Arthroscopic revision cuff repair: do tendons have a second chance to heal?

Hallucination Rates and Reference Accuracy of ChatGPT and Bard for Systematic Reviews: Comparative Analysis

BACKGROUND

Large language models (LLMs) have raised both interest and concern in the academic community. They offer the potential for automating literature search and synthesis for systematic reviews but raise concerns regarding their reliability, as the tendency to generate unsupported (hallucinated) content persist.

OBJECTIVE

The aim of the study is to assess the performance of LLMs such as ChatGPT and Bard (subsequently rebranded Gemini) to produce references in the context of scientific writing.

METHODS

The performance of ChatGPT and Bard in replicating the results of human-conducted systematic reviews was assessed. Using systematic reviews pertaining to shoulder rotator cuff pathology, these LLMs were tested by providing the same inclusion criteria and comparing the results with original systematic review references, serving as gold standards. The study used 3 key performance metrics: recall, precision, and F1-score, alongside the hallucination rate. Papers were considered "hallucinated" if any 2 of the following information were wrong: title, first author, or year of publication.

RESULTS

In total, 11 systematic reviews across 4 fields yielded 33 prompts to LLMs (3 LLMs×11 reviews), with 471 references analyzed. Precision rates for GPT-3.5, GPT-4, and Bard were 9.4% (13/139), 13.4% (16/119), and 0% (0/104) respectively (P<.001). Recall rates were 11.9% (13/109) for GPT-3.5 and 13.7% (15/109) for GPT-4, with Bard failing to retrieve any relevant papers (P<.001). Hallucination rates stood at 39.6% (55/139) for GPT-3.5, 28.6% (34/119) for GPT-4, and 91.4% (95/104) for Bard (P<.001). Further analysis of nonhallucinated papers retrieved by GPT models revealed significant differences in identifying various criteria, such as randomized studies, participant criteria, and intervention criteria. The study also noted the geographical and open-access biases in the papers retrieved by the LLMs.

CONCLUSIONS

Given their current performance, it is not recommended for LLMs to be deployed as the primary or exclusive tool for conducting systematic reviews. Any references generated by such models warrant thorough validation by researchers. The high occurrence of hallucinations in LLMs highlights the necessity for refining their training and functionality before confidently using them for rigorous academic purposes.

Outcomes After Revision Arthroscopic Rotator Cuff Repair: A Systematic Review

BACKGROUND

Arthroscopic rotator cuff repair (ARCR) is one of the most commonly performed orthopaedic procedures; however, it is unclear how to manage patients with symptomatic retears.

PURPOSE

To perform a systematic review to evaluate the outcomes of revision ARCR in the literature.

STUDY DESIGN

Systematic review; Level of evidence, 4.

METHODS

A systematic literature search based on PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines was performed utilizing the Embase, MEDLINE, and Cochrane Library databases. Eligible for inclusion were clinical studies reporting on revision rotator cuff repair. Clinical outcomes were recorded.

RESULTS

The 17 included studies were composed of 888 shoulders in 885 patients. The mean age of the 885 patients was 58.6 ± 9.1 years, with a mean follow-up of 28.1 ± 22.1 months. The mean visual analog scale score was 2.1 ± 1.8, the mean American Shoulder and Elbow Surgeons score was 75.0 ± 18.3, and the mean Constant score was 67.4 ± 16.6. The overall rate of retears after revision was 25.3%, with an 8.0% reoperation rate including 2.7% undergoing reverse shoulder arthroplasty and 2.0% undergoing subsequent rotator cuff repair. Finally, the complication rate was 8.6%.

CONCLUSION

Revision ARCR improved patient-reported outcomes, with 92% of patients free from reoperations at a mean follow-up of 2 years. Overall, 78.4% of patients were satisfied at final follow-up; however, the rate of retears was found to be higher than after primary ARCR.

Outcomes of arthroscopic single-row repair alone vs. repair with human dermal allograft patch augmentation in patients with large to massive, posterosuperior rotator cuff tears: a retrospective comparative study

BACKGROUND

Large to massive rotator cuff tears (RCTs) affect shoulder functions profoundly with unmanageable disability without intervention. The retear rates with arthroscopic rotator cuff repair (ARCR) in these patients are abysmal. Patch augmentation has been credited for preventing retears, improving functions by increasing the strength, and acting as a bioconductive scaffold. This study aimed to assess the retear rates and compare the clinical and radiological outcomes between the ARCR with and without acellular human dermal allograft (HDA) augmentation.

METHODS

This is a retrospective comparative study among patients diagnosed with large to massive, posterosuperior RCTs, operated between January 2020 and December 2021, including 36 patients (group I) with and 131 patients (group II) without HDA augmentation, with a mean follow-up of 20 (range, 12-35) months. The average age was 64 (range, 49-80) and 66 (range, 41-81) years in groups I and II, respectively. In group I, there were 16 male and 20 female patients, whereas in group II, there were 58 male and 73 female patients.

RESULTS

The visual analog scale score improved to 1.1 ± 1.7 in group I and 2.1 ± 1.7 in group II (P = .005). There was a greater improvement in the University of California, Los Angeles shoulder score to 30.1 ± 4.2 in group I compared with 23.2 ± 3.9 in group II (P = .046). Forward flexion (degrees) improved from a mean of 103.2 ± 18.6 to 138.9 ± 23.5 in group I and from 106.4 ± 21.3 to 127.0 ± 19.5 in group II (P = .004). The acromiohumeral interval (mm) measured in anteroposterior radiographs increased to 8.4 ± 1.8 in group I and 8.2 ± 2.0 in group II (P = .006). The satisfaction after the procedure was 4.4 ± 0.6 in group I and 3.1 ± 1.1 in group II (P = .044). The retear rate in the HDA-augmented group was 5.6% as compared with 29.1% in the nonaugmented group, which was statistically significant (P = .007). There were no complications or adverse tissue reactions against HDA seen in any patients.

CONCLUSION

In patients with large to massive, posterosuperior RCTs, patch augmentation with acellular HDA significantly averted the retears after ARCR without any graft-related complications. The augmentation also resulted in improved shoulder function and greater range of motion compared with the nonaugmented group.

Dermal Allograft Augmentation for Large and Massive Rotator Cuff Tears and Revisions: A Case Series

Introduction Arthroscopic advancements have notably improved rotator cuff repair, yet the healing of tendons and retear risks still pose significant challenges, especially with massive and large tears. These larger tears have shown a quadruple increase in retear rate when the tear size grows from 1-3 cm. The prevalent cause of failure, particularly in tears with fatty degeneration, is suture tearing. As such, techniques that enhance tendon healing are pivotal due to their correlation with improved functional outcomes. Dermal allograft augmentation (DAA) is one method that has demonstrated improvements in rotator cuff repair biomechanics, as well as the promotion of vascularization and neotendon formation. This study evaluates the efficacy of DAA in the postoperative healing of large and massive rotator cuff tears, assessed through patient-reported outcomes and functional range of motion. Methods This retrospective study reviewed a single surgeon's database, selecting patients who underwent arthroscopic surgery for full-thickness rotator cuff tears measuring at least 3 cm, or those with maximum stage II fatty infiltration of the supraspinatus or infraspinatus muscle on MRI. We used three validated patient-reported outcome (PRO) measures, American Shoulder and Elbow Surgeon Score (ASES), Single Assessment Numeric Evaluation (SANE), and Simple Shoulder Test (SST), alongside a range of motion metrics, external rotation (ER), and forward flexion (FF). These were recorded pre-operatively and, at minimum, one year post-operatively, with a standard student t-test employed for comparative analysis. Results Out of 18 total patients, the average age was 61 years, and 83% were male. The cause of injury was mostly traumatic in nature (n=11), with the remainder being chronic (n=7). All three PROs (n=11) demonstrated a statistically significant increase in score including ASES (p=0.003), SANE (p=0.004), and SST (p=0.039). External rotation also significantly improved pre- to post-operative function (46 vs 58, p=0.049). Three patients (17%) suffered from a retear within three months of their rotator cuff repair. Two patients (11%) required a reverse shoulder arthroplasty within a year. Conclusion Our study demonstrated that DAA in rotator cuff repair yielded notable functional and clinical advancements, with a moderately low retear rate as confirmed by ultrasound. These outcomes suggest that DAA is a promising intervention for large and historically difficult rotator cuff tears, offering significant implications for future treatment protocols.

Long-term results of revision rotator cuff repair for failed cuff repair: a minimum 10-year follow-up study

HYPOTHESIS

Rotator cuff repair remains associated with high retear rates, which range from 13% to 79%. The objective of this study was to evaluate the long-term clinical and structural results after revision rotator cuff repair at a minimum 10-year follow-up.

METHODS

We retrospectively studied the records of all patients who underwent revision rotator cuff repair in 3 different institutions between July 2001 and December 2007 with a minimum 10-year follow-up. A total of 54 patients (61% males, mean age 52 ± 6 years old) met the inclusion criteria. Outcome measures included pain (visual analog scale [VAS]), range of motion (ROM), Subjective Shoulder Value (SSV), and the Constant score. Superior migration, osteoarthritis, and acromiohumeral interval (AHI) were assessed on standard radiographs. Fatty infiltration and structural integrity of the repaired tendon were evaluated on magnetic resonance imaging or computed tomographic arthrogram.

RESULTS

At a mean 14.1 years (10.4-20.5), range of motion did not progress significantly in elevation and internal rotation between pre- and postoperation (158° [range, 100°-180°] to 164° [range, 60°-180°], P = .33, and L3 [range, sacrum-T12] to T12 [range, buttocks-T7], P = .34, respectively) and decreased in active external rotation from 45° (range, 10°-80°) to 39° (range, 10°-80°) (P = .02). However, VAS, SSV, and Constant score were all significantly improved at last follow-up (P < .001). AHI decreased significantly (P = .002) from 10 mm (7-14 mm) to 8 mm (0-12 mm). Two percent of the supraspinatus/infraspinatus tendons were Sugaya 1, 24% were Sugaya 2, 35% were Sugaya 3, 12% were Sugaya 4, and 27% were Sugaya 5. Goutallier score progressed for all muscles, but this did not reach significance and mean Goutallier remained <2 for all 4 muscles at last follow-up. Hamada score progressed from 0% >grade 2 preoperatively to 6% >grade 2 at last follow-up.

CONCLUSION

Revision rotator cuff repair provides significant pain relief and improvement in functional scores at long-term follow-up. The mild progression of fatty infiltration, AHI, and Hamada score suggests that despite high retear rates (39% of stage 4 and 5 in the Sugaya classification), revision repair could possibly have a protective role on the evolution toward cuff tear arthropathy.

Bone marrow edema in the acromioclavicular joint after arthroscopic rotator cuff repair

BACKGROUND

Bone marrow edema (BME) in the acromioclavicular joint (ACj) may occur after arthroscopic rotator cuff repair (ARCR), resulting in persistent postoperative ACj pain. We investigated the incidence and clinical characteristics of BME in ACj after ARCR.

METHODS

Patients exhibiting ACj-related symptoms prior to ARCR surgery, no BME in the ACj on MRI, and an ability to undergo 2-year postoperative physical and MRI examinations of the ACj were included. Patients who underwent distal clavicle resection or osteophyte resection below the ACj and those unable to achieve primary repair of the rotator cuff were excluded. MRI evaluation was performed preoperatively and at 3, 6, 12, and 24 months postoperatively. Patients who showed BME in ACj on fat-suppressed T2 MRI at 24 months postoperatively were classified into the BME + group, and those with no BME were classified into the BME- group. Patient background, operation time, Japanese Orthopedic Association (JOA) score, re-tear rate, ACj tenderness, and cross-body adduction test were compared between groups. In addition, we examined the time of occurrence of BME.

RESULTS

A total of 345 ARCRs were performed during the study period. After the exclusion of 114 shoulders, the remaining 231 shoulders were included in this study. There were 208 cases in the BME- group and 23 cases in the BME + group. The incidence of BME was 9.96 %. Although the JOA scores of both groups showed a significant improvement postoperatively compared to preoperatively, there was no significant difference between the two groups. In terms of re-tear, there was 1 case in the BME + group and 7 in the BME- group. Positive ACj tenderness and cross-body adduction test were significantly higher in the BME + group (P < 0.001). BME occurred in 6 cases at 3 months, 9 cases at 6 months, and 8 cases at 1 year postoperatively. None of the patients developed BME at more than 1 year postoperatively. BME was observed on MRI in all cases at 2 years postoperatively.

CONCLUSIONS

The rate of occurrence of BME in the ACj after ARCR was 9.96 %. Patients with BME were significantly more likely to have ACj tenderness and positive cross-body adduction test. BME in the ACj often occurs within 6 months to 1 year after ARCR of small-to-medium rotator cuff tears, suggesting a relationship with postoperative functional improvement of the shoulder joint. The ACj should be considered as a potential site of persistent pain after ARCR for small-to-medium rotator cuff tears.

Combined Biologic Augmentation Strategies with Collagen Patch Graft, Microfractures, and Platelet Concentrate Injections Improve Functional and Structural Outcomes of Arthroscopic Revision Rotator Cuff Repair

BACKGROUND

Arthroscopic revision rotator cuff repair (ARRCR) is challenging. Biologic strategies seem to be promising. The aim was to evaluate the effectiveness of the combination of microfractures of the greater tuberosity, augmentation with collagen patch graft, and platelet concentrate injections in ARRCR.

METHODS

A retrospective comparative study was conducted on patients that underwent ARRCR with a minimum follow-up of two years. Patients in the augmentation group underwent ARRCR combined with microfractures, collagen patch graft, and postoperative subacromial injections of platelet concentrate. A standard rotator cuff repair was performed in the control group.

PRIMARY OUTCOME

Constant-Murley score (CMS).

SECONDARY OUTCOMES

disease-specific, health-related quality of life using the Disabilities of the Arm, Shoulder, and Hand (DASH) score; assessment of tendon integrity with magnetic resonance at least six months after surgery. Significance was set at p < 0.05.

RESULTS

Forty patients were included. Mean follow-up was 36.2 ± 8.7 months. The mean CMS was greater in the augmentation group (p = 0.022). No differences could be found for DASH score. Healing failure rate was higher in the control group (p = 0.002).

CONCLUSION

Biologic augmentation of ARRCR using a combination of microfractures, collagen patch graft, and subacromial injections of platelet concentrate is an effective strategy in improving tendon healing rate.

LEVEL OF EVIDENCE

retrospective cohort study, level III.

Capsular remnant in the rotator cuff footprint is a novel arthroscopic finding may indicate the etiology of the tear

PURPOSE

The aim of the study was that the capsule remnant is a common indicator of sub-acromial impingement syndrome and bursal side onset rotator cuff tears.

METHODS

Sixty-three patients with capsule remnants in the rotator cuff footprint (Study group) were detected, while the 53 patients did not have any remnant on the tendon footprint (Control group) between 2015 and 2020 were included. Demographic data of patients, such as age, gender, and operated side information were obtained from the archive files. Acromion type, presence of osteophytes in the acromioclavicular joint, Acromiohumeral Distance (AHD), Acromial Index (AI), Critical Shoulder Angle (CSA), and Coracoacromial Ligament (CAL) degeneration values were evaluated from preoperative MRI, radiographic images, and arthroscopic video recordings.

RESULTS

Severe stages of CAL degeneration were observed in 82.5% of the patients who had capsule remnant (p: 0.001). While type 2 acromion was found in 61.9%, and also acromioclavicular joint osteophyte was found in 58.7% of the patients in the study group. The mean AHD was 8.22 ± 1.56 mm in the study group and 9.2 ± 1.3 mm in the control group. The mean CSA was 43.3 ± 4.9 in the study group and 40.8 ± 4.2 in the control group. The AI was measured as 0.8 ± 0.1 in the study group and 0.8 ± 0.01 in the control group. As a result of these measurements, a statistical difference was found between the two groups in terms of type 2 acromion ratio (p < 0.001), presence of osteophytes in the acromioclavicular joint (p < 0.001), mean acromio-humeral distance (p < 0.001), critical shoulder angle (p = 0.004), and acromial index values (p < 0.001).

CONCLUSION

The findings of sub-acromial impingement syndrome were found to be more prominent in patients with full-thickness degenerative tear and findings of capsular remnant in the footprint. If the presence of the current finding is detected during glenohumeral arthroscopy, sub-acromial impingement syndrome should be examined in more detail to reveal the underlying cause and prevent a recurrence.