Although degenerative rotator cuff tears are often initially asymptomatic, more than 50% will eventually become symptomatic.1,2 When conservative management fails, rotator cuff repair (RCR) can significantly improve shoulder function.3 Historically, the decision to proceed with surgical repair has largely been determined by physicians, based on factors such as range of motion, shoulder strength, and size of tear.4 Currently, treatment decision-making has shifted from a physician-dominated model to a shared decision-making model.5
Although determinants of physician decision-making for the management of rotator cuff tears have been identified, patient decision-making factors have not been thoroughly investigated. In a previous study, patients who switched from conservative management for rotator cuff tears to RCR reported long durations of intense pain and failed conservative treatments as factors that motivated their decision.6 An increased understanding of patient decision-making factors for RCR will allow physicians to better understand patients' treatment goals and postoperative expectations. This understanding may improve patient–physician collaboration and patient counseling, ultimately leading to increased patient satisfaction and better functional outcomes.7
The primary aim of this study was to investigate the factors that influenced patients' decision to undergo RCR. It also aimed to investigate factors that influenced physicians' decision to recommend RCR, as well as the relationship between these factors and functional shoulder outcomes. Finally, this study sought to identify the frequency of preoperative conservative treatments and determine if a relationship exists between these factors and shoulder outcomes.
Materials and Methods
This was a prospective cohort study. Institutional review board approval was obtained prior to patient enrollment. Enrollment occurred between February 2017 and October 2017. Three board-certified orthopedic surgeons (M.D.P., B.S.T., F.P.T.) performed all surgeries at one private orthopedic institution.
A 13-question survey was developed to assess the factors that influenced patients' decision to proceed with RCR. This survey was developed through the use of an evidence-based review of RCR literature and the Delphi technique.8,9 First, an in-depth literary search of the factors that influence patient decision-making was performed using the PubMed, Cochrane, and EMBASE databases. Once several factors were identified, the list underwent 3 rounds of orthopedic expert review by the authors to achieve consensus to develop the final survey.
Patients were asked to rate each factor regarding its importance in their decision to proceed with RCR on a 5-point Likert scale, ranging from 1 (high importance) to 5 (no importance). Patients were also surveyed on the use of physical therapy, corticosteroid injections, nonsteroidal anti-inflammatory drugs, and opioid medications before surgery.
All patients older than 18 years electing arthroscopic RCR for a full- or partial-thickness rotator cuff tear were eligible to participate. Exclusion criteria were revision shoulder procedures, irreparable rotator cuff tears, and concomitant biceps tenodesis procedures.
A similar survey was developed using the Delphi technique to assess what factors were critical in surgeons' decision to recommend RCR. This questionnaire consisted of 5 questions on a 5-point Likert scale. The first question assessed strength of recommendation for RCR, while the other 4 questions assessed importance of specific patient parameters in decision-making. Patient and surgeon survey items are listed in Table 1.
Factors Assessed in Patient and Surgeon Decision-making
Functional outcome scores of all patients were recorded using the American Shoulder and Elbow Surgeons (ASES) score. The minimal clinically important difference (MCID) for the ASES score in patients with rotator cuff pathology is 12.0.10 This indicates that a mean difference of 12 points on the ASES score is needed for patients to observe meaningful changes in shoulder function. American Shoulder and Elbow Surgeons questionnaires were administered preoperatively and at 2, 6, 12, 26, and 52 weeks postoperatively. Preoperative demographics were also collected.
A power analysis was performed to estimate the number of participants needed to observe statistically significant differences between the 13 decision-making factors. This power analysis indicated that 79 patients would need to be enrolled. Enrollment was increased to account for potential patient attrition and ensure adequate power.
Likert questionnaire responses were converted to binary high-importance or low-importance scores, with a score of “1” or “2” equating to high importance. Conversion to binary categories was based on recommendations from 2 independent professional statisticians after reviewing the distribution of the questionnaire responses. Chi-square and Fisher exact tests were used to analyze the relationship between categorical data. Numeric data were analyzed using t tests, Mann–Whitney U tests, and Pearson or Spearman correlations. Because 13 patient decision-making factors were included, a Bonferroni corrected alpha level of 0.004 was used when analyzing the relationship between patient decision-making factors and demographics/ASES scores. The MCID was used to determine if statistically significant results were clinically meaningful.
A total of 144 patients were prospectively enrolled in this study. Patient demographics are presented in Table 2.
The mean 1-year postoperative ASES score for the total cohort increased compared with the preoperative score (84.2±16.9 vs 42.7±16.7; P<.01) (Figure 1).
Line graph displaying the mean American Shoulder and Elbow Surgeons (ASES) scores during the first postoperative year. Error bars represent mean 95% confidence interval.
Of the 13 patient decision-making factors, the following factors received the greatest frequency of high-importance scores: limited shoulder function (80.6%), surgeon recommendation (79.2%), daily chronic pain (76.4%), concern for tear enlargement (76.4%), and inability to sleep (71.5%). The frequency of high-importance rating for each factor is displayed in Figure 2.
Bar graph displaying the frequency of high-importance rating for each patient decision-making factor.
Patient responses were stratified by sex and age (<55 years or ≥55 years) to assess for differences in the importance of each decision-making factor. Neither sex nor age significantly affected the frequency of high-importance scores (P>.05). Figure 3 displays the frequency of high-importance ratings stratified by sex and age.
Bar graphs displaying the frequency of high-importance rating for each patient decision-making factor stratified by sex (A) and age (B).
Surgeons recommended RCR for 47.2% of patients, and highly recommended RCR for 38.2% of patients. Surgeons recommended neither for nor against RCR in the remaining 14.6% of patients. Among surgeons, the factors that had the highest importance scores were patient activity level (83.3%), risk of tear progression (73.6%), size of tear and atrophy on magnetic resonance imaging (69.4%), and patient age (68.8%).
Surgeons were more likely to rate patient level of activity as a highly important factor for males, compared with females (males, 90.9%; females, 74.6%; P=.01). Surgeons were more likely to rate patient age as a highly important factor for younger patients (<55 years, 89.4%; ≥55 years, 58.8%; P<.01).
Physical therapy, oral nonsteroidal anti-inflammatory drugs, and oral opioid medications were used preoperatively by 45.8%, 78.5%, and 19.4% of patients, respectively. Preoperatively, at least 1 corticosteroid injection was administered to 42.4% of patients, and 26.4% received multiple corticosteroid injections. Seven percent of patients did not attempt any conservative treatments before undergoing RCR. Males were less likely than females to have received 1 corticosteroid injection (males, 32.9%; females, 55.2%; P=.01) or multiple corticosteroid injections (males, 17.8%; females, 36.4%; P=.01). Males also attempted fewer preoperative treatments compared with females (males, 2.0; interquartile range, 1.0–2.0; females, 2.0; interquartile range, 2.0–3.0; P=.03). There were no differences in the use of conservative treatments between younger and older patients (younger, 2.0; interquartile range, 1.0–3.0; older, 2.0; interquartile range, 1.0–3.0; P=.92).
Relationship With Preoperative Function
Patient demographics and decision-making factors were analyzed for their association with preoperative shoulder function. Preoperatively, male ASES scores were higher than female ASES scores (49.3±16.8 vs 35.6±15.0; P<.01). Other demographics, including workers' compensation status, did not affect preoperative ASES scores.
Patients who reported the following factors as highly important in their decision to undergo RCR had significantly worse preoperative ASES scores: limited shoulder function (40.4±16.6 vs 52.8±17.0; P=.001), daily chronic pain (38.3±15.4 vs 58.4±14.6; P<.001), concern for tear enlargement (40.5±17.8 vs 50.7±13.2; P=.001), inability to sleep (37.6±15.0 vs 56.1±16.0; P<.001), and failure of conservative management (37.6±14.9 vs 48.1±18.2; P<.001).
Patients had better preoperative ASES scores when their age (45.8±15.6 vs 36.0±19.1; P<.01) and activity level (44.5±16.8 vs 33.9±17.6; P<.01) were reported by surgeons as highly important in their decision to recommend RCR.
Finally, use of the following treatments before RCR was associated with worse preoperative ASES scores: at least 1 corticosteroid injection (35.8±15.0 vs 47.8±17.3; P<.01), multiple corticosteroid injections (33.6±14.9 vs 46.0±17.0; P<.01), and oral opioids (30.1±14.4 vs 45.7±16.6; P<.01). Furthermore, increased number of attempted conservative treatments correlated with worse preoperative shoulder function (P<.01, r=−0.36).
Relationship With Postoperative Function
The same factors were analyzed for their association with 1-year postoperative shoulder function. Patients with a workers' compensation claim demonstrated worse ASES scores after RCR (68.7±16.6 vs 85.1±16.8; P<.01). Other demographics did not affect postoperative ASES scores.
Patients had worse postoperative ASES scores when they reported “work-related injury” as highly important in their decision to undergo RCR (67.1±12.4 vs 85.6±16.8; P=.001). There was no association between surgeon decision-making factors and postoperative outcomes.
The preoperative use of physical therapy (80.8±17.0 vs 87.1±16.5; P=.045) and multiple corticosteroid injections (78.9±18.8 vs 86.4±15.8; P=.04) were associated with worse postoperative ASES scores. Furthermore, increased number of attempted conservative treatments weakly correlated with worse postoperative shoulder function (P=.03, r=−0.21).
There was a weak correlation between worse preoperative shoulder function and worse postoperative shoulder function (P=.01, r=0.23).
There is limited knowledge about the factors that motivate patients to elect to undergo RCR. The results of this study indicated that limited shoulder function, surgeon recommendation, daily chronic pain, and concern for rotator cuff tear enlargement were the most common factors affecting patients' decisions to undergo RCR, regardless of sex or age. Surgeon recommendation was the second most common patient determinate for RCR and did not statistically vary based on sex or age, indicating that surgeon counseling has a significant effect on all patients' decision-making.
The patient decision-making factors identified as highly important in this study are similar to those previously identified by Minns Lowe et al6 while investigating why patients switched from conservative treatment to surgical treatment during the United Kingdom Rotator Cuff Tear trial. Participants who requested RCR reported the daily impact of symptoms, long duration of intense pain, and failed conservative treatments as reasons for seeking operative management in that study.
In the current cohort, multiple patient decision-making factors were associated with worse preoperative shoulder function; however, the mean difference between them was greater than the MCID for only 3: limited shoulder function, daily chronic pain, and inability to sleep. These results are expected, as patients with these symptoms are likely to report lower shoulder outcome scores. The only decision-making factor associated with worse postoperative outcome scores was work-related injury. Interestingly, patients who rated this as an important decision-making factor did not display worse preoperative function. Similar to this study's results, Henn et al11 found that patients with workers' compensation claims reported worse shoulder function after RCR; however, they found that these patients had worse preoperative function as well.
Patient activity level was rated as a highly important decision-making factor by surgeons in more than 80% of cases. In contrast, patient age was rated as highly important in the least number of cases, indicating that surgeons view other factors, such as patient activity level and size of tear, as more indicative of a patient's need for RCR. Previous authors have stated that patient activity level and size of tear should be taken into consideration when deciding to proceed with RCR, while chronological age should not be a determining factor.4 This indicates that orthopedic surgeons have incorporated evidence-based recommendations into their decision-making. Surgeon decision-making factors for RCR recommendation were also associated with differences in preoperative shoulder function; however, the mean difference between them was not greater than the MCID, indicating that these differences are not clinically significant.
Ninety-three percent of patients attempted at least one form of conservative treatment; however, fewer than half attempted physical therapy before undergoing RCR. This is concerning, given the results of a recent meta-analysis, which indicated that 1-year outcomes following surgical management of rotator cuff tears are not clinically superior to those following physical therapy.12 Furthermore, the American Academy of Orthopaedic Surgeons offers a “strong” recommendation for treatment of symptomatic full-thickness tears with physical therapy.13 The underutilization of preoperative physical therapy is one aspect of patient care that can be improved. Because physical therapy has the potential to successfully manage the symptoms of full-thickness rotator cuff tears, increased use of physical therapy in this patient cohort may have resulted in fewer of these patients failing conservative management. Surgeons should examine preoperative compliance with physical therapy and emphasize its potential benefits as a nonoperative treatment for rotator cuff tears.
Interestingly, female patients attempted more preoperative treatments and were more likely to use corticosteroid injections before RCR than male patients. This relationship may explain the better preoperative shoulder function scores in the male cohort compared with the female cohort. This relationship between sex and conservative treatments warrants further investigation. A potential reason for this difference may be that men are more willing to undergo surgery before exhausting all possible conservative treatments. Although Modi et al14 demonstrated that sex does not affect willingness to undergo shoulder or elbow surgery, Somerson et al15 observed that men have a lower shoulder impairment threshold for electing to undergo shoulder arthroplasty than women. Previous studies have also indicated that women are less willing to undergo elective orthopedic surgery.14,16–18 Regardless of the reason, physicians should be cognizant of the relationship between sex and the use of conservative treatment when counseling patients about treatment options.
The use of corticosteroid injections and opioid medications was also associated with worse preoperative shoulder function. Furthermore, patients who attempted more conservative treatments had worse shoulder function. This may indicate that postponing RCR to attempt conservative treatments may result in worse shoulder function compared with proceeding with RCR after initial rotator cuff tear diagnosis or failed physical therapy. Another possibility is that patients with worse shoulder function may have been more willing to attempt conservative treatments. Unfortunately, the authors did not query patients regarding their decision-making for conservative treatments.
Preoperative physical therapy and multiple corticosteroid injections were associated with worse postoperative shoulder function; however, the mean difference was not greater than the MCID for either analysis. This indicates that the worse postoperative outcomes associated with preoperative physical therapy are not likely to manifest as a clinically noticeable difference.
This study had multiple strengths. The use of the Delphi technique increased the validity of the patient and surgeon surveys used in this study. This study was also strengthened by the prospective collection of shoulder function outcomes. This allowed for the identification of relationships between specific factors and patient outcomes.
A limitation of this study was the exclusion of patients with rotator cuff tears who were not scheduled to undergo RCR. This prevented comparison between patients who underwent RCR and those who did not. Another limitation was the conversion of patient and surgeon responses into binary scores. Although this was not ideal, the questionnaire responses were not appropriately distributed for ordinal statistical analysis. Also, when assessing the surgeon decision-making factor “patient age,” the authors did not record if it was a young age or an old age that influenced decision-making. The surgeon decision-making factor “patient activity level” was limited in a similar fashion. Also, the inclusion of multiple surgeons may have decreased the internal validity of the surgeon questionnaire. Finally, the recruitment of patients from a private community orthopedic practice may limit the generalizability of the results to other orthopedic patient populations.
Limitations in shoulder function and surgeon recommendation were the most common motives involved in patients' decisions to undergo RCR. Multiple factors were associated with worse preoperative shoulder function; however, only patients with work-related injuries demonstrated clinically worse postoperative shoulder function. While counseling patients, surgeons should be cognizant of patient decision-making factors and their relationship with shoulder function.
- Sher JS, Uribe JW, Posada A, Murphy BJ, Zlatkin MB. Abnormal findings on magnetic resonance images of asymptomatic shoulders. J Bone Joint Surg Am. 1995;77(1):10–15. https://doi.org/10.2106/00004623-199501000-00002 PMID: doi:10.2106/00004623-199501000-00002 [CrossRef]7822341
- 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(3):199–203. https://doi.org/10.1067/mse.2001.113086 PMID: doi:10.1067/mse.2001.113086 [CrossRef]11408898
- Huang R, Wang S, Wang Y, Qin X, Sun Y. Systematic review of all-arthroscopic versus mini-open repair of rotator cuff tears: a meta-analysis. Sci Rep. 2016;6(1):22857. https://doi.org/10.1038/srep22857 PMID: doi:10.1038/srep22857 [CrossRef]26947557
- Oh LS, Wolf BR, Hall MP, Levy BA, Marx RG. Indications for rotator cuff repair: a systematic review. Clin Orthop Relat Res. 2007;455(455):52–63. https://doi.org/10.1097/BLO.0b013e31802fc175 PMID: doi:10.1097/BLO.0b013e31802fc175 [CrossRef]
- Youm J, Chenok K, Belkora J, Chan V, Bozic K. The emerging case for shared decision making in orthopaedics. J Bone Joint Surg Am. 2012;94(20):1907–1912. https://doi.org/10.2106/00004623-201210170-00011 doi:10.2106/00004623-201210170-00011 [CrossRef]
- Minns Lowe CJ, Moser J, Barker KL. Why participants in the United Kingdom Rotator Cuff Tear (UKUFF) trial did not remain in their allocated treatment arm: a qualitative study. Physiotherapy. 2018;104(2):224–231. https://doi.org/10.1016/j.physio.2017.09.002 PMID: doi:10.1016/j.physio.2017.09.002 [CrossRef]29361297
- Sepucha KR, Atlas SJ, Chang Y, et al. Informed, patient-centered decisions associated with better health outcomes in orthopedics: prospective cohort study. Med Decis Making.2018;38(8):1018–1026. https://doi.org/10.1177/0272989X18801308 PMID: doi:10.1177/0272989X18801308 [CrossRef]30403575
- Hsu C, Ohio T. Delphi techniques. Pract Assess, Res Eval. 2007;12(10). https://doi.org/10.1576/toag.220.127.116.11071
- Thangaratinam S, Redman CW. The Delphi technique. Obstet Gynaecol. 2005;7(2):120–125. https://doi.org/10.1576/toag.18.104.22.168071
- Tashjian RZ, Deloach J, Green A, Porucznik CA, Powell AP. Minimal clinically important differences in ASES and simple shoulder test scores after nonoperative treatment of rotator cuff disease. J Bone Joint Surg Am. 2010;92(2):296–303. https://doi.org/10.2106/JBJS.H.01296 PMID: doi:10.2106/JBJS.H.01296 [CrossRef]20124055
- Henn RF III, Tashjian RZ, Kang L, Green A. Patients with workers' compensation claims have worse outcomes after rotator cuff repair. J Bone Joint Surg Am. 2008;90(10):2105–2113. https://doi.org/10.2106/JBJS.F.00260 PMID: doi:10.2106/JBJS.F.00260 [CrossRef]18829907
- Ryösä A, Laimi K, Äärimaa V, Lehtimäki K, Kukkonen J, Saltychev M. Surgery or conservative treatment for rotator cuff tear: a meta-analysis. Disabil Rehabil. 2017;39(14):1357–1363. https://doi.org/10.1080/09638288.2016.1198431 PMID: doi:10.1080/09638288.2016.1198431 [CrossRef]
- American Academy of Orthopaedic Surgeons. Management of rotator cuff injuries clinical practice guideline. https://www.aaos.org/rotatorcuffinjuriescpg. Accessed December 8, 2019.
- Modi CS, Veillette CJH, Gandhi R, Perruccio AV, Rampersaud YR. Factors that influence the choice to undergo surgery for shoulder and elbow conditions. Clin Orthop Relat Res. 2014;472(3):883–891. https://doi.org/10.1007/s11999-013-3357-0 PMID: doi:10.1007/s11999-013-3357-0 [CrossRef]
- Somerson JS, Hsu JE, Neradilek MB, Matsen FA III, . The “tipping point” for 931 elective shoulder arthroplasties. J Shoulder Elbow Surg. 2018;27(9):1614–1621. https://doi.org/10.1016/j.jse.2018.03.008 PMID: doi:10.1016/j.jse.2018.03.008 [CrossRef]29748122
- Hawker GA, Wright JG, Coyte PC, et al. Differences between men and women in the rate of use of hip and knee arthroplasty. N Engl J Med. 2000;342(14):1016–1022. https://doi.org/10.1056/NEJM200004063421405 PMID: doi:10.1056/NEJM200004063421405 [CrossRef]10749964
- Gandhi R, Perruccio AV, Rampersaud YR. Predictors of willingness to undergo elective musculoskeletal surgery. Patient Prefer Adherence. 2013;7:191–197. https://doi.org/10.2147/PPA.S41852 PMID: doi:10.2147/PPA.S41852 [CrossRef]23493231
- Lavernia CJ, Contreras JS, Parvizi J, Sharkey PF, Barrack R, Rossi MD. Do patient expectations about arthroplasty at initial presentation for hip or knee pain differ by sex and ethnicity?Clin Orthop Relat Res.2012;470(10):2843–2853. https://doi.org/10.1007/s11999-012-2431-3 PMID: doi:10.1007/s11999-012-2431-3 [CrossRef]22733183
Factors Assessed in Patient and Surgeon Decision-making
|Patient Decision-making Factor||Surgeon Decision-making Factor|
|Inability to sleep (pain at night)||Size of tear and atrophy (on magnetic resonance imaging)|
|Limited function (inability to use the arm)||Activity level of patient|
|Surgeon recommendation||Age of patient|
|Daily, chronic pain||Risk of tear progression|
|Duration of symptoms (symptoms lasting more than 6 months)|
|Concerns for demands of job|
|Insurance coverage issues (concerned there may be a change in insured status)|
|Physical therapy/injections/pain medication not working|
|Concerned the tear might become larger|
|Can no longer play my favorite sport or hobby|
|Age, mean±SD, y||59.3±10.4|
|Body mass index, mean±SD, kg/m2||29.9±5.9|
| Male||53.5% (77)|
| Female||46.5% (67)|
| White||85.4% (123)|
| Black||4.9% (7)|
| Native American||0.7% (1)|
| Other||9.0% (13)|
|Involvement of dominant arm (No.)||61.1% (88)|
|Workers' compensation claim (No.)||9.7% (14)|