Article Text

Download PDFPDF

Systematic review
Double-row rotator cuff repair yields improved tendon structural integrity, but no difference in clinical outcomes compared with single-row and triple-row repair: a systematic review
  1. Lane N Rush,
  2. Felix Henry Savoie III,
  3. Eiji Itoi
  1. Department of Orthopaedic Surgery, Tulane University School of Medicine, New Orleans, Louisiana, USA
  1. Correspondence to Dr Felix Henry Savoie III, Department of Orthopaedic Surgery, Tulane University School of Medicine, 1430 Tulane Ave, #8632, New Orleans, LA 70112, USA; fsavoie{at}


Importance Techniques for arthroscopic rotator cuff repair are rapidly evolving. Single-row (SR) and double-row (DR) repair techniques have been well described, but recent studies have suggested the biomechanical superiority of transosseous-equivalent (TOE) DR repairs and triple-row (TR) repairs. However, there is conflicting literature regarding the clinical superiority of one technique over the other.

Objective To systematically review the best available evidence for arthroscopic repair of rotator cuff tears using SR, DR and TR techniques, and to identify predictors of patient functional outcomes and retear rates in each group.

Evidence review PubMed, Embase, Google Scholar and the Cochrane Database of Systematic Reviews were searched to identify all clinical papers describing arthroscopic repair of rotator cuffs using SR, DR or TR techniques.

Findings There were 10 papers that met our study criteria; they involved 580 patients. There was no significant difference in clinical outcomes between SR and DR repairs (p=0.57). Studies involving TR repairs were too underpowered to detect any difference in clinical outcomes between SR and DR repairs. The imaging confirmed retear rate of SR repairs was 30.3% (71/234), whereas DR repairs demonstrated a retear rate of 19.3% (41/212). The retear rate of TR repairs was 23.5% (8/34), but the low number of reported TR repairs did not allow us to make any conclusions regarding its superiority or inferiority compared with SR or DR. DR repairs demonstrated a statically significant decrease in retear rates compared with SR repairs (p=0.001). A subgroup analysis of retear rates in SR repairs versus TOE DR repairs demonstrated a trend towards improved structural integrity of the TOE DR repairs, but this was not statistically significant (p=0.07).

Conclusions and relevance There is no clear difference in clinical outcomes between SR, DR and TR techniques, but DR techniques demonstrate improved structural integrity at short-term follow-up.

Level of evidence IV.

  • rotator cuff
  • shoulder
  • rotator cuff repair
  • double-row TOE repair
View Full Text

Statistics from

What is already known

  • Prior meta-analyses of single-row and double-row repair techniques have failed to reveal differences in clinical outcomes.

  • There is evidence to suggest that healing rates and tendon structural integrity may be improved with double-row techniques.

What are the new findings

  • There is a paucity of evidence to support the use of the triple-row repair.

  • Double-row repairs demonstrate improved structural integrity compared with single-row constructs, but functional outcomes are similar between the groups.


Rotator cuff tears are among the most common disorders of the shoulder, and these injuries can result in significant pain, disability and dysfunction. The current improvements in technique and technology have made arthroscopic repair of rotator cuff tears the most common method of repair. Numerous time-zero biomechanical studies have been performed to determine the ideal repair construct. The ideal construct is one that provides strong initial fixation strength and minimal tissue gapping without overconstraining or strangulating the tissue.1 Despite the numerous biomechanical studies favouring double-row (DR) fixation, clinical results have not supported any difference in outcome between various fixation techniques, indicating that there may be additional factors that play a role in outcome.2–12

Single-row (SR) repair has long been considered the standard treatment option for medium to large rotator cuff tears, but biomechanical studies have questioned the ability of this construct to fully recreate the rotator cuff footprint. The DR technique addresses the shortcomings of the SR repair by providing a medial and lateral contact area to create a larger footprint and to increase the overall tendon–bone contact area.1 13 The restoration of the anatomic footprint of the rotator cuff is believed to provide improved healing and mechanical strength to the repair, and multiple biomechanical and clinical studies have shown good outcomes with this technique. The DR technique has been further modified to include the DR suture bridge technique, or the so-called ‘transosseous-equivalent’ (TOE).14–16 This repair construct has been proven to be more biomechanically sound than the traditional DR repair, providing a higher tendon–bone contact area, higher contact pressures and a higher load-to-failure.14–16

More recent studies have advocated for the use of triple-row (TR) rotator cuff repair. Proponents of this technique cite greater footprint contact area and stronger contact pressures compared with SR and DR repairs.17 Despite the biomechanical advantages of such a construct, there is very little clinical evidence to support the use of a TR repair.18 19

Several studies, systematic reviews and meta-analyses have attempted to compare SR and DR techniques.20–27 However, there have been no studies to collectively examine SR, DR and TR repair techniques. The purpose of this study was to perform a systematic review of the best available literature regarding these three methods of arthroscopic rotator cuff repair in order to compare their clinical outcomes, retear rates and complications. We hypothesised that the improved biomechanical strength of the TR repair technique would translate into improved functional outcomes and decreased retear rates compared with traditional DR techniques, TOE DR techniques and SR techniques.


Study design

The primary objective of this study was to perform a systematic review to evaluate the functional outcomes, radiographic outcomes, and retear rates of SR, DR and TR rotator cuff repairs. The search was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement. In June 2016, an electronic literature review was performed independently by two of the authors (LNR, FHS). All abstracts of arthroscopic SR, DR and TR repairs were reviewed. Searches were performed on PubMed, Embase, Google Scholar and the Cochrane Database of Systematic Reviews using the search criteria ‘single-row rotator cuff’, ‘double-row rotator cuff’ and ‘triple-row rotator cuff’ to identify all articles published from January 1980 to June 2016. The resulting abstracts were screened for inclusion, and the references from these articles were also screened for potential studies. We sought to identify clinical outcome studies that stratified their cohorts by repair type, tear size and operative technique, as well as those studies containing clinical outcome scores and postoperative imaging studies to evaluate the presence or absence of rotator cuff healing.

Study eligibility

Studies were eligible for inclusion if they presented the highest level of evidence for a given repair technique. For SR and DR repair techniques, this included eight randomised controlled trials. No level I or level II studies had detailed results of the TR technique, therefore making level IV studies the highest level of evidence available in the literature. Other inclusion criteria were arthroscopic repair and greater than 10-month minimum follow-up.

Exclusion criteria included studies with non-human subjects, biomechanical studies and studies with less than 10 months of follow-up.

Data extraction

All relevant data from the included studies were independently examined and extracted by the three authors. Data included study characteristics, patient demographics, surgical technique and details, functional outcomes, outcomes, retear rates and complications. Study characteristics included study type, enrolment numbers, year of publication and level of evidence. Patient demographics included age, sex and handedness. Surgical details included suture anchor configuration (SR, DR or TR), knot type, anchor type and number of anchors used. Rehabilitation details included duration of pillow-sling immobilisation and study-specific rehabilitation protocols, when available. Functional outcome scores were collected and included American Shoulder and Elbow Surgeons (ASES), Constant-Murley (Constant) and University of California Los Angeles (UCLA) scores. Retear rates and imaging outcomes included all imaging that reported retears at final radiological follow-up. Imaging studies included MRI, magnetic resonance arthrography, CT or ultrasound (U/S).

Risk of bias

A quality appraisal of the literature was undertaken to evaluate for a potential risk of bias in each study. The 10 studies included in this review were independently reviewed by the authors and were screened for performance bias, selection bias, detection bias, attrition bias and publication bias.

Statistical analysis

The data were pooled from 10 studies and a meta-analysis was performed using the RevMan software (Review Manager V.5.3, Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2011). Continuous data (UCLA, ASES scores and Constant scores) were reported as standardised mean differences and forest plots were generated. Retear rates were reported as ORs using a fixed-effect model. Statistical heterogeneity was calculated from patient demographic data. p Values <0.05 were considered statistically significant.


Study selection

A PRISMA flow chart is shown in figure 1. Using the aforementioned search method, 1097 titles were identified. A large majority of these citations were screened by two of the authors and were excluded, as these studies did not meet the inclusion criteria. The majority of articles documenting SR and DR techniques were excluded due to their low level of evidence, leaving eight level I studies comparing SR and DR techniques. The two studies of TR repair constructs were of level IV evidence; however, these studies represented the highest level of evidence regarding TR repairs and were included in this review. Overall, the meta-analysis was performed on 10 studies with 580 patients.

Figure 1

PRISMA 2009 flow diagram. PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses.

Patient demographics and study characteristics are listed in table 1. The surgical techniques, tear size, number of suture anchors, type of suture anchors and postoperative rehabilitation protocols are detailed in table 2, and table 3 classifies studies according to retear rate. The initial tear sizes, repair configurations, outcome measures and rehabilitation protocols varied significantly between studies.

Table 1

Patient demographics and study characteristics

Table 2

Surgical and rehabilitation protocols

Table 3

Studies by retear rate


A variety of outcome scores were recorded in the aforementioned studies (table 1). The Constant, ASES and UCLA scores were the most commonly used outcome scores across the studies, and as such were analysed to determine functional outcomes. Five studies examined differences in postoperative Constant scores between SR and DR repairs (table 4 and figure 2). There was no statistically significant difference in outcome scores between the groups (mean difference, −0.06, 95% CI −0.26 to 0.14, p=0.57).

Table 4

Clinical outcomes of single row and double row repairs

Figure 2

Constant scores. DR, double-row; SR, single-row.

Five studies examined differences in postoperative ASES scores between SR and DR repairs (figure 3). The five individual studies and meta-analysis did not demonstrate a statistically significant difference between the groups (mean difference, −0.09, 95% CI −0.30 to 0.11, p=0.40).

Figure 3

ASES scores. ASES, American Shoulder and Elbow Surgeons; DR, double-row; SR, single-row.

Three studies evaluated postoperative UCLA scores, and meta-analysis of these studies demonstrated a statistically significant improvement in postoperative UCLA scores in the DR repair cohort (mean difference, −0.30, 95% CI −0.55 to −0.06, p=0.01). Furthermore, no significant heterogeneity was found among the studies (Constant, p=0.61; ASES, p=0.40; UCLA, p=0.48) (figure 4).

Figure 4

UCLA scores. DR, double-row; SR, single-row; UCLA, University of California Los Angeles.

None of the TR repair studies used Constant, ASES or UCLA functional outcome scores, which did not allow us to make comparisons across groups for this cohort.

Retear rates

Table 3 and figure 5 detail the retear rate following SR or DR repair. Nine of the 10 studies reported the retear rate following rotator cuff repair. Of these studies, seven used MRI to verify cuff integrity postoperatively, one used ultrasound  and one used MRI and ultrasound. The pooled results demonstrate a retear rate of 30.3% in SR repairs (71/234), 19.3% in DR repairs (41/212) and 23.5% in TR repairs (8/34). Meta-analysis of SR and DR repairs demonstrated a statistically significant difference in tear recurrence between the two groups (OR, 2.12, 95% CI 1.35 to 3.32, p=0.001).

Figure 5

Retear rates of single row versus double row.

A subgroup analysis was performed to evaluate tear recurrences in traditional DR repairs and TOE DR repairs. Five of the seven studies reviewed used a traditional DR repair configuration, whereas Barber and Gartsman et al used a TOE DR repair.6 9 The retear rate was 20% (35/175) in the traditional DR cohort and 9.5% (6/63) in the TOE DR group (OR, 0.50, 95% CI 0.30 to 0.82, p=0.006; OR, 0.39, 95% CI 0.14 to 1.09, p=0.07) (figure 6 and figure 7, respectively).

Figure 6

Retear rates of single row (SR) versus traditional double row (DR).

Figure 7

Retear rates of SR versus TOE DR. DR, double row; SR, single row; TOE, transosseous-equivalent.

TR retear rates were 23.5% (8/34). These studies were too underpowered to detect any significant difference compared with SR or DR repairs.


We performed a systematic review of the best available evidence to compare SR, DR and TR rotator cuff repair techniques. The results of this meta-analysis demonstrated superiority of the DR construct over the SR in terms of maintenance of cuff structural integrity at 2-year follow-up, but no overall difference between SR and DR in terms of functional outcomes. The lack of consistent clinical outcome scores and the paucity of data regarding TR repairs did not allow for a thorough analysis of TR repairs, and no conclusions could be drawn regarding that repair technique.

There were nine studies that examined the structural integrity of the rotator cuff postoperatively.2–9 Most of these studies (7/9) used MRI evaluation.2–5 7–9 The results of this analysis demonstrated a significantly lower retear rate in the DR repair group. These results are consistent with recent studies that have also compared healing rates between the two groups. Biomechanical studies of DR repairs have demonstrated improved re-creation of the rotator cuff footprint, tighter contact forces at the tendon–bone interface and decreased gap formation.1 14 15 27 These biomechanical advantages appear to have translated into improved tendon healing rates at short-term follow-up. However, it must be noted that only two studies had a greater than 2-year follow-up, and longer-term studies may demonstrate a much greater incidence of tear recurrence.5–9

Recent biomechanical studies have proven the superiority of the transosseous DR repair over the traditional DR construct, demonstrating improved yield load and ultimate load in the TOE repair.13 However, no studies have demonstrated any difference in clinical results between the two techniques.28 A subgroup analysis was performed to evaluate these two methods of repair in terms of cuff integrity. TOE repairs demonstrated a lower total percentage of reruptures compared with traditional DR (9.5% vs 20%), but compared with SR repairs they did not demonstrate a statistically significant difference in retear rates (p=0.07). Standard DR repairs did, however, demonstrate superiority to SR repairs in terms of retear rates (p=0.006).

Functional outcomes varied greatly between studies. Only five studies recorded both ASES and Constant scores.2 5 7–9 ASES and Constant scores were similar between SR and DR repairs; however, the results of postoperative UCLA scores demonstrated a statistically significant advantage in DR repairs compared with SR repairs. These data could be skewed by the larger effect size of Carbonel et al 8 data among the three cohorts in this subgroup (160/262 patients, 60.6%). In that study, tears were stratified by size (either 1–3 cm or >3 cm). SR and DR groups demonstrated no difference in outcomes with tears between 1 and 3 cm, whereas the postoperative UCLA scores favoured the DR group in tears >3 cm.8

This method of reporting outcomes by tear size likely played a role in detecting a statistically significant difference between the two groups. A recent meta-analysis by Millett et al 21 also looked at functional outcome scores comparing SR and DR repairs. They found that studies that stratified results by initial tear size demonstrated differences in functional outcome scores between SR and DR repairs, but when all tear sizes were compared between SR and DR repairs there was no difference. They concluded that the evaluation of SR and DR repairs in tears of all sizes may be inadequate in detecting differences in clinical outcomes between groups.21

Hein et al 29 performed a recent systematic review of retear rates after arthroscopic SR, DR and TOE repairs. Their data collection included 2048 rotator cuff tears across 32 studies; these tears were then classified as small (<1 cm), medium (1–3 cm), large (3–5 cm) and massive (>5 cm). They found that both DR and TOE repairs had decreased retear rates compared with SR repairs in medium and massive tears. DR repairs had lower retear rates than SR for small tears, but TOE repairs did not differ significantly from SR repairs. They also found no statistically significant difference in retear rate in large tears (3–5 cm) when comparing SR, DR and TOE repairs. Based on these findings, they concluded that when comparing tears of any size, the DR and TOE had significantly lower retear rates than SR. However, when tear size was accounted for, only medium and massive tears treated with DR or TOE demonstrated superiority over SR repairs, and DR had lower retear rates than SR for tears sized less than 1 cm.29 These findings underscore the importance of consistent tear size reporting in future studies, as most reports vary greatly in their documentation and classification of tear size.

Due to the lack of similar recorded outcome measures between studies, we were unable to investigate clinical outcomes between traditional DR and TOE DR repairs. This is not surprising given the relative paucity of data comparing transosseous repairs with standard DR repairs. A level IV study from Toussaint et al 28 in 2011 demonstrated favourable short-term clinical outcomes and cuff integrity in TOE repairs, but this study lacked a control group. Park et al 30 reported improved pain and functional scores following TOE rotator cuff repairs, but they did not find any statistically significant difference between TOE and the standard DR repair group.


This study has several limitations. Only 10 studies were analysed, resulting in a relatively low number of patients for inclusion. Despite the high level of evidence available to examine SR and DR repairs, there are sparse data in regard to TR repairs; this left us unable to draw any conclusions in regard to TR repairs.

Another limitation was the lack of uniform objective outcome measurements and differing follow-up periods (between 10 months and 32 months). Longer follow-up could potentially lead to an increased number of retears and a decrease in clinical outcome score. Longer follow-up times are clearly needed.


Despite the fact that DR repair techniques clearly demonstrate a decreased rate of rerupture at early follow-up, there is no consistent difference in clinical outcomes between SR and DR repairs. Also, there is no superiority of TOE DR techniques over standard DR techniques in terms of clinical outcomes. At this time, there is little clinical evidence regarding TR repair techniques.


View Abstract


  • Competing interests None declared.

  • Provenance and peer review Commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.