n Feature Article
Agreement Among ASES Members on the AAOS Clinical Practice Guidelines E. Scott Paxton, MD; Jonas L. Matzon, MD; Alexa C. Narzikul, BA; Pedro K. Beredjiklian, MD; Joseph A. Abboud, MD
abstract The American Academy of Orthopaedic Surgeons (AAOS) has recently developed several clinical practice guidelines (CPG) involving upper extremity conditions. The purpose of the current study was to evaluate the practice patterns of members of the American Shoulder and Elbow Society (ASES) with regard to the CPGs. An e-mail survey was sent to the 340 members of the ASES. The survey contained 40 questions involving the subject matter of the 2 existing AAOS CPGs pertaining specifically to the shoulder: Optimizing the Management of Rotator Cuff Problems and the Treatment of Glenohumeral Joint Arthritis. Overall, 98 responses were obtained, for a response rate of 29%. Only 19 of 47 CPGs were not “inconclusive” and a recommendation was actually made. A majority (more than 50%) of surgeons agreed with 17 (90%) of 19 of these AAOS recommendations. A strong majority (more than 80%) adhered to 13 (68%) of 19 recommendations. There were 4 consensus recommendations, and more than 50% agreed with all of them. Of the 5 moderate recommendations, more than 50% agreed with 4 of them. There were 10 weak recommendations, and more than 50% of surgeons agreed with 9 of them. There was more than 80% agreement on 18 of 28 inconclusive recommendations. Although the AAOS CPGs are not meant to be fixed protocols, they are intended to unify treatment and/or diagnosis of common problems based on the best evidence available. Despite the majority of the AAOS CPG recommendations for rotator cuff problems and glenohumeral arthritis being inconclusive, most surgeons agree with most of the CPG recommendations. [Orthopedics. 2015; 38(3):e169-e177.]
The authors are from the Warren Alpert Medical School of Brown University (ESP), Providence, Rhode Island; and the Rothman Institute (JLM, ACN, PKB, JAA), Philadelphia, Pennsylvania. Dr Matzon, Ms Narzikul, and Dr Beredjiklian have no relevant financial relationships to disclose. Dr Paxton is a paid consultant for Tornier. Dr Abboud is a paid consultant for DePuy, Integra, and Tornier and receives royalties from Integra. Correspondence should be addressed to: Joseph A. Abboud, MD, Rothman Institute, 925 Chestnut St, Philadelphia, PA 19107 ([email protected]). Received: January 30, 2014; Accepted: May 13, 2014. doi: 10.3928/01477447-20150305-53
MARCH 2015 | Volume 38 • Number 3
e169
n Feature Article
T
he current standards of evidencebased medicine (EBM) require physicians to use the best available evidence to guide their clinical decision making. To simplify this process for their members, the American Academy of Orthopaedic Surgeons (AAOS) has recently developed clinical practice guidelines (CPGs) involving many orthopedic conditions. These CPGs were put forth by AAOS physician volunteer work groups based on systematic reviews of the current scientific and clinical data. Per the AAOS CPG publication, the implications for practice of each recommendation are listed in Table 1.1,2 The rationale was to improve patient care, educate qualified physicians, and develop a standard of evidence for common clinical conditions. Since 2009, two CPGs have been introduced specifically for shoulder conditions: Optimizing the Management of Rotator Cuff Problems and the Treatment of Glenohumeral Joint Arthritis. Since these guidelines have been introduced, the rate of adherence to them by practicing physicians is not known. The purpose of the current study was to evaluate the adherence of members of the American Shoulder and Elbow Society (ASES) to these CPGs involving shoulder conditions. The authors’ hypothesis was that more than 50% of respondents would agree with the majority of the noninconclusive recommendations.
Table 1
American Academy of Orthopaedic Surgeons Clinical Practice Guidelines Strengths of Recommendation Statement Rating
Description of Evidence Strength
Implication for Practice
Strong
Evidence is based on 2 or more “High” strength studies with consistent findings for recommending for or against the intervention. A Strong recommendation means that the benefits of the recommended approach clearly exceed the potential harm (or that the potential harm clearly exceeds the benefits in the case of a strong negative recommendation), and that the strength of the supporting evidence is high.
Practitioners should follow a Strong recommendation unless a clear and compelling rationale for an alternative approach is present.
Moderate
Evidence from 2 or more “Moderate” strength studies with consistent findings, or evidence from a single “High” quality study for recommending for or against the intervention. A Moderate recommendation means that the benefits exceed the potential harm (or that the potential harm clearly exceeds the benefits in the case of a negative recommendation), but the strength of the supporting evidence is not as strong.
Practitioners should generally follow a Moderate recommendation but remain alert to new information and be sensitive to patient preferences.
Limited
Evidence from 2 or more “Low” strength studies with consistent findings, or evidence from a single “Moderate” quality study recommending for or against the intervention or diagnostic. A Limited recommendation means the quality of the supporting evidence that exists is unconvincing or that well-conducted studies show little clear advantage to one approach vs another.
Practitioners should be cautious in deciding whether to follow a recommendation classified as Limited and should exercise judgment and be alert to emerging publications that report evidence. Patient preference should have a substantial influencing role.
Inconclusive
Evidence from a single “Low” quality study or conflicting finding that does not allow a recommendation for or against the intervention.
An Inconclusive recommendation means that there is a lack of compelling evidence resulting in an unclear balance between benefits and potential harm.
Consensus
The supporting evidence is lacking and requires the work group to make a recommendation based on expert opinion by considering the known potential harm and benefits associated with the treatment. A Consensus recommendation means that expert opinion supports the guideline recommendation although there is no available empirical evidence that meets the inclusion criteria.
Practitioners should be flexible in deciding whether to follow a recommendation classified as Consensus, although they may set boundaries on alternatives. Patient preference should have a substantial influencing role.
Materials and Methods An online survey (SurveyMonkey, Portland, Oregon) was created to assess the practice patterns related to the AAOS rotator cuff and glenohumeral arthritis CPGs. Each recommendation from the CPGs was translated into a question. The wording used was either identical to that in the AAOS report or as close as possible. A link to the survey was sent to the 340 members of the ASES with the e-mail address on file as of October 2012. The e-mail contained a brief study description and a link from which the participating surgeon could access and complete the survey.
e170
The survey contained 40 questions, with 4 questions related to the participant’s type of practice, practice envi-
ronment, number of rotator cuff repairs performed yearly, and number of total shoulder arthroplasties (TSAs) performed
Copyright © SLACK Incorporated
only increase in importance as the population ages. The large volume of clinical and research interest published in the orthopaedic literature in the last 10 years recognizes the relative importance of rotator cuff disease. These studies have attempted to address a multitude of important questions regarding rotator cuff treatment, including the following
issues: • The timing and role of non-surgical treatment such as steroid injections, physical therapy, or modalities • The indications for chronic rotator cuff repair • The surgical indications for acute traumatic tears • The effect of multiple confounding factors such as age, diabetes,
or smoking on surgical prognosis • The most effective or appropriate surgical strategy, including débridement versus rotator cuff repair • The most effective of the many available postoperative rehabilitation protocols As with many orthopaedic problems, a multitude of differ-
ent treatment options exist for the patient with a painful rotator cuff. Many of these treatment options have been controversial. Not surprisingly, available research finds little consensus among orthopaedic surgeons on rotator cuff treatment options. For example, 2005 study that examined the surgical indication patterns and the role of physical therapy in New York state
n Feature Article
Table 1: Optimizing the management of rotator cuff problems clinical practice guideline recommendations
yearly. The remaining questions related to the subject matter of the 2 existing shoulder AAOS CPGs (Figure 1). Each question represented a recommendation regardless of the strength of the recommendation (inconclusive, limited, moderate, strong, or consensus). Responses to the study were recorded using a webbased database (SurveyMonkey) without identifiers. Overall, 340 survey invitation e-mails were sent, of which none were returned owing to delivery errors. Therefore, the authors considered all e-mail addresses to be valid and that these recipients had received the study. Descriptive statistics were used to evaluate the responses. The moderate, weak, and consensus recommendations were evaluated for more than 50% of respondents agreeing/adhering to the recommendation. These recommendations were also assessed for a strong majority agreement of more than 80%. The inconclusive recommendations were only assessed for a strong majority agreement of more than 80%. This was due to the fact that a recommendation had not actually been made and many of the questions had only 2 answer choices (ie, agree or disagree). Therefore, there would always be a more than 50% answer choice due to the binary nature of these questions. A multivariate analysis was performed using the 4 demographic questions asked. The Kruskal-Wallis test was used to compare ordinal variables with nonordinal variables, and Spearman’s rho was used to check for association.
Recommendation
Strength of recommendation
1. In the absence of reliable evidence, it is the opinion of the work group that surgery not be performed for asymptomatic, full thickness rotator cuff tears.
Consensus
2. Rotator cuff repair is an option for patients with chronic, symptomatic full thickness tears.
Weak
3. a. We cannot recommend for or against exercise programs (supervised or unsupervised) for patients with rotator cuff tears.
Inconclusive
b. We cannot recommend for or against subacromial injections for patients with rotator cuff tears.
Inconclusive
c. We cannot recommend for or against the use of nonsteroidal anti-inflammatory drugs (NSAIDs), activity modification, ice, heat, iontophoresis, massage, Transcutaneous Electrical Nerve Stimulation (TENS), Pulsed Electromagnetic Field (PEMF), or phonophoresis (ultrasound) for nonoperative management of rotator cuff tears.
Inconclusive
4. a. We suggest that patients who have rotator cuff-related symptoms in the absence of a full thickness tear be initially treated non-surgically using exercise and/or NSAIDs.
Moderate
b. We cannot recommend for or against subacromial corticosteroid injection or PEMF in the treatment of rotator cuff-related symptoms in the absence of a full thickness tear.
Inconclusive
c. We cannot recommend for or against the use of iontophoresis, phonophoresis, TENS, ice, heat, massage, or activity modification for patients who have rotator cuff related symptoms in the absence of a full thickness tear.
Inconclusive
5. Early surgical repair after acute injury is an option for patients with a rotator cuff tear.
Weak
6. We cannot recommend for or against the use of perioperative subacromial corticosteroid injections or non-steroidal anti-inflammatory medications in patients undergoing rotator cuff surgery.
Inconclusive
7. a. It is an option for physicians to advise patients that the following factors correlate with less favorable outcomes after rotator cuff surgery: Increasing age
Weak
MRI Tear Characteristics
Weak
Worker’s Compensation Status
Moderate
b. We cannot recommend for or against advising patients in regard to the following factors related to rotator cuff surgery: Diabetes
Inconclusive
Comorbidities (multiple)
Inconclusive
Smoking
Inconclusive
Prior shoulder infection
Inconclusive
Cervical disease (neck pain and myelopathy)
Inconclusive
8. We suggest that routine acromioplasty is not required at the time of rotator cuff repair.
Moderate
9. It is an option to perform partial rotator cuff repair, debridement, or muscle transfers for patients with irreparable rotator cuff tears when surgery is indicated.
Weak
10. a. It is an option for surgeons to attempt to achieve tendon to bone healing of the cuff in all patients undergoing rotator cuff repair.
Weak
b. We cannot recommend for or against the preferential use of suture anchors versus bone tunnels for repair of full thickness rotator cuff tears.
Inconclusive
c. We cannot recommend for or against a specific technique (arthroscopic, mini-open or open repair) when surgery is indicated for full thickness rotator cuff tears.
Inconclusive
11. a. We suggest surgeons not use a non-crosslinked, porcine small intestine submucosal xenograft patch to treat patients with rotator cuff tears. b. We cannot recommend for or against the use of soft tissue allografts or other xenografts to treat patients with rotator cuff tears.
Moderate Inconclusive
12. In the absence of reliable evidence, it is the opinion of the work group that local cold therapy is beneficial to relieve pain after rotator cuff surgery.
Consensus
13. a. We cannot recommend for or against the preferential use of an abduction pillow versus a standard sling after rotator cuff repair.
Inconclusive
b. We cannot recommend for or against a specific time frame of shoulder immobilization without range of motion exercises after rotator cuff repair.
Inconclusive
c. We cannot recommend for or against a specific time interval prior to initiation of active resistance exercises after rotator cuff repair.
Inconclusive
d. We cannot recommend for or against home-based exercise programs versus facility-based rehabilitation after rotator cuff surgery.
Inconclusive
14. We cannot recommend for or against the use of an indwelling subacromial infusion catheter for pain management after rotator cuff repair.
Inconclusive
Note: This summary does not contain rationales that explain how and why these recommendations were developed nor does it contain the evidence supporting these recommendations. All readers of this summary are strongly urged to consult the full guideline and evidence report for this information.We are confident that those who read the full guideline and evidence report will see that the recommendations were developed using systematic evidence-based processes designed to combat bias, enhance transparency, and promote reproducibility. This summary of recommendations is not intended to stand alone.
Figure 1: Optimizing the Management of Rotator Cuff Problems clinical practice guideline recommendations. (Reprinted with permission. © American Academy of Orthopaedic Surgeons.)
AAOS Now_January 2011.indd 46
Results At the time of final analysis, 98 surveys had been completed, for a response rate of 29%. Of the completing surgeons, most (64%) identified themselves as shoulder and elbow surgeons (Table 2). The participants were relatively evenly split in terms of their current practice environment: 22% “private practice,” 41% “academic practice,” and 37% “private
MARCH 2015 | Volume 38 • Number 3
and academic practice.” A strong majority (83%) of surgeons performed more than 40 rotator cuff repairs per year, with 49% performing more than 100 year. The majority (67%) of surgeons performed more than 20 TSAs per year, with 47% performing more than 50 per year and 24% performing more than 100 per year. Of note, 14% of respondents reported that they do not perform TSA. A majority (more than 50%) of surgeons agreed with 17 (90%) of 19 AAOS recommendations that were not inconclusive (28 were inconclusive). A strong majority (more than 80%) adhered to only 13 (68%) recommendations. There were no strong recommendations. Of the 28 inconclusive recommendations, a strong majority (more than 80%) agreed with 18 (64%). Figure 2 provides a complete list of questions and answer distributions.
12/22/2010 2:57:25 PM
Table 2
Respondent Specialties Specialty
%
Shoulder and elbow surgery
64
Orthopedic sports medicine
29
Upper extremity (including hand)
4
General practice
3
Recommendations With a Moderate Strength of Recommendation There were 5 guidelines with a moderate strength of recommendation, and more than 50% of respondents adhered to 4 of them. More than half (53.1%) of respondents reported that they do not discuss workers’ compensation status as correlating with a less favorable outcome.
e171
earCh and teChnology at n Feature Article
ngth indistudies on ous treatof none treatment unclear, rgical techh support
owing areas quality
herapy in as well as litation of
therapy ent of
corticol treatment
viscosupurgical meral oa py in the meral
idement, plasty/interreatment of
us thromulder need for
as a or after
r subshoulder
areas ance our ent opoa and ecommene future. roup hope dentify and etter qualclinical chers to in current treatment oa. NOW
out
s
rved as roup that ractice ment of eoarthritis. g disclore, Wyeth.
hester.edu
Work
January 2010
now
aaos
41
Table 1: The Treatment of Glenohumeral Joint Osteoarthritis Clinical Practice Guideline Recommendations Recommendation
Strength of recommendation
We are unable to recommend for or against physical therapy for the initial treatment of patients with osteoarthritis (OA) of the glenohumeral (GH) joint.
Inconclusive
We are unable to recommend for or against the use of pharmacotherapy in the initial treatment of patients with GH joint OA.
Inconclusive
We are unable to recommend for or against the use of injectable corticosteroids when treating patients with GH joint OA.
Inconclusive
The use of injectable viscosupplementation is an option when treating patients with GH joint OA.
Weak
We are unable to recommend for or against the use of arthroscopic treatments for patients with GH joint OA. These treatments include débridement, capsular release, chondroplasty, microfracture, removal of loose bodies, biologic and interpositional grafts, subacromial decompression, distal clavicle resection, acromioclavicular joint resection, biceps tenotomy or tenodesis, and labral repair or advancement.
Inconclusive
We are unable to recommend for or against open débridement and/or nonprosthetic or biologic interposition arthroplasty in patients with GH joint OA. These treatments include allograft, biologic and interpositional grafts, and autograft.
Inconclusive
Total shoulder arthroplasty and hemiarthroplasty are options when treating patients with GH joint OA.
Weak
We suggest total shoulder arthroplasty (TSA) over hemiarthroplasty when treating patients with GH joint OA.
Moderate
An option for reducing immediate postoperative complication rates is for patients to avoid TSA by surgeons who perform less than two TSAs per year.
Weak
In the absence of reliable evidence, it is the opinion of this work group that physicians use peri-operative mechanical and/or chemical VTE (venous thromboembolism) prophylaxis for TSA patients.
Consensus
The use of either keeled or pegged all-polyethylene cemented glenoid components are options when performing TSA.
Weak
In the absence of reliable evidence, it is the opinion of this work group that TSA not be performed in patients with GH joint OA who have an irreparable rotator cuff tear.
Consensus
We are unable to recommend for or against biceps tenotomy or tenodesis when performing TSA in patients who have GH joint OA.
Inconclusive
We are unable to recommend for or against a subscapularis transtendonous approach or a lesser tuberosity osteotomy when performing TSA in patients who have GH joint OA.
Inconclusive
We are unable to recommend for or against a specific type of humeral prosthetic design or method of fixation when performing TSA in patients with GH joint OA.
Inconclusive
We are unable to recommend for or against physical therapy following TSA.
Inconclusive
Figure 2: The Treatment of Glenohumeral Joint Osteoarthritis clinical practice guideline recommendaTable 2: Understanding AAOS language tions. (Reprinted with permission. © guideline American Academy of Orthopaedic Surgeons.) Strength
Overall Quality of Evidence
Strong
Good
Description of Evidence Level I evidence from more than one study with consistent for recommending for or against the intervention or Tablefindings 3 diagnostic.
Moderate
Fair
Level II or III evidence from more than one study with Moderate Recommendations
Clinical Practice Guideline Optimizing Weak
consistent findings, or Level I evidence from a single study for recommending for or against the intervention or Agreement diagnostic.
the Management of Rotator Cuff Problems Poor Level IV or V evidence from more than one study with
consistent findings, or Level II or III evidence from a single
The authors suggest that patients who have rotator study forcuff–related recommending for or against the 87.8% intervention or symptoms in the absence of a full-thickness diagnostic. tear be initially treated (NSAIDs); 98% Inconclusive No evidence or conflicting evidence The evidence is insufficient or conflicting(exercise) and does not nonoperatively using exercise and/or NSAIDs.
allow a recommendation for or against the intervention or
It is an option for physicians to advise patients that the following fac46.9% diagnostic. tors correlate less favorable outcomes after cuff surgery: Consensus Nowith evidence There rotator is no supporting evidence. In the absence of reliable evidence, the work group is making a recommendation workers’ compensation. based on their clinical opinion and considering the known
harms and benefits associated The authors suggest that routine acromioplasty is not required at the with the treatment. 63.9% time of rotator cuff repair.
The authors suggest that surgeons not use a non–cross-linked, porcine small intestine submucosal xenograft patch to treat patients with rotator cuff tears.
98% (primary repair); 96.9% (revision repair)
Treatment of Glenohumeral Arthritis The authors suggest total shoulder arthroplasty over hemiarthroplasty when treating patients with glenohumeral joint arthritis. Abbreviation: NSAIDs, nonsteroidal anti-inflammatory drugs.
e172
94.8%
Three of the 5 recommendations were adhered to by more than 80% of respondents, with 36.1% performing routine acromioplasty with all rotator cuff repairs (Table 3). Recommendations With a Weak Strength of Recommendation There were 10 guidelines with a weak strength of recommendation, and more than 50% of respondents adhered to 9 of them, with 72.4% of surgeons not using viscosupplementation to treat glenohumeral arthritis. Seven of the 10 recommendations were adhered to by more than 80% of respondents, with 20.4% reporting that they do not discuss increased age as relating to a less favorable outcome after rotator cuff repair and 20.4% not believing that immediate postoperative complications are reduced when the surgeon performs more than 2 TSAs per year (Table 4). Recommendations With a Consensus Strength of Recommendation There were 4 consensus recommendations, and a majority (more than 50%) of respondents adhered to all 4. However, 3 of these 4 recommendations did not achieve a strong majority (more than 80%) of respondents (Table 5). Routine cold therapy after rotator cuff surgery is offered by 69.4% of respondents. Mechanical or chemical venous thromboembolism prophylaxis after TSA is used by 77.8% of respondents. Total shoulder arthroplasty is not considered a treatment option for patients with glenohumeral arthritis and an irreparable rotator cuff tear by 79.3% of respondents. Recommendations With an Inconclusive Strength of Recommendation There were 28 inconclusive recommendations, and there was more than 80% agreement on 18 of them (Table 6). Confounding Variables No correlation was found between the 4 demographic questions. The only cor-
Copyright © SLACK Incorporated
n Feature Article
relation between demographics and the CPGs was found in upper extremity surgeons (with practices including and excluding hand) preferring arthroscopic rotator cuff repair as opposed to open rotator cuff repair more than general orthopedists and sports medicine surgeons (P6 wk)
The authors cannot recommend for or against a specific time interval prior to initiation of active resistance exercises after rotator cuff repair.
7.1% (depends on repair); 0% (0-2 wk); 1.0% (at or after 2-4 wk); 7.1% (at or after 4-6 wk); 22.4% (at or after 6-8 wk); 12.2% (at or after 8-10 wk); 21.4% (at or after 10-12 wk); 25.5% (at or after 12-16 wk); 3.1% (at or after 16 wk)
The authors cannot recommend for or against home-based exercise programs vs facility-based rehabilitation after rotator cuff surgery.
8.2% (100% home-based); 15.5% (1%-24% home-based); 18.6% (25%49% home-based); 17.5% (50%-74% home-based); 20.6% (75%-90% home-based); 19.6% (91%-100% home-based)
The authors cannot recommend for or against the use of an indwelling subacromial infusion catheter for pain management after rotator cuff repair.
9.2%
Treatment of Glenohumeral Arthritis The authors are unable to recommend for or against physical therapy for the initial treatment of patients with GJA. The authors are unable to recommend for or against the use of pharmacotherapy in the initial treatment of patients with GJA.
e174
61% 93.9%
Copyright © SLACK Incorporated
n Feature Article
Table 6 (cont’d)
Inconclusive Recommendations Clinical Practice Guideline
Agreement
The authors are unable to recommend for or against the use of injectable corticosteroids when treating patients with GJA.
82.7%
The authors are unable to recommend for or against the use of arthroscopic treatments for patients with GJA. These treatments include debridement, capsular release, chondroplasty, microfracture, removal of loose bodies, biologic and interpositional grafts, subacromial decompression, distal clavicle resection, acromioclavicular joint resection, biceps tenotomy or tenodesis, and labral repair or advancement.
70.1% (treat arthroscopically); 67.0% (use debridement); 53.6% (use capsular release); 40.2% (use chondroplasty); 25.8% (use microfracture); 63.9% (use loose body removal); 5.2% (use interpositions); 21.6% (use subacromial decompression); 14.4 (use distal clavical resection); 9.3% (use acromioclavicular joint resection); 43.3% (use biceps tenotomy/tenodesis); 5.2% (use labral repair/advancement)
The authors are unable to recommend for or against open debridement and/or nonprosthetic or biologic interposition arthroplasty in patients with GJA. These treatments include allograft, biologic and interpositional grafts, and autograft.
5.3% (open debridement); 7.4% (interpositions)
The authors are unable to recommend for or against biceps tenotomy or tenodesis when performing TSA in patients who have GJA.
12.2% (preserve); 54.4% (tenodesis); 7.8% (tenotomy); 25.6% (patient specific)
The authors are unable to recommend for or against a subscapularis transtendinous approach or a lesser tuberosity osteotomy when performing TSA in patients who have GJA.
76.4% (transtendon); 23.6% (lesser tuberosity osteotomy)
The authors are unable to recommend for or against a specific type of humeral prosthetic design or method of fixation when performing TSA in patients with GJA.
The authors are unable to recommend for or against physical therapy following TSA.
1.1% (monoblock); 34.1% (modular with variable neck-shaft angle); 70.5% (modular with variable head sizes); 89.8% (modular with eccentric heads); 3.4% (other) 85.7 % (use physical therapy)
Abbreviations: GJA, glenohumeral joint arthritis; NSAIDs, nonsteroidal anti-inflammatory drugs; PEMF, pulsed electromagnetic field; TENS, transcutaneous electrical nerve stimulation; TSA, total shoulder arthroplasty.
higher-level studies. This suggests that a subset of shoulder specialists rely on different information for clinical decision making and not solely on the studies the CPG work groups used or the AAOS CPGs themselves. Despite this, the current authors found that most shoulder surgeons practice in largely comparable manners. Of the 5 moderate recommendations, a majority of surgeons agreed with 4. Multiple preoperative factors were assessed by the work group as relating to outcome after rotator cuff repair. Although it was a moderate recommendation, only 46.9% of surgeons discuss workers’ compensation status as correlating to a worse outcome after rotator cuff repair, and this was the only moderate recommendation
MARCH 2015 | Volume 38 • Number 3
that was not adhered to by more than 50% of respondents. Workers’ compensation status has been documented for multiple orthopedic procedures as having a negative effect on outcome.3-9 A recent meta-analysis of 20 prospective studies evaluating the effect of workers’ compensation status on outcomes of orthopedic procedures found that the overall risk ratio for experiencing an unsatisfactory result for patients with vs without compensation is 2.08 (95% confidence interval, 1.54-2.82).10 Cuff and Pupello11 prospectively evaluated patients after rotator cuff repair stratified to compensated and noncompensated patients. Only 4% of patients without a compensation claim were found to be noncompliant, whereas 52% of compensated patients
were noncompliant. Overall, the compensated patients had worse outcomes than the noncompensated patients. Not surprisingly, the noncompliant compensated patients had worse scores than the compliant compensated patients. Healing rates also differed in the groups but did not reach statistical significance, with 84% healed in the noncompensated group vs 75% in the compliant compensated group and 59% in the noncompliant compensated group.11 These findings provide some foundation for preoperative counseling of workers’ compensation patients and postoperative surveillance of these patients for signs of noncompliance. Physicians may feel uncomfortable talking with workers’ compensation patients preoperatively
e175
n Feature Article
about their worse prognosis; this may explain why many orthopedic surgeons do not. The current authors did not ask the respondents about their reasoning for their practices, so they can only speculate that this may be the case. Alternatively, surgeons may not believe that workers’ compensation patients have poorer outcomes, although the data are convincing. Unlike workers’ compensation status, most respondents discuss magnetic resonance imaging characteristics (92.9%) and patient age (79.6%) in relation to outcomes after rotator cuff repair, although both are weak recommendations. In addition, 93.9% discuss smoking, 78.6% discuss history of shoulder infection, and 73.5% discuss diabetes mellitus as portending a worse outcome after rotator cuff repair, despite the fact that the CPGs for these 3 factors were inconclusive. A small percentage (31.6%) discuss cervical spine disease as a predictor of outcome. Although given the way the recommendation is worded, it is difficult to know whether this includes simply a history of cervical spine problems that have been adequately treated, or whether it more specifically means an acute process with neurologic findings. It would be expected that many surgeons would be more cognizant of a patient with an acute radiculopathy and may wait for resolution prior to surgery but may be less aware of a history of a single-level fusion or some recalcitrant pain. The work group made a moderate recommendation against routine acromioplasty performed with rotator cuff repair. Despite this, 36.1% of respondents perform an acromioplasty with all rotator cuff repairs. This was not surprising because these procedures have historically been performed together; however, only approximately one-third of surgeons perform them routinely, which seems to provide evidence that most surgeons are leaning away from routine acromioplasty. Continued belief in acromial impingement as the primary cause of rotator cuff tears
e176
may explain why more than one-third of the ASES surgeons surveyed still perform routine acromioplasty with a rotator cuff repair, despite multiple studies reporting no significant benefit.12-15 This is evidenced by a letter to the editor regarding this CPG by Lubowitz et al16 discussing the role of acromioplasty with rotator cuff repair and postulating that it may reduce the need for reoperation. Although many of the work group’s recommendations were inconclusive, they relate to important topics in shoulder surgery. Surprisingly, almost half (45.4%) of surgeons reported that they use soft tissue allografts or xenografts in revision rotator cuff repairs. Biomechanical studies evaluating patch augmentation is promising, with decreased gap formation by up to 48%.17 Also, small studies have shown improvement over the preoperative state with the use of an augmented patch during repair.18,19 However, inflammatory reactions and increased pain have been reported with the use of certain products.20 Surgeons are interpreting these results in different ways, with approximately half believing that the current products have benefit in the revision setting and half not. Only 11.2% of surgeons reported that they use a patch in a primary setting. Most (80.4%) surgeons surveyed agreed with the AAOS work group consensus that rotator cuff repair should not be performed for asymptomatic tears; however, 19.6% reported that they perform repairs for symptomatic tears. Natural history studies have shown that approximately half of asymptomatic tears will progress to symptomatic tears within 3 years, and half of these will increase in size.21 In addition, 4% to 20% of asymptomatic patients will have silent progression of their tear.21,22 Therefore, an argument could be made for surgical fixation in these patients because nonoperative treatment may result in tear progression, fatty atrophy, and tendon retraction, resulting in the inability to repair the tendon if it were to become symptomatic or increasing the difficulty
of tendon mobilization and repair. Successful repair may ultimately result in less improvement in function if fatty atrophy progresses significantly prior to repair. In addition, the age of the patient may come into consideration, with those answering in disagreement referring to a younger population where the risk of an irreparable rotator cuff tear is more worrisome. There are several possible sources of bias in the current authors’ results. The overall response rate was 29%. Therefore, a large portion of the ASES did not respond, and the results of the study are susceptible to respondent bias. It is possible that the nonresponders are members who adhere to the CPGs differently. In addition, by using a web-based survey, the authors may have introduced a selection bias in which more computer-savvy ASES members may dominate the survey. Also, the recommendations for glenohumeral arthritis and rotator cuff tears were made using studies published prior to January 2009 and October 2008, respectively. This survey was sent in January 2013, leaving a gap of more than 4 years. It is possible that newer data have been published in this time period that have altered surgeons’ practices. Finally, this study assumes that how the ASES members respond and how they practice are identical, and that may not be the case.
Conclusion The results of this study show that despite the lack of high-level evidence, most shoulder surgeons practice in comparable ways and in agreement with the AAOS CPGs on rotator cuff disease and glenohumeral arthritis.
References 1. Izquierdo R, Voloshin I, Edwards S, et al. American Academy of Orthopaedic Surgeons Clinical Practice Guideline on the treatment of glenohumeral joint osteoarthritis. J Bone Joint Surg Am. 2011; 93:203-205. 2. Pedowitz RA, Yamaguchi K, Ahmad CS, et al. American Academy of Orthopaedic Surgeons Clinical Practice Guideline on optimizing the management of rotator cuff problems.
Copyright © SLACK Incorporated
n Feature Article
J Bone Joint Surg Am. 2012; 94:163-167. 3. Anderson PA, Subach BR, Riew KD. Predictors of outcome after anterior cervical discectomy and fusion: a multivariate analysis. Spine (Phila Pa 1976). 2009; 34:161-166. 4. Bhatia S, Piasecki DP, Nho SJ, et al. Early return to work in workers’ compensation patients after arthroscopic full-thickness rotator cuff repair. Arthroscopy. 2010; 26:10271034. 5. Brinker MR, Savory CG, Weeden SH, Aucoin HC, Curd DT. The results of total knee arthroplasty in workers’ compensation patients. Bull Hosp Jt Dis. 1998; 57:80-83. 6. Denard PJ, Ladermann A, Burkhart SS. Long-term outcome after arthroscopic repair of type II SLAP lesions: results according to age and workers’ compensation status. Arthroscopy. 2012; 28:451-457. 7. Holtby R, Razmjou H. Impact of work-related compensation claims on surgical outcome of patients with rotator cuff related pathologies: a matched case-control study. J Shoulder Elbow Surg. 2010; 19:452-460. 8. Hou WH, Tsauo JY, Lin CH, Liang HW, Du CL. Worker’s compensation and return-towork following orthopaedic injury to extremities. J Rehabil Med. 2008; 40:440-445. 9. Wexler G, Bach BR Jr, Bush-Joseph CA, Smink D, Ferrari JD, Bojchuk J. Outcomes of anterior cruciate ligament reconstruction in patients with workers’ compensation claims. Arthroscopy. 2000; 16:49-58.
MARCH 2015 | Volume 38 • Number 3
10. de Moraes VY, Godin K, Tamaoki MJ, Faloppa F, Bhandari M, Belloti JC. Workers’ compensation status: does it affect orthopaedic surgery outcomes? A meta-analysis. PLoS One. 2012; 7:e50251. 11. Cuff DJ, Pupello DR. Prospective evaluation of postoperative compliance and outcomes after rotator cuff repair in patients with and without workers’ compensation claims. J Shoulder Elbow Surg. 2012; 21:1728-1733. 12. Chahal J, Mall N, MacDonald PB, et al. The role of subacromial decompression in patients undergoing arthroscopic repair of fullthickness tears of the rotator cuff: a systematic review and meta-analysis. Arthroscopy. 2012; 28:720-727. 13. Gartsman GM, Khan M, Hammerman SM. Arthroscopic repair of full-thickness tears of the rotator cuff. J Bone Joint Surg Am. 1998; 80:832-840. 14. Milano G, Grasso A, Salvatore M, Zarelli D, Deriu L, Fabbriciani C. Arthroscopic rotator cuff repair with and without subacromial decompression: a prospective randomized study. Arthroscopy. 2007; 23:81-88. 15. Shin SJ, Oh JH, Chung SW, Song MH. The efficacy of acromioplasty in the arthroscopic repair of small- to medium-sized rotator cuff tears without acromial spur: prospective comparative study. Arthroscopy. 2012; 28:628-635. 16. Lubowitz JH, McIntyre LF, Provencher MT, Poehling GG. AAOS rotator cuff clinical
practice guideline misses the mark. Arthroscopy. 2012; 28:589-592. 17. McCarron JA, Milks RA, Mesiha M, et al. Reinforced fascia patch limits cyclic gapping of rotator cuff repairs in a human cadaveric model. J Shoulder Elbow Surg. 2012; 21:1680-1686. 18. Badhe SP, Lawrence TM, Smith FD, Lunn PG. An assessment of porcine dermal xenograft as an augmentation graft in the treatment of extensive rotator cuff tears. J Shoulder Elbow Surg. 2008; 17:35S-39S. 19. Bond JL, Dopirak RM, Higgins J, Burns J, Snyder SJ. Arthroscopic replacement of massive, irreparable rotator cuff tears using a GraftJacket allograft: technique and preliminary results. Arthroscopy. 2008; 24:403-409. 20. Walton JR, Bowman NK, Khatib Y, Lin klater J, Murrell GA. Restore orthobiologic implant: not recommended for augmentation of rotator cuff repairs. J Bone Joint Surg Am. 2007; 89:786-791. 21. Yamaguchi K, Tetro AM, Blam O, Evanoff BA, Teefey SA, Middleton WD. Natural history of asymptomatic rotator cuff tears: a longitudinal analysis of asymptomatic tears detected sonographically. J Shoulder Elbow Surg. 2001; 10:199-203. 22. Mall NA, Kim HM, Keener JD, et al. Symptomatic progression of asymptomatic rotator cuff tears: a prospective study of clinical and sonographic variables. J Bone Joint Surg Am. 2010; 92:2623-2633.
e177
Agreement Among ASES Members on the AAOS Clinical Practice Guidelines E. Scott Paxton, MD; Jonas L. Matzon, MD; Alexa C. Narzikul, BA; Pedro K. Beredjiklian, MD; Joseph A. Abboud, MD
abstract The American Academy of Orthopaedic Surgeons (AAOS) has recently developed several clinical practice guidelines (CPG) involving upper extremity conditions. The purpose of the current study was to evaluate the practice patterns of members of the American Shoulder and Elbow Society (ASES) with regard to the CPGs. An e-mail survey was sent to the 340 members of the ASES. The survey contained 40 questions involving the subject matter of the 2 existing AAOS CPGs pertaining specifically to the shoulder: Optimizing the Management of Rotator Cuff Problems and the Treatment of Glenohumeral Joint Arthritis. Overall, 98 responses were obtained, for a response rate of 29%. Only 19 of 47 CPGs were not “inconclusive” and a recommendation was actually made. A majority (more than 50%) of surgeons agreed with 17 (90%) of 19 of these AAOS recommendations. A strong majority (more than 80%) adhered to 13 (68%) of 19 recommendations. There were 4 consensus recommendations, and more than 50% agreed with all of them. Of the 5 moderate recommendations, more than 50% agreed with 4 of them. There were 10 weak recommendations, and more than 50% of surgeons agreed with 9 of them. There was more than 80% agreement on 18 of 28 inconclusive recommendations. Although the AAOS CPGs are not meant to be fixed protocols, they are intended to unify treatment and/or diagnosis of common problems based on the best evidence available. Despite the majority of the AAOS CPG recommendations for rotator cuff problems and glenohumeral arthritis being inconclusive, most surgeons agree with most of the CPG recommendations. [Orthopedics. 2015; 38(3):e169-e177.]
The authors are from the Warren Alpert Medical School of Brown University (ESP), Providence, Rhode Island; and the Rothman Institute (JLM, ACN, PKB, JAA), Philadelphia, Pennsylvania. Dr Matzon, Ms Narzikul, and Dr Beredjiklian have no relevant financial relationships to disclose. Dr Paxton is a paid consultant for Tornier. Dr Abboud is a paid consultant for DePuy, Integra, and Tornier and receives royalties from Integra. Correspondence should be addressed to: Joseph A. Abboud, MD, Rothman Institute, 925 Chestnut St, Philadelphia, PA 19107 ([email protected]). Received: January 30, 2014; Accepted: May 13, 2014. doi: 10.3928/01477447-20150305-53
MARCH 2015 | Volume 38 • Number 3
e169
n Feature Article
T
he current standards of evidencebased medicine (EBM) require physicians to use the best available evidence to guide their clinical decision making. To simplify this process for their members, the American Academy of Orthopaedic Surgeons (AAOS) has recently developed clinical practice guidelines (CPGs) involving many orthopedic conditions. These CPGs were put forth by AAOS physician volunteer work groups based on systematic reviews of the current scientific and clinical data. Per the AAOS CPG publication, the implications for practice of each recommendation are listed in Table 1.1,2 The rationale was to improve patient care, educate qualified physicians, and develop a standard of evidence for common clinical conditions. Since 2009, two CPGs have been introduced specifically for shoulder conditions: Optimizing the Management of Rotator Cuff Problems and the Treatment of Glenohumeral Joint Arthritis. Since these guidelines have been introduced, the rate of adherence to them by practicing physicians is not known. The purpose of the current study was to evaluate the adherence of members of the American Shoulder and Elbow Society (ASES) to these CPGs involving shoulder conditions. The authors’ hypothesis was that more than 50% of respondents would agree with the majority of the noninconclusive recommendations.
Table 1
American Academy of Orthopaedic Surgeons Clinical Practice Guidelines Strengths of Recommendation Statement Rating
Description of Evidence Strength
Implication for Practice
Strong
Evidence is based on 2 or more “High” strength studies with consistent findings for recommending for or against the intervention. A Strong recommendation means that the benefits of the recommended approach clearly exceed the potential harm (or that the potential harm clearly exceeds the benefits in the case of a strong negative recommendation), and that the strength of the supporting evidence is high.
Practitioners should follow a Strong recommendation unless a clear and compelling rationale for an alternative approach is present.
Moderate
Evidence from 2 or more “Moderate” strength studies with consistent findings, or evidence from a single “High” quality study for recommending for or against the intervention. A Moderate recommendation means that the benefits exceed the potential harm (or that the potential harm clearly exceeds the benefits in the case of a negative recommendation), but the strength of the supporting evidence is not as strong.
Practitioners should generally follow a Moderate recommendation but remain alert to new information and be sensitive to patient preferences.
Limited
Evidence from 2 or more “Low” strength studies with consistent findings, or evidence from a single “Moderate” quality study recommending for or against the intervention or diagnostic. A Limited recommendation means the quality of the supporting evidence that exists is unconvincing or that well-conducted studies show little clear advantage to one approach vs another.
Practitioners should be cautious in deciding whether to follow a recommendation classified as Limited and should exercise judgment and be alert to emerging publications that report evidence. Patient preference should have a substantial influencing role.
Inconclusive
Evidence from a single “Low” quality study or conflicting finding that does not allow a recommendation for or against the intervention.
An Inconclusive recommendation means that there is a lack of compelling evidence resulting in an unclear balance between benefits and potential harm.
Consensus
The supporting evidence is lacking and requires the work group to make a recommendation based on expert opinion by considering the known potential harm and benefits associated with the treatment. A Consensus recommendation means that expert opinion supports the guideline recommendation although there is no available empirical evidence that meets the inclusion criteria.
Practitioners should be flexible in deciding whether to follow a recommendation classified as Consensus, although they may set boundaries on alternatives. Patient preference should have a substantial influencing role.
Materials and Methods An online survey (SurveyMonkey, Portland, Oregon) was created to assess the practice patterns related to the AAOS rotator cuff and glenohumeral arthritis CPGs. Each recommendation from the CPGs was translated into a question. The wording used was either identical to that in the AAOS report or as close as possible. A link to the survey was sent to the 340 members of the ASES with the e-mail address on file as of October 2012. The e-mail contained a brief study description and a link from which the participating surgeon could access and complete the survey.
e170
The survey contained 40 questions, with 4 questions related to the participant’s type of practice, practice envi-
ronment, number of rotator cuff repairs performed yearly, and number of total shoulder arthroplasties (TSAs) performed
Copyright © SLACK Incorporated
only increase in importance as the population ages. The large volume of clinical and research interest published in the orthopaedic literature in the last 10 years recognizes the relative importance of rotator cuff disease. These studies have attempted to address a multitude of important questions regarding rotator cuff treatment, including the following
issues: • The timing and role of non-surgical treatment such as steroid injections, physical therapy, or modalities • The indications for chronic rotator cuff repair • The surgical indications for acute traumatic tears • The effect of multiple confounding factors such as age, diabetes,
or smoking on surgical prognosis • The most effective or appropriate surgical strategy, including débridement versus rotator cuff repair • The most effective of the many available postoperative rehabilitation protocols As with many orthopaedic problems, a multitude of differ-
ent treatment options exist for the patient with a painful rotator cuff. Many of these treatment options have been controversial. Not surprisingly, available research finds little consensus among orthopaedic surgeons on rotator cuff treatment options. For example, 2005 study that examined the surgical indication patterns and the role of physical therapy in New York state
n Feature Article
Table 1: Optimizing the management of rotator cuff problems clinical practice guideline recommendations
yearly. The remaining questions related to the subject matter of the 2 existing shoulder AAOS CPGs (Figure 1). Each question represented a recommendation regardless of the strength of the recommendation (inconclusive, limited, moderate, strong, or consensus). Responses to the study were recorded using a webbased database (SurveyMonkey) without identifiers. Overall, 340 survey invitation e-mails were sent, of which none were returned owing to delivery errors. Therefore, the authors considered all e-mail addresses to be valid and that these recipients had received the study. Descriptive statistics were used to evaluate the responses. The moderate, weak, and consensus recommendations were evaluated for more than 50% of respondents agreeing/adhering to the recommendation. These recommendations were also assessed for a strong majority agreement of more than 80%. The inconclusive recommendations were only assessed for a strong majority agreement of more than 80%. This was due to the fact that a recommendation had not actually been made and many of the questions had only 2 answer choices (ie, agree or disagree). Therefore, there would always be a more than 50% answer choice due to the binary nature of these questions. A multivariate analysis was performed using the 4 demographic questions asked. The Kruskal-Wallis test was used to compare ordinal variables with nonordinal variables, and Spearman’s rho was used to check for association.
Recommendation
Strength of recommendation
1. In the absence of reliable evidence, it is the opinion of the work group that surgery not be performed for asymptomatic, full thickness rotator cuff tears.
Consensus
2. Rotator cuff repair is an option for patients with chronic, symptomatic full thickness tears.
Weak
3. a. We cannot recommend for or against exercise programs (supervised or unsupervised) for patients with rotator cuff tears.
Inconclusive
b. We cannot recommend for or against subacromial injections for patients with rotator cuff tears.
Inconclusive
c. We cannot recommend for or against the use of nonsteroidal anti-inflammatory drugs (NSAIDs), activity modification, ice, heat, iontophoresis, massage, Transcutaneous Electrical Nerve Stimulation (TENS), Pulsed Electromagnetic Field (PEMF), or phonophoresis (ultrasound) for nonoperative management of rotator cuff tears.
Inconclusive
4. a. We suggest that patients who have rotator cuff-related symptoms in the absence of a full thickness tear be initially treated non-surgically using exercise and/or NSAIDs.
Moderate
b. We cannot recommend for or against subacromial corticosteroid injection or PEMF in the treatment of rotator cuff-related symptoms in the absence of a full thickness tear.
Inconclusive
c. We cannot recommend for or against the use of iontophoresis, phonophoresis, TENS, ice, heat, massage, or activity modification for patients who have rotator cuff related symptoms in the absence of a full thickness tear.
Inconclusive
5. Early surgical repair after acute injury is an option for patients with a rotator cuff tear.
Weak
6. We cannot recommend for or against the use of perioperative subacromial corticosteroid injections or non-steroidal anti-inflammatory medications in patients undergoing rotator cuff surgery.
Inconclusive
7. a. It is an option for physicians to advise patients that the following factors correlate with less favorable outcomes after rotator cuff surgery: Increasing age
Weak
MRI Tear Characteristics
Weak
Worker’s Compensation Status
Moderate
b. We cannot recommend for or against advising patients in regard to the following factors related to rotator cuff surgery: Diabetes
Inconclusive
Comorbidities (multiple)
Inconclusive
Smoking
Inconclusive
Prior shoulder infection
Inconclusive
Cervical disease (neck pain and myelopathy)
Inconclusive
8. We suggest that routine acromioplasty is not required at the time of rotator cuff repair.
Moderate
9. It is an option to perform partial rotator cuff repair, debridement, or muscle transfers for patients with irreparable rotator cuff tears when surgery is indicated.
Weak
10. a. It is an option for surgeons to attempt to achieve tendon to bone healing of the cuff in all patients undergoing rotator cuff repair.
Weak
b. We cannot recommend for or against the preferential use of suture anchors versus bone tunnels for repair of full thickness rotator cuff tears.
Inconclusive
c. We cannot recommend for or against a specific technique (arthroscopic, mini-open or open repair) when surgery is indicated for full thickness rotator cuff tears.
Inconclusive
11. a. We suggest surgeons not use a non-crosslinked, porcine small intestine submucosal xenograft patch to treat patients with rotator cuff tears. b. We cannot recommend for or against the use of soft tissue allografts or other xenografts to treat patients with rotator cuff tears.
Moderate Inconclusive
12. In the absence of reliable evidence, it is the opinion of the work group that local cold therapy is beneficial to relieve pain after rotator cuff surgery.
Consensus
13. a. We cannot recommend for or against the preferential use of an abduction pillow versus a standard sling after rotator cuff repair.
Inconclusive
b. We cannot recommend for or against a specific time frame of shoulder immobilization without range of motion exercises after rotator cuff repair.
Inconclusive
c. We cannot recommend for or against a specific time interval prior to initiation of active resistance exercises after rotator cuff repair.
Inconclusive
d. We cannot recommend for or against home-based exercise programs versus facility-based rehabilitation after rotator cuff surgery.
Inconclusive
14. We cannot recommend for or against the use of an indwelling subacromial infusion catheter for pain management after rotator cuff repair.
Inconclusive
Note: This summary does not contain rationales that explain how and why these recommendations were developed nor does it contain the evidence supporting these recommendations. All readers of this summary are strongly urged to consult the full guideline and evidence report for this information.We are confident that those who read the full guideline and evidence report will see that the recommendations were developed using systematic evidence-based processes designed to combat bias, enhance transparency, and promote reproducibility. This summary of recommendations is not intended to stand alone.
Figure 1: Optimizing the Management of Rotator Cuff Problems clinical practice guideline recommendations. (Reprinted with permission. © American Academy of Orthopaedic Surgeons.)
AAOS Now_January 2011.indd 46
Results At the time of final analysis, 98 surveys had been completed, for a response rate of 29%. Of the completing surgeons, most (64%) identified themselves as shoulder and elbow surgeons (Table 2). The participants were relatively evenly split in terms of their current practice environment: 22% “private practice,” 41% “academic practice,” and 37% “private
MARCH 2015 | Volume 38 • Number 3
and academic practice.” A strong majority (83%) of surgeons performed more than 40 rotator cuff repairs per year, with 49% performing more than 100 year. The majority (67%) of surgeons performed more than 20 TSAs per year, with 47% performing more than 50 per year and 24% performing more than 100 per year. Of note, 14% of respondents reported that they do not perform TSA. A majority (more than 50%) of surgeons agreed with 17 (90%) of 19 AAOS recommendations that were not inconclusive (28 were inconclusive). A strong majority (more than 80%) adhered to only 13 (68%) recommendations. There were no strong recommendations. Of the 28 inconclusive recommendations, a strong majority (more than 80%) agreed with 18 (64%). Figure 2 provides a complete list of questions and answer distributions.
12/22/2010 2:57:25 PM
Table 2
Respondent Specialties Specialty
%
Shoulder and elbow surgery
64
Orthopedic sports medicine
29
Upper extremity (including hand)
4
General practice
3
Recommendations With a Moderate Strength of Recommendation There were 5 guidelines with a moderate strength of recommendation, and more than 50% of respondents adhered to 4 of them. More than half (53.1%) of respondents reported that they do not discuss workers’ compensation status as correlating with a less favorable outcome.
e171
earCh and teChnology at n Feature Article
ngth indistudies on ous treatof none treatment unclear, rgical techh support
owing areas quality
herapy in as well as litation of
therapy ent of
corticol treatment
viscosupurgical meral oa py in the meral
idement, plasty/interreatment of
us thromulder need for
as a or after
r subshoulder
areas ance our ent opoa and ecommene future. roup hope dentify and etter qualclinical chers to in current treatment oa. NOW
out
s
rved as roup that ractice ment of eoarthritis. g disclore, Wyeth.
hester.edu
Work
January 2010
now
aaos
41
Table 1: The Treatment of Glenohumeral Joint Osteoarthritis Clinical Practice Guideline Recommendations Recommendation
Strength of recommendation
We are unable to recommend for or against physical therapy for the initial treatment of patients with osteoarthritis (OA) of the glenohumeral (GH) joint.
Inconclusive
We are unable to recommend for or against the use of pharmacotherapy in the initial treatment of patients with GH joint OA.
Inconclusive
We are unable to recommend for or against the use of injectable corticosteroids when treating patients with GH joint OA.
Inconclusive
The use of injectable viscosupplementation is an option when treating patients with GH joint OA.
Weak
We are unable to recommend for or against the use of arthroscopic treatments for patients with GH joint OA. These treatments include débridement, capsular release, chondroplasty, microfracture, removal of loose bodies, biologic and interpositional grafts, subacromial decompression, distal clavicle resection, acromioclavicular joint resection, biceps tenotomy or tenodesis, and labral repair or advancement.
Inconclusive
We are unable to recommend for or against open débridement and/or nonprosthetic or biologic interposition arthroplasty in patients with GH joint OA. These treatments include allograft, biologic and interpositional grafts, and autograft.
Inconclusive
Total shoulder arthroplasty and hemiarthroplasty are options when treating patients with GH joint OA.
Weak
We suggest total shoulder arthroplasty (TSA) over hemiarthroplasty when treating patients with GH joint OA.
Moderate
An option for reducing immediate postoperative complication rates is for patients to avoid TSA by surgeons who perform less than two TSAs per year.
Weak
In the absence of reliable evidence, it is the opinion of this work group that physicians use peri-operative mechanical and/or chemical VTE (venous thromboembolism) prophylaxis for TSA patients.
Consensus
The use of either keeled or pegged all-polyethylene cemented glenoid components are options when performing TSA.
Weak
In the absence of reliable evidence, it is the opinion of this work group that TSA not be performed in patients with GH joint OA who have an irreparable rotator cuff tear.
Consensus
We are unable to recommend for or against biceps tenotomy or tenodesis when performing TSA in patients who have GH joint OA.
Inconclusive
We are unable to recommend for or against a subscapularis transtendonous approach or a lesser tuberosity osteotomy when performing TSA in patients who have GH joint OA.
Inconclusive
We are unable to recommend for or against a specific type of humeral prosthetic design or method of fixation when performing TSA in patients with GH joint OA.
Inconclusive
We are unable to recommend for or against physical therapy following TSA.
Inconclusive
Figure 2: The Treatment of Glenohumeral Joint Osteoarthritis clinical practice guideline recommendaTable 2: Understanding AAOS language tions. (Reprinted with permission. © guideline American Academy of Orthopaedic Surgeons.) Strength
Overall Quality of Evidence
Strong
Good
Description of Evidence Level I evidence from more than one study with consistent for recommending for or against the intervention or Tablefindings 3 diagnostic.
Moderate
Fair
Level II or III evidence from more than one study with Moderate Recommendations
Clinical Practice Guideline Optimizing Weak
consistent findings, or Level I evidence from a single study for recommending for or against the intervention or Agreement diagnostic.
the Management of Rotator Cuff Problems Poor Level IV or V evidence from more than one study with
consistent findings, or Level II or III evidence from a single
The authors suggest that patients who have rotator study forcuff–related recommending for or against the 87.8% intervention or symptoms in the absence of a full-thickness diagnostic. tear be initially treated (NSAIDs); 98% Inconclusive No evidence or conflicting evidence The evidence is insufficient or conflicting(exercise) and does not nonoperatively using exercise and/or NSAIDs.
allow a recommendation for or against the intervention or
It is an option for physicians to advise patients that the following fac46.9% diagnostic. tors correlate less favorable outcomes after cuff surgery: Consensus Nowith evidence There rotator is no supporting evidence. In the absence of reliable evidence, the work group is making a recommendation workers’ compensation. based on their clinical opinion and considering the known
harms and benefits associated The authors suggest that routine acromioplasty is not required at the with the treatment. 63.9% time of rotator cuff repair.
The authors suggest that surgeons not use a non–cross-linked, porcine small intestine submucosal xenograft patch to treat patients with rotator cuff tears.
98% (primary repair); 96.9% (revision repair)
Treatment of Glenohumeral Arthritis The authors suggest total shoulder arthroplasty over hemiarthroplasty when treating patients with glenohumeral joint arthritis. Abbreviation: NSAIDs, nonsteroidal anti-inflammatory drugs.
e172
94.8%
Three of the 5 recommendations were adhered to by more than 80% of respondents, with 36.1% performing routine acromioplasty with all rotator cuff repairs (Table 3). Recommendations With a Weak Strength of Recommendation There were 10 guidelines with a weak strength of recommendation, and more than 50% of respondents adhered to 9 of them, with 72.4% of surgeons not using viscosupplementation to treat glenohumeral arthritis. Seven of the 10 recommendations were adhered to by more than 80% of respondents, with 20.4% reporting that they do not discuss increased age as relating to a less favorable outcome after rotator cuff repair and 20.4% not believing that immediate postoperative complications are reduced when the surgeon performs more than 2 TSAs per year (Table 4). Recommendations With a Consensus Strength of Recommendation There were 4 consensus recommendations, and a majority (more than 50%) of respondents adhered to all 4. However, 3 of these 4 recommendations did not achieve a strong majority (more than 80%) of respondents (Table 5). Routine cold therapy after rotator cuff surgery is offered by 69.4% of respondents. Mechanical or chemical venous thromboembolism prophylaxis after TSA is used by 77.8% of respondents. Total shoulder arthroplasty is not considered a treatment option for patients with glenohumeral arthritis and an irreparable rotator cuff tear by 79.3% of respondents. Recommendations With an Inconclusive Strength of Recommendation There were 28 inconclusive recommendations, and there was more than 80% agreement on 18 of them (Table 6). Confounding Variables No correlation was found between the 4 demographic questions. The only cor-
Copyright © SLACK Incorporated
n Feature Article
relation between demographics and the CPGs was found in upper extremity surgeons (with practices including and excluding hand) preferring arthroscopic rotator cuff repair as opposed to open rotator cuff repair more than general orthopedists and sports medicine surgeons (P6 wk)
The authors cannot recommend for or against a specific time interval prior to initiation of active resistance exercises after rotator cuff repair.
7.1% (depends on repair); 0% (0-2 wk); 1.0% (at or after 2-4 wk); 7.1% (at or after 4-6 wk); 22.4% (at or after 6-8 wk); 12.2% (at or after 8-10 wk); 21.4% (at or after 10-12 wk); 25.5% (at or after 12-16 wk); 3.1% (at or after 16 wk)
The authors cannot recommend for or against home-based exercise programs vs facility-based rehabilitation after rotator cuff surgery.
8.2% (100% home-based); 15.5% (1%-24% home-based); 18.6% (25%49% home-based); 17.5% (50%-74% home-based); 20.6% (75%-90% home-based); 19.6% (91%-100% home-based)
The authors cannot recommend for or against the use of an indwelling subacromial infusion catheter for pain management after rotator cuff repair.
9.2%
Treatment of Glenohumeral Arthritis The authors are unable to recommend for or against physical therapy for the initial treatment of patients with GJA. The authors are unable to recommend for or against the use of pharmacotherapy in the initial treatment of patients with GJA.
e174
61% 93.9%
Copyright © SLACK Incorporated
n Feature Article
Table 6 (cont’d)
Inconclusive Recommendations Clinical Practice Guideline
Agreement
The authors are unable to recommend for or against the use of injectable corticosteroids when treating patients with GJA.
82.7%
The authors are unable to recommend for or against the use of arthroscopic treatments for patients with GJA. These treatments include debridement, capsular release, chondroplasty, microfracture, removal of loose bodies, biologic and interpositional grafts, subacromial decompression, distal clavicle resection, acromioclavicular joint resection, biceps tenotomy or tenodesis, and labral repair or advancement.
70.1% (treat arthroscopically); 67.0% (use debridement); 53.6% (use capsular release); 40.2% (use chondroplasty); 25.8% (use microfracture); 63.9% (use loose body removal); 5.2% (use interpositions); 21.6% (use subacromial decompression); 14.4 (use distal clavical resection); 9.3% (use acromioclavicular joint resection); 43.3% (use biceps tenotomy/tenodesis); 5.2% (use labral repair/advancement)
The authors are unable to recommend for or against open debridement and/or nonprosthetic or biologic interposition arthroplasty in patients with GJA. These treatments include allograft, biologic and interpositional grafts, and autograft.
5.3% (open debridement); 7.4% (interpositions)
The authors are unable to recommend for or against biceps tenotomy or tenodesis when performing TSA in patients who have GJA.
12.2% (preserve); 54.4% (tenodesis); 7.8% (tenotomy); 25.6% (patient specific)
The authors are unable to recommend for or against a subscapularis transtendinous approach or a lesser tuberosity osteotomy when performing TSA in patients who have GJA.
76.4% (transtendon); 23.6% (lesser tuberosity osteotomy)
The authors are unable to recommend for or against a specific type of humeral prosthetic design or method of fixation when performing TSA in patients with GJA.
The authors are unable to recommend for or against physical therapy following TSA.
1.1% (monoblock); 34.1% (modular with variable neck-shaft angle); 70.5% (modular with variable head sizes); 89.8% (modular with eccentric heads); 3.4% (other) 85.7 % (use physical therapy)
Abbreviations: GJA, glenohumeral joint arthritis; NSAIDs, nonsteroidal anti-inflammatory drugs; PEMF, pulsed electromagnetic field; TENS, transcutaneous electrical nerve stimulation; TSA, total shoulder arthroplasty.
higher-level studies. This suggests that a subset of shoulder specialists rely on different information for clinical decision making and not solely on the studies the CPG work groups used or the AAOS CPGs themselves. Despite this, the current authors found that most shoulder surgeons practice in largely comparable manners. Of the 5 moderate recommendations, a majority of surgeons agreed with 4. Multiple preoperative factors were assessed by the work group as relating to outcome after rotator cuff repair. Although it was a moderate recommendation, only 46.9% of surgeons discuss workers’ compensation status as correlating to a worse outcome after rotator cuff repair, and this was the only moderate recommendation
MARCH 2015 | Volume 38 • Number 3
that was not adhered to by more than 50% of respondents. Workers’ compensation status has been documented for multiple orthopedic procedures as having a negative effect on outcome.3-9 A recent meta-analysis of 20 prospective studies evaluating the effect of workers’ compensation status on outcomes of orthopedic procedures found that the overall risk ratio for experiencing an unsatisfactory result for patients with vs without compensation is 2.08 (95% confidence interval, 1.54-2.82).10 Cuff and Pupello11 prospectively evaluated patients after rotator cuff repair stratified to compensated and noncompensated patients. Only 4% of patients without a compensation claim were found to be noncompliant, whereas 52% of compensated patients
were noncompliant. Overall, the compensated patients had worse outcomes than the noncompensated patients. Not surprisingly, the noncompliant compensated patients had worse scores than the compliant compensated patients. Healing rates also differed in the groups but did not reach statistical significance, with 84% healed in the noncompensated group vs 75% in the compliant compensated group and 59% in the noncompliant compensated group.11 These findings provide some foundation for preoperative counseling of workers’ compensation patients and postoperative surveillance of these patients for signs of noncompliance. Physicians may feel uncomfortable talking with workers’ compensation patients preoperatively
e175
n Feature Article
about their worse prognosis; this may explain why many orthopedic surgeons do not. The current authors did not ask the respondents about their reasoning for their practices, so they can only speculate that this may be the case. Alternatively, surgeons may not believe that workers’ compensation patients have poorer outcomes, although the data are convincing. Unlike workers’ compensation status, most respondents discuss magnetic resonance imaging characteristics (92.9%) and patient age (79.6%) in relation to outcomes after rotator cuff repair, although both are weak recommendations. In addition, 93.9% discuss smoking, 78.6% discuss history of shoulder infection, and 73.5% discuss diabetes mellitus as portending a worse outcome after rotator cuff repair, despite the fact that the CPGs for these 3 factors were inconclusive. A small percentage (31.6%) discuss cervical spine disease as a predictor of outcome. Although given the way the recommendation is worded, it is difficult to know whether this includes simply a history of cervical spine problems that have been adequately treated, or whether it more specifically means an acute process with neurologic findings. It would be expected that many surgeons would be more cognizant of a patient with an acute radiculopathy and may wait for resolution prior to surgery but may be less aware of a history of a single-level fusion or some recalcitrant pain. The work group made a moderate recommendation against routine acromioplasty performed with rotator cuff repair. Despite this, 36.1% of respondents perform an acromioplasty with all rotator cuff repairs. This was not surprising because these procedures have historically been performed together; however, only approximately one-third of surgeons perform them routinely, which seems to provide evidence that most surgeons are leaning away from routine acromioplasty. Continued belief in acromial impingement as the primary cause of rotator cuff tears
e176
may explain why more than one-third of the ASES surgeons surveyed still perform routine acromioplasty with a rotator cuff repair, despite multiple studies reporting no significant benefit.12-15 This is evidenced by a letter to the editor regarding this CPG by Lubowitz et al16 discussing the role of acromioplasty with rotator cuff repair and postulating that it may reduce the need for reoperation. Although many of the work group’s recommendations were inconclusive, they relate to important topics in shoulder surgery. Surprisingly, almost half (45.4%) of surgeons reported that they use soft tissue allografts or xenografts in revision rotator cuff repairs. Biomechanical studies evaluating patch augmentation is promising, with decreased gap formation by up to 48%.17 Also, small studies have shown improvement over the preoperative state with the use of an augmented patch during repair.18,19 However, inflammatory reactions and increased pain have been reported with the use of certain products.20 Surgeons are interpreting these results in different ways, with approximately half believing that the current products have benefit in the revision setting and half not. Only 11.2% of surgeons reported that they use a patch in a primary setting. Most (80.4%) surgeons surveyed agreed with the AAOS work group consensus that rotator cuff repair should not be performed for asymptomatic tears; however, 19.6% reported that they perform repairs for symptomatic tears. Natural history studies have shown that approximately half of asymptomatic tears will progress to symptomatic tears within 3 years, and half of these will increase in size.21 In addition, 4% to 20% of asymptomatic patients will have silent progression of their tear.21,22 Therefore, an argument could be made for surgical fixation in these patients because nonoperative treatment may result in tear progression, fatty atrophy, and tendon retraction, resulting in the inability to repair the tendon if it were to become symptomatic or increasing the difficulty
of tendon mobilization and repair. Successful repair may ultimately result in less improvement in function if fatty atrophy progresses significantly prior to repair. In addition, the age of the patient may come into consideration, with those answering in disagreement referring to a younger population where the risk of an irreparable rotator cuff tear is more worrisome. There are several possible sources of bias in the current authors’ results. The overall response rate was 29%. Therefore, a large portion of the ASES did not respond, and the results of the study are susceptible to respondent bias. It is possible that the nonresponders are members who adhere to the CPGs differently. In addition, by using a web-based survey, the authors may have introduced a selection bias in which more computer-savvy ASES members may dominate the survey. Also, the recommendations for glenohumeral arthritis and rotator cuff tears were made using studies published prior to January 2009 and October 2008, respectively. This survey was sent in January 2013, leaving a gap of more than 4 years. It is possible that newer data have been published in this time period that have altered surgeons’ practices. Finally, this study assumes that how the ASES members respond and how they practice are identical, and that may not be the case.
Conclusion The results of this study show that despite the lack of high-level evidence, most shoulder surgeons practice in comparable ways and in agreement with the AAOS CPGs on rotator cuff disease and glenohumeral arthritis.
References 1. Izquierdo R, Voloshin I, Edwards S, et al. American Academy of Orthopaedic Surgeons Clinical Practice Guideline on the treatment of glenohumeral joint osteoarthritis. J Bone Joint Surg Am. 2011; 93:203-205. 2. Pedowitz RA, Yamaguchi K, Ahmad CS, et al. American Academy of Orthopaedic Surgeons Clinical Practice Guideline on optimizing the management of rotator cuff problems.
Copyright © SLACK Incorporated
n Feature Article
J Bone Joint Surg Am. 2012; 94:163-167. 3. Anderson PA, Subach BR, Riew KD. Predictors of outcome after anterior cervical discectomy and fusion: a multivariate analysis. Spine (Phila Pa 1976). 2009; 34:161-166. 4. Bhatia S, Piasecki DP, Nho SJ, et al. Early return to work in workers’ compensation patients after arthroscopic full-thickness rotator cuff repair. Arthroscopy. 2010; 26:10271034. 5. Brinker MR, Savory CG, Weeden SH, Aucoin HC, Curd DT. The results of total knee arthroplasty in workers’ compensation patients. Bull Hosp Jt Dis. 1998; 57:80-83. 6. Denard PJ, Ladermann A, Burkhart SS. Long-term outcome after arthroscopic repair of type II SLAP lesions: results according to age and workers’ compensation status. Arthroscopy. 2012; 28:451-457. 7. Holtby R, Razmjou H. Impact of work-related compensation claims on surgical outcome of patients with rotator cuff related pathologies: a matched case-control study. J Shoulder Elbow Surg. 2010; 19:452-460. 8. Hou WH, Tsauo JY, Lin CH, Liang HW, Du CL. Worker’s compensation and return-towork following orthopaedic injury to extremities. J Rehabil Med. 2008; 40:440-445. 9. Wexler G, Bach BR Jr, Bush-Joseph CA, Smink D, Ferrari JD, Bojchuk J. Outcomes of anterior cruciate ligament reconstruction in patients with workers’ compensation claims. Arthroscopy. 2000; 16:49-58.
MARCH 2015 | Volume 38 • Number 3
10. de Moraes VY, Godin K, Tamaoki MJ, Faloppa F, Bhandari M, Belloti JC. Workers’ compensation status: does it affect orthopaedic surgery outcomes? A meta-analysis. PLoS One. 2012; 7:e50251. 11. Cuff DJ, Pupello DR. Prospective evaluation of postoperative compliance and outcomes after rotator cuff repair in patients with and without workers’ compensation claims. J Shoulder Elbow Surg. 2012; 21:1728-1733. 12. Chahal J, Mall N, MacDonald PB, et al. The role of subacromial decompression in patients undergoing arthroscopic repair of fullthickness tears of the rotator cuff: a systematic review and meta-analysis. Arthroscopy. 2012; 28:720-727. 13. Gartsman GM, Khan M, Hammerman SM. Arthroscopic repair of full-thickness tears of the rotator cuff. J Bone Joint Surg Am. 1998; 80:832-840. 14. Milano G, Grasso A, Salvatore M, Zarelli D, Deriu L, Fabbriciani C. Arthroscopic rotator cuff repair with and without subacromial decompression: a prospective randomized study. Arthroscopy. 2007; 23:81-88. 15. Shin SJ, Oh JH, Chung SW, Song MH. The efficacy of acromioplasty in the arthroscopic repair of small- to medium-sized rotator cuff tears without acromial spur: prospective comparative study. Arthroscopy. 2012; 28:628-635. 16. Lubowitz JH, McIntyre LF, Provencher MT, Poehling GG. AAOS rotator cuff clinical
practice guideline misses the mark. Arthroscopy. 2012; 28:589-592. 17. McCarron JA, Milks RA, Mesiha M, et al. Reinforced fascia patch limits cyclic gapping of rotator cuff repairs in a human cadaveric model. J Shoulder Elbow Surg. 2012; 21:1680-1686. 18. Badhe SP, Lawrence TM, Smith FD, Lunn PG. An assessment of porcine dermal xenograft as an augmentation graft in the treatment of extensive rotator cuff tears. J Shoulder Elbow Surg. 2008; 17:35S-39S. 19. Bond JL, Dopirak RM, Higgins J, Burns J, Snyder SJ. Arthroscopic replacement of massive, irreparable rotator cuff tears using a GraftJacket allograft: technique and preliminary results. Arthroscopy. 2008; 24:403-409. 20. Walton JR, Bowman NK, Khatib Y, Lin klater J, Murrell GA. Restore orthobiologic implant: not recommended for augmentation of rotator cuff repairs. J Bone Joint Surg Am. 2007; 89:786-791. 21. Yamaguchi K, Tetro AM, Blam O, Evanoff BA, Teefey SA, Middleton WD. Natural history of asymptomatic rotator cuff tears: a longitudinal analysis of asymptomatic tears detected sonographically. J Shoulder Elbow Surg. 2001; 10:199-203. 22. Mall NA, Kim HM, Keener JD, et al. Symptomatic progression of asymptomatic rotator cuff tears: a prospective study of clinical and sonographic variables. J Bone Joint Surg Am. 2010; 92:2623-2633.
e177
