short- and long-term outcomes of TSA.

The purpose of this study is to detail the postoperative complications that occur following TSA using a large national database. Specifically, our goals are to determine the incidence and types of complications after shoulder arthroplasty, determine the patient factors that are associated with these complications, and evaluate the effects of these complications on postoperative in-hospital mortality and length of stay (LOS). Our hypothesis is that there would be a correlation between specific patient factors and complications and that these complications would adversely correlate to patient postoperative outcomes.

METHODS

DESIGN

We conducted a retrospective analysis of TSAs captured by the Nationwide Inpatient Sample (NIS) database between 2006 and 2010. The NIS is the largest all-payer inpatient database that is currently available to the public in the United States.⁷

The NIS is a part of the Healthcare Cost and Utilization Project funded by the Agency for Healthcare Research and Quality (AHRQ) and the US Department of Health and Human Services. The NIS database is designed to approximate a 20% sample of US hospitals and the patients they serve, including community, academic, general, and specialty-specific hospitals such as orthopedic hospitals.⁷ The 2010 update of the NIS database contains discharge data from 1051 hospitals across 45 states, with a representative sample of >39 million inpatient hospital stays.⁷ The NIS database and its data sources have been independently validated and assessed for quality each year since 1988.⁸Furthermore, comparative analysis of multiple database elements and distributions has been validated against standard norms, including the National Hospital Discharge Survey.⁹ The NIS database has been used in numerous published studies.^2,10,11

PATIENT SELECTION

The yearly NIS databases from 2006 to 2010 were compiled. Patients aged ≥40 years who underwent a TSA were identified using the International Classification of Diseases, 9th Revision (ICD-9), procedural code 81.80. Exclusion criteria were patients with a primary or a secondary diagnosis of humeral or scapular fracture, chronic osteomyelitis, rheumatologic diseases, or evidence of concurrent malignancy (Figure 1).

Native to NIS are patient demographics, including age, sex, and race. Patient comorbidities as described by Elixhauser and colleagues¹² are also included in the database.

Continue to: OUTCOMES...

OUTCOMES

The primary outcome of this study was a description of the type and frequency of postoperative complications of TSA. To conduct this analysis, we queried the TSA cohort for specific ICD-9 codes representing acute cardiac, central nervous system, infectious, gastrointestinal, genitourinary, postoperative shock, renal, respiratory, surgical, vascular, and wound complications. The ICD-9 codes used to identify complications were modeled according to previous literature on various surgical applications and were further parsed to reflect only acute postoperative diagnoses^13-15(see the Appendix for the comprehensive list of ICD-9 codes).

Two additional outcomes were analyzed, including postoperative mortality and LOS. Postoperative mortality was defined as death occurring prior to discharge. We calculated the average LOS among the complication and the noncomplication cohort.

STATISTICAL ANALYSIS

Patient demographics and target outcomes of the study were analyzed by frequency distribution. Where applicable, the chi-square and the Student’s t tests were used to confirm the statistical difference for dichotomous and continuous variables, respectively. Multivariate regressions were performed after controlling for possible clustering of the data using a generalized estimating equation following a previous analytical methodology.^16-20 The results are reported with odds ratios and 95% confidence intervals where applicable, all statistical tests with P ≤ 0.05 were considered to be significant, and all statistical tests were two-sided. We conducted all analyses using SAS, version 9.2 (SAS Institute).

RESULTS

From 2006 to 2010, a weighted sample of 141,973 patients was found to undergo a TSA. After applying our inclusion and exclusion criteria, our study cohort consisted of 125,766 patients (Figure 1).

Continue to: OVERALL TSA COHORT DEMOGRAPHICS...

OVERALL TSA COHORT DEMOGRAPHICS

The average age of the TSA cohort was 69.4 years (standard deviation [SD], 21.20), and 54.1% were females. The cohort had significant comorbidities, with 83.3% of them having at least 1 comorbidity at the time of surgery. Specifically, 31.3% of the patients had 1 comorbidity, 26.5% had 2 comorbidities, and 25.4% had ≥3 comorbidities. Hypertension was the most common comorbidity present in 66.2% of patients, and diabetes was the second most common comorbidity with a prevalence of 16.8%.

COMPLICATION COHORT DEMOGRAPHICS

An overall postoperative complication rate of 6.7% (weighted sample of 8457 patients) was noted in the overall TSA cohort. The TSA cohort was dichotomized into patients who suffered at least 1 complication (weighted, n = 8457) and patients undergoing routine TSAs (weighted, n = 117,308). The average age was significantly higher in the complication vs routine cohort (71.38 vs 69.27 years, P < 0.0001). Similarly, there were significantly more comorbidities (2.51 vs 1.71, P < 0.0001) in the complication cohort.

COMPLICATIONS

We noted a complication rate of 6.7% (weighted sample of 8457 patients). A single complication was noted in 5% of these patients, whereas 1.3% and 0.4% of the patients had 2 and ≥3 complications, respectively. Respiratory abnormalities (2.9%), acute renal failure (0.8%), and cardiac complications (0.8%) were the most prevalent complications after TSA. The list of complications is detailed in Figure 2. Logistic regression analysis of patient characteristics predicting complications showed that advanced age (odds ratio [OR], 2.1 in those aged ≥85 years) and increasing number of comorbidities (≥3; OR, 3.5) were most significant in predicting complications (all P < 0.0001) (Figure 3). Despite the ubiquity of hypertension in this patient population, it was not a significant predictor of complication (OR, 0.9); in contrast, pulmonary disorders (OR, 5.1) and fluid and electrolyte disorders (4.0) were most strongly associated with the development of a postoperative complication after surgery (Figure 4).

EFFECT OF COMPLICATIONS ON LOS

The average length of hospitalization was 2.3 days (95% confidence interval, 2.22-2.25) among the entire cohort. The average LOS was longer in the complication cohort (3.9 days) than in patients who did not have a complication (2.1 days, P < 0.0001). Of the specific complications noted, hemodynamic shock (11.8 days); infectious, most commonly pneumonia (7.6 days); and vascular complications (6.9 days) were associated with the longest hospitalizations. This result is summarized in Figure 5.

MORTALITY

An overall postoperative (in-house) mortality rate of 0.07% was noted (weighted, n = 88). Comparison between the patient cohort that died vs those who survived TSA resulted in significant differences in the rates of complications. Complications that were most significantly different between the cohorts included cardiac (60.47% vs 0.75%, P < 0.0001), postoperative shock (26.61% vs 0.04%, P < 0.0001), and respiratory complications (43.1% vs 2.8%, P < 0.0001). It is important to note that the overall rate of postoperative shock was exceedingly low in the TSA cohort, but it was highly prevalent in the mortality cohort, occurring in 26.61% of patients. A summary of the mortality statistics is presented in Figure 6.

Continue to: DISCUSSION...

DISCUSSION

TSA continues to be associated with high levels of satisfaction;¹ as a result, its incidence is increasing.² As our understanding and efficiency improves nationally, it is imperative that we determine the short-term and longer-term outcomes and complications. In addition, the factors that may affect prognosis must be elucidated to provide a more individualized and effective standard of care. To date, most of the outcome studies of TSA have evaluated long-term outcomes and specific implant-related complications.^1,5,6,21,22 Our intent was to evaluate the complications that occur in the postoperative period and their effect on unique “patient care” outcomes. With knowledge of these complications and the predisposing factors, we can better assess patients, risk-stratify, and provide appropriate guidelines.

We noted that complications occurring after TSA are not uncommon, with >6% of patients suffering a postoperative complication. In this study, the number of complications noted was associated with worse patient outcomes. In addition, we noted that patients undergoing a TSA have a significant burden of comorbidities; however, hematologic and fluid disorders (eg, iron deficiency anemia, pulmonary circulatory disorders, and fluid imbalances) were most important in predicting postoperative complications.

Increased LOS in the hospital after TSA was associated with the occurrence of complications. Of all noted complications, shock and infectious and vascular complications led to the longest hospitalizations. Hospital-acquired pneumonia was the most common infectious etiology, while pulmonary embolism and deep vein thrombosis were the most consistent vascular complications. Although seldom studied in the TSA population, a similar finding has been noted in patients after THA. O’Malley and colleagues,²³ using the American College of Surgeon’s National Surgical Quality Improvement Program database, identified independent factors that were associated with complications and average prolonged LOS. They noted that the occurrence of major complications was associated with a prolonged LOS. Some, but not all the major complications, included organ space infection, cardiac events, pneumonia, and venous thromboembolic events.²³ Therefore, attempts to limit the amount of time spent in hospitals and control the associated costs must focus on managing the incidence of complications.

Postoperative mortality after TSA was uncommon, occurring in 0.07% of the patients in this study. The low incidence of mortality noted in this study is probably related to the fact that our data represent mortality, whereas in the hospital and, unlike most mortality studies, it does not account for patient demise that may occur in the months after surgery. Other reports have noted that mortality occurs in <1.5% of these patients.^24-28 Singh and colleagues²⁵ observed in their evaluation of perioperative mortality after TSA a mortality rate of 0.8% with 90 days after 4380 shoulder replacements performed at their institution. Using multivariate analysis, they were able to identify associations between mortality and increasing American Society of Anesthesiology (ASA) class and Charlson Comorbidity Index. These results in relation to ours would indicate that the majority of patients who die after shoulder arthroplasty do so after initial discharge. Although we could not determine a causal relationship between mortality and patient comorbidities, we noted that certain complications strongly correlated with mortality. In patients who died, there was a relatively high incidence of cardiac (60.5%) and respiratory (43.1%) complications. Similarly, although postoperative shock was almost nonexistent in the patients who survived surgery (0.04%), it was much more common in the patients who suffered mortality (26.6%).

This study is not without limitations. Data were extracted from a national database, therefore precluding the inclusion of specific details of surgery and functional assessment. Inherent to ICD-9 coding, we were unable to assess the exact detail and severity of complications. For instance, we cannot be certain what criteria were used to define “acute renal failure” for each patient. This study is retrospective in nature and therefore adequate randomization and standardization of patients is not possible. Similarly, the nature of the database may not allow for exacting our inclusion and exclusion criteria. However, the large sample size of the patient population lessens the chance of potential biases and type 2 errors. Prior to October 2010, reverse shoulder arthroplasty was coded under the ICD-9procedural code 81.80 as TSA. Therefore, there is some overlap between TSA and reverse shoulder arthroplasty in our data. Reverse shoulder arthroplasty is now coded under ICD-9 procedural code 81.88. It is possible that results may differ if reverse shoulder arthroplasty were excluded from our patient cohort. This can be an area of future research.

CONCLUSION

Although much is known about the long-term hardware and functional complications after TSA, in this study, we have attempted to broaden the understanding of perioperative complications and the associated sequelae. Complications are common after TSA surgery and are related to adverse outcomes. In the setting of healthcare changes, the surgeon and the patient must understand the cause, types, incidence, and outcomes of medical and surgical complications after surgery. This allows for more accurate “standard of care” metrics. Further large-volume multicenter studies are needed to gain further insight into the short- and long-term outcomes of TSA.

The purpose of this study is to detail the postoperative complications that occur following TSA using a large national database. Specifically, our goals are to determine the incidence and types of complications after shoulder arthroplasty, determine the patient factors that are associated with these complications, and evaluate the effects of these complications on postoperative in-hospital mortality and length of stay (LOS). Our hypothesis is that there would be a correlation between specific patient factors and complications and that these complications would adversely correlate to patient postoperative outcomes.

METHODS

DESIGN

We conducted a retrospective analysis of TSAs captured by the Nationwide Inpatient Sample (NIS) database between 2006 and 2010. The NIS is the largest all-payer inpatient database that is currently available to the public in the United States.⁷

The NIS is a part of the Healthcare Cost and Utilization Project funded by the Agency for Healthcare Research and Quality (AHRQ) and the US Department of Health and Human Services. The NIS database is designed to approximate a 20% sample of US hospitals and the patients they serve, including community, academic, general, and specialty-specific hospitals such as orthopedic hospitals.⁷ The 2010 update of the NIS database contains discharge data from 1051 hospitals across 45 states, with a representative sample of >39 million inpatient hospital stays.⁷ The NIS database and its data sources have been independently validated and assessed for quality each year since 1988.⁸Furthermore, comparative analysis of multiple database elements and distributions has been validated against standard norms, including the National Hospital Discharge Survey.⁹ The NIS database has been used in numerous published studies.^2,10,11

PATIENT SELECTION

The yearly NIS databases from 2006 to 2010 were compiled. Patients aged ≥40 years who underwent a TSA were identified using the International Classification of Diseases, 9th Revision (ICD-9), procedural code 81.80. Exclusion criteria were patients with a primary or a secondary diagnosis of humeral or scapular fracture, chronic osteomyelitis, rheumatologic diseases, or evidence of concurrent malignancy (Figure 1).

Native to NIS are patient demographics, including age, sex, and race. Patient comorbidities as described by Elixhauser and colleagues¹² are also included in the database.

Continue to: OUTCOMES...

OUTCOMES

The primary outcome of this study was a description of the type and frequency of postoperative complications of TSA. To conduct this analysis, we queried the TSA cohort for specific ICD-9 codes representing acute cardiac, central nervous system, infectious, gastrointestinal, genitourinary, postoperative shock, renal, respiratory, surgical, vascular, and wound complications. The ICD-9 codes used to identify complications were modeled according to previous literature on various surgical applications and were further parsed to reflect only acute postoperative diagnoses^13-15(see the Appendix for the comprehensive list of ICD-9 codes).

Two additional outcomes were analyzed, including postoperative mortality and LOS. Postoperative mortality was defined as death occurring prior to discharge. We calculated the average LOS among the complication and the noncomplication cohort.

STATISTICAL ANALYSIS

Patient demographics and target outcomes of the study were analyzed by frequency distribution. Where applicable, the chi-square and the Student’s t tests were used to confirm the statistical difference for dichotomous and continuous variables, respectively. Multivariate regressions were performed after controlling for possible clustering of the data using a generalized estimating equation following a previous analytical methodology.^16-20 The results are reported with odds ratios and 95% confidence intervals where applicable, all statistical tests with P ≤ 0.05 were considered to be significant, and all statistical tests were two-sided. We conducted all analyses using SAS, version 9.2 (SAS Institute).

RESULTS

From 2006 to 2010, a weighted sample of 141,973 patients was found to undergo a TSA. After applying our inclusion and exclusion criteria, our study cohort consisted of 125,766 patients (Figure 1).

Continue to: OVERALL TSA COHORT DEMOGRAPHICS...

OVERALL TSA COHORT DEMOGRAPHICS

The average age of the TSA cohort was 69.4 years (standard deviation [SD], 21.20), and 54.1% were females. The cohort had significant comorbidities, with 83.3% of them having at least 1 comorbidity at the time of surgery. Specifically, 31.3% of the patients had 1 comorbidity, 26.5% had 2 comorbidities, and 25.4% had ≥3 comorbidities. Hypertension was the most common comorbidity present in 66.2% of patients, and diabetes was the second most common comorbidity with a prevalence of 16.8%.

COMPLICATION COHORT DEMOGRAPHICS

An overall postoperative complication rate of 6.7% (weighted sample of 8457 patients) was noted in the overall TSA cohort. The TSA cohort was dichotomized into patients who suffered at least 1 complication (weighted, n = 8457) and patients undergoing routine TSAs (weighted, n = 117,308). The average age was significantly higher in the complication vs routine cohort (71.38 vs 69.27 years, P < 0.0001). Similarly, there were significantly more comorbidities (2.51 vs 1.71, P < 0.0001) in the complication cohort.

COMPLICATIONS

We noted a complication rate of 6.7% (weighted sample of 8457 patients). A single complication was noted in 5% of these patients, whereas 1.3% and 0.4% of the patients had 2 and ≥3 complications, respectively. Respiratory abnormalities (2.9%), acute renal failure (0.8%), and cardiac complications (0.8%) were the most prevalent complications after TSA. The list of complications is detailed in Figure 2. Logistic regression analysis of patient characteristics predicting complications showed that advanced age (odds ratio [OR], 2.1 in those aged ≥85 years) and increasing number of comorbidities (≥3; OR, 3.5) were most significant in predicting complications (all P < 0.0001) (Figure 3). Despite the ubiquity of hypertension in this patient population, it was not a significant predictor of complication (OR, 0.9); in contrast, pulmonary disorders (OR, 5.1) and fluid and electrolyte disorders (4.0) were most strongly associated with the development of a postoperative complication after surgery (Figure 4).

EFFECT OF COMPLICATIONS ON LOS

The average length of hospitalization was 2.3 days (95% confidence interval, 2.22-2.25) among the entire cohort. The average LOS was longer in the complication cohort (3.9 days) than in patients who did not have a complication (2.1 days, P < 0.0001). Of the specific complications noted, hemodynamic shock (11.8 days); infectious, most commonly pneumonia (7.6 days); and vascular complications (6.9 days) were associated with the longest hospitalizations. This result is summarized in Figure 5.

MORTALITY

An overall postoperative (in-house) mortality rate of 0.07% was noted (weighted, n = 88). Comparison between the patient cohort that died vs those who survived TSA resulted in significant differences in the rates of complications. Complications that were most significantly different between the cohorts included cardiac (60.47% vs 0.75%, P < 0.0001), postoperative shock (26.61% vs 0.04%, P < 0.0001), and respiratory complications (43.1% vs 2.8%, P < 0.0001). It is important to note that the overall rate of postoperative shock was exceedingly low in the TSA cohort, but it was highly prevalent in the mortality cohort, occurring in 26.61% of patients. A summary of the mortality statistics is presented in Figure 6.

Continue to: DISCUSSION...

DISCUSSION

TSA continues to be associated with high levels of satisfaction;¹ as a result, its incidence is increasing.² As our understanding and efficiency improves nationally, it is imperative that we determine the short-term and longer-term outcomes and complications. In addition, the factors that may affect prognosis must be elucidated to provide a more individualized and effective standard of care. To date, most of the outcome studies of TSA have evaluated long-term outcomes and specific implant-related complications.^1,5,6,21,22 Our intent was to evaluate the complications that occur in the postoperative period and their effect on unique “patient care” outcomes. With knowledge of these complications and the predisposing factors, we can better assess patients, risk-stratify, and provide appropriate guidelines.

We noted that complications occurring after TSA are not uncommon, with >6% of patients suffering a postoperative complication. In this study, the number of complications noted was associated with worse patient outcomes. In addition, we noted that patients undergoing a TSA have a significant burden of comorbidities; however, hematologic and fluid disorders (eg, iron deficiency anemia, pulmonary circulatory disorders, and fluid imbalances) were most important in predicting postoperative complications.

Increased LOS in the hospital after TSA was associated with the occurrence of complications. Of all noted complications, shock and infectious and vascular complications led to the longest hospitalizations. Hospital-acquired pneumonia was the most common infectious etiology, while pulmonary embolism and deep vein thrombosis were the most consistent vascular complications. Although seldom studied in the TSA population, a similar finding has been noted in patients after THA. O’Malley and colleagues,²³ using the American College of Surgeon’s National Surgical Quality Improvement Program database, identified independent factors that were associated with complications and average prolonged LOS. They noted that the occurrence of major complications was associated with a prolonged LOS. Some, but not all the major complications, included organ space infection, cardiac events, pneumonia, and venous thromboembolic events.²³ Therefore, attempts to limit the amount of time spent in hospitals and control the associated costs must focus on managing the incidence of complications.

Postoperative mortality after TSA was uncommon, occurring in 0.07% of the patients in this study. The low incidence of mortality noted in this study is probably related to the fact that our data represent mortality, whereas in the hospital and, unlike most mortality studies, it does not account for patient demise that may occur in the months after surgery. Other reports have noted that mortality occurs in <1.5% of these patients.^24-28 Singh and colleagues²⁵ observed in their evaluation of perioperative mortality after TSA a mortality rate of 0.8% with 90 days after 4380 shoulder replacements performed at their institution. Using multivariate analysis, they were able to identify associations between mortality and increasing American Society of Anesthesiology (ASA) class and Charlson Comorbidity Index. These results in relation to ours would indicate that the majority of patients who die after shoulder arthroplasty do so after initial discharge. Although we could not determine a causal relationship between mortality and patient comorbidities, we noted that certain complications strongly correlated with mortality. In patients who died, there was a relatively high incidence of cardiac (60.5%) and respiratory (43.1%) complications. Similarly, although postoperative shock was almost nonexistent in the patients who survived surgery (0.04%), it was much more common in the patients who suffered mortality (26.6%).

This study is not without limitations. Data were extracted from a national database, therefore precluding the inclusion of specific details of surgery and functional assessment. Inherent to ICD-9 coding, we were unable to assess the exact detail and severity of complications. For instance, we cannot be certain what criteria were used to define “acute renal failure” for each patient. This study is retrospective in nature and therefore adequate randomization and standardization of patients is not possible. Similarly, the nature of the database may not allow for exacting our inclusion and exclusion criteria. However, the large sample size of the patient population lessens the chance of potential biases and type 2 errors. Prior to October 2010, reverse shoulder arthroplasty was coded under the ICD-9procedural code 81.80 as TSA. Therefore, there is some overlap between TSA and reverse shoulder arthroplasty in our data. Reverse shoulder arthroplasty is now coded under ICD-9 procedural code 81.88. It is possible that results may differ if reverse shoulder arthroplasty were excluded from our patient cohort. This can be an area of future research.

CONCLUSION

Although much is known about the long-term hardware and functional complications after TSA, in this study, we have attempted to broaden the understanding of perioperative complications and the associated sequelae. Complications are common after TSA surgery and are related to adverse outcomes. In the setting of healthcare changes, the surgeon and the patient must understand the cause, types, incidence, and outcomes of medical and surgical complications after surgery. This allows for more accurate “standard of care” metrics. Further large-volume multicenter studies are needed to gain further insight into the short- and long-term outcomes of TSA.

The number of RTSAs performed has steadily increased over the past 20 years, with recent estimates of approximately 20,000 RTSAs performed in the United States in 2011.^10,11 However, there is little information about the similarities and differences between those patients undergoing RTSA in various parts of the world regarding surgical indications, patient demographics, and outcomes. The purpose of this study is to perform a systematic review and meta-analysis of the RTSA body of literature to both identify and compare characteristics of studies published (level of evidence, whether a conflict of interest existed), patients analyzed (age, gender), and surgical indications performed across both continents and countries. Essentially, the study aims to answer the question, "Across the world, are we treating the same patients?" The authors hypothesized that there would be no significant differences in RTSA publications, subjects, and indications based on both the continent and country of publication.

METHODS

A systematic review was conducted according to PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) guidelines using a PRISMA checklist.¹² A systematic review registration was performed using PROSPERO, the international prospective register of systematic reviews (registration number CRD42014010578).¹³Two reviewers independently conducted the search on March 25, 2014, using the following databases: Medline, Cochrane Central Register of Controlled Trials, SportDiscus, and CINAHL. The electronic search citation algorithm utilized was: (((((reverse[Title/Abstract]) AND shoulder[Title/Abstract]) AND arthroplasty[Title/Abstract]) NOT arthroscopic[Title/Abstract]) NOT cadaver[Title/Abstract]) NOT biomechanical[Title/Abstract]. English language Level I to IV evidence (2011 update by the Oxford Centre for Evidence-Based Medicine¹⁴) clinical studies were eligible. Medical conference abstracts were ineligible for inclusion. All references within included studies were cross-referenced for inclusion if missed by the initial search with any additionally located studies screened for inclusion. Duplicate subject publications within separate unique studies were not reported twice, but rather the study with longer duration follow-up or, if follow-up was equal, the study with the greater number of patients was included. Level V evidence reviews, letters to the editor, basic science, biomechanical and cadaver studies, total shoulder arthroplasty (TSA) papers, arthroscopic shoulder surgery papers, imaging, surgical techniques, and classification studies were excluded.

A total of 255 studies were identified, and, after implementation of the exclusion criteria, 103 studies were included in the final analysis (Figure 1). Subjects of interest in this systematic review underwent RTSA for one of many indications including rotator cuff tear arthropathy, osteoarthritis, rheumatoid arthritis, posttraumatic arthritis, instability, revision from a previous RTSA for instability, infection, acute proximal humerus fracture, revision from a prior proximal humerus fracture, revision from a prior hemiarthroplasty, revision from a prior TSA, osteonecrosis, pseudoparalysis, tumor, and a locked shoulder dislocation. There was no minimum follow-up or rehabilitation requirement. Study and subject demographic parameters analyzed included year of publication, years of subject enrollment, presence of study financial conflict of interest, number of subjects and shoulders, gender, age, body mass index, diagnoses treated, and surgical positioning. Clinical outcome scores sought were the DASH (Disability of the Arm, Shoulder, and Hand), SPADI (Shoulder Pain And Disability Index), Absolute Constant, ASES (American Shoulder and Elbow Score), KSS (Korean Shoulder Score), SST-12 (Simple Shoulder Test), SF-12 (12-item Short Form), SF-36 (36-item Short Form), SSV (Subjective Shoulder Value), EQ-5D (EuroQol-5 Dimension), SANE (Single Assessment Numeric Evaluation), Rowe Score for Instability, Oxford Instability Score, UCLA (University of California, Los Angeles) activity score, Penn Shoulder Score, and VAS (visual analog scale). In addition, ROM (forward elevation, abduction, external rotation, internal rotation) was analyzed. Radiographs and magnetic resonance imaging data were extracted when available. The methodological quality of the study was evaluated using the MCMS (Modified Coleman Methodology Score).¹⁵

STATISTICAL ANALYSIS

First, the number of publications per year, level of evidence, and Modified Coleman Methodology Score were tested for association with the calendar year using linear regression. Second, demographic data were tested for association with the continent using Pearson’s chi-square test or ANOVA. Third, indications were tested for association with the continent using Fisher’s exact test. Finally, clinical outcome scores and ROM were tested for association with the continent using ANOVA. Statistical significance was extracted from studies when available. Statistical significance was defined as P < .05.

Continue to: RESULTS...

RESULTS

There were 103 studies included in the analysis (Figure 1). A total of 8973 patients were included, 62% of whom were female with a mean age of 70.9 ± 6.7 years (Table 1). The average follow-up was 34.3 ± 19.3 months. North America had the overall greatest total number of publications on RTSA, followed by Europe (Figure 2). The total yearly number of publications increased by a mean of 1.95 publications each year (P < .001). There was no association between the mean level of evidence with the year of publication (P = .296) (Figure 3). Overall, the rating of studies was poor for the MCMS (mean 36.9 ± 8.7). The MCMS decreased each year by a mean of 0.76 points (P = .037) (Figure 4).

Table 1. Demographic Data by Continent

Abbreviations: MCMS, Modified Coleman Methodology Score; N/A, not available.

In studies that reported press-fit vs cemented prostheses, the highest percentage of press-fit prostheses compared with cemented prostheses was seen in Australia (84% press-fit), whereas the highest percentage of cemented prostheses was seen in North America (89% cemented). A higher percentage of studies from North America had a financial conflict of interest (COI) than did those from other countries (54% had a COI).

Continue to: Rotator cuff tear arthropathy...

Rotator cuff tear arthropathy was the most common indication for RTSA overall in 5459 patients, followed by pseudoparalysis in 1352 patients (Tables 2 and 3). While studies in North America reported rotator cuff tear arthropathy as the indication for RTSA in 4418 (75.8%) patients, and pseudoparalysis as the next most common indication in 535 (9.2%) patients, studies from Europe reported rotator cuff tear arthropathy as the indication in 895 (33.5%) patients, and pseudoparalysis as the indication in 795 (29.7%) patients. Studies from Asia also had a relatively even split between rotator cuff tear arthropathy and pseudoparalysis (45.3% vs 37.8%), whereas those from Australia were mostly rotator cuff tear arthropathy (77.7%).

Table 2. Number (Percent) of Studies With Each Indication by Continent

Abbreviations: CRPP, closed reduction and percutaneous pinning; ORIF, open reduction internal fixation; RTSA, reverse total shoulder arthroplasty; TSA, total shoulder arthroplasty.

Table 3. Number of Patients With Each Indication as Reported by Individual Studies by Continent

Abbreviations: CRPP, closed reduction and percutaneous pinning; ORIF, open reduction internal fixation; RTSA, reverse total shoulder arthroplasty; TSA, total shoulder arthroplasty.

The ASES, SST-12, and VAS scores were the most frequently reported outcome scores in studies from North America, whereas the Absolute Constant score was the most common score reported in studies from Europe (Table 4). Studies from North America reported significantly higher postoperative external rotation (34.1° ± 13.3° vs 19.3° ± 8.9°) (P < .001) and a greater change in flexion (69.0° ± 24.5° vs 56.3° +/- 11.3°) (P = .004) compared with studies from Europe (Table 5).

Table 4. Outcomes by Continent