These authors reported the 10-year outcomes of isolated supraspinatus repairs in 288 of 511 patients who, in 2003, underwent repair of full-thickness isolated supraspinatus tears at a mean age of 56.5 ± 8.3 years; half were men.
The repair was open (anterosuperior approach) in 113 shoulders (39%) and arthroscopic in 175 shoulders (61%). Acromioplasty was performed in 273 shoulders (95%), while biceps tenodesis or tenotomy was performed in 92 (32%) and 58 shoulders (20%), respectively. All tendons were repaired using a single row of suture anchors (149 anchor screws, including 116 that were metallic and 33 that were resorbable), interference fit anchors (34, including 16 that were metallic and 18 that were resorbable), or transosseous repairs (83 shoulders, with material unspecified in 22 shoulders). All repairs were “watertight” at the end of the intervention.
Following surgery, the arm was supported in a sling with the arm at the side (n = 138) or at 20 degrees of abduction (n = 150) for a mean of 5.6 ± 1.0 weeks (median, 6 weeks; range, 1 to 8 weeks). All centers followed the same rehabilitation protocol, with passive-motion exercises initiated on the first postoperative day, and when possible, hydrotherapy after skin healing. Active shoulder motion was allowed after a mean of 8.2 ± 6.8 weeks (median, 6 weeks; range, 3 to 50 weeks). Patients were not allowed to perform any strengthening or strenuous work for 6 months after the surgery. Light sports and demanding activities were allowed after 6 months.
188 of the 511 could not be reached, and 35 were excluded because they had a reoperation (17 had a retear, 7 had conversion to an arthroplasty, and 11 had other causes). 210 patients were also evaluated using magnetic resonance imaging (MRI).
Thirty shoulders (10.4%) had complications, including stiffness (20 shoulders), infection (1 shoulder), and other complications (9 shoulders).
Thirty shoulders (10.4%) had complications, including stiffness (20 shoulders), infection (1 shoulder), and other complications (9 shoulders).
The total Constant score improved from a mean of 51.8 ± 13.6 points (range, 19 to 87 points) preoperatively to 77.7 ± 12.1 points (range, 37 to 100 points) at 10 years.
At the 10-year follow-up evaluation, the mean Subjective Shoulder Value (SSV) was 84.9 ± 14.8 (range, 20 to 100), and the mean Simple Shoulder Test (SST) was 10.1 ± 2.2 (range, 3 to 12).
Of the 210 shoulders evaluated using MRI, the repair integrity was Sugaya type I (normal appearing) in 26 shoulders (12%), type II (normal thickness, high signal) in 85 (41%), type III (less than 50% normal thickness) in 59 (28%), type IV (small full thickness defect) in 27 (13%), and type V (medium or large defect) in 13 (6%).
The total Constant score at the final follow-up was significantly associated with tendon healing (p < 0.005) and was inversely associated with preoperative fatty infiltration (p < 0.001). Neither the surgical approach nor the preoperative retraction influenced the outcomes.
The rate of retears was significantly higher (32%) for patients whowere >65 years old than in those who were 55 to 65 years old (20%) or <55 years old (15%).
The total Constant score at the final follow-up was significantly associated with tendon healing (p < 0.005) and was inversely associated with preoperative fatty infiltration (p < 0.001). Neither the surgical approach nor the preoperative retraction influenced the outcomes.
The rate of retears was significantly higher (32%) for patients whowere >65 years old than in those who were 55 to 65 years old (20%) or <55 years old (15%).
Comment:
Recognizing that their paper did not have a control group, the authors contrasted their results with those of another recent study:
Treatment of Nontraumatic Rotator Cuff Tears: A Randomized Controlled Trial with Two Years of Clinical and Imaging Follow-up that reviewed 180 shoulders with symptomatic, nontraumatic, supraspinatus tears that were randomized into one of three cumulatively designed intervention groups: the physiotherapy-only group (denoted as Group 1), the acromioplasty and physiotherapy group (denoted as Group 2), and the rotator cuff repair, acromioplasty, and physiotherapy group (denoted as Group 3). That study found no significant differences (p = 0.38) in the mean change of Constant score: 18.4 points (95% confidence interval, 14.2 to 22.6 points) in Group 1, 20.5 points (95% confidence interval, 16.4 to 24.6 points) in Group 2, and 22.6 points (95% confidence interval, 18.4 to 26.8 points) in Group 3. There were no significant differences in visual analog scale for pain scores (p = 0.45) and patient satisfaction (p = 0.28) between the groups. At two years, the mean sagittal size of the tendon tear was significantly smaller (p < 0.01) in Group 3 (4.2 mm) compared with Groups 1 and 2 (11.0 mm).
Rotator cuff repair and acromioplasty were significantly more expensive than physiotherapy only (p < 0.01).There was no significant difference in clinical outcome between the three interventions at the two-year follow-up.
Conclusion: From these studies it seems clear that for many patients with supraspinatus tears, non-operative management can be a cost-effective management strategy, not only because the treatment costs less, but because it avoids the "patients were not allowed to perform any strengthening or strenuous work for 6 months after the surgery -light sports and demanding activities were allowed after 6 months" restriction that are part of a repair.
So, when considering the management of a rotator cuff tear that did not arise from a traumatic episode, there is surely time to try non-operative management to avoid the cost, risks, and downtime of surgery. It is of interest that an article attempting to model the societal and economic value of rotator cuff repair is often quoted without consideration of that simple fact.
In the absence of real data, the authors' goal was to estimate the value of surgical treatment for full-thickness rotator cuff tears from a societal perspective using a Markov decision model of the lifetime direct and indirect costs (e.g. inability to work, lower wages, missed workdays, disability payments) associated with surgical and continued nonoperative treatment for symptomatic full-thickness rotator cuff tears. Patients with a symptomatic full-thickness rotator cuff tear 'underwent' either open or arthroscopic rotator cuff repair or continued to receive nonoperative treatment. After one year, all rotator cuff repairs resulted in either (1) healed rotator cuff repair (symptomatic and asymptomatic), (2) asymptomatic retear, (3) symptomatic retear, or (4) death; shoulders treated without surgery resulted in either (1) symptomatic tear, (2) asymptomatic tear, or (3) death.
The model indicated that surgical treatment results in an average improvement of 0.62 QALY. The model suggested that the age-weighted mean total societal savings from rotator cuff repair compared with nonoperative treatment was $13,771 over the lifetime of the patient. Savings ranged from + $77,662 for patients who are thirty to thirty-nine years old to - $11,997 for those who are seventy to seventy-nine years old.
The model concluded that "rotator cuff repair is cost-effective for all populations" and "The estimated lifetime societal savings of the approximately 250,000 rotator cuff repairs performed in the U.S. each year was $3.44 billion." This is a staggering figure.
Because actual data were not available for many of the key elements in the model, values were based on Level V assumptions. As the authors point out 'evidence to support some of the model assumptions is limited.'These assumptions had a strong influence on the results. Here are some considerations of the assumptions used:
(1) It was assumed that all symptomatic full-thickness rotator cuff tears assigned to the repair group were repairable, whereas this does not seem to be the common experience
(2) The model recognizes that healed repairs can be symptomatic or asymptomatic, but does not assume any disability if the healed repair is symptomatic, whereas we have all seen patients who are unable to return to work after a symptomatic "successful" cuff repair.
(3) The long term retear rate after cuff repair was assumed to be 2%, whereas a recent post found retear rates after one year approximating 10%.
(4) Patient outcomes for all repairs were assumed to be the same as those of seventy-three patients who underwent surgery at a large orthopaedic surgery group; presumably those cases represented careful selection of patients, of reparable cuff tears, and excellent surgical technique by a well trained shoulder surgeon, whereas many patients with cuff tears are not prime surgical candidates, have irreparable cuff tears and may be cared for by surgeons who are not specialized in shoulder (see the 4Ps: problem, patient, procedure, physician).
(5) It was assumed that workers lost an average of twenty-eight additional days as a result of rotator cuff repair compared with those undergoing non-operative treatments, whereas recent data suggests that repairs should be protected from loading for up to six months after surgery. In that many cuff tears occur in those with physical laboring jobs, the 28 days seems like a short interval for return to work. Furthermore, it is not clear that cuff tears treated with non-operative management need to miss work at all.
(6) The model assumed complication rates of stiffness (2.5%) and of infection (0.1%) following rotator cuff repair, whereas some studies have reported a complication rate of 10.6% of which the most common was persistent stiffness.
(7) The model used an expensive approach to non-operative management costing $1802, whereas there is substantial evidence that an inexpensive home program provides an effective method for non-operative management.
(8) The model assumes that only 5% of asymptomatic retears following repair become symptomatic annually, whereas it assumes that 8.8% of asymptomatic tears managed non-operatively become symptomatic per year.
These and other assumptions led the model to conclude that "rotator cuff repair produces societal cost savings for patients under the age of sixty-one years and is cost-effective for all patients". However, it is easy to see that different assumptions might have led to a different answer.
(4) Patient outcomes for all repairs were assumed to be the same as those of seventy-three patients who underwent surgery at a large orthopaedic surgery group; presumably those cases represented careful selection of patients, of reparable cuff tears, and excellent surgical technique by a well trained shoulder surgeon, whereas many patients with cuff tears are not prime surgical candidates, have irreparable cuff tears and may be cared for by surgeons who are not specialized in shoulder (see the 4Ps: problem, patient, procedure, physician).
(5) It was assumed that workers lost an average of twenty-eight additional days as a result of rotator cuff repair compared with those undergoing non-operative treatments, whereas recent data suggests that repairs should be protected from loading for up to six months after surgery. In that many cuff tears occur in those with physical laboring jobs, the 28 days seems like a short interval for return to work. Furthermore, it is not clear that cuff tears treated with non-operative management need to miss work at all.
(6) The model assumed complication rates of stiffness (2.5%) and of infection (0.1%) following rotator cuff repair, whereas some studies have reported a complication rate of 10.6% of which the most common was persistent stiffness.
(7) The model used an expensive approach to non-operative management costing $1802, whereas there is substantial evidence that an inexpensive home program provides an effective method for non-operative management.
(8) The model assumes that only 5% of asymptomatic retears following repair become symptomatic annually, whereas it assumes that 8.8% of asymptomatic tears managed non-operatively become symptomatic per year.
It may be worthwhile to reflect on whether this model addresses the right question. No one would doubt that a successful rotator cuff repair that returns a worker to work is a good thing for that person and for society. On the other hand non-operative management can be of value to many patients. The real question is 'how do we decide which rotator cuff tears in which patients will benefit from rotator cuff repair and the critical post-operative rehabilitation period that must follow?'. The model's conclusion that 'rotator cuff repair is cost-saving across all patients' does not change the fact that many cuff tears are not reparable and that many patients with cuff tears are not good candidates for cuff repair.
In terms of guidance, the model leaves us with the quizzical statement "Although rotator cuff repair is cost-saving across all patients, nonoperative treatment is the preferred strategy for a large number of patients. "
In terms of guidance, the model leaves us with the quizzical statement "Although rotator cuff repair is cost-saving across all patients, nonoperative treatment is the preferred strategy for a large number of patients. "
===
The reader may also be interested in these posts:
Healing through joint replacement
Supporting progress in shoulder surgery
Consultation for those who live a distance away from Seattle.
Click here to see the new Shoulder Arthritis Book.
Click here to see the new Rotator Cuff Book
Information about shoulder exercises can be found at this link.
Use the "Search" box to the right to find other topics of interest to you.
You may be interested in some of our most visited web pages including:shoulder arthritis, total shoulder, ream and run, reverse total shoulder, CTA arthroplasty, and rotator cuff surgery as well as the 'ream and run essentials'See from which cities our patients come.
