How often is “bone on bone” bogus? (Member Post)

Bares bones touching bare bones

“Bone on bone” can be a bona fide accurate description. Sometimes severe cases of arthritis really do result in the loss of most or all cartilage, and BOB is literally true, not even an exaggeration. Cartilage can in fact go away. Naked joint surfaces can slide on each other.

There are various grading/scoring systems for osteoarthritis, and they all have a “worst” tier that shows things like “marked narrowing of joint space, severe sclerosis [scarring, hardening of tissues], and definite deformity of bone ends.” The Kellgren-Lawrence (KL) system is probably the most common.

Knee joint X-rays for each KL grade according to OA severity. The KL grade system categorizes grades according to two aspects of OA: the extent of joint space narrowing & the size of osteophytes (spurs). From Ensemble deep-learning networks for automated osteoarthritis grading in knee X-ray images, by Pi et al., CC BY 4.0.

Weirdly, cartilage may be a little more optional than we think, because not all of this severe osteoarthritis is painful, or not very. This is partly because the sclerosis can probably be effective, much like calluses on hard-working fingers. It’s also probably a function of complex metabolism and neurology, with some people being much less likely to hurt in exactly the same structural circumstances that are agonizing for others.

So wouldn’t it be tragic if “bone on bone” scared such a patient who’s actually mostly fine? More pain, and unnecessary surgery? If “imaging is not needed to diagnose osteoarthritis” (Hunter) — because it’s actually just pain and disability that matter, not the tissue state — then do we really need to say what the joint looks like? Why bother, when it’s apt to make their symptoms worse, or discourage xx exercise?

It’s important to consider the possibility that those words should never be said even to patients who actually do have severe osteoarthritis. Is it kind or helpful to describe these cases in that way? More about that in part 2. For now I just want to make it clear that the factual accuracy of BOB is not the only consideration here.

Sometimes, however, BOB is not even true. We need to consider that first.

“Bone on bone” also isn’t always accurate

How often do professionals exaggerate the severity of osteoarthritis by invoking “bone on bone” when it isn’t actually true? Or not nearly true enough for those dramatic words?

Unfortunately, we don’t know the “exaggeration rate” without some research that will probably never happen. But we can make an educated guess based on some clues:

• There’s plenty of expert and patient opinion about this.
• Not all BOB is created equal, so it can obviously be technically true without being true in spirit.
• Osteoarthritis grading is known to be error-prone and imperfect, and therefore vulnerable to bias and “spin.”
• There’s a clear profit-motive for exaggerating, and we know that many orthopaedic surgeries are over-prescribed and so likelier to be sold by inflating the need.

And now in more detail…

The anecdotal and soft evidence that BOB is overstated

• It’s a common clinical impression that BOB is often an exaggeration. “It happens far too often,” writes Dr. Howard Luks, an orthopedic surgeon, responding directly to my question about this. “If I had a dollar for every time an orthopedist told a patient it was the worst knee OA they ever saw I would be rich,” writes Jeffrey Fusilier, Doctor of Physical Therapy. (Of course, cynicism can also be exaggerated! But these clinical suspicions probably reflect at least some truth.)
• Patients often report that they have been told they were BOB only to substantially recover. An example I know well: my father. He refused surgery despite occasional episodes of severe pain. Since then, with regular exercise and some help from canes and scooters, his knees have been better — flare-ups are now rarer and milder! His original x-rays from don’t look like KL 4 knees to me: either he was not very BOB in the first place, or he adapted to it better than most people think is possible, or a little of both.
• The sheer variety of obviously nocebic statements of all kinds reported by patients is damning. It’s clear that melodramatic clinical communication happens. Exaggeration is something humans do quite a bit of for many reasons!
• Bunzli et al reported that all of the 27 people in their study of beliefs about knee osteoarthritis were sure that their knee pain was “bone on bone,” which shows at least that it’s a near-universal mental image (in people headed for joint replacement). Most of that belief probably came from doctors, and it’s unlikely to be equally true in all cases. The more universal the idea, the more it probably differs from messy reality.
• We tend to talk about BOB like it’s all or nothing, but the cartilage might be gone only in a tiny spot, rather than a large area, like the difference between a little bald spot on the top of your head versus Vin Diesel bald. This is the main mechanism by which BOB can be an exaggeration without actually being a “lie”: that is, it can be technically true, but just not in the same league as the next patient. Surely formal, expert grading systems take care of all this? Um, about that…

Surgery isn’t always a bad option, but surgeons & patients alike reach for it too readily. “Pigs oink & surgeons cut!” The words “bone on bone” seem finely tuned to emphasize the putative necessity of surgery & to teach patients to fear the very thing that we most need, even with severe osteoarthritis: exercise.

Harder evidence that BOB is often overstated

• Osteoarthritis grading is as much art as science. It “depends on the clinician’s subjective assessment” and the accuracy of that assessment “varies significantly depending on the clinician’s experience and can be particularly low” (Pi et al). A 2018 study showed lousy agreement between clinicians about the same images: “None of the studied OA grading scales showed acceptable reliability. The evaluation of patients with OA should not be dependent on radiographic findings alone; clinical findings should also guide the treatment and follow-up” (Köse et al).
• Extremes are easier though, right? 100% of people would agree that Vin Diesel is bald, because he’s super bald. So surely arthritis grading is more reliable with severe osteoarthritis? Yes, but still well short of perfect. Klara et al showed that even lowly medical students could usually see the same knee nastiness as an experienced surgeon. But not always.
• Other joints are much harder to grade than knees. For instance, KL grading “is not applicable to the subtalar joint” (Fawzy et al).
• If we “follow the money,” it’s obvious that there’s a profit motive for exaggeration. One of the best ways to sell treatments is to dramatize the seriousness of the need, to warn people. Many medical warnings are justified, of course — but the unjustified ones benefit from that legitimacy. Clinicians can easily rationalize getting overzealous with their warnings (“better safe than sorry”). And so it’s likelier that moderately severe arthritis probably gets “rounded up” to BOB. It’s probably just a matter of time before it becomes true, amiright? Or patients are “frankly” told that BOB is inevitable.
• Speaking of profitable procedures: surgeries for arthritis are definitely over-prescribed, particularly in the United States, where they are most likely to be profit-motivated. And if you’re recommending too many surgeries, you’re probably also trying to justify them… and a bit of BOB hyperbole is a handy way to do it. But is there good evidence of over-prescription? Whether surgery is “over” prescribed is relative to how much it should be prescribed, which is a squishy comparison! However, when good clinical trial evidence clearly shows that a procedure just doesn’t work, or not very well, and probably should only be prescribed quite rarely if it all, then we should see the industry respond by doing much less of that procedure — and if it doesn’t, that constitutes strong, direct evidence of over-prescription. And we do indeed see that kind of evidence for all arthroscopic procedures in general, because they are virtually all “disproven” for arthritis for years now (Siemieniuk et al). Persistently high rates of prescription are especially clear for meniscectomy (Degen et al), debridement (Wasserburger et al), and “knee lube” injections (viscosupplementation; Zhu et al). The most important surgical prescription for BOB is knee replacement, and we lack evidence of that kind for that specific procedure. That is probably just a matter of time, however …
• It’s much harder to know if a treatment is over-prescribed when we know that it is truly helpful for many patients, and that is probably the case for replacements. However — and this is subtle but important — many orthopaedic surgeries have never really been tested properly … and there’s a very strong pattern of such surgeries failing good clinical trials when they are finally conducted (Louw et al). Knee replacements do not actually rest on a foundation of controlled clinical trials, and the only major partially controlled trial to date, in 2015, was a mix of good and bad news (Skou et al). It’s a safe bet that proper testing of replacement, when it finally happens, will produce at least some bad news, showing that less helpful than many surgeons currently believe. And so some non-trivial amount of over-prescription also seems likely.

That’s a fair amount of indirect evidence. While we cannot know for sure, I do feel safe proceeding with the assume that BOB is “often” an exaggeration.

So now what about the value of BOB’ing when it is not an exaggeration? Does accuracy justify it?

To be continued.