Return-to-sport criteria is a list of accomplishments a person must “check-off” in order for me to recommend they return to sport without restrictions. As simple as it sounds, there’s still a lot of disagreement in the sports medicine community. So, the purpose of my post is to share what I do, and why I think it’s important; not to tell you how you should test your athletes. My protocol sometimes deviate from my mentors’, meaning I may hold more weight in certain tests than they do. Although I have these preferences, I run through my work’s standardized protocol to be consistent among my colleagues; that’s important.

I am very open to discussing and changing my thoughts as research comes out and is presented. In fact, it’s pretty refreshing that my current job has several meetings a year to go over protocols and our return to sport criteria — they’re always seeking to be the best and open to changes as evidence emerges.

Obviously before this final test, they would hit other goals, test return-to-run criteria, and get some graded exposure to practices and skill drills before they reach this point. So, here is what I would test for their final greenlight to return with no restrictions.

Time

• I’ll follow the specific surgeon’s timeline-based testing permission, but below are ranges I use when surgeons trust me to determine when to test. If I get a generic protocol, I would talk to the surgeon directly to see what they think of my own personal timelines. If you’re a PT, remember, the surgeon is the only one who truly saw what was inside the knee joint, the tissue quality, and what was done to the knee. Don’t assume all knees are the same. Work as a team and discuss things with your surgeon. They are truly masters of their craft and the anatomy of the human body!
• Below are the most common injuries/surgeries and time criteria:
  • Meniscal repairs: 3-6 months
  • Menisectomy: 6-8 weeks
  • ACL reconstructions: 9-12 months, at least 12 months for athletes under 18
• Other’s not mentioned: MPFL reconstruction, non-op PCL, LPFL reconstruction (due to bipartite patella), multiligamentous reconstruction, MCL reconstruction, patellar tendon repair, ACI, OAT, microfracture, and many more.
• Also, if they had a significant bone bruise or had some chondroplasty performed during surgery, I educate them that their pain and effusion may make things slightly longer and/or more irritable throughout rehab.

ROM

• Equal bilateral
  • Some say within 10% is good, but I’m a stickler on this. There’s many studies showing that a lack of equal extension puts a person at risk for early onset osteoarthristis. Additionally, a flexed knee puts more low load prolonged stress on the patellofemoral joint during standing, walking, running, and jumping. Combine that with a weak quad (common after knee injury), and it may explain why anterior knee structures get irritated when full ROM has not been obtained post-operatively. For flexion, I like them to have at least 135 (“normal”), but prefer they get 140-145 since most athletes get put in tough positions or may require deep squatting. I’m more lenient with flexion since they often get it back later regardless…but 135 being the minimum requirement.
  • Also if you don’t have full ROM by the time you do return-to-sport testing, you likely have other issues going on.
  • If there was a long duration between injury and surgery, it may be hard to get good ROM quickly if they lacked quite a bit before. Ask your surgeon or read the operative notes because surgeons often move the limb in full ROM to double check the surgery while the person is under anesthesia…Prehab is a beautiful thing.

Effusion

• Less than 10% girth between limbs before, during, and after exercise
  • Effusion often means the joint is irritated due to too much load. If they have increased effusion, it wouldn’t make sense for me to let them return to the demands of sport let alone even expose them to the stresses during return-to-sport testing.
  • It’s hard to determine outside joint swelling versus inside joint effusion, but I often feel they have joint effusion when it’s months out whereas swelling occurs closer to the operation. My biggest treatment for them is education to minimize load on the joint; ice can help with pain, but won’t help much with inside joint effusion.

Balance

• Single leg stance with knee bent at 60 degrees flexion for 30 seconds
  • 60-degree knee flexion is very demanding of the quadricep muscle. I like to see how well they maintain balance qualitatively, through time, and also under quad fatigue. I wouldn’t mind going for a full minute, but I feel 30 seconds is an accurate look. I feel 10 seconds is not enough to induce mild fatigue.
• Y-balance anterior reach <4 cm difference
  • This is consistent with the literature that showed patients with knee injury reported high function (not reduced injury risk) with < 4 cm difference. People are different sizes, so it’s hard to say 4 cm for everyone, however I have not come across other research stating a percentage is better.
• Y-balance posteromedial and posterolateral reach <6 cm difference
  • I do this for the same reason as the anterior reach. With knee injuries I hold much more weight in the anterior reach just because I feel it demands more quad effort, and more research finds consistency in anterior reach compared to other directions. Most people also go with 4 cm since anterior reach has been studied more, but I think there’s many other factors when doing frontal/transverse reaches so I am a bit more lenient.

Strength

• Manual muscle test 5/5
  • Just kidding. Manual muscle test is a test of weakness, not strength. They should have hit 5/5 long ago. There’s literature that shows there could still be up to a 25% difference in strength when  a “5/5” MMT is compared to handheld dynamometer values on an injured and non-injured shoulder. Now translate that to a bigger quad muscle and a better reference standard of a isokinetics. I’m willing to bet the difference that a MMT can NOT detect is much bigger in the quad than the shoulder.
• Hip strength <10% difference measured with handheld dynamometer
  • Knee pathology and injury often shows hip weakness (not clear on whether it was the chicken or the egg), but I find it’s worth checking. This isn’t a make or break for me unless I also know balance deficits, or there’s a huge deficit (>25%).
• Isokinetic testing
  • Gundersen is the home of George Davies, a sports PT who pioneered isokinetics testing in rehab. Obviously me and the people of Gundersen are very bias towards isokinetics, but I also believe it gives me a lot of meaningful information that helps me decide on returning a person to sport and exercises selection. Below are definitions and norms in each area.
  • 
  • Peak Torque: I want to see that they are <10% equal in the max amount of force they can produce. Regardless if they’re male or female, I encourage them to hit 100% equal although I would clear them if they meet above criteria
  • Peak Torque to Bodyweight Ratio: I hold the most value here. In order for someone to cut, change directions, and/or decelerate, they must be able to stop their own body. There’s some research that shows that people often injure their ACL when their leg is less flexed and their center of mass is far from the foot contact. This tells me that they should be strong enough to stop the weight of their moving body in a flexed knee position (eccentric knee flexion). The quads need a ton of strength to allow good knee flexion while pumping the brakes. Again, I would pass them if they hit the criteria, but regardless of gender, I tell them keep working their tail off. If their UNINVOLVED limb doesn’t meet criteria, I tell them frankly they need to get stronger overall…and tell them how of course.
  • Ham quad ratio: This is important because the hamstrings are a dynamic stabilizer to the ACL. I only hold weight in this value once their peak torque is within 10% because it may be falsely high (When quads are weak it falsely shows a higher ratio).
  • Acceleration time: There is no normative data or literature to this, but I like to see this equal and below 100 msec. They say ACL injuries often happen within 100 msec, so I like to see it below that time. Although most people see “acceleration” as pushing and going, I think of it as quickly developing enough force to stop the body (yes, technically rate of force development, but we only get that with isometric testing). I’m likely wrong here, but I won’t ever let it hold someone back…just something I monitor.
  • If you do not have a isokinetic machine to test, you can use a belt-fixed handheld dynamometer to measure a isometric contraction. When compared to isokinetics, the handheld dynamometer can overestimate by 20%. This means for them to pass, they should have 100% equal or more strength than their uninvolved limb.
  • You can get weakness in the uninvolved limb within 2 weeks of injury. Just a consideration since many compare involved limb strength to the uninvolved  limb even if it may be weaker than pre-injury level.

Functional Tests/Movement Quality

• Single leg step down x20 each side
  • Some people do double limb force plate observation, but most sports tasks are single leg (running, hopping, cutting, etc). I feel I get the same information, if not more, when forcing them to perform on one leg. I also like the 20 reps because, again, they’re moving into higher reps which may induce mild fatigue both physical and mentally (to consciously keep proper form). I look for heels staying down, dynamic knee valgus, and contralateral hip drop; those are what I consider “bad”
• Single leg hop tests <10% difference in LSI and 90% of their height
  • From what I’ve read, these hop tests do no predict risk of injury, they just correlate well with how a person functions (patient reported outcome measures). I use this more to see their willingness to do a single leg hop. It is a single hop, which is fairly tame and controllable.
• Crossover hop tests
  • There are studies that show more than 2 single leg hop test variations don’t tell you much more information about a person’s current functional level or risk of injury so I chose the crossover and the single hop because the single leg hop is a good intro and warm up, while the crossover adds a lateral component. I look symmetry of distance and quality of landing. I use 2-Dimensional slow motion camera maybe 10% of the time because it can be useful, however there can be many errors and flaws with using a 2-D system.
• Maximal controlled leap test (not yet added permanently)
  • Dr. Michael Reiman presented this test at CSM this year during his hip course. It involved jumping with one leg, then landing with the opposite leg. Since you uninvolved leg should be able to propel you further, your involved leg will be challenged in controlling your “normal” forces. On the other side, If your involved leg does not propel you as far as your uninvolved, it may indicate issues in either strength or power of the injured leg. There was also another study presented during CSM that showed a triple hop test is more of an indicator of ankle propulsion rather than knee function, meaning the more common hop test may not actually even test true performance measures in the involved knee. Although I have yet to review Reiman’s cited article from JOSPT, I may be adding this test in, or replacing the crossover hop test once I do. I’ve already added it in while testing a few patients this week.
• Single leg vertical jump test (not yet added permanently)
  • There was a single recent study that showed single leg vertical jump was correlated with quadricep isokinetic strength. No other functional hop test has been shown to correlate with quad strength specifically to my awareness. I have only used this once, and measured through the MyJump2 app which has been validated against the Vertec.

Cardiovascular 

• Lower Extremity Functional Test (LEFT)
  • This test is used to observe movement, but I feel people are moving quite fast to observe it accurately. I like it because it involves pivoting, cutting, and twisting, all during a max effort task. I like that there are norms too (high school and division 3 athletes). It’s unlikely that you may have a baseline of this test, so having a test with norms is useful. Other test like beep test, shuttle run, and T-Test are good, but I like how this incorporates all the difference movements. Also, George Davies made this which is why we at Gundersen are bias towards this test. No, I don’t have many linear sport athletes do this.

Fear

• ACL-RSI greater than 80%
  • Fear has been shown over and over to be a limiting factor in return to sport yet less than 10% of PTs (of 2000 ortho and sports section members surveyed) measure it in any formal way. I did not use a questionnaire until this year, which I admittedly had to force myself to use to stay consistent with evidence. I could not find any good numbers, so I chose a score of at least 80%. Generally the score improves as rehab progresses anyways, so no extra intervention may be needed. However, if it is “low”, I put them through the ringer in rehab and educate them on what they’ve accomplished in rehab. I’ll also then stay positive about any concerns they might have without giving them the feeling that I am pushing them back into sport. Sometimes you may have to take one for the athlete if the athlete is mentally not ready.

Other

• Acute on Chronic Workload Ratio
  • There is a very big difference in performing 100% for a day vs. several days a week at practice and during a game. From day one of their injury, athletes are quite challenged by me in rehab. By about the 2nd month, they are working frequently with their strength coaches. It’s important to continue the athletes workload to ensure they can return to the high demands of practices and games once they’re “cleared”
• Normative Data
  • Allometric scaling means all of a person’s score on test such as strength, distance hopped, and other measurements should be relative to their body attributes (height, weight, etc). Although it is a recommendation, it by no means indicates a person is at less risk of an injury.
  • For certain populations, you can find normative data in performance measures for athletes sorted by competition level, gender, age, sport, and many more. If an athlete falls short of those normative data, it is not a make or break, but I do tell them they need to work harder to match the level of their peers. If a division 1 linebacker reps a 400 pound squat, it would be smart to still encourage them to work hard in the offseason to reach those numbers. Many athletes return to sport, but not prior level of performance, therefore I feel it is important to keep this in mind along with collaborating with their strength and conditioning coach. Why aren’t I a lot stricter? Most often times I don’t have their baseline strength, meaning what if they were weak even before injury? If they were only squatting 200 pounds before their injury, would I hold them back if they couldn’t squat 400 pounds AFTER surgery? Lastly, normative data does have it’s flaws. Sometimes the cohort you are looking for isn’t available. Sometimes there’s factors in cohort data that makes it invalid. These are considerations, not recommendations.

I hope this was helpful for you all to either consider tests you were not aware of, or at least find some confirmation or challenges to reasoning on how you developed your return to sport criteria. It was a fun way for me to critique myself as I typed all this. Mine isn’t the correct way to do it, but I wanted to give a glimpse into this head of mine. Do you feel I have some flaws? Gaps in evidence? Poor logic? Outdated norms? Let me know in the comments, or you likely know how to contact me through email or phone. I’m happy to discuss anytime…I oughtta find a hobby sometime :).