Translate this page into:
Reverse shoulder arthroplasty for rotator cuff tears with and without prior failed rotator cuff repair: A large-scale comparative analysis
, Andrew D. Carbonea, Nicole Zubizarretaa, Jashvant Poerana, Bradford O. Parsonsa, Olivier Verborgtb,c, Leesa M. Galatza, Paul J. Caglea
∗Corresponding author: Alexander J. Vervaecke. alexander.vervaecke@gmail.com
-
Received: ,
Accepted: ,
This article was originally published by Reed Elsevier India Pvt. Ltd. and was migrated to Scientific Scholar after the change of Publisher.
Abstract
Abstract
Large-scale data assessing the effect of a prior failed rotator cuff repair (RCR) on the outcome of reverse shoulder arthroplasty (RSA) is currently lacking. Therefore, this study aimed (1) to assess the course of patients undergoing RCR, specifically focusing on the need for conversion to RSA within two years, and (2) to compare outcomes following RSA performed for rotator cuff tears (RCTs) with and without prior RCR.
This retrospective cohort study included data from the CMS Data Set (2016–2018). For the first study objective, we included patients undergoing an RCR; these were followed for 24 months to identify a conversion to RSA. For the second study objective, we included RSAs for RCTs, stratified by those with and without a prior RCR (preceding 24 months). Outcomes (hospitalization cost, institutional post-acute care discharge, 90-day readmission and health resource utilization up to 6 months post-RSA) were compared between propensity score-matched groups.
Out of 33,244 RCRs, 433 (1.3%) patients underwent RSA conversion within two years. Among 7534 RSA cases for RCTs, 245 (3.3%) had an RCR in the preceding two years. In the propensity score analysis, except for a minimal increase in the number of physical rehabilitation visits (RR 1.10; p = 0.0009), no differences were observed between those with and without prior RCR in terms of other RSA outcomes. These included hospitalization cost, discharge to institutional post-acute care facility, 90-day readmission and 6-month post-op cost
Rotator cuff repair in elderly patients, when utilizing currently employed indication criteria, results in low conversion rates to RSA within 2 years postoperatively. Furthermore, large dataset outcomes after RSA for RCT such as cost, post-acute care discharge, physical rehabilitation, and readmission rates appear not to be negatively affected by the presence of a prior RCR.
Level 3 evidence; Retrospective cohort study.
Keywords
Rotator cuff repair
Reverse shoulder arthroplasty
Decision-making
Rotator cuff tear
Elderly population
Revision rotator cuff repair
Failed repair
1 Introduction
The treatment of symptomatic rotator cuff tears (RCT) without concomitant osteoarthritis in the elderly population remains challenging and patient selection for operative care, if indicated, is critical. Even after large or massive tears, high patient satisfaction and improvements in both pain- and function scores can be obtained with rotator cuff repairs (RCR).2,9 Older age however, is a significant and independent predictor of a higher rate of early retears following this procedure.12 Reportedly, 6-month post-RCR retear rates are 15%, 25%, and 34% in patients aged 60–69, 70–79, and >80 years old, respectively.6
In addition to RCR, reverse shoulder arthroplasty (RSA) offers a treatment alternative for rotator cuff tears in elderly patients with RSA in this context frequently reserved for those with irreparable rotator cuff tears or massive tears with pseudo-paralysis.20 Furthermore, in case of a failed rotator cuff repair, RSA may also be performed as a salvage procedure. However, there is no consensus on whether a previous repair negatively influences the outcome compared to a primary RSA, and contradictory data is published.4,8,19,21
Currently, a large-scale assessment investigating the effect of a prior failed rotator cuff repair on the outcome of reverse shoulder arthroplasty performed for RCTs is lacking. Specifically, information on the conversion rates after primary RCRs in an elderly population and a comparison between outcomes of RSA for RCTs with and without preceding failed RCR may contribute to counseling patients in their decision-making when considering operative treatment.
2 Methods
2.1 Study design
This retrospective cohort study addressed two separate study aims: (1) To assess the clinical course of elderly patients after RCR (explicitly looking at conversion to RSA within 2-years post-op). (2) To compare outcomes (including health care utilization) following RSA for RCTs between patients who had a previous repair versus those who did not and to investigate possible associations between those outcomes and shorter time intervals between the two procedures. Data were extracted from US national Medicare claims (‘Limited Dataset’ 2016–2018). These claims data are subdivided into various datasets related to the types of coverage. They contain information on federal health insurance beneficiaries who are ≥65 years old or <65 years old with specific disabilities or comorbidities. Information includes International Classification of Diseases, 10th edition (ICD-10) diagnosis and procedure codes, Current Procedural Terminology (CPT) codes, and detailed information on resource utilization, including Medicare payments and institutional post-acute care utilization. The dataset was specifically selected due to the availability of longitudinal information on healthcare utilization of approximately 60 million individual patients (±98% of individuals ≥65 years old in the US).15 Additionally, using a large dataset maximizes the ability to identify meaningful differences between groups, particularly in terms of infrequent events or outcomes. Moreover, observational data improve external validity by incorporating real-world data from various hospitals and providers. This study was exempt from Institutional Review Board approval due to the use of HIPAA-compliant anonymized data and meeting the FDA guidelines for Not Human Subjects Research.
2.2 Study sample
To address the first study objective, we included patients with continuous Medicare enrollment (only eligible through their age) undergoing an RCR (ICD-10 procedure codes 0LQ13ZZ; 0LQ14ZZ; 0LQ20ZZ; 0LQ23ZZ; 0LQ24ZZ and CPT codes 23412; 23410; 23420; 29827, with lateralization modifiers) and a diagnosis code for RCT (ICD-10 code M75.1xx) from January to December 2016. Subsequently, longitudinal data for up to two years after the index repair of each patient was analyzed to identify cases that underwent a conversion to RSA (ICD-10 codes 0RRK00Z; 0RRJ00Z, same lateralization). For the second study objective, we included patients undergoing an RSA with a primary diagnosis of RCT in an inpatient setting between January 2017 to June 2018. They were stratified by those that received an RCR in the 2-year timeframe prior to the RSA and those that did not. Patients with a concomitant diagnosis of a proximal humerus fracture (S42.2) were excluded.
2.3 Study variables
The main variables for the first study objective were the need for and time to RSA within two years after a primary RCR. The main variables for the second study objective were a history of a prior RCR in patients receiving an RSA for RCT and the time between the potential RCR and conversion to RSA. Outcomes of interest were the total cost of the RSA hospitalization, discharge to institutional post-acute care, 6-month health resource utilization (overall health care cost for ambulatory visits, inpatient treatments, and skilled nursing facility stays, both continuous and dichotomized by using the 75th percentile), 90-day all-cause readmission rates and the number of ambulatory health care and physical therapy visits (both continuous and dichotomized by using the 75th percentile) during the early 6-month postoperative period.
Other variables of interest were patient demographics (age, sex, race), Charlson and non-Charlson (depression, thyroid disease, obesity, smoking status and hyperlipidemia) comorbidities and tear classification (full-thickness, partial thickness or unspecified).1,5,10,11,13,16,18 These were defined using ICD-10 diagnosis codes. Charlson comorbidities were categorized into four groups based on the index score (0, 1, 2, and >2); a higher score indicates a greater burden. In addition, the annual number of RSA procedures per hospital was included as a hospital-specific variable.
2.4 Statistical analysis
Group differences are assessed by p-values and standardized differences (SDD). The latter was chosen as group differences in large datasets easily reach statistical significance; an SDD ≥0.10 (10%) is considered to reflect a meaningful difference in variable distribution between groups.3 For the second study objective, a propensity score analysis was applied. First, propensity scores were calculated in a regression model using all available variables with the outcome of the presence or absence of a prior RCR before an RSA. Then, patients treated with RSA for RCT with a prior RCR (‘cases’) were matched to three patients without a prior RCR (‘controls’); 1:3 matching. In this ‘matched’ study sample, regression models measured the differences in outcomes between RSAs with and without a previous RCR. Additional model adjustments were carried out for all variables that remained statistically different after propensity matching. We report odds ratios (OR) for binary outcomes, % change for continuous outcomes, and rate ratios (RR) for count outcomes such as the number of physical rehabilitation visits, all with 95% confidence intervals (CI). Spearman correlation models measured the (linear) association of time between RCR and conversion to RSA for the continuous outcome measurements. The Youden's index was applied to try and establish an ideal timeframe between both procedures. All analyses were done using SAS v.9.4 statistical software (SAS Institute, Cary, NC).
3 Results
3.1 Study objective 1
Out of 33,244 RCRs, 12,034 (36.2%) were performed for full-thickness and 3880 (11.7%) for partial thickness tears. In 17,330 (52.1%) cases, the extent of the tear was unspecified. Within two years of follow-up, 433 (1.3%) patients underwent revision surgery to an RSA. The median time between the index and secondary surgery was 9.0 months (interquartile range 7.9) (Fig. 1).
3.2 Study objective 2
Among 7534 RSA cases, 245 (3.3%) had an RCR in the preceding two years. Patients with a prior failed RCR were significantly more likely to be younger (mean of 73.0 years versus 75.3 years; SDD = 44.7%), female (65.7% versus 59.8%; SDD = 12.3%) and have a prior diagnosis of depressive disorders (21.2% versus 15.4%; SDD = 15.1%) and hyperlipidemia (62.5% versus 53.3%; SDD = 18.5%) (Table 1). The median time between RCR and conversion to RSA was 8.1 months (interquartile range 8.1) (Fig. 2). In the propensity score-matched cohort (Table 2), regression analysis revealed that patients with a prior RCR had a modest increase in the total number of postoperative ambulatory physical therapy visits (RR 1.10 CI 1.04; 1.16, p = 0.0009/mean 7.9 ± 11.8 versus 6.7 ± 10.8 visits). No differences were found for other outcomes (Table 3). In patients with a prior RCR, there were no trends observed between worse linear outcomes (following RSA) and shorter timeframes between the RSA and previous RCR (Table 4). Finally, by using the Youden's index, no optimal time interval between the two procedures was found (Supplemental Table I).
| Unmatched cohort | ||||
| Prior RCR | p-value | SDD | ||
| Study variables | Yes (n = 244) | No (n = 7289) | ||
| Hospital characteristics | ||||
| Yearly RSA-volume | 18.3 ± 15.5 | 23.3 ± 20.3 | 0.0145 | 27.9% |
| Sex | 0.0614 | 12.3% | ||
| Male | 84 (34.3%) | 2933 (40.2%) | ||
| Female | 161 (65.7%) | 4356 (59.8%) | ||
| Age (years) | 73.0 ± 4.7 | 75.3 ± 5.4 | <0.0001 | 44.7% |
| Race | 0.3348 | 10.7% | ||
| White | 235 (95.9%) | 6831 (93.7%) | ||
| Black | 6 (2.4%) | 233 (3.2%) | ||
| Other | 4 (1.6%) | 225 (3.1%) | ||
| Charlson-Deyo Index | 0.3629 | 11.9% | ||
| 0 | 124 (50.6%) | 3373 (46.3%) | ||
| 1 | 62 (25.3%) | 2013 (27.6%) | ||
| 2 | 26 (10.6%) | 993 (13.6%) | ||
| >2 | 33 (13.5%) | 910 (12.5%) | ||
| Comorbidities | ||||
| Osteoporosis | 154 (62.9%) | 5256 (72.1%) | 0.0016 | 19.9% |
| Uncomplicated diabetes | 44 (18.0%) | 1209 (16.6%) | 0.5704 | 3.6% |
| Complicated diabetes | 24 (9.8%) | 512 (7.0%) | 0.0969 | 10.0% |
| Congestive heart failure | 12 (4.9%) | 455 (6.2%) | 0.3907 | 5.7% |
| Myocardial infarction | 14 (5.7%) | 416 (5.7%) | 0.9963 | 0.3% |
| Cerebrovascular disease | 3 (1.2%) | 117 (1.6%) | 0.6397 | 3.2% |
| COPD | 44 (18.0%) | 1426 (19.6%) | 0.5330 | 4.1% |
| Renal failure | 18 (7.3%) | 802 (11.0%) | 0.0707 | 12.7% |
| Peripheral vascular disease | 10 (4.1%) | 305 (4.2%) | 0.9370 | 0.5% |
| Rheumatic disease | 13 (5.3%) | 523 (7.2%) | 0.2630 | 7.7% |
| Dementia | 4 (1.6%) | 120 (1.6%) | 0.9868 | 0.1% |
| Depression | 52 (21.2%) | 1123 (15.4) | 0.0136 | 15.1% |
| Thyroid disease | 49 (20.0%) | 1609 (22.1%) | 0.4408 | 5.2% |
| Smoking | 79 (32.2%) | 2437 (33.4%) | 0.6979 | 5.1% |
| Obesity | 39 (15.9%) | 1360 (18.7%) | 0.2780 | 7.3% |
| Hyperlipidemia | 153 (62.4%) | 3889 (53.4%) | 0.0050 | 18.5% |
| Propensity-score matched cohort | ||||
| Prior RCR | p-value | SDD | ||
| Study variables | Yes (n = 244) | No (n = 732) | ||
| Hospital characteristics | ||||
| Yearly RSA-volume | 18.3 ± 15.5 | 19.8 ± 17.1 | – | 9.1% |
| Sex | 0.8448 | 1.5% | ||
| Male | 83 (34.3%) | 244 (33.3%) | ||
| Female | 161 (65.7%) | 488 (66.7%) | ||
| Age (years) | 73.0 ± 4.7 | 73.1 ± 4.7 | – | 1.3% |
| Race | 0.8617 | 4.0% | ||
| White | 235 (95.9%) | 707 (96.6%) | ||
| Black | 6 (2.4%) | 14 (1.9%) | ||
| Other | 4 (1.6%) | 11 (1.5% | ||
| Charlson-Deyo Index | 0.5886 | 10.1% | ||
| 0 | 124 (50.6%) | 378 (51.6%) | ||
| 1 | 62 (25.3%) | 187 (25.5%) | ||
| 2 | 26 (10.6%) | 89 (12.2%) | ||
| >2 | 33 (13.5%) | 112 (7.7%) | ||
| Comorbidities | ||||
| Osteoporosis | 154 (62.9%) | 1091 (74.7%) | <0.0001 | 20.6% |
| Uncomplicated diabetes | 44 (18.0%) | 216 (14.8%) | 0.0102 | 13.1% |
| Complicated diabetes | 24 (9.8%) | 79 (5.4%) | 0.1981 | 6.5% |
| Congestive heart failure | 12 (4.9%) | 64 (4.4%) | 0.5148 | 3.5% |
| Myocardial infarction | 14 (5.7%) | 66 (4.5%) | 0.7085 | 1.9% |
| Cerebrovascular disease | 3 (1.2%) | 14 (1.0%) | 0.8941 | 0.7% |
| COPD | 44 (18.0%) | 300 (20.5%) | 0.8462 | 1.0% |
| Renal failure | 18 (7.3%) | 110 (7.5%) | 0.8435 | 1.0% |
| Peripheral vascular disease | 10 (4.1%) | 42 (2.9%) | 0.8159 | 1.2% |
| Rheumatic disease | 13 (5.3%) | 86 (5.9%) | 0.6530 | 2.4% |
| Dementia | 4 (1.6%) | 15 (1.0%) | 1.0000 | 0% |
| Depression | 52 (21.2%) | 196 (13.4%) | <0.0001 | 19.6% |
| Thyroid disease | 49 (20.0%) | 260 (17.8%) | 0.0652 | 9.5% |
| Smoking | 79 (32.2%) | 481 (32.9%) | 0.9778 | 0.2% |
| Obesity | 39 (15.9%) | 257 (17.6%) | 0.2656 | 5.9% |
| Hyperlipidemia | 153 (62.4%) | 753 (51.5%) | 0.0338 | 11.2% |
| Continuous outcomes | Prior RCR with 2-year lookback period | |||||||
| Yes (n = 244) | No (n = 732) | |||||||
| Mean | SD | Mean | SD | % | 95% CI | p-value | ||
| Hospitalization cost ($) | 14,972 | 4149 | 14,815 | 3993 | 1.3 | −2.5 | 5.3 | 0.5013 |
| 6-month post-op cost ($) | 5414 | 11,532 | 5311 | 10,966 | −1.5 | −21.2 | 23.2 | 0.8972 |
| RR | 95% CI | p-value | ||||||
| 6-month Ambulatory visits (#) | 12.3 | 13.8 | 11.5 | 14.7 | 1.03 | 0.99 | 1.08 | 0.1395 |
| 6-month Physical therapy visits (#) | 7.9 | 11.8 | 6.7 | 10.8 | 1.10 | 1.04 | 1.16 | 0.0009 |
| Binary outcomes | N | % | N | % | OR | 95 CI | p-value | |
| Discharge to post-acute care | 18 | 7.4 | 80 | 10.9 | 0.60 | 0.35 | 1.04 | 0.0673 |
| 90-day readmission | 19 | 7.8 | 56 | 7.7 | 0.99 | 0.57 | 1.70 | 0.9693 |
| 6-month post-op cost >75 percentile | 67 | 27.5 | 185 | 25.3 | 1.12 | 0.80 | 1.55 | 0.5112 |
| 6-month Ambulatory visits >75% percentile (>18 visits) | 75 | 30.7 | 200 | 27.3 | 1.18 | 0.86 | 1.62 | 0.3080 |
| 6-month Physical therapy visits >75% percentile (>13 visits) | 73 | 29.9 | 207 | 28.3 | 1.20 | 0.86 | 1.68 | 0.2832 |
| Continuous outcomes | R | CI | R2 | p-value | |
| Hospitalization cost ($) | −0.02 | −0.15 | 0.10 | 0.0006 | 0.7114 |
| 6-month post-op cost ($) | −0.06 | −0.18 | 0.07 | 0.0032 | 0.3770 |
| 6-month Ambulatory visits (#) | −0.05 | −0.18 | 0.07 | 0.0030 | 0.3964 |
| 6-month Physical therapy visits (#) | −0.01 | −0.14 | 0.11 | 0.0002 | 0.8441 |
4 Discussion
In this study, we found that the rate of RCRs necessitating subsequent conversion to RSA within two years after the index repair in the Medicare population is low (1.3%). However, when conversion did occur, it fairly consistently occurred in the early acute postoperative period (median of 9 months post-op). Furthermore, except for a minimal increase in postoperative physical therapy visits, our results demonstrate that acute and other short-term database outcomes following RSA were not negatively affected by the presence of a prior recent RCR. Finally, no clear associations were found between shorter time intervals between the initial RCR and conversion to RSA and the outcomes following RSA.
Deciding on the optimal surgical treatment for elderly patients with RCTs is challenging. Higher age is an independent predictor of structural failures following RCR and is associated with multiple factors that may negatively affect functional outcomes. In contrast, RSA offers a treatment alternative with reliable clinical results in older patients. However, our findings suggest that, when following the currently employed indication criteria, an RCR is unlikely to result in conversion to RSA in the early 2-year postoperative period despite the high retear rates published in the literature.6 This discrepancy may be explained by the reported variability in the functional impact of retears as some patients experience improvement compared to their preoperative condition despite structural failure. Furthermore, the etiology of failures in RCR and the need to undergo secondary surgery are multifactorial and not solely based on the presence of a retear or structural failure.
Although speculative, one could expect older patients with more comorbidities to be more likely selected for RSA as initial treatment for their RCT whereas, for younger, healthier patients, an RCR will be preferred. However, the lack of variation between patient characteristics and comorbidities of RSA cases with and without a prior RCR in our second study objective, implies that those variables do not predominately drive the selection of surgical treatment. Specialty group surveys suggest a significant between-surgeon variation regarding treatment preferences of RCTs.7,14 Marked inconsistencies were found between how important surgeons perceived specific determinants of outcome of RCR to be, such as age, fatty infiltration, and muscle atrophy, and when they would consider RSA as treatment of choice.17 These dissimilarities in daily practice may be mitigated by future prospective studies aiming to achieve consensus in clinical dilemmas regarding surgical treatment of RCTs in the elderly.
There is no clear unanimity on how the presence of a prior RCR affects outcomes after RSA.4,8,19,21 Insight on this topic is essential to correctly counsel elderly patients about the potential downsides associated with conversion to RSA following failure after RCR as compared with primary RSA. Shields et al. found that patients with a prior repair exhibited significantly worse postoperative ASES (10% decline) and pain (16% decline) scores and overall decreased functional gains compared to patients without previous shoulder surgery. Similar unfavorable findings were described by Boileau et al. who attributed lower Constant scores and active anterior elevation to the same subgroup of patients. The reasoning behind these inferior results is unclear but may be linked to limited deltoid atrophy and scar formation related to the RCR. In contrast, Sadoghi et al. and Erickson et al., after matching for baseline differences between groups, found that a prior RCR did not negatively affect clinical outcomes following RSA, thus disputing these earlier conclusions.
These discrepancies in the literature may impede adequate patient selection for operative treatment, and agreement is needed as to whether attempting an RCR precludes functional consequences further down the line. In our large-scale assessment, patients with a prior RCR were found to have a negligible increase in the amount of postoperative physical rehabilitation visits. However, no differences in hospitalization or 6-month postoperative cost were found between groups when assessing for the presence of an RCR within 2 years prior to the RSA procedure. Moreover, the presence of a prior RCR did not increase the readmission rate, the overall amount of ambulatory health care visits, nor the percentage of patients receiving more than standard (>75th percentile) physical rehabilitation sessions within six months postoperatively.
In order to investigate a possible time-interval related detrimental effect of a prior RCR on the outcome of RSA, as appears to be the case for total knee arthroplasty (TKA) within six months after a prior knee arthroscopy within, we assessed for time-interval and outcome correlations.22 However, no significant linear associations were found between shorter time intervals between RCR and conversion to RSA and worse outcomes. This indicates that attempting an RCR does not result in a long delay if a conversion is necessary. Lastly, in our study, the Youden's index failed to provide a time-interval cut-off with a reliable sensitivity and/or specificity after which superior outcomes were found for either of the dichotomized outcomes further stipulating that, based upon our study population, there is no time-interval related effect present between a prior RCR and conversion to RSA.
Our study has several limitations. Given that our conversion rates are based upon a large administrative database, we were unable to assess rates and outcomes relative to significant tear characteristics such as the chronicity, width, multi-tendon involvement and retraction of the tear and muscle fatty infiltration. In addition, we were limited by the coding accuracy of the inputted data which is an inherent limitation of database research. Due to the nature of a retrospective cohort study, we cannot reliably infer what the reasoning of the respective physician was on whether to treat a patient with an RCR or RSA or to proceed to a conversion procedure. Elderly patients are possibly less likely to have reoperations due to comorbidities, low operation rates are not fully equivalent to high rates of successful surgery. Finally, while the number of postoperative physical therapy visits related to the RSA may be a potential surrogate at best, it should not be interpreted as a replacement for clinical function assessments. Patient-reported and clinically-measured outcome parameters are beyond the possibility of the CMS dataset.
5 Conclusion
Rotator cuff repair in elderly patients, when utilizing currently employed indication criteria, results in low conversion rates to RSA within two years postoperatively. Subsequent conversion of a failed RCR tot RSA does not seem to negatively impact cost, readmission and health resource utilization compared to primary RSA for RCT. Furthermore, there appears to be no time-interval-related detrimental effect between a prior RCR and conversion to RSA.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors.
Author contributions
Alexander Vervaecke: Conceptualization, methodology, writing – original draft, writing review.
Andrew Carbone: Conceptualization, methodology, writing – review.
Nicole Zubizarreta: Formal analysis, investigation, validation.
Jashvant Poeran: Formal analysis, validation, writing – review, supervision, project administration.
Bradford Parsons: Writing – review.
Olivier Verborgt: Conceptualization, writing – review.
Leesa Galatz: Writing – review, supervision.
Paul Cagle: Methodology, writing – review, supervision.
References
- Factors affecting healing after arthroscopic rotator cuff repair. World J Orthoped. 2015;6(2):211-220.
- [Google Scholar]
- Repair of rotator cuff tears in the elderly: does it make sense? A systematic review. Am J Sports Med. 2020;48(3):744-753.
- [Google Scholar]
- An introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivariate Behav Res. 2011;46(3):399-424.
- [Google Scholar]
- Reverse total shoulder arthroplasty after failed rotator cuff surgery. J Shoulder Elbow Surg. 2009;18(4):600-606.
- [Google Scholar]
- Factors affecting rotator cuff healing after arthroscopic repair: osteoporosis as one of the independent risk factors. Am J Sports Med. 2011;39(10):2099-2107.
- [Google Scholar]
- Relationship between age and rotator cuff retear. J Bone Jt Surg - Am. 2017;99(14):1198-1205.
- [Google Scholar]
- Variation in orthopaedic surgeons' perceptions about the indications for rotator cuff surgery. J Bone Jt Surg Am. 2005;87(9):1978-1984.
- [Google Scholar]
- Does having a rotator cuff repair prior to reverse total shoulder arthroplasty influence the outcome? Bone Jt J. 2019;101B(1):63-67.
- [Google Scholar]
- Rotator cuff surgery in patients older than 75 years with large and massive tears. J Shoulder Elbow Surg. 2017;26(2):265-272.
- [Google Scholar]
- Depression increases the risk of rotator cuff tear and rotator cuff repair surgery: a nationwide population-based study. PLoS One. 2019;14(11)
- [Google Scholar]
- The rotator cuff healing index: a new scoring system to predict rotator cuff healing after surgical repair. Am J Sports Med. 2019;47(1):173-180.
- [Google Scholar]
- Factors predicting rotator cuff retears: an analysis of 1000 consecutive rotator cuff repairs. Am J Sports Med. 2014;42(5):1134-1142.
- [Google Scholar]
- The effect of diabetes, hyperlipidemia, and statins on the development of rotator cuff disease: a nationwide, 11-year, longitudinal, opulation-based follow-up study. Am J Sports Med. 2015;43(9):2126-2132.
- [Google Scholar]
- Management of full thickness rotator cuff tears. A survey amongst members of the Flemish Elbow and Shoulder Surgeons Society (FLESSS) Acta Orthop Belg. 2010;76(1):14-21.
- [Google Scholar]
- Use of the Medicare database in epidemiologic and health services research: a valuable source of real-world evidence on the older and disabled populations in the US. Clin Epidemiol. 2017;9:267-277.
- [Google Scholar]
- Effect of smoking on healing failure after rotator cuff repair. Am J Sports Med. 2018;46(12):2960-2968.
- [Google Scholar]
- The Management of large and massive rotator cuff tears- Current trends amongst UK shoulder surgeons. Open J Orthop Rheumatol. 2018;3:5-11.
- [Google Scholar]
- Impact of previous rotator cuff repair on the outcome of reverse shoulder arthroplasty. J Shoulder Elbow Surg. 2011;20(7):1138-1146.
- [Google Scholar]
- Reverse shoulder arthroplasty for irreparable massive rotator cuff tears: a systematic review with meta-analysis and meta-regression. J Shoulder Elbow Surg. 2017;26(9):e265-e277.
- [Google Scholar]
- Previous rotator cuff repair is associated with inferior clinical outcomes after reverse total shoulder arthroplasty. Orthop J Sport Med. 2017;5(10):1-6.
- [Google Scholar]
- Total knee arthroplasty within six months after knee arthroscopy is associated with increased postoperative complications. J Arthroplasty. 2015;30(8):1313-1316.
- [Google Scholar]