Should you ice your knee after reconstructive ACL surgery?

Posted on July 19, 2017

(Last Updated On: July 19, 2017)

Six years ago I wrote Should you ice an injury?

  1. Icing doesn’t do anything for the thing it’s most purported to help, swelling
  2. Icing can decrease pain

That wasn’t new. Twelve years ago in,

Cryotherapy after ACL reconstruction: a meta-analysis

The same results were found. No benefit for swelling, no benefit for range of motion, but can help decrease pain. Yes, attempting to numb the hell out of an area can lessen what you’re able to feel, pain included. So can jumping your naked body in a mound of snow. That doesn’t mean it’s worth doing.

Three years ago, as if we needed another review, in,

Effectiveness and safety of cryotherapy after arthroscopic anterior cruciate ligament reconstruction. A systematic review of the literature

same results.

Despite this, ice seems as prescribed as ever. The inevitable rationale can be found in the abstract of the first study above,

“As the cryotherapy apparatus is fairly inexpensive, easy to use, has a high level of patient satisfaction, and is rarely associated with adverse events, we believe that cryotherapy is justified in the postoperative management of knee surgery.”

That’s been my thought process for a long time now too. People ask should they ice whatever ailment, and I always tell them “Don’t feel like you need to do it. If you like icing, then sure. It can help decrease pain a bit, making an area numb tends to do that :), but it’s not doing anything else.” Where if controlling swelling is the goal, 1) look at your training / rehab 2) wrapping is where you want to go to. (How to wrap the knee.)

Then I wasn’t so sure.

Quick primer on ACL surgery

Multiple holes are drilled to make room for the new anterior cruciate ligament.

ACl surgery drill holes

We then pull our new ACL, the graft, in the holes, which you can see above in the middle.

To get the graft snug and keep it in the hole, we often double it so we can loop it and button it on the top at the side of the femur.

This way we can pull the graft through the bottom at the tibia, but it will stay put at the top.

acl surgery endobutton with labeling

To get this bottom part snug and keep it from moving, a button won’t get the job done. Hopefully you can imagine if we button both ends similarly, then we won’t keep tension in the graft. In the picture above the top of the graft (by the loop) is tensioned, but that’s only because we’re pulling the bottom. We need to keep the bottom position where it is until, over time, bone can grab the graft and keep it there.

So we place an inteference screw to do just that: interfere with the graft from moving.

ACL Surgery Screw 1 ACL Surgery Screw with graft ACL Surgery Screw with graft with lines ACL tibia tunnel simulation

This is crucial. The graft cannot be too loose, or loosen too much, because that effectively cancels out the surgery. When we tear our ACL it’s now so lax as to be useless, whether it’s truly torn or not. We don’t want the graft to end up useless as well.

-> Some may actually injure their ACL but not tear it, but it ends up so stretched out it might as well be torn. (Plastic deformation.) This happened to me, and provided a nightmare in getting an accurate diagnosis. Orthopedist “Your ACL is torn.” Radiologist “Your ACL is intact.” Fifth doctor finally gives correct phrasing “Your ACL is intact, but stretched out so much it might as well be torn.”

Thermal effects

Thermal expansion is a material response to heat. For the most part, when material(s) get warmer, they expand. Engineers build this into structures like bridges. They for instance may give a little extra space for those days when it gets hot.

concern for thermal expansion bridge

When things get colder the opposite tends to happen- constriction.

Because of this, there is something called shrink fitting. From Wikipedia,

“a technique in which an interference fit is achieved by a relative size change after assembly. This is usually achieved by heating or cooling one component before assembly and allowing it to return to the ambient temperature after assembly, employing the phenomenon of thermal expansion to make a joint. For example, the thermal expansion of a piece of a metallic drainpipe allows a builder to fit the cooler piece to it. As the adjoined pieces reach the same temperature, the joint becomes strained and stronger.

Other examples are the fitting of a wrought iron tyre around the rim of a wooden cart wheel by a wheelwright, or of a steel tyre to the wheel of a railway engine or rolling stock. In both cases the tyre will be heated and expands to slightly greater than the wheel’s diameter, and is fitted around it. After cooling, the tyre contracts, binding tightly in place.”

In the book Surgical Atlas of Sports Orthopaedics and Sports Traumatology I caught a picture of this, when press fitting a new piece of bone into a defect area.

Very much recommend this book.

Very much recommend this book.

Idea being you cool the to be inserted bone plug so it constricts. Once inside the defect area though, once it gains some heat back, it will then expand to have a more snug fit.

Back to ACL surgery

Different materials have different coefficients of thermal expansion. Not all things expand or constrict the same amount for a given temperature increase.

Hopefully we can see the potential problem now with icing. If we go changing the temperature of an area, like our knee after surgery, we give rise for the potential to expand or constrict our knee AND the interference screw, but not at the same rates.

This presents an issue because what if the screw, in response to icing and a lowering of temperature, is constricting more than the bone tunnel? That’s going to increase the amount of space in the tunnel. Not good.

Expansion seems less an issue-

  1. People don’t really heat an area, such as using a hot pad, after a surgery
  2. If the screw expands more than the bone, it will expand into the graft and bone. Probably a good thing.
  3. If the bone expands more than the screw, the bone may still expand into the screw. (The bone expands all directions, including into the tunnel.) Probably a good thing.

The one caveat I can see here is if the screw is not completely in the bone. If it’s more like this,

Acl inteference screw and graft diagram copy screw lower

In that case, if the bone expands a good deal, could it start to push the screw out a bit?

Let’s try to stop some of the guessing here and get more specific. We’re going to get into weeds for a minute but come back out of it.

Regardless, right now we can see, and likely expected, the values of plastic and bone to not be the same.

Let’s take a worst case scenario and say we’re using a PEEK screw with a thermal expansion coeffcient of 50. This gets a little hazy, but this means 50 *10^-6 meters per degree kelvin. Checking unit conversions, this is equal to 50 * 10^-6 meters per degree celsius.

Again, for bone we’re at 27.5 *10^-6 meters per degree celsius.

This begs the question, how many degrees does our knee cool down when icing? Amazingly, there is research on this,

Cryotherapy Decreases Intraarticular Temperature after ACL Reconstruction

The skin goes down much more than inside the joint. This study found a 26 degree decrease from using an ice pack after only 15 minutes-

Skin temperature response to cryotherapy

The screw is often right by the skin, but inside the knee to some extent. We’ll go with the in the joint study for now, where a 6 degree celsius decrease in the temperature of the knee joint after a one hour session was found.

  • PEEK screw: (50 * 10^-6) * 6 degrees celsius = 0.0003 meters
  • Bone: (27.5 * 10^-6) * 6 degrees celsius = 0.000165 meters

So now we know how much the screw and bone constrict based on a one hour session of icing. Relative to one another,

  • 0.0003 meters – 0.000165 meters = 0.000135 meters

That’s 0.0053 inches or 0.14 millimeters.

However, the above is only linear expansion. We’re talking about tunnels here, so we’re interested in the volume difference. Volumes expand or contract three times as much as lengths.

  • 0.14 millimeters * 3 = 0.42 millimeters

Meaning when we ice our knee for an hour we’d be talking about potentially enlarging the tunnel by 0.42 millimeters.

The money question of course is does this amount matter?

According to the Surgical Atlas book referenced earlier, for the tibial tunnel, a screw 1 millimeter larger than the drilled channel will be inserted. We then need to consider the graft is also in this tunnel. It looks like we already have some margin for error built in. To where unless we are causing a 1 millimeter difference in size from temperature change, we’re probably fine.

For kicks, let’s see what that temperature change would have to be.

Just earlier we had:

[(50 * 10^-6) – (27.5 * 10^-6)] * 6 degrees = 0.000135 meters

Now: 1 millimeter is 0.001 meters. Some algebra:

[(50 * 10^-6) – (27.5 * 10^-6)] * 3 * Temperature Change = 0.001 meters

-> 3 is for volume

3 * Temperature Change = 0.001 meters / [(50 * 10^-6) – (27.5 * 10^-6)]

3 * Temperature Change = 44.4 degrees celsius

Temperature Change = 44.4 / 3

Temperature Change = 14.8 degrees celsius

That’s a pretty big amount. Though again, a 26 degree difference on the surface of the skin has been found. If the screw is poking the skin out like this (this is not uncommon)…

Acl inteference screw and graft diagram copy screw lower

When we consider this appears to be a worst case scenario as we used the version of PEEK plastic with the most disparate coefficient from bone, that the screw should be a millimeter bigger than the tunnel anyways, with a graft snug in there too, that icing is rarely found to have adverse events, we’re probably ok. Though I think this is reason enough to more thoroughly study this. Part of me wonders if the screw is poking out like that is it as resilient to temperature as the body? The body has a blood supply and mechanisms to keep things warmer. The screw does not. In vivo, studied together, things could change.

One wants to tilt every possible probability in favor of success with this surgery, since it’s such a hard one to come back from. While we haven’t taken a ton away from icing here, we certainly haven’t added any more reason to do it. Considering there was barely any reason to do it to begin with, if it’s my knee, I’m not icing after this surgery (which I didn’t.)


Want more help with ACL recovery? Take a look at The most important phase of ACL rehab.


Subscribe to Blog via Email

Enter your email address to subscribe to this blog and receive notifications of new posts by email.