Clinical Evaluation Techniques for Injury Risk Assessment in Elite Swimmers: A Systematic Review

Abstract

Context:

Evidence concerning a systematic, comprehensive injury risk assessment in the elite swimming population is scarce.

Objective:

To evaluate the quality of current literature regarding clinical assessment techniques used to evaluate the presence and/or development of pain/injury in elite swimmers and to categorize objective clinical assessment tools into relevant predictors (constructs) that should consistently be evaluated in injury risk screens of elite swimmers.

Data Sources:

PubMed, Embase, Scopus, CINAHL, SPORTDiscus, PEDro, and the Cochrane Library Reviews were searched through September 2018.

Study Selection:

Studies were included for review if they assessed a correlation between clinic-based objective measures and the presence and/or development of acute or chronic pain/injury in elite swimmers. All body regions were included. Elite swimmers were defined as National Collegiate Athletic Association, collegiate, and junior-, senior-, or national-level swimmers. Only cohort and cross-sectional studies were included (both prospective and retrospective); randomized controlled trials, expert opinion, and case reports were excluded, along with studies that focused on interventions, performance, or specific swim-stroke equipment or technology.

Study Design:

Systematic review and qualitative analysis.

Level of Evidence:

Level 3.

Data Extraction:

PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines were utilized at each phase of review by 2 reviewers; a third reviewer was utilized for tie breaking purposes. Qualitative analysis was performed using the Methodological Items for Non-Randomized Studies (MINORS) assessment tool.

Results:

A total of 21 studies assessed the presence and/or development of injury/pain in 3 different body regions: upper extremity, lower extremity, and spine. Calculated average MINORS scores for comparative (n = 17) and noncomparative (n = 4) studies were 18.1 of 24 and 10.5 of 16, respectively. Modifiable, objectively measurable injury risk factors in elite swimmers were categorized into 4 constructs: (1) strength/endurance, (2) mobility, (3) static/dynamic posture, and (4) patient-report regardless of body region.

Conclusion:

Limited evidence exists to draw specific correlations between identified clinical objective measures and the development of pain and/or injury in elite swimmers.

Elite swimmers typically specialize in 1 or more of 4 recognized competitive strokes: freestyle, backstroke, butterfly, and breaststroke, with variety in event distance. Yet even with stroke specialization within competition, elite swimmers spend the majority of their training volume swimming freestyle. The repetitive nature of the normal freestyle stroke cycle can predispose elite swimmers to musculoskeletal injuries of the upper extremity, knee, and spine.⁴⁶ Based on a survey performed by the National Collegiate Athletic Association, overall injury rates in elite swimmers were 4 injuries per 1000 hours of training for men and 3.78 injuries per 1000 hours of training for women.⁴⁸ The most recent incidence data from the FINA World Championships indicated that the majority of injuries were related to overuse (68.1%) and affected the shoulder (26.3%), knee (10.1%), lower back (9.8%), and hip/groin region (9.6%).³¹ Average training time lost was 2 days, and more than two-thirds (69%) of affected athletes modified their training routine.^31,47 Modifications included decreased training frequency, decreased training volume, and/or altered stroke selection for training. The one-third of swimmers who did not modify their training routine as a result of reported pain speaks to the attitudes surrounding pain and its “normalcy” in the elite swimming population.¹⁸

Shoulder injuries are significantly more common than other musculoskeletal injuries in elite swimmers, with prevalence ranging from 40% to 91%.^{3,10,20,23,33,40} The knee is another significant source of pain and injury in the elite swimming population, with 34% to 86% of elite swimmers reporting at least 1 episode of knee pain or injury in their career.^21,32,44 Multiple studies have identified a greater incidence of knee pain in breaststroke specialists.^32,44 Finally, the spine has also been recognized as a body region at risk of injury in elite swimmers, with prevalence ranging from 22.2% to 47% in breaststroke specialists and 33.3% to 50% in butterfly specialists.^9,14,29

In the context of sport participation and athletic performance, screening has been defined as a strategy used to detect conditions in individuals without signs or symptoms of that condition.^11,42 The goal is to identify risk factors early and enable timely intervention and management to reduce future morbidity, missed practice and competition time, and progression of injury. Other research advocates that risk factors should be identified to better understand why injuries occur and to predict which athletes are at increased risk for injury.^2,40 Identified modifiable risk factors in elite swimmers include deficits and asymmetries in rotator cuff muscle strength; asymmetries in quadriceps and hamstrings muscle strength; impaired scapular control; lack of glenohumeral stability; abnormal posture; extreme ranges of hip, knee, and glenohumeral mobility; impaired motor control; faulty stroke technique/biomechanics; and heavy training volume.¹⁸ These modifiable risk factors interact in a dynamic way with nonmodifiable risk factors, including sex, age, stroke specialization, and years of training experience. The dynamic interaction between these risk factors can limit the ability and validity of a single clinical assessment tool to identify athletes at risk of injury.¹¹

Information regarding the most appropriate and evidence-based clinical assessment tools to identify injury risk in high-level swimmers is limited. A single systematic review has been performed in an attempt to identify risk factors and classify the level of certainty of these risk factors for identifying shoulder pain alone in swimmers.¹⁹ No comprehensive systematic review has been performed to assimilate objective tools to use in injury risk assessment beyond the shoulder region.

The primary purpose of this study was to include all body regions in an assessment of pain and injury risk in elite swimmers and to evaluate the quality of the current literature regarding clinical assessment techniques used to assess this risk. Our goal was to assess the content of the included studies to support the development of recommendations regarding specific objective measures to screen for at-risk athletes. The secondary purpose of this study was to identify categories of objective clinical assessment tools and build the foundation for objective test clusters (constructs) that should consistently be evaluated in injury risk screens of elite swimmers to guide injury risk reduction recommendations, the development of injury prevention programs, and interventions in elite swimmers.

Methods

This systematic review was registered through PROSPERO (CRD42018086936). Seven databases (PubMed, CINAHL, SPORTdiscus, Embase, Scopus, PEDro, and the Cochrane Library Reviews) were searched systematically from the earliest available date through September 2018. The search strategy was developed in consultation with a medical librarian and adapted to each database (Table 1). This systematic review utilized the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines in both search and report phases of the research process.²⁷ The search yield was imported into Mendeley,³⁵ a web-based organization tool allowing for consolidation of literature searches from multiple databases, duplicate elimination, and precise tracking for construction of a PRISMA flow sheet (Figure 1).

Table 1.

Search strategy

Database	Search Terms
PubMed, Embase, Scopus, CINAHL, SPORTDiscus	(Exam* OR Screen* OR Test* OR Measure* OR Assess* OR Predict* OR Tool* OR Technique*) AND (injur* OR return to sport OR pain*) AND swim* AND (elite* OR compet*) NOT (triathlet* OR div* OR polo OR synchroniz*) Filters: Humans; English
Cochrane Library Reviews	swim* AND (elite* OR compet*)
PEDro	swim* AND (elite* OR compet*)

Figure 1.

PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flowchart. RCT, randomized controlled trial.

Titles and abstracts were first screened independently by 2 reviewers using the outlined inclusion and exclusion criteria (Table 2). Specifically, studies were included in this review if the study population included elite swimmers, defined as competing at junior national, national, and collegiate levels. Included studies had to examine the relationship between objective outcomes, tests or measures, and pain/injury in the study population. Pain/injury was defined by this author group as either subjective or objective discomfort or pathology associated with swim-specific training and competition. The authors of this study chose not to include in-the-water swim stroke analysis as a component of injury risk assessment to promote the development of a clinic-based screening cluster to identify at-risk elite swimmers. Randomized controlled trials were intentionally excluded from this systematic review, as the authors wanted to strictly assess the relationship between objective findings and pain/injury in this population without associated interventions.

Table 2.

Study inclusion and exclusion criteria

Inclusion criteria
Study type	• Retrospective and prospective studies • Cohort • Cross-sectional studies
Population	• Junior-, senior-, or national-level swimmers • NCAA collegiate swimmers
Assessment techniques	• Objective clinical assessment tool to measure the following: strength, endurance, ROM, motor control, static posture, dynamic posture, instability, patient questionnaires
Outcomes	• Discussion regarding objective outcomes with any of the following: pain, injury, injury risk, injury development, missed practice/competition time
Exclusion criteria
Study type	• RCT • Expert opinion or case reports • Focused on intervention or treatment • Performance-related results • Focused on swim stroke equipment or high-cost technology
Population	• Masters age-group swimmers • Para-athlete, triathlete/open-water, and water polo sport focused

NCAA, National Collegiate Athletic Association; RCT, randomized controlled trial; ROM, range of motion.

After the initial title and abstract screens, full texts of remaining articles were then obtained and screened independently by the same 2 reviewers. Differences in opinion between reviewers were resolved by independent appraisal from a third reviewer. Citation tracking using Mendeley and reference checking of the included articles was also performed independently by these 2 reviewers.

To address the secondary purpose of this study, manuscripts were analyzed by 2 authors who independently categorized the articles into constructs of objective assessment components (regardless of body region) based on prior research studies and associated clinical commentary. The development of these constructs aimed to outline a foundation for test clusters to enhance the efficacy of injury risk assessment in elite swimmers.^16,18,25,26

Data Analysis

Quality assessment was independently performed by 2 reviewers based on the Methodological Items for Non-Randomized Studies (MINORS) quality assessment tool³⁶ (see Table A1 in the Appendix, available in the online version of this article). The MINORS tool was selected because of its validity in the quality assessment of nonrandomized trials. The tool was originally created for quality assessment of nonrandomized surgical trials and includes 12 items scored as 0 (not reported), 1 (reported but inadequate), and 2 (reported and adequate).³⁶ Scored items included study aim, prospective data collection, bias, and statistical analysis—each used to characterize the methodological and scientific value of the associated publication.

Relevant data and information were extracted from included articles by 2 independent reviewers. Unfortunately, the diversity among studies with regard to injury (definition, location, and chronicity), selected clinical assessment tools (range of motion, strength, endurance, posture, symmetry, and quality of movement), training volume (club, collegiate, and national level swimmers), and overall methodological quality precluded performance of a quantitative meta-analysis. As such, a level of certainty indicator, as previously outlined by Hill et al,¹⁹ was used to appraise the association between an identified risk factor and the development of injury/pain in this population. Included studies did not reference relative risk or odds ratios in their evaluation of the utility of certain clinical assessment tools in association with injury risk.

Results

Ultimately, 21 articles were considered relevant to our research objectives and were included in this systematic review (see Table A2 in the Appendix, available online). For ease of interpretation, these were divided into body region: upper extremity (UE), lower extremity (LE), and spine. The methodological quality scores of the included articles are outlined in Table A1 in the Appendix (available online). All 21 studies discussed the competitive elite swimming population: 17 discussed UE pain/injury, 3 discussed LE pain/injury, and 1 discussed spine pain/injury. No high-quality predictive validity studies exist to correlate objective measurements and modifiable risk factors with the development of injury in elite swimmers.

Relevant Objective Clinical Assessment and Screening Tools

Upper Extremity

A total of 17 nonrandomized studies met the inclusion criteria for the UE, with the shoulder cited as the injured UE body region in all 17 studies.^{4-8,13,15,16,17,25,26,30,38,39,41,45,49} The average MINORS quality scores for relevant UE comparative and noncomparative studies were 18.1 of 24, and 10.5 of 16, respectively. The study populations included elite club, collegiate, and national-level swimmers. The number of participants in the included studies varied, ranging from 15 to 236.

Clinical assessment tools utilized to assess for the presence and risk of shoulder pain/injury varied, but the following were consistently referenced when comparing involved versus uninvolved UE: shoulder external rotation (ER) range of motion (ROM),^41,45 shoulder ER and internal rotation (IR) strength, endurance and associated ER:IR ratio,^5,6 scapular control and mobility,^4,38,41 shoulder complex muscle extensibility,^16,39 and glenohumeral instability.^4,26

Included studies suggest that associations between an increased risk of shoulder pain and (1) impaired muscle extensibility in pectoralis minor^16,39 and latissimus dorsi musculature,^30,39 (2) limited or excess glenohumeral external rotation range of motion at 90° of abduction,^41,45 and (3) limited scapular upward rotation with humeral elevation^38,41 (see Table A2 in the Appendix, available online, for proposed cutoff values). Current research exists to support the relationship between shoulder IR and ER ROM and risk for shoulder injury/risk development in swimmers. Evidence available to support the relationship between glenohumeral rotation mobility and shoulder injury risk is sufficient to support a risk association, but confidence is limited by sample size and quality of included studies and inconsistent findings between studies.

Moderate-quality evidence exists to support the relationship between shoulder IR and ER muscular strength and endurance ratios and the presence of injury/pain in elite swimmers (see Table A2 in the Appendix, available online, for proposed values).^5,7 Bak and Magnusson⁵ proposed a cutoff value of functional ER:IR strength ratio ≥1.08 as related to shoulder injury/pain; however, again, this cutoff value has not been comprehensively validated in a prospective study.

Low-quality evidence exists to support the relationship between static and dynamic scapular positioning and the presence of shoulder injury/pain in elite swimmers. Conflicting relationships have been found between scapular dyskinesia^4,25,38,41 and shoulder pain/injury. Diverse and inconclusive methods exist to assess these scapular postures with both static and dynamic UE tasks. Limited evidence exists to support the association between glenohumeral instability^5,37,39 and shoulder pain/injury risk. Generalized joint hypermobility, as assessed by a Beighton score of ≥7 (of a possible 9), has been correlated with significant strength and fatigue deficits in shoulder medial rotation.⁴⁹

Low-quality evidence also exists relating the results of patient-reported outcomes to shoulder pain/injury risk in elite swimmers. The Swimmer’s Functional Pain Scale¹³ and the Kerlan-Jobe Orthopaedic Clinic Shoulder and Elbow score (KJOC)⁴⁹ have both been validated in the assessment of shoulder pain in elite swimmers (see Table A2 in the Appendix, available online, for proposed cutoff values).

Patient-reported stroke specialty has also been shown to be associated with at-risk body region^14,29,34,48: freestyle and backstroke specialties were most commonly associated with shoulder injury risk. Insufficient study number and population size in included studies related to patient-reported outcomes and limits the efficacy of these tools to assess risk of shoulder pain/injury in a predictive manner.

Lower Extremity

Three nonrandomized studies met the criteria for the LE, with the knee as the most commonly cited body region of LE pain/injury.^32,34,44 The average MINORS quality score for relevant LE comparative studies was 15.6 of 24. The study populations included competitive age group, elite, and international swimmers. Two of the 3 studies included only breaststroke and individual medley specialists but did not specify competition distance.^32,44 The number of athletes in the included studies varied, ranging from 19 to 36.

One of the 3 studies discussed the assessment of hip IR ROM in breaststroke swimmers to assess for the presence of knee pain and/or injury.³² Limited evidence exists to support the relationship between impaired hip IR active range of motion (see Table A2 in the Appendix, available online, for proposed values) and the development of knee pain or injury in breaststroke specialists.³²

A weak relationship exists between the presence of knee pain/injury in elite swimmers and quadriceps/hamstrings strength asymmetry (knee flexion/extension ratio), defined as >10% difference between sides.³⁴ No formal cutoff values have been identified to stratify low, normal, and high values for knee agonist/antagonist strength and endurance ratios in elite swimmers as they relate to knee pain/injury development. It has been hypothesized that stroke specialty may also affect these values.³² Insufficient study number and population size in included studies related to each of the above-outlined objective assessment tools limits the efficacy of these tools to assess risk of knee pain/injury in a predictive manner.

Spine

A single, nonrandomized study met the criteria to be included for the spine,⁵⁰ with thoracic and lumbar regions cited as the most common areas of spinal pain/injury. The MINORS quality score was 18 of 24. This study included 329 elite adolescent swimmers.

The recommended clinical assessment tools used to screen for hyperlordosis in the lumbar region and hyperkyphosis in the thoracic region were lordosis plumbline and kyphosis plumbline distances, respectively. Low-quality evidence supports the relationship between abnormal, static thoracic, and lumbar plumbline distances⁵⁰ and the presence of back pain in elite swimmers. Association between these static postural assessments and associated back pain/injury has not been validated.

Based on comparable systematic reviews^1,19,28 and relevant clinical commentary,^11,42 referenced outcome tools were divided into clinical constructs to comprehensively screen for injury risk in elite swimmers. These constructs include: (1) mobility, (2) strength/endurance, (3) static and dynamic postures, and (4) patient report (Figure 2).

Figure 2.

Clinic-based objective assessment constructs associated with injury risk in elite swimmers (*biomechanical analysis of swim stroke should also be performed). ER, external rotation; GH, glenohumeral; IR, internal rotation; KJOC, Kerlan-Jobe Orthopaedic Clinic Shoulder and Elbow score; SFPS, Swimmer’s Functional Pain Scale.

Discussion

The primary purpose of this systematic review was to evaluate the evidence in an attempt to make recommendations regarding the utility of clinic-based assessment techniques in the identification of elite swimmers at risk of pain/injury development. In doing so, the authors attempted to stratify these objective assessment tools into the following constructs: (1) mobility, (2) strength/endurance, (3) static and dynamic postures, and (4) patient report (Figure 2). This article comprehensively evaluated the current literature and assessed the relationship between modifiable risk factors and injury development, across body regions, in elite swimmers.

Data assimilated by Hill et al¹⁹ in a previous systematic review indicated that shoulder IR/ER strength, pectoralis length and scapular kinematics, strength, and dyskinesis currently lack sufficient evidence to support their ability to assess risk of shoulder pain/injury. These findings were supported by this systematic review. Current literature does support a relationship between shoulder pain and (1) glenohumeral rotation mobility and (2) shoulder ER:IR endurance ratios.

This systematic review also examined lower extremity and spine injuries in elite swimmers to provide novel and comprehensive recommendations for injury risk screening in elite swimmers, regardless of body region. Currently, insufficient evidence exists to support associations between knee pain/injury and identified objective lower extremity outcomes, including hip IR mobility and knee flexion to extension side-to-side symmetry and strength ratios. Furthermore, this study identified a single study that weakly supports the relationship between spinal pain and abnormal thoracic and lumbar plumbline measurements.

Significant variability exists in outcome measure selection, the methods/performance of associated assessment techniques within each construct, and validated cutoff values, which makes the selection of true injury prevention “screening tools” difficult. For an objective measure to be validated as an effective screening tool in the construct of injury prevention, the following criteria must be met: (1) perform a prospective cohort study to identify risk factor(s) and define cutoff values, (2) validate test and cutoff value in multiple cohorts, and (3) perform a randomized controlled trial to test effect of a combined screening and intervention program.² Based on these criteria and current available literature, evidence only exists to support the foundational development of these constructs.

In the context of injury risk assessment in elite swimmers, it is worth noting the relationship between the timing of objective assessment and associated objective findings. For example, one study demonstrated that over the course of a competitive swim season, shoulder IR strength increased at both mid- and postseason time points and was associated with the development of shoulder pain secondary to its effect on the shoulder ER:IR strength ratio.⁶ Another study demonstrated a significant difference in pre- and postpractice scapular upward rotation in elite swimmers with and without shoulder impingement,³⁸ and yet another demonstrated a relationship between reduced shoulder ER range and joint position sense with increased fatigue.^12,22,24 The intentional assessment of elite swimmers in a fatigued state may provide more effective and relevant objective data in the context of injury risk assessment.¹¹

Limitations

We were unable to calculate effect sizes or perform a meta-analysis secondary to the heterogeneity of studies, clinical assessment techniques, and objective outcome tools utilized. Therefore, the conclusions are founded on the quality of the included cohort, cross-sectional, and case-control studies as assessed by the MINORS tool.⁴³ A second study limitation was the extrapolation of evidence quality provided by the MINORS tool. The tool only provided maximal/“ideal” values, and therefore the authors were not able to objectively define ranges of study quality. Finally, only 2 studies^25,45 of the total included were prospective studies but did not use relative risk or odds ratios to assess the predictive validity of the selected clinical assessment tools. The remaining 19 studies assessed clinical outcomes at a single time point or with only 1 short-term (pre- vs posttraining session) follow-up. Included study designs and lack of predictive validity extrapolations limited our ability to make direct recommendations regarding the above-identified clinical outcomes and their stand-alone utility in injury prediction in elite swimmers.

Conclusion

Limited evidence exists to draw specific correlations between clinical objective measures and the development of pain and/or injury in elite swimmers. Grouping the best and most appropriate clinical outcomes into the following constructs may allow clinicians to begin to account for the dynamic interactions between modifiable and nonmodifiable risk factors in the development of pain/injury in elite swimmers (mobility, strength/endurance, static/dynamic postures, and patient report) at the shoulder, knee, and spine.