Return to Preinjury Levels of Participation After Superior Labral Repair in Overhead Athletes: A Systematic Review

Data Acquisition

Search Strategy

We accessed the CINAHL, MEDLINE, and SPORTDiscus online databases to search the terms listed in Table 1. Terms were searched individually and then combined into distinct Boolean phrases. The search limits were human studies, English-language articles only, and years 1972 to 2013. In some cases, review of a retrieved article revealed potential citations not originally identified in the electronic search. A total of 215 articles were retrieved from the systematic search (Figure).

Table 1.  Systematic Review Search Terms Using Patient or Problem, Intervention, Comparison, Outcome Method

View Fullscreen

View Figure

• Download as Powerpoint
• Download Figure

Data Extraction and Analysis

For this systematic review, we collected and recorded from each article the percentage of athletes who returned to full participation so that we could calculate the primary measure of interest, odds of return to full participation, for this systematic review. For studies that provided data regarding return to participation for dichotomous populations or surgical procedures, we constructed 2 × 2 contingency tables in Excel (version 2007; Microsoft Corporation, Redman, WA) to calculate the odds of return to full activity for each comparative group. We calculated the odds of full return to condense the reported rates of return into a more definable chance of success or failure after surgery. When return to participation was reported only as a percentage (without the raw data), we multiplied the percentage of participants who returned by the total number of participants in each subgroup to complete the 2 × 2 table. Odds for a distinct group were computed by dividing 1 return-to-participation variable by another.

For example, in 1 study, 4 overhead athletes returned to full participation, and 14 overhead athletes returned to limited activity; however, 10 nonoverhead athletes returned to full participation, with only 6 returning in a limited capacity.⁵ To calculate the odds for each group of athletes, we simply divided the occurrences for each group (overhead athletes: 4 with full return divided by 14 with limited return = 0.3; nonoverhead athletes: 10 with full return divided by 6 with limited return = 1.7). When the calculated odds were 1.0 or less, the resultant interpretation was that the chance of not returning to full activity after surgery was greater. Conversely, when the odds were greater than 1.0, the interpretation was that the chance of returning to full activity after surgery was greater.⁶

The odds were needed to calculate the final odds ratios (ORs) that would compare the odds of the event of interest (returning to full activity) between the specified groups (eg, overhead versus nonoverhead athletes, throwing versus nonthrowing athletes) for each study. Therefore, using the example given, the OR calculation of 1.7/0.3 = 5.8 would be interpreted as nonoverhead athletes being 5.8 times more likely to return to full activity after surgery than overhead athletes. An OR greater than 1.0 would indicate a possible statistical relationship or association between the variables.⁶ To confirm if a statistical association existed, 95% confidence intervals (CIs) were constructed. If the 95% CI contained the value 1.0, then no statistical association was present at an α level of .05.

Assessment of Quality and Bias for Retained Articles

We used the intervention study appraisal score sheet described by MacDermid⁷ to assess the quality of each article retained for the systematic review. This instrument was selected because it aligned well with case series reports that characterized most of the retained articles' designs. The assessment sheet comprised 24 questions divided among 7 subheadings (study question, study design, subjects, intervention, outcomes, analysis, and recommendations); each question could receive a score ranging from 0 to 2 for a maximum score of 48 points. This scoring sheet was modified to a binary (yes = 1 or no = 0) scoring system, yielding a possible 24 points to better illustrate commonalities among retained studies. Two of the authors (A. S., N. M.) individually reviewed and appraised each retained article. Upon completion of all appraisals, the 2 authors met to compare their results. When the authors agreed on an individual article's score, they accepted it. When the authors disagreed on a score, they revisited the article and discussed the discrepancy in score until agreement was reached. If agreement could not be reached, a third author (T.L.U.) was consulted to serve as the final authority. After the critical appraisal, the appropriate strength of recommendation was selected using the Strength of Recommendation Taxonomy, which consists of ratings of A, B, or C (high to low).⁸

We determined the risk of bias per the recommendations from the PRISMA guidelines.^9,10 The retained articles were assessed for different types of bias, including selection, performance, attrition, detection, reporting, and recall.¹¹ Each type of bias was determined by the primary author per the definitions provided by the Agency for Healthcare Research and Quality (Table 2).

Table 2.  Assessment of Risk Criteria for Case Series Reports

View Fullscreen

Return to Preinjury Levels of Participation

Isolated Superior Labral Repair

The proportion of athletes who returned to their preinjury levels of participation after isolated superior labral repair was reported to range from 22% to 92% for the 5 studies reviewed.^5,12–15 In 2 studies,^5,12 the authors subdivided return to participation into 2 levels (full return, limited return) based on perceived functional limitation; full return for overhead or throwing athletes ranged from 22% to 37%, but 63% to 78% of athletes returned with limitation. These authors also reported that a higher percentage (48% to 63%) of nonoverhead or nonthrowing athletes returned in full, whereas only 38% to 45% returned with limitation.^5,12 In 3 other studies,^13–15 investigators divided return to participation into full return and no return. When comparing baseball players with other overhead athletes, return to full participation was 38% for the former and 75% for the latter.¹⁵ When comparing the return to full activity for 2 repair techniques,¹⁴ we observed that suture anchor repairs were superior to transglenoid suture repairs; however, little difference existed in the rate of return to full activity for overhead athletes between the techniques (56% for suture anchor and 67% for transglenoid suture repair). When comparing specific sports, Yung et al¹³ demonstrated a 92% return to participation for tennis, handball, and badminton athletes.

Using the originally reported return-to-participation rates, we generated ORs for full return for 4 of 5 studies (Table 5).^5,12,14,15 Authors of 3 studies found that nonbaseball players (OR = 5.0; 95% CI = 0.8, 33.2),¹⁵ nonthrowers (OR = 2.3; 95% CI = 0.4, 12.5),¹² and nonoverhead athletes (OR = 5.8; 95% CI = 1.3, 26.2; P < .05)⁵ had at least a 2 times greater chance of returning to full activity after isolated superior labral repair. Yung et al¹³ noted return-to-participation rates of 92% for 12 of 13 overhead athletes and 100% for 3 of 3 patients classified as nonoverhead athletes (n = 2) or nonathletes (n = 1). We could not calculate the odds because the denominator was zero, nullifying the calculation and limiting the comparison between overhead and nonoverhead athletes.

Table 5.  Return-to-Participation Odds and Interpretations for Athletes With Isolated Superior Labral Repair

View Fullscreen

Concurrent Shoulder Procedures

Researchers^16–21 examining various types of overhead athletes after superior labral repair with concurrent procedures reported return to preinjury levels of participation ranging from 41% to 86%. Concurrent procedures consisted of debridement of partial-thickness rotator cuff tear or the labrum, subacromial decompression, bursectomy, or distal clavicle excision. Follow-up time ranged from 12 to 120 months, with the populations from all articles including baseball players. Researchers who categorized return to participation with limitation reported that 14% to 37% of athletes returned with limitation,^16,17,21 whereas those who focused on no return reported that 17% to 100% of athletes did not return to sport after surgery.^18–21

Using the originally reported return-to-participation rates, we generated ORs for 4 of the 6 studies that focused on labral repair with concurrent soft tissue debridement (Table 6).^17–19,21 The subgroupings for each study varied: authors comparing baseball and “other” sports,¹⁷ overhead and nonoverhead athletes,^18,19 and only overhead athletes who underwent surgery for superior labral tears and superior labral tears with concomitant rotator cuff injury.²¹ Ide et al¹⁷ reported that both baseball and “other” athletes had positive chances of returning to full participation (all odds > 1.0); however, the other group had a greater chance of successful return to full activity (OR = 3.5; 95% CI = 0.8, 16.3) than did baseball athletes. The 2 investigations also showed that overhead athletes (odds = 1.3–2.5) and nonoverhead/other athletes (odds = 2.0–5.0) both had positive odds for returning to full activity after superior labral repair with soft tissue debridement, but again the ORs favored the nonoverhead (OR = 1.5; 95% CI = 0.4, 6.0)¹⁹ and other (OR = 2.0; 95% CI = 0.2, 19.9) groups.¹⁸ These observations were not different, as each 95% CI contained the value 1.0. When comparing surgical procedures among only overhead athletes, most of whom were baseball players (20 of 23 patients), Neri et al²¹ observed an association for athletes who underwent isolated superior labral repair. The patients with isolated repairs had a much greater chance of returning to full participation (OR = 28; 95% CI = 2.4, 323.7; P < .05) than those who underwent concurrent labral repair and soft tissue debridement.²¹

Table 6.  Return-to-Participation Odds and Interpretations for Athletes With Concurrent Superior Labral Repair and Soft Tissue Debridement

View Fullscreen

Critical Appraisal Results

The results of the critical appraisal outcomes are presented in Table 7. Of 24 possible points available as part of the binary assessment, the scores of the 5 studies specific to isolated superior labral repair ranged from 10 to 15 total points (42% to 63% of available points). The 6 articles that described superior labral repair with concurrent soft tissue debridement scored from 11 to 17 points (42% to 70%) of the 24 possible points. The wide variation of return-to-participation findings among the 11 studies suggests that the Strength of Recommendation Taxonomy recommendation would be C due to the quality of the studies with level 3 (case series) evidence.⁸

Table 7.  Critical Appraisal Results for Studies Retained for Systematic Review

View Fullscreen

Risk of Bias Results

The risk-of-bias assessment revealed that recall bias was the most prevalent across all studies (100%) because return to participation was assessed 1 to 2 years after patients were discharged from formal care. Both detection bias and selection bias were evident in 73% of the studies reviewed (Table 8).

Table 8.  Possible Risk of Bias Results

View Fullscreen

Population Demographics

The mean age of the participants across all studies ranged from 24 to 36 years. This range of ages indicated that the reported return-to-participation rates cannot be extrapolated to younger age groups, which is a concern because younger athletes (age < 18 years) have a greater chance of advancing to additional stages of participation (collegiate or professional levels) than athletes already performing at those upper levels of competition. If operative intervention is recommended for a younger athlete, supplying preoperative return-to-participation prognoses may be helpful in determining if undergoing the corrective procedure will allow return to the current level of activity and advancement to higher stages of participation. However, it is possible that factors beyond patient age, such as variations in healing rates, skill levels, financial considerations, and internal or external motivators, can influence the decision to have surgery.^22–24 Eisner et al²⁵ reported a return-to-participation rate of 70% after partial-thickness rotator cuff repair in patients less than age 19 years; however, the article could not be included in this review because it did not meet all inclusion criteria.

The populations across all studies reviewed were heterogeneous, as they comprised multiple types of athletes from various sports with subgroups defined as overhead/nonoverhead, overhead/other, or throwing/nonthrowing athletes. Grouping throwing/overhead sports, such as baseball, softball, swimming, tennis, and volleyball, and nonthrowing/nonoverhead sports, such as football, karate, and weightlifting did not provide a clear estimate of return to participation because individual sports have different biomechanical requirements and demands. Combining athletes from multiple sports into 1 distinct group limits a clinician's ability to delineate return-to-participation prognoses for individual athletes, creating a gap in the knowledge. As such, the calculated ORs need to be interpreted in context for each study.

Another concern that we identified was an overwhelming lack of prospective data collection related to athletic performance and playing status and a lack of reporting of these same factors in the postoperative data. Detection, recall, and selection bias were the most common types of bias present within the studies, likely due to the retrospective case series design. Outcome measures were rarely selected prospectively, and the assessors who performed follow-up examinations were not blinded to the intervention, creating the possibility of detection bias. All individuals were asked orally at postoperative follow-up if they had returned to their preinjury levels of participation, which subjected the responses to recall bias and individual patient perception given that postoperative follow-up occurred between 1 and 10 years. Selection bias was also evident due to a lack of matching or stratification. The identified biases weaken the strength of the information derived from each article because the level of return to participation was not consistently known (full return, limited return, no return, or return with or without pain), thus lessen the usefulness of the reported return-to-participation rates.

Surgical Considerations

Orthopaedic surgeons are charged with determining if apparent shoulder pathologic conditions are clinically important. The decision to recommend or not recommend surgical intervention depends on the information derived from subjective history and objective testing, as well as patient goals and expectations. However, we observed that the likelihood of returning to preinjured levels of participation was inconsistent across studies, indicating that preoperative discussions focused on the rate of return to activity would be difficult for clinicians to have with their patients.

The ORs for returning to full, preinjured levels of participation after isolated superior labral repair favored nonoverhead athletes; however, some overhead athletes did return to full preinjured levels. This observation suggested that some type of return is possible, either in a full or limited capacity, for some overhead athletes after repair, although the exact technique of repair was not standardized. However, only 1 study⁵ demonstrated an association, with researchers finding that nonoverhead athletes were 5.8 times more likely to return to full activity after superior labral repair. However, clinicians should interpret this result with caution, as the 95% CI was rather wide (1.3, 26.2). Furthermore, surgically debriding the surrounding soft tissue appeared to confound the return-to-participation rates for overhead athletes, as the odds for returning to full participation after these procedures were all greater than 1, meaning more likely to return to full participation. This observation is in contrast to the isolated repair studies, for which all odds were less than 1 for the overhead-athlete groups. To further examine why greater odds of return to participation occurred with more surgery, 3 possible explanations were identified: unequal group sizes, surgical details, and clinical experience. First, unequal group sizes were identified within and between the studies of isolated repair and the studies that included concurrent debridement (from which odds could be calculated). The concurrent-debridement studies had an average of 8.5 more patients in the overhead-athlete groups than the same patient groups in the isolated-repair studies.^16–21 The increased number of patients undergoing labral repair with concurrent debridement inflated the odds of successful return, which could explain why the chance of full return was greater after a procedure that involved treatment to more tissue. Conversely, the data from Neri et al²¹ suggested a 28 times greater chance of returning to full activity after isolated superior labral repair than after concurrent procedures. However, in that study, twice as many patients were in the isolated superior labral repair group and relatively few patients did not return to full activity, which could account for the difference in observations between the procedures. Second, clinical details, including anatomical considerations (eg, severity of injury, tissue integrity), surgical considerations (placement and amount of hardware, amount of tissue debrided), individual patient differences, and postoperative rehabilitation details, were either not reported or were limited in description. The lack of consistent details among studies possibly limited the ability to draw accurate conclusions regarding return to activity after superior labral repair. Third, differences in clinician experiences and skills are neither standardized nor measurable. Clinicians do not universally agree about how to optimally repair superior labral injuries, suggesting that the optimum method for treating labral pathologic conditions in the symptomatic overhead athlete is not fully understood.²⁶

To properly compare results across various studies, we must know the surgical details that were employed to address the superior labral lesions in each cohort of patients. In this review, all researchers reported the techniques used in the surgical procedures; however, gaps among the studies included failure to report the number of anchors used and their placement on the glenoid.^12,14,16,20

Anchor quantity typically is determined during surgery based on the size of the lesion. Of the 9 studies in which authors reported anchor quantity, the total number of anchors used per patient ranged from 1^5,15,17–19 to 4,^12–14 with a mean of 2.3.²¹ The variation in the number of anchors used within and between studies confounded the return-to-participation rates because not all superior labral lesions were the same size and, therefore, were not fixated similarly.

Of similar importance is anchor placement as it relates to both anatomical fixation and restoration of biomechanical capabilities.²⁷ Inappropriately placed anchors can potentially “strangle the biceps tendon”¹⁹ and not eliminate the dysfunction caused by the lesion.²⁷ Similar to the discrepancy noted about the quantity of anchors used, variations were present among the 8 studies^{5,12,13,16–19,21} in which researchers reported the location of anchor placement. With return-to-participation rates ranging from 22% to 92%, functional restoration for overhead athletes was possibly affected by the location of anchor placement.

Additionally, surgical portal placement has been shown to affect the outcomes for superior labral repairs. Researchers²⁸ recently reported that poor portal placement during primary superior labral repair can result in full-thickness rotator cuff tears of the supraspinatus. Anchor quantity and placement and portal placement are thought to contribute to restoration of labral function²⁹ and are necessary information for critically analyzing outcomes, such as patient satisfaction and return to participation.

Rehabilitation

When exploring the reasons why return-to-participation rates for overhead athletes were not more consistent or successful, we should also examine the rehabilitation components. The postoperative rehabilitation guidelines used for each case series were consistent in design, with progressive strengthening and range-of-motion exercises used throughout the rehabilitation process. However, no researchers reported the critical details of the rehabilitation program, including the specific exercises used, patient compliance with the protocol, the duration (total number of visits) or frequency of rehabilitation, exercise dosage, or use of a home program. A detailed outline describing each of these rehabilitation components would allow for better dissemination of these confounding factors, as 1 or more of the components could potentially have affected the postoperative outcome.

Limitations

Our study had limitations, so clinicians need to exercise caution in interpreting these results. The data were derived from retrospective studies with only moderate quality and evident biases. Whereas retrospective studies permit reviewers to establish relatively clear inclusion and exclusion criteria, adequate follow-up time, and definitions of primary and secondary outcomes, each case series was limited by (1) not having a prospective assessment of playing status or functional performance before or immediately after injury diagnosis, (2) no determination of adequate sample size, (3) lack of a thorough statistical analysis, and (4) merging groups of athletes from a variety of sports with a large range of ages. Researchers reported the postoperative rehabilitation as general guidelines and did not provide the specific exercises that were implemented, duration and intensity of the exercises, or number of treatment visits. Given these shortcomings, the best available evidence is not strong enough to concretely supply clinicians with global ORs for return to participation at preinjured levels for either overhead or nonoverhead athletes after operative repair of the superior labrum with or without concurrent soft tissue debridement. Instead, we calculated a range of approximate ORs, showing an apparent trend for nonoverhead athletes to have more success returning to full competition after arthroscopic labral repair. Despite the results favoring nonoverhead athletes, operative treatment for labral pathologic conditions should not be abandoned as a viable option for overhead athletes because operative restoration of the disrupted anatomy may be appropriate based on the functional deficits and demands of each patient. The identified gaps in reporting methods suggested that in future clinical outcomes reports, researchers should control particular areas, such as specific activity, age, surgical details and technique, and rehabilitation. Additionally, future researchers should use prospective data-collection methods related to preoperative and postoperative subjective and objective clinical measures in addition to a preinjury assessment of sport-specific performance. Using these measures would help control for potential biases and add an integrated, measurable component of function, merging the patient-specific subjective considerations with the objective performance of each athlete.