Upper limb functional testing in athletes: A Delphi study

Abstract

Background

Functional testing has recently become more and more popular to assess athletes, both for injury prevention, as well as in an objective of performance. However, the relationship between the results of these tests and performances (or injuries) or their interpretation remains unclear.

Objective

The aim of this study is to explore the usefulness, the characteristics, and the interpretation of the most frequently used upper-limb functional test.

Methods

Twenty-two experts with an excellent knowledge of upper limb functional tests and an expertise in sports medicine and/or sports training of at least 5 years were recruited. They answered to qualitative and quantitative questions about functional testing trough structured questionnaires (online).

Results

Four rounds were needed to reach a consensus about the usefulness as well as the characteristics of each test. Different sports-specific batteries of tests were also suggested by the experts and reached consensus. However, concerning the interpretation of the test, a consensus was only found for half of the tests considered.

Conclusion

The current study summarizes the characteristics and the usefulness of the most popular upper-limb functional tests. However, the interpretation of some tests will have to be further explored since no consensus was found for them.

Introduction

Overhead athletes load their shoulder with high velocity and in extreme range of motion in their sport. These repetitive solicitations may cause medium- to long-term musculoskeletal adaptations. With the aim of preventing injuries, preseason screening (as well as regular re-assessments) has become essential for the management of overhead athletes in order to detect possible imbalances or dysfunctions that could enhance the risk of shoulder injury.^1,2 As for performance, regular evaluations allow trainers to better appreciate athletes’ progress, which in turn helps to guide training.^3,4

In addition to mobility and strength assessments, functional testing has recently become more and more popular to assess athletes, both for injury prevention, as well as to improve performances.^2,4–6 Moreover, after an injury, functional testing could provide objective criteria to help make return-to-sport decisions. ⁵

Functional tests like the upper quarter Y balance test,^7,8 the single arm medicine ball throw, ⁴ the seated medicine ball throw, ⁹ or the closed kinetic upper extremity stability test ¹⁰ were the first to appear in the scientific literature and were followed by other tests, such as the athletic shoulder test, ¹¹ the posterior shoulder endurance test ¹² or recently, the upper limb rotation test. ¹³ The ever-increasing number of functional tests described in the literature underline the need to categorize them. Indeed, because of timing or fatigue, it is never possible to perform all of the tests during a single session with an athlete or a patient. Additionally, some tests may be more appropriate for certain categories of patients, for certain sports or to assess certain specific qualities. All of these aspects are yet to be clearly defined or summarized in the literature.

The aim of this Delphi study (a research technique aimed at finding a consensus based on the opinion of experts, through a series of structured questionnaires)^14,15 was to explore the usefulness, the characteristics, and the interpretation of the most frequently used upper-limb functional tests based on the scientific and clinical experience of a group of experts.

Methods

Participants

The recruitment included physiotherapists, medical doctors, orthopedic surgeons as well as strength and conditioning coaches that (1) had an excellent knowledge of upper limb functional tests, (2) practiced these tests on a regular basis with athletes, (3) had an expertise in sports medicine and/or sports training of at least 5 years, and (4) were not involved in the elaboration of the questionnaires. Thirty-five clinicians and researchers corresponding to the criteria were identified by the research team (that included practitioners and researchers with a high expertise in shoulder rehabilitation). Special care was given when recruiting experts, to make sure, that a wide variety of countries was represented and the number of experts per country was limited so as not to be disproportioned. They were contacted and informed about the procedure of the study by email. Among them, 22 agreed to take part in the study. This number seemed sufficient as 15 participating subjects are required based on the guidelines by Hasson et al. ¹⁵

Procedure

The different functional tests included were chosen based on the literature and discussions with the research team. For that purpose, a search was performed on PubMed and Scopus database. A combination of keywords and MeSH terms related to shoulder functional testing. A detailed description as well as illustrations were provided in the questionnaire for each test to make sure that everyone had the same information at the beginning of the study. The first of the four rounds included both quantitative and qualitative questions on usefulness, validity, reliability, advantages, and disadvantages of the different upper limb functional tests described in the literature. The majority were open-ended questions. For each one, experts gave a mark out of 100, based on their knowledge and experience (Table 1).

Table 1.

Description of the participants and the sports they are involved in.

Profession	Country	Experience	Sports population
Physiotherapist (N = 16) Medical doctor (N = 3) Orthopedic surgeon (N = 1) Fitness coach (N = 2)	France (N = 4) Quatar (N = 3) USA (N = 2) Belgium (N = 2) Canada (N = 2) Italy (N = 2) UK (N = 2) Ireland (N = 1) Japan (N = 1) Switzerland (N = 1) Norway (N = 1) Sweden (N = 1)	≤10 years (N = 5) 10–20 years (N = 10) 20–30 years (N = 3) ≥30 years (N = 4)	Handball (N = 9) Swimming (N = 9) Rugby (N = 7) Tennis (N = 6) Volleyball (N = 6) Baseball (N = 2) Others: football goalkeeper (N = 2), fencing (N = 2), squash (N = 2), water polo (N = 1), table tennis (N = 1), gaelic football (N = 1), basketball (N = 1), ski (N = 1)

The results were summarized and sent to each participant with a second questionnaire. Qualitative data of the first questionnaire was expressed as quantitative data (% of agreement). Each expert had the option of revising their answer, based on the opinion of the other participants and/or to give additional information. Questions about the applicability of the test for a specific sport, and whether including the test would be useful as part of a more global battery of tests were also added in the second stage (Table 2). The objective was to create sport-specific batteries of tests that could be used by clinicians, physical trainers, or coaches, both to assess performance but also risk factors of injury or return-to-play capacity. The same procedure was conducted for the third and fourth rounds. The open-ended questions were progressively changed into closed-end questions to reach a consensus.

Table 2.

Questions asked to the experts during the study.

What's your definition of a functional test? (1st questionnaire)

According to you, what are the qualities of a good functional test? (1st questionnaire)

Do you think … test has a value in assessing shoulder function in overhead athletes? (1st questionnaire)

Do you think …test has a value in assessing kinetic chain function in overhead athletes? (1st questionnaire)

Do you think … test has good validity? (1st questionnaire)

Which “qualities” do you assess … test with? (1st questionnaire)

What are the advantages and disadvantages of … test? (1st questionnaire)

How do you interpret the results of … test? (1st questionnaire)

Would you recommend … test? (1st questionnaire)

How much difference in test performance would you consider as clinically relevant for each test? (1st questionnaire)

Which of these tests would you recommend for a battery of tests to assess performance in … (sport)? (2nd questionnaire)

Which of these tests would you recommend for a battery of tests to assess risk factor or return to play capacity in … (sport)? (2nd questionnaire)

Based on the recommendations by Diamond et al., ¹⁴ the level required for consensus had to be equal or superior to 75% of agreement. When 75% of agreement was obtained for a question, a consensus was considered to be reached and the question was not repeated in the next round. Questions that did not reach the consensus of 75% were rephrased in a different way in order to enhance their understanding and had to be answered again during the subsequent stage. Each questionnaire was sent by email to the participants and answered anonymously. The survey was hosted by the SurveyMonkey website. A timeframe of 1 month was provided to complete each questionnaire. A delay of approximately 6–8 weeks was necessary between the questionnaires to analyze the answers and to prepare the next survey.

The different tests are illustrated in Figure 1 and detailed in Appendix 1. The questions asked are presented in Table 2.

Figure 1.

Illustration of the functional tests included in the study.

1: upper quarter Y balance test; 2: closed kinetic chain upper extremity stability test; 3: one-arm hop test; 4: upper limb rotation test; 5: countermovements push-ups; 6: posterior shoulder endurance test; 7: seated single arm shot put test; 8: seated medicine ball throw; 9: single arm medicine ball throw; 10: athletic shoulder test.

Results

All 22 experts answered the first survey while only 19 took part in the second and third stages. A final fourth stage was organized in order to have a better consensus and 18 experts participated in this final stage (18.2% dropout).

Initially, eight functional tests were included in the study, based on the discussion with the research team. Two other tests were added after the first questionnaire, based on participants’ suggestions (countermovement push-ups and posterior shoulder endurance test).

Functional testing: Definition

First, the experts were asked to give their definition of functional testing. The answers given by the experts were summarized using the following definition: a functional test can be considered as a “valid and reliable, qualitative or quantitative test aimed to assess at a certain point in time, joint's, kinetic chains or individual's performance that is as close as possible to daily living- or sport activities.”

Analysis of specific characteristics of the target functional tests

For each of the tests included, experts were asked to determine whether the test was appropriate to assess shoulder function and/or kinetic chain function. Then, they gave their opinion on the validity (“Does the test measure what it claims to measure?”) and the reproducibility of the tests. These questions were presented as “yes-no” questions in the first stage. In the second round, the experts had the opportunity, if needed, to revise their opinion based on the opinion of the other experts. The information was then expressed in percentages. For example, a value of 55% means that 55% of the experts believed that the test is reliable. The results of each test are presented in Figure 2.

Figure 2.

The usefulness of the functional tests to assess shoulder or kinetic chain function as well as validity of the tests (expressed in percents).

UQYBT: upper quarter Y balance test; CKCUEST: closed kinetic chain upper extremity stability test; SMBT: seated medicine ball throw; SSASPT: seated single arm shot put test; OAHT: one arm hop test; SAMBT: single arm medicine ball throw; ULRT: upper limb rotation test; AST: athletic shoulder test; CMP: countermovement push-ups; PSET: posterior shoulder endurance test.

The experts were then instructed to define, for each test, the characteristics as well as the advantages and the disadvantages of each test. Only the variables that reached at least 75% were retained. The percentages mentioned in the text correspond to the percentage of experts who agreed with the quality, advantages, or disadvantages. The main qualities assessed by each test are summarized in Table 3.

Table 3.

Main qualities assessed by each of the tests.

	Strength	Power	GH/KC stability	Coordination	Endurance
UQYBT			x
CKCUEST			x	x
SMBT		x
SASPT		x
OAHT		x	x
SAMBT		x
AST	x
ULRT			x	x
CMP	x	x
PSET					x

UQYBT: upper quarter Y balance test; CKCUEST: closed kinetic chain upper extremity stability test; SMBT: seated medicine ball throw; SSASPT: seated single arm shot put test; OAHT: one arm hop test; SAMBT: single arm medicine ball throw; ULRT: upper limb rotation test; AST: athletic shoulder test; CMP: countermovement push-ups; PSET: posterior shoulder endurance test.

The upper quarter Y balance test

According to the experts, this test mainly assesses the stability of the gleno-humeral joint (78.9%) and the stability of the kinetic chain (84.21%). The main advantage of this test is that it is inexpensive (78.9%) and involves both the shoulder and the trunk (84.2%). This test may be used for closed chain sports like wrestling, football, and rugby but not for open chain sports like tennis, baseball, volleyball, or handball (89.5%).

The closed kinetic chain upper extremity stability test

The main qualities assessed by this test are the stability of the kinetic chain (94.74%), muscular endurance (78.95%), and coordination (73.68%). This test has the advantages of being easy to implement (100%), inexpensive (100%), quick to do (100%), and does not require any specific material (100%). Moreover, this test combines shoulder and trunk stability, which is appropriate to assess both functions (100%). However, this test is inadequate to assess shoulder function in the first months following surgery (94.74%) and is not functional for open chain sports like tennis, baseball, handball or volleyball (84.21%).

The seated medicine ball throw

According to the experts, this test mainly assesses power (89.47%) and global upper limb strength (84.21%). It has the advantages of being inexpensive (100%), easy to perform (94.74%), explosive (94.74%), and quick to do (84.21%). However, this test does not evaluate shoulder function in the end range of throwing motion (100%) and does not allow comparisons between limbs (89.47%). It is also sometimes difficult to be precise in recording the throwing distance (78.95%) and to monitor compensations of the non-injured side (94.74%).

The seated single arm shot put test

This test mainly assesses power (94.74%). It is an inexpensive test (100%), easy to perform (94.74%), explosive (88.89%), and comparisons between limbs are possible (94.44%). As for the disadvantages of this test, some experts reported difficulty of standardization, the difficulty to precisely measure the distance thrown, or the inability to perform this test in the first months after a surgery. However, none of these propositions reached the consensus of 75%.

The one arm hop test

The main qualities assessed by this test would be evaluating plyometric qualities (100%), power (94.74%), kinetic chain function (89.47%), and shoulder strength (89.47%). This test is inexpensive (100%), requires motor control and coordination (100%) and allows comparison between sides (100%). However, this test is very challenging (94.74%) and may be difficult to realize for a certain cohort of athletes (94.74%). Similar to the closed kinetic chain upper extremity stability test, this test does not reproduce an overhead task and the gesture is different from the throwing motion (94.74%).

The single arm medicine ball throw

According to the experts, this test mainly assesses power (100%), kinetic chain function (89.47%), shoulder strength (78.95%), and coordination (78.95%). The main advantages are that it is inexpensive (100%), easy to perform (89.47%), and quick to do (84.21%). Moreover, this test assesses shoulder function in a throwing position (94.44%) and it allows comparison between sides (94.47%). However, this test cannot be performed in the first months after surgery (89.47%) and it is sometimes difficult to be precise with the measurement (73.68%). Finally, the weight of the ball used in this test is heavier than the ball usually used in overhead activities (73.68%).

The athletic shoulder test

This test mainly assesses shoulder strength (88.89%) as well as the rate of force development (94.74%). It gives an indication of rate of force development in a higher range of motion (100%) and not in a rotational plane (100%). And, according to the experts, this test would be useful for following the state of fatigue in elite players (84.21%). However, there are currently no reference values for this test (89.47%) and it reproduces overhead task in an unnatural context (78.95%). Moreover, it requires expensive equipment (force plates and software for analysis) (89.47%) and it is longer than other tests to set up (78.95%).

The upper limb rotation test

The main qualities assessed by this test are trunk stability (100%), trunk mobility (89.47%), coordination (89.47%), and shoulder proprioception (73.68%). The test has the advantage of not requiring specific material (94.74%), and that it is easy to perform in the clinical setting or in the field (89.47%). Moreover, one limb works in a closed chain and the other in an open chain, which is interesting from a clinical perspective (89.47%). However, there are no reference values for this test at the moment (94.74%) and patients with pathologies that do not tolerate weightbearing on the shoulder may be unable to perform this test (94.74%).

Countermovement push-ups

According to the experts, the main qualities assessed by countermovement push-ups are power (100%), plyometric capacity (100%), and global shoulder strength (92,31%). The main advantages are that it is easy to perform (85.71%) and it does not require specific material (100%). Moreover, this test is dynamic (83.33%) and challenging (92.31%). For additional data, this test can be performed on a force platform (92.86%). However, this test has also some disadvantages. It is not possible to perform this test in the first months after surgery (100%), it is sometimes complicated to avoid compensations (78.57%) and bilateral comparisons are not possible (76.92%). Finally, although appropriate for contact sports in a closed chain, such as rugby (78.57%), this test does not resemble a throwing motion (83.33%).

Posterior shoulder endurance test

The posterior shoulder endurance test aims to assess the endurance of posterior shoulder muscles (100%). As for other previously described tests, this one is quick (92.86%) and easy (100%) to perform. It does not require specific material (85.71%) and allows comparisons between limbs (100%). However, this test is not performed in the end of range throwing position (92.31%) and does not include the kinetic chain (92.31%).

Global interpretation of the tests

Unilateral tests like the UQYBT, SSASPT, SMBT, OAHT, AST, ULRT, or PSET allow bilateral comparisons. A percentage of difference can be calculated between the two arms and then interpreted. For bilateral tests like the CKCUEST, SMBT, or CMP the interpretation is much more complicated since it requires previous data from the athlete or normative data of athletes of the same age and level. Normative data can also be interesting in unilateral tests to assess the capacity of the athlete in comparison to others with similar characteristics.

The experts gave their opinion on how much difference in the test performance they considered as clinically relevant for each test. All these values are summarized in Table 4. In practice, a difference in percentage (between two test sessions, between two athletes, or between dominant and non-dominant arms) lower than the value mentioned in the table below cannot be considered as clinically relevant while a difference superior to the value mentioned is. The percentage mentioned in brackets represents the percentage of agreement.

Table 4.

Interpretation of the tests suggested by the experts.

Test	Interpretation
UQYBT	No consensus
CKCUEST	15% (75%)
SMBT	20% (78.57%)
SASPT	20% (84.62%)
OAHT	20% (86.67%)
SAMBT	20% (84.62%)
AST	No consensus
ULRT	No consensus
CMP	No consensus
PSET	No consensus

UQYBT: upper quarter Y balance test; CKCUEST: closed kinetic chain upper extremity stability test; SMBT: seated medicine ball throw; SSASPT: seated single arm shot put test; OAHT: one arm hop test; SAMBT: single arm medicine ball throw; ULRT: upper limb rotation test; AST: athletic shoulder test; CMP: countermovement push-ups; PSET: posterior shoulder endurance test.

Different values were suggested for the interpretation of UQYBT, AST, ULRT, CMP, and PSET but none of them reached the consensus of 75%.

Sport-specific batteries of tests

The last two stages were aimed at establishing batteries of tests based on the characteristics and the specificities of each sport. A distinction between performance and risk factors/return to play assessment was made for each sport. The batteries of tests aiming to assess performance are summarized in Table 5. The ones with the objective to assess risk factors or return to play capacity are presented in Table 6. The tests that did not reached consensus are not presented in the tables.

Table 5.

Tests batteries to assess performance.

Handball	Upper limb rotation test a Single arm medicine ball throw b
Volleyball	Single arm medicine ball throw b Upper limb rotation test b
Swimming	Upper limb rotation test b Posterior shoulder endurance test b
Tennis	Single arm shot put test a The athletic shoulder test a Single arm medicine ball throw b Upper limb rotation test b
Baseball	Single arm medicine ball throw b Upper limb rotation test b Posterior shoulder endurance test b
Rugby	Upper quarter Y balance test a CKCUEST a The athletic shoulder test b Countermovements push-ups b

^a between 75 and 85% of consensus.

^b > 85% of consensus.

Table 6.

Tests batteries to assess risk factors or return to play capacity.

Handball	Upper limb rotation test a Athletic shoulder test a CKCUEST a Single arm medicine ball throw b
Volleyball	Single arm medicine ball throw a Athletic shoulder test b Upper limb rotation test b
Swimming	Upper limb rotation test b Posterior shoulder endurance test a
Tennis	Single arm shot put test a Athletic shoulder test a Single arm medicine ball throw b Upper limb rotation test b
Baseball	Single arm medicine ball throw b Upper limb rotation test b Posterior shoulder endurance test b
Rugby	CKCUEST a Athletic shoulder test b Countermovements push-ups b Upper quarter Y balance test b

^a between 75 and 85% of consensus.

^b > 85% of consensus.

Discussion

The objective of this Delphi study was to obtain a consensus on the usefulness and the interpretation of upper limb functional tests that have been described in literature. Experts from different professions (medical doctors, fitness trainers, physiotherapists) and working in a large variety of upper limb sports were recruited for this purpose. They all had an expertise in sports medicine and/or sports training for at least 5 years and were familiar with upper limb functional tests in their clinical practice. In the first survey, they listed the criteria that were required for a good functional test.

The first criteria was the reliability (quality of performing consistently well) of the test. Since this variable has been previously explored in literature, the current study did not focus on that. The reliability of each test is presented in Appendix 1.

Then, the tests needed to be valid (measures what it claims to measure) and sport-specific. The PSET as well as the SSASPT and the SAMBT were considered as the most valid tests according to the experts. The PSET has been demonstrated to specifically assess the endurance of posterior shoulder muscles (posterior deltoid, middle trapezius, infraspinatus) ¹⁶ and would be appropriate for swimmers as repetitive movements with moderate to high intensity are performed for a long period in this sport. This test is also performed in the same plane as swimming movements, which makes it quite specific to the gesture. However, this test does not appear to be appropriate for assessing shoulder endurance in overhead sports which solicit the shoulder in a cocked position. For that purpose, Declève et al. ¹⁷ recently developed a new functional test aimed at assessing shoulder endurance in a cocked position: The Shoulder Endurance Test. The validity of this test will have to be further explored throughout the upcoming years to define whether this test can be considered (or not) as the reference test for assessing shoulder endurance in sports requiring repetitive cocking movements. The SSASPT and the SAMBT specifically assess shoulder power in a functional plane, close to the sports gesture. As the SAMBT is performed from a cocked position, this test is recommended by the experts for all sports requiring a cocking movement such as handball, volleyball, tennis, and baseball. In the literature, a correlation between 0.76 and 0.79 was found between the SAMBT and field performance in javelin throwers, ⁴ which shows the sport-specificity of this test. A significant and important correlation was also found between this test and isokinetic internal rotators strength in the same population (r range ≤ 0.93). For the SSAPST, only a moderate correlation was found with softball performance (0.45–0.46). However, this correlation is less important than the one found for the SAMBT, which makes this test less sport-specific than the other. Nevertheless, the gesture of this test appears to be close to some components of the tennis serve.^18,19 That is why this test was recommended for tennis players by the experts. But the relationship between this test and tennis performance will hopefully be confirmed by the literature in the future. A good validity was also reported for the ULRT, a test performed both in closed- and open-chain and which includes trunk rotation and cocking movements. Since trunk rotation and cocking movements have to be performed in overhead sports, this test seems to be appropriate to assess shoulder/trunk stability as well as coordination in overhead athletes. Furthermore, this test was suggested for tennis, handball, and volleyball by the experts. The AST was considered as moderately valid by the experts (53%). Even if practiced in closed chain and only explored in rugby players up until now, this test has been recommended by experts in tennis, volleyball, and handball too. In volleyball or in handball, the test can be assimilated to the blocks or the contacts with the opponent and assesses the stability of the shoulder in a high and instable position. In tennis, it can be related to the impact with the ball during the serve. Future studies are necessary to determine the usefulness of this test in other sports.

On the other hand, the OAHT and the UQYBT were considered as the tests with the lowest validity (inferior to 50%). Effectively, those two tests assess shoulder and kinetic chain stability in a closed chain task but are far away from any sport gesture. Westrick et al. ⁷ demonstrated a moderate correlation between the UQYBT and lateral trunk endurance (r range = 0.35–0.45). Borms et al. ²⁰ found a correlation between the supero-lateral score of this test and external rotators strength at a speed of 180° per second. However, no other relationships have been found between the two tests and shoulder or kinetic chain performance as well as with shoulder injury, which makes the usefulness and the interpretation of these tests limited. Moreover, the OAHT is a very challenging test, which makes its use only possible with a certain category of people (with a high physical capacity).

Finally, a good functional test has to be easy to implement in practice. This is the case of the majority of the tests suggested in this Delphi study. The first exception is the AST. This test requires a force platform to measure the rate of force development, which is difficult on the field. An alternative test with the use of a handheld isometric dynamometer instead of the platform was suggested by the experts but this option has yet to be validated. The second exception is the CMP. Even if this test can be performed on the field, data cannot be collected without a force platform or tri-dimensional markers. The data collected this way is more accurate but these tools are not readily available to everyone.

From a clinical point of view, the majority of the tests described in the study seems to be useful in practice but have to be chosen correctly based on the objective of the assessment. According to the experts, the SAMBT, the CMP, and the AST would be the most appropriate tests to assess performance in athletes while the CKCUEST, the AST, and the ULRT would be the most appropriate to assess injury risk/return-to-play. If the objective is to assess upper limb power, the SMBT, the SASPT, the SAMBT, the OAHT, and the CMP can be performed. Depending on the level of athlete, the characteristics of the test or the sport (closed- or open-chain sport) one or two of these tests will be chosen. Since the SMBT does not allow bilateral comparisons, ⁹ and the OAHT has a low validity, ²¹ the SSASPT ²² and the SAMBT ⁴ should be chosen in priority to measure plyometric qualities in open-chain sports. As for the CMP, this test is interesting to assess power in closed-chain sports ²³ but requires specific material to analyze data. The SAMBT as well as the ATS will be useful in assessing global upper limb strength too. ¹¹ The CKCUEST, the ULRT, the AST, and the UQYBT are aimed at assessing shoulder stability in different planes of motion. Since all of these tests are performed in closed-chain, they are more appropriate for closed-chain or contact sport. The first exception is the ULRT, which includes both closed- and open-chain movements and can be appropriate for both closed- and open-chain sports. The second test is the AST, which assesses the shoulder in a position close to a “blocking” gesture in volleyball or the position of the impact with the ball in tennis serve. However, a lack of validity and of sport-specificity of the UQYBT was highlighted by the experts. Its use may therefore be reconsidered in practice. The PSET is used to assess the endurance of the posterior muscles of the shoulder, especially in swimmers. The Shoulder Endurance Test ¹⁷ should be further explored to assess shoulder endurance in volleyball, handball, and tennis players.

For an overall assessment of performance and/or risk of injury for an athlete, all of the previously described aspects (power, strength, endurance…) have to be considered. Since no test combines all of these aspects, the implementation of batteries of tests is obvious. Olds et al. ²⁴ were the first to suggest a “return to sport” battery of tests to assess upper limb function. Though this battery includes varied exercises, in open- and closed chain, the later was, according to the experts, not sport-specific and some exercises were considered as too easy for some athletes. That is why sport-specific batteries of test, closer to sports gesture (like those described previously in the manuscript) will have greater use in practice. Suggestions for batteries of tests were described in the current study. In handball and volleyball players, the ULRT and the SAMBT would be recommended to be performed to assess performance. To assess injury risk or return-to-play, these two tests as well as the CKCUEST and the AST would have to be done. In swimming, the experts would recommend performing the ULRT and the PSET both for performance and risk of injury /return-to-play. In tennis, the SSASPT, the SAMBT, the AST, and the ULRT will have to be done for performance and risk of injury /return-to-play too. In baseball, the SAMBT, the ULRT, and the PSET would be recommended. For rugby, closed kinetic chain tests such as the CKCUEST, the UQYBT, the AST and the CMP would be preferred, both for performance or risk of injury /return-to-play.

Although the characteristics of the previously described tests seem to be consensual among the experts, their interpretation still remains controversial for some of them. In the current study, experts reached a consensus on the “percentage of difference considered as clinically relevant (between limbs or between sessions)” for some of the tests presented. However, it was not the case for the UQYBT, the AST, the ULRT, the CMP, and the PSET. The relationship between the results of these tests and performance (or injuries) is not clear, which makes their use complicated in practice. Is it relevant to consider a certain percentage of bilateral difference, a percentage relative to normative values of athletes at the same age and same discipline or should comparisons be done with an absolute value? The debate is ongoing. However, each test is different and considers different variables. So, the answer may be different for each one. Moreover, as the tests are quite new, further studies will be necessary to provide normative values (based on the age, the gender, and the sport) for each test to make their interpretation easier in practice.

Limitations

Firstly, a consensus was not found for all the questions and considerations, especially for the interpretation of some tests. A meeting could have been organized between the experts to answer to the items that lacked consensus. However, since worldwide experts were recruited and the Delphi Study process was conducted during the Covid-19 pandemic, this would have been complicated to organize.

Secondly, the panel of the experts solicited was only chosen by the research team, based on the inclusion criteria. However, some qualified experts may have been missed because they were not known by the members of the research team.

Lastly, the process of the Delphi Study was quite long. The first two questionnaires included a lot of questions and took more than 30 min to be answered. This was mentioned to all experts at the beginning of the study, before their inclusion. However, some of the identified experts may have refused to participate in the study because it was too time consuming.

Conclusion

Functional testing has been increasingly used in screening athletes, both for injury prevention and performance objectives. The objective of this Delphi study was to find a consensus on the usefulness, the qualities, and the specificities of each test as well as the best way to interpret their results.

The tests suggested in this Delphi study can be divided into two categories: the open- and the closed-chain tests. The SMBT, the SSASPT, the SAMBT, and the PSET are considered as open-chain tests. The UQYBT, the CKCUEST, the OAHT, the AST, and the CMP can be considered as closed-chain tests. The ULRT cannot be assimilated into one of those two categories as one arm is working in a closed-chain and the other in open-chain during the test. In terms of sport-specificity, open-chain tests would be more appropriate for open-chain sports like volleyball, baseball, or handball while closed-chain tests would be more useful for closed chain sports like swimming or rugby. The first exception is for the AST where the way in which the shoulder is solicited is close to blocks in volleyball, contact in handball, or impact in tennis serves. The second test is the PSET which is interesting for swimmers as good posterior shoulder muscle endurance is required in this population. As for the ULRT, this test would be appropriate both for open-chained sports that require a cocking position (volleyball, handball, tennis, baseball) or closed-chain sports like swimming. However, since each test has its specificities and assesses different qualities, when evaluating performance, risk of injury or return-to-play, the combination of multiple evaluations in sports-specific batteries of tests is more pertinent than the use of one single test.

Prior to their use, the experimenter must ensure that the test(s) chosen is(are) not too difficult/challenging for the athlete. Tests like the OAHT or the CMP may not be suitable for all categories of athletes. The metrological qualities (validity, reliability) of the tests will also have to be considered before implementation. Finally, in a case of return-to-play decision after a surgery, a sufficient rehabilitation time will have to be given before performing functional tests as they may be too soliciting in the first months following a surgery.