Systematic Review: What Surgical Technique Provides the Best Outcome for Symptomatic Partial Articular-Sided Rotator Cuff Tears?

Abstract

Purpose

There is no consensus in the literature regarding the optimal surgical treatment of symptomatic partial rotator cuff tears. We attempted to determine the optimal surgical treatment for partial articular-sided rotator cuff tears through a systematic review of appropriate studies.

Methods

Medline®, PubMed, Ovid, and the cochrane register of controlled trials were searched for all studies published between January 1991 to March 2010 that used the key words “shoulder”, “partial rotator cuff tear”, “PASTA”, “articularsided rotator cuff tear”, “incomplete rotator cuff tear”, “arthroscopic” and “repair”. Inclusion criteria were studies (Level I to IV) that reported clinical outcomes in patients who had arthroscopic evaluation and arthroscopic or mini-open treatment of a symptomatic partial articular-sided rotator cuff tear. One of three surgical treatments was used: debridement with or without acromioplasty; transtendon arthroscopic repair; or tear completion with repair. Exclusion criteria included studies with over 50% overhead throwers or athletes, studies that involved an open approach to the rotator cuff without arthroscopy, and data presented in technical notes or review papers. Data abstracted from the studies included patient demographics, tear characteristics, surgical procedure(s), and clinical outcomes.

Results

Of 588 studies involving partial rotator cuff tears, 14 studies were identified which met our inclusion and exclusion criteria. All studies were Level IV retrospective case-series studies. Seven studies reported outcomes after rotator cuff debridement. Tear completion and repair was performed in three studies. Transtendon repair of a partial articular-sided rotator cuff tear was performed in three studies. Although different outcome measures were used, each study reported subjective and objective improvement postoperatively. One study compared outcomes in patients who underwent arthroscopic debridement versus another group where patients had tear completion and mini-open repair. Improved long-term results and decreased reoperation rates were reported in the tear completion and repair group.

Conclusion

On the basis of the available evidence, no single technique provides superior clinical outcomes. Level I and II comparison studies are needed to determine the optimal treatment of partial articular-sided rotator cuff tears.

Introduction

Although partial rotator cuff tears were first described by Codman in 1934¹, very little was known about their recognition and treatment until shoulder arthroscopy became popular in the 1990's. With improved pre-operative MRI imaging, our understanding of partial rotator cuff tears is increasing. Partial rotator cuff tears can be articular-sided (Partial Articular Sided Rotator Cuff Tears or PASTA lesions), bursal side or intra-tendinous (seen only on imaging studies). These tears can be secondary to internal impingement in young throwing and overhead athletes, or be due to trauma, and most commonly represent the early stages of degenerative rotator cuff tendon tearing. While there are many studies in the literature that describe the treatment of these tears, there is no general consensus on the optimal surgical treatment of symptomatic partial rotator cuff tears. We elected to limit our review to articular- sided tears.

Several factors are felt to be important in determining the treatment of partial tears. Partial tears in overhead athletes are felt to be secondary to either occult instability or loss of humeral rotation and thus require their own separate analysis and treatment. In the 1980's and early 1990's, surgical treatment of all types of partial rotator cuff tears consisted primarily of rotator cuff debridement.^2-4 In 1990, Ellman described a classification system that graded the partial-thickness tear by determining the amount of exposed articular footprint.⁵ Grade I tears were less than 3 mm deep, grade II tears were between 3 and 6 mm, and grade III tears were greater than 6 mm and involved more than half of the tendon thickness. As the popularity of arthroscopic management grew, several authors recommended debridement for tears less than 50% of the tendon thickness and repair for tears greater than 50% of the tendon thickness.^{3, 5, 6} Cadaveric studies have shown some variability in the medial-to-lateral distance of the supraspinatus footprint from 6.9 to 21 mm, and controversy exists regarding the amount of exposed footprint needed for a tear to be classified as a 50% partial-thickness tear.^7-9

Surgical treatment has evolved from simple arthroscopic debridement to transtendon arthroscopic repair and completion of the partial tear and repair of the newly created full-thickness tear. It is unclear which surgical treatment option provides the best clinical and functional outcome.^{2, 10-12} We designed a systematic review of published studies examining clinical outcomes after surgical treatment (debridement, completion and repair, or transtendon repair) of partial articular-sided rotator cuff tears in an attempt to answer this question. Our hypothesis was that clinical studies would not show a difference in subjective and objective clinical outcomes when comparing among three different surgical techniques.

Methods

Literature Search

We searched Medline®, PubMed, Ovid, and the Cochrane Register of Controlled Trials for all published literature from January 1991 to March 2010 using the following key words: “shoulder”, “partial rotator cuff tears”, “PASTA”, “articular-sided rotator cuff tear”, “incomplete rotator cuff tear”, “arthroscopic” and “repair”. General terms were used to avoid missing any potential studies. All studies were cross-referenced, and all abstracts and presentations were eliminated. Inclusion criteria included studies (Level I to IV) that reported clinical outcomes in patients who had arthroscopic evaluation and arthroscopic or mini-open treatment of a symptomatic partial articular-sided rotator cuff tear. In addition, studies needed to separate articular-sided tears from bursal-sided tears and use one of three surgical treatment options: debridement with or without acromioplasty, transtendon arthroscopic repair, or tear completion and repair. Exclusion criteria included bursal-sided partial tears, studies with over 50% overhead throwers or athletes, studies that involved an open approach to the rotator cuff without an arthroscopic component, and data presented in technical notes or review papers. Data abstracted from the studies included patient demographics, tear characteristics, surgical procedure, and clinical outcomes.

Data Abstraction

Data was abstracted from each of the 14 studies that met our inclusion and exclusion criteria. Demographic data included type of study, level of evidence, number of patients enrolled, mean age, number of shoulders, gender and arm dominance. The Ellman tear grade⁵ was collected for all studies using this classification system. Associated procedures were identified. Preoperative and postoperative clinical outcome scales used were the University of California, Los Angeles (UCLA)¹³, American Shoulder and Elbow Surgeons (ASES)¹⁴, Constant-Murley score¹⁵, L’Insalata score¹⁶, satisfactory Neer score¹⁷, and/or a visual analog pain score. We did not include postoperative imaging as only two studies reported these outcomes. The data was collected in table format using Excel computer data basing program (Microsoft, Redmond, WA). No statistical analysis was performed.

Results

Literature Search

Using the general search terms, 39,942 articles were found. When performing a search for keyword “partial rotator cuff”, 588 studies were identified and the abstracts of each of these articles were reviewed to identify those applicable to the study. Forty-eight studies were thought to be appropriate for analysis. After a full-text review of inclusion and exclusion criteria, ten studies were excluded because they were technique articles.^{12, 18-26}. Thirteen studies were excluded because they were review articles.^{5, 10, 11, 27-36} Biomechanical studies (2) of partial rotator cuff tears were also excluded.^{37, 38} Four studies involved over 50% athletes and were excluded.^{2, 39-41} One study was excluded because it compared classification systems⁹ and one study was excluded because it did not separate articular-sided partial tears from bursal-sided partial tears.⁴² Four studies were excluded because they involved an open approach to the rotator cuff and didn't involve arthroscopic or mini-open treatment of the partial rotator cuff tear.^{17, 28, 43, 44} There were 14 studies that met the final inclusion and exclusion criteria and were included in the final data analysis.^{3, 4, 21, 45-55}

Outcomes

Study design, level of evidence, total number of shoulders treated, mean age, tear grade, average follow-up time, and outcome measures used were include in our analysis (Tables 1-4). All studies were Level IV evidence and consisted of retrospective case series reviews. There were 497 partial articular-sided rotator cuff tears in the studies that met our inclusion criteria. Three-hundred and three (61%) shoulders had cuff debridement without repair. The average age of patients in the debridement group was 52 years and 94.5% of patients had a subacromial decompression. Transtendinous repair was performed in 64 (13%) shoulders with an average age of 47.5 years. Thirty of these patients (47%) had a concomitant subacromial decompression. Completion of the partial tear and repair was performed in 102 patients (21%) with an average age of 54 years. Subacromial decompression was performed in 100 patients (98%).

Table 1.

Surgical Transtendon Repair of Partial Articular-Sided Rotator Cuff Tears								Assessment of Clinical Outcomes
Source	Study	Level of Evidence	No. of Partial Tears	No. of Artic-Sided Tears	Mean age	Tear Grade	Average Follow-up	UCLA score pre-op/post-op	Analog pain scale pre-op/post-op	ASES score pre-op/post-op	Constant score pre-op/post-op	L'Insalata score post-op	Satisfactory Neer score post-op
Snyder et al.	Case Series	IV	31	31	42	NR	23 months	32 (post-op)	NR	NR	NR	NR	93%
Gartsmann et al.	Case Series	IV	111	85	52	Ellman I, II, III	32 months	NR	6.7/1.2	NR	NR	NR	NR
Cordaso et al.	Case Series	IV	25	19	57	Ellman I	53 months	NR	NR	NR	NR	89	NR
Cordaso et al.	Case Series	IV	52	44	53	Ellman II	53 months	NR	NR	NR	NR	90	NR
Park et al.	Case Series	IV	37	24	52	Ellman I, II	42 months	NR	6.2/1.1	38/88	NR	NR	NR
Kartus et al.	Case Series	IV	26	13	52	Ellman II	101 months	NR	NR	NR	72 (post-op)	NR	NR
Ozbaydar et al.	Case Series	IV	19	12	49	< 50%	NR	16.8/29 (p < 0.05)	NR	NR	NR	NR	NR
Liem et al.	Case Series	IV	46	46	59	Ellman I, II	50 months	NR	NR	37.4/86.6 (p < 0.001)	NR 87.7 (post-op)	NR	NR

Surgical Debridement of Partial Rotator Cuff Tears

There were seven studies that met our inclusion criteria in which the partial tear was treated with arthroscopic debridement of partial articular-sided rotator cuff tear (Table 1). All studies were case series and Level IV evidence. Five studies included only Ellman grade I or II tears (less than 50% rotator cuff insertion thickness). One study did not grade tears and one study included Ellman grade III tears.

In 1991, Snyder⁴ followed 31 partial articular-sided rotator cuff tears treated with debridement. Tear grade and depth were not recorded. Mean follow-up was 23 months (range, 10-43 months). Twenty-six patients (84%) had satisfactory outcomes (UCLA score greater than 28) and the average UCLA score was 33 (0-35). Twenty-nine patients (93%) had satisfactory results based on Neer's criteria. Results were similar in patients who had a subacromial decompression compared to those without.

In 1995, Gartsman and Milne³ reported on 111 articular-sided partial rotator cuff tears with an average age of 42.5 years. They divided patients into three groups: impingement, instability, and trauma. The partial tear was thought to be secondary to impingement in 85 shoulders. Forty-four of these tears were Ellman grade I, 29 were grade II, and 12 were grade III. The tear was arthroscopically debrided and an acromioplasty was performed in every case. Average follow-up was 32.3 months (range, 26 to 84 months.). Subjective pain scores improved in this group from 6.7 to 1.2, but no shoulder outcome scores were reported.

In 2002, Cordasco⁵⁰ reported on 162 patients who had arthroscopic debridement and acromioplasty for a symptomatic partial rotator cuff tear. Included were 44 shoulders with Ellman grade II partial articular-sided tears and 19 articular-sided Ellman grade I tears. Mean follow-up was 4.5 years (range, 2-10 years). Failure of treatment was defined by a L'Insalata score of less than 70 at follow-up. No failures were noted in the grade I tears while 2 out of 44 (5%) of the grade II tears were considered failures.

In 2003, Park⁵² followed 24 shoulders with Ellman grade I and II tears. At the time of arthroscopy, the average articular-sided tear depth was 4.5 mm. All tears were treated with debridement and subacromial decompression. Average follow-up was 3.5 years (range, 2-6 years). ASES scores improved from an average of 38 preoperatively to 88 postoperatively. Pain scores improved in all patients.

In 2006, Ozbaydar⁴⁸ reported on 12 patients with partial articular-sided tears with less than 50% of the footprint, treated with arthroscopic debridement and bursectomy. Fourteen of 19 patients had a subacromial decompression (if a subacromial spur was present). Average follow-up was not recorded. Mean UCLA scores improved from 16.8 preoperatively to 29.0 postoperatively. They did not separate partial bursal-sided tears from partial articular-sided tears when reporting outcomes.

In 2006, Kartus⁵¹ examined 13 Ellman grade II articular-sided rotator cuff tears. All patients underwent debridement and acromioplasty. The mean follow-up was 101 months (range, 60-128). The average Constant score at follow-up was 72 (range 35-97). No pre-op score was recorded. Abduction strength was similar on the operative side compared to the contralateral side.

In 2008, Liem⁵³ evaluated 46 patients with Ellman grade I or II tears. Twenty-six Ellman grade I tears were managed with acromioplasty alone and 20 Ellman grade II tears were treated with debridement and acromioplasty. At an average follow-up of 50 months, the ASES score improved from 37.4 to 86.6. Mean postoperative Constant score was 87.6 points. No significant differences were found when comparing the outcomes of grade I and grade II tear treatment.

Transtendon Repair of Partial Rotator Cuff Tears

There were three studies that met our inclusion criteria and involved an arthroscopic transtendon repair of a partial articular-sided rotator cuff tear (Table 2).^{45, 46, 54} All studies were Level IV evidence. One study only included Ellman grade III tears and two studies included both Ellman grade II and III tears.

Table 2.

Surgical Transtendon Repair of Partial Articular-Sided Rotator Cuff Tears								Assessment of Clinical Outcomes
Source	Study	Level of Evidence	No. of Partial Tears	No. of Artic-Sided Tears	Mean age	Tear Grade	Average Follow-up	UCLA score pre-op/post-op	Analog pain scale pre-op/post-op	ASES score pre-op/post-op	Constant score pre-op/post-op	L'Insalata score pre-op/post-op	Satisfactory Neer score post-op
Ide et al.	Case Series	IV	17	17	42	Ellman III	39 months	17.3/32.9 (p < 0.01)	NR	NR	NR	NR	NR
Tauber et al.	Case Series	IV	16	16	NR	Ellman II, III	NR	15.8/32.8 (p < 0.01)	7.9/1.2 (p < 0.01)	NR	NR	NR	NR
Castricini et al.	Case Series	IV	31	31	53	Ellman II, III	33 months	NR	NR	NR	44.4/91.6 (p < 0.01)	NR	NR

In 2005, Ide⁵⁴ evaluated 17 patients who underwent arthroscopic transtendon repair of Ellman grade III partial articular-sided tears using suture anchors. Subacromial decompression was not recorded. Mean follow-up was 39 months (range, 25-57 months). UCLA scores improved from 17.3 preoperatively to 32.9 postoperatively. Japanese Orthopaedic Association Shoulder Scores improved from 68.4 preoperatively to 94.8 postoperatively. No complications were reported and only one patient had a fair or poor result.

In 2008, Tauber⁴⁶ reported on 16 patients who underwent arthroscopic transtendon repair of partial articular-sided tears using trans-osseus tunnels. Seven patients had an Ellman grade II tear and nine patients had a grade III tear. Average follow-up was not recorded, but all patients were evaluated at a minimum of 18 months. Mean UCLA scores improved from 15.8 preoperatively to 32.8 at follow-up. Pain scores decreased from 7.9 to 1.2. There were no differences in outcome when comparing grade II and grade III tears.

In 2009, Castricini⁴⁵ examined 33 patients who underwent trantendon arthroscopic repair of a partial articular-sided rotator cuff tear. They included both Ellman grade II and III tears and used suture anchors for repair as described by Lo and Burkhart.²⁵ Seven of 33 patients underwent subacromial decompression and 8 patients had either biceps tenodesis or tenotomy. Mean follow-up was 33 months (range, 26-45). The Constant score increased from 48.2 preoperatively to 91.6 postoperatively.

Surgical Completion and Repair of Partial Rotator Cuff Tears

There were three studies which met our inclusion criteria and involved arthroscopic completion and repair of a partial articular-sided rotator cuff tear (Table 3).^{47, 49, 55} All studies were Level IV evidence. All three studies included only partial tears greater than 50% of the insertion thickness.

Table 3.

Surgical Takedown and Repair of Partial Rotator Cuff Tears								Assessment of Clinical Outcomes
Source	Study	Level of Evidence	No. of Partial Tears	No. of Artic-Sided Tears	Mean age	Tear Grade	Average Follow-up	UCLA score pre-op/ post-op	Analog pain scale pre-op/ post-op	ASES score pre-op/ post-op	Constant score pre-op/ post-op	L'Insalata score post-op	Satisfactory Neer score post-op
Deutsch A	Case Series	IV	41	33	49	Ellman III	38 months	NR	6.5/0.8 (p < 0.001)	42/93 (p < 0.001)	NR	NR	NR
Porat et al	Case Series	IV	36	36	60	>50%	42 months	17.2/31.5 (p <0.05)	NR	NR	NR	NR	NR
Kamath et al.	Case Series	IV	42	33	53	>50%	39 months	NR	6.5/2.7 (p < 0.001)	47.0/82.7 (p < 0.001)	NR	NR	NR

In 2007, Deutsch⁵⁵ evaluated the outcome of 41 Ellman grade III partial-thickness tears; 33 were articular-sided and eight were bursal-sided. Tear thickness was between 60-90% of the tendon thickness with a mean tear depth of 75% or 9 mm. All patients underwent completion of the partial tear and suture anchor repair. Thirty-nine of 41 patients had a subacromial decompression. Mean follow-up was 38 months (range, 24-50). Average ASES scores improved from 42 preoperatively to 93 postoperatively. Pain level and satisfaction scores were significantly improved in all patients.

In 2008, Porat and colleagues⁴⁹ examined 36 patients with partial articular-sided tears greater than 50% who were treated with arthroscopic completion and repair. They used 14 mm as the average footprint thickness and measured the exposed footprint intra-articularly. In these tears, a marking suture was placed and the tear was then evaluated in the subacromial space. Once the marking suture was discovered, a shaver was used to complete the tear. Single- or double-row suture anchor repair was then performed depending on how the tear reduced to the footprint. Thirty-four of 36 patients received a subacromial decompression. Patients were followed for a minimum of two years (range, 24-73 months). They reported 83% good to excellent outcomes. The UCLA score improved from 17.25 preoperatively to 31.47 postoperatively.

In 2009, Kamath et al⁴⁷ reported on 42 cases which underwent arthroscopic completion of a partial-thickness tear greater than 50% (5-6 mm) to a full-thickness tear with subsequent repair. There were 33 articular-sided tears and nine bursal-sided tears. All patients had a subacromial decompression. 17 shoulders had concomitant procedures in addition to the rotator cuff repair. At the time of arthroscopy, if the tear involved greater than 50% of the tendon thickness, the tendon was split longitudinally to verify the thickness. The tear was completed and either a single-row or double-row suture anchor repair was performed. The average ASES score improved from 46.1 preoperatively to 82.1 postoperatively. Overall patient satisfaction was 93%.

Debridement vs. Mini-open Repair

In 1999, Weber⁵⁶ compared 32 patients with Ellman grade III tears who had arthroscopic debridement and acromioplasty to 33 patients with Ellman grade III tears who had acromioplasty, tear completion and mini-open repair. Treatment groups were not randomized but were determined by preoperative doctor-patient discussion. There were 29 articular-sided tears and three bursal-sided tears in the debridement group, and 28 articular-sided tears and five bursal-sided tears in the repair group. All patients had a subacromial decompression. Follow-up averaged 47.7 months in the debridement group and 38.1 months in the repair group. The average postoperative UCLA scores in the debridement group were statistically less (22.7) when compared to the repair group (31.6). There were six re-operations in the debridement group because of post-operative pain, and none in the repair group. Weber concluded that debridement and acromioplasty are not adequate treatment of most grade III partial tears.

Discussion

When analyzing partial rotator cuff tears, there are wide ranges in patient age, tear pattern, tear depth, location, and grade. Younger overhead athletes often develop articular-sided partial rotator-cuff tears from internal impingement, and middle-aged patients are more likely to have partial tears secondary to external impingement or intrinsic tendon degeneration.^{10, 11, 36, 40} In order to focus our review, we identified studies that reported clinical outcomes in patients who had arthroscopic evaluation and arthroscopic or mini-open treatment of a symptomatic partial articular-sided rotator cuff tear. We excluded studies that included overhead athletes or bursal-sided tears. The three primary surgical techniques for managing partial articular-sided rotator cuff tears are debridement, transtendon repair, or takedown and repair. This systematic review focused on comparing clinical outcomes among these groups. All studies were Level IV evidence and consisted of retrospective case series. However, we did identify one retrospective non-randomized comparison study.

Table 4.

Surgical Takedown and Repair of Partial Rotator Cuff Tears						Assessment of Clinical Outcomes
Source	Comparison	Level of Evidence	No. of Shoulders	Mean age	Tear Grade	Articular-Sided Partial Tear	Average Time to Clinical Assessment	UCLA post-op	Reoperation
SC Weber	Debridement of Partial Rotator Cuff Tear	IV	32	49	III	91%	48 months	22.7	6
	Mini-open Repair of Partial Rotator Cuff Tear	IV	33	46	III	85%	38 months	31.6	0

Comparing clinical outcomes among the three groups was difficult because different outcome instruments were used. In the debridement group, two of seven studies used UCLA scores for outcome assessment, two studies used ASES scores, two studies reported Constant scores and two studies used visual analog pain scores. One study used the L'Insalata shoulder outcome questionnaire. In the transtendon repair group, two of three studies reported UCLA scores postoperatively and one of three studies used the Constant score. In the studies involving takedown and repair, one of the three studies reported UCLA scores and two of three studies used ASES and visual pain scores.

When comparing the three groups, UCLA, ASES, and Constant scores were similar (Tables 1-3). The average UCLA score was 27.9 (range, 22.7-32) in the debridement group, 32.85 (range, 32.8-32.9) in the transtendon repair group, and 31.5 in two studies of takedown and repair. The mean Constant score was 79.9 in the debridement group. Only one of seven studies reported Constant scores after either transtendon repair or rotator cuff takedown and repair. Average ASES score was 87.3 in the debridement group and 87.9 in the takedown and repair group. No study reported ASES scores in the transtendon repair group. No single surgical technique provided superior clinical outcomes. In Weber's comparison study, he reported significantly higher UCLA scores in the repair group compared to the debridement group for Ellman grade III partial tears.⁵⁶ However, patients were not randomized and the type of surgical treatment was determined preoperatively by the surgeon.

Current teaching on arthroscopic treatment of partial articular-sided rotator cuff tears suggests that takedown and repair or transtendon repair should be considered when the tear depth exceeds 50%, and debridement can be performed if the tear is less than 50%. Five of six studies in the rotator cuff debridement group included only Ellman grade I and II tears (less than 50% thickness). In contrast to the debridement group, the studies in the two repair groups almost exclusively consisted of grade III partial tears (greater than 50%).^{47, 49, 54, 55} Selection bias showed studies with rotator cuff repair were done on higher-grade partial tears and studies examining debridement were performed on lower-grade tears.

In addition, ten of the 14 studies used the classification system described by Ellman.⁵ Grade I tears were less than 3 mm deep, grade II were between 3 and 6 mm, while grade III were greater than 6 mm and involved more than half of the tendon. Most studies made the assumption that grade III tears involved over 50% of the tendon insertion. However, cadaveric studies have shown variability in the medial-to-lateral distance of the supraspinatus footprint from 6.9 to 21 mm.^7-9 Equating a tear depth of 50% with 6 mm of exposed footprint may not be accurate in all cases.

Recommendation for Partial Articular-Sided Tears <50%

Despite the selection bias in the literature and lack of Level I, II, or III studies, arthroscopic debridement and subacromial decompression do appear to provide improved clinical outcomes and decreased pain for partial articular-sided tears of less than 50%.^{48, 50-53} Unfortunately, there are no studies comparing debridement versus repair in this group. Although some surgeons anecdotally report greater strength improvement with repair compared to debridement, there is no evidence at the current time to support this.

Recommendations for Partial Articular-Sided Tears >50%

In one study, improved clinical outcomes were achieved with repair compared to debridement for partial articular-sided tears greater than 50%.⁵⁶ There were three studies involving transtendon repair and three studies involving takedown and repair of partial articular-sided tears greater than 50%.^{45-47, 49, 54, 55} Each case series reported satisfactory clinical outcomes postoperatively. There have been a considerable number of surgical technique articles describing various ways to perform a transtendon rotator cuff repair for a partial articular-sided tear.^{12, 25} Unfortunately, there are no prospective studies comparing transtendon repair versus takedown and repair. Although theoretical advantages exist in performing a transtendon repair, we are unable to make any conclusions that superior clinical outcomes are achieved when compared to takedown of the tear and repair.

Conclusions

On the basis of the available evidence, we were unable to conclude which of the three operative procedures for symptomatic partial articular sided rotator cuff tears results in the most favorable outcome. Every study that met our inclusion/exclusion criteria was a retrospective case series with differing outcome measures that did not allow making comparisons between the studies. Most studies reported satisfactory outcomes after debridement, transtendon repair, or tear takedown and repair. However, every study was a retrospective case series. In addition, different clinical outcome measures were used in each study making it difficult to draw conclusions. Future studies are recommended that prospectively evaluate repair versus debridement, or transtendon repair versus takedown and repair in a homogenous group of partial rotator cuff tears of the same depth and location.