Hip impingement surgery – can you return to sports and other activities?

A 2018 study in the American Journal of Sports Medicine claims that surgery for hip impingement is VERY successful. It says that arthroscopic hip surgery for femoroacetabular impingement reliably produces excellent results for patients. They claim it helps the overwhelming majority of hip pain patients return to sports quickly.

Other studies about FAI surgery don’t show these utopian results – like this one or this one. We’ve also seen a recent study on hip impingement surgery versus physical therapy. Surgery led to patient disappointment in that study too.

This new hip impingement study, Systematic Review and Meta-analysis of Outcomes After Hip Arthroscopy in Femoroacetabular Impingement, makes a bold claim.

It claims that 87.7% of patients return to sport activities after hip arthroscopy for FAI.

What makes this study so different? Let’s take a closer look.

 

What did they do in this hip impingement study?

This was a meta-analysis. It’s a review of studies on hip arthroscopy for femoroacetabular impingement.

In case you aren’t familiar with it, arthroscopy is where the surgeon inserts a small camera into the hip joint. The surgeon can perform his “fixes” with a small incision and small instruments. This is in contrast to open surgery, in which a much larger incision is made and the femur is totally pulled out of the socket.

The researchers gathered a total of 31 studies. This encompassed 1911 patients in total and 1981 hips. That’s a big sample size.

All the studies had to:

• be about arthroscopic surgery for FAI and ONLY arthroscopic surgery.
• be less than 12 years old.
• include patient outcomes, with at least 6-months of follow-up.

A meta-analysis summarizes all the information in all the studies with statistics.

As we look at this study, it’s important to keep this German aphorism in mind (often incorrectly attributed to Winston Churchill): “Trust no statistics that you didn’t falsify yourself.”

Let’s take a look at why and how we should be suspicious about the findings of this hip impingement study.

“Trust no statistics that you didn’t falsify yourself.”

 

This hip impingement study has an extremely high risk of bias

What does it mean to be biased? It means the authors behind the study may have incentives that predispose them to certain conclusions. In this case, that the study has a high risk of pro-surgery bias.

There are three major factors contributing to this high risk of bias.

• The authors are surgeons who make money performing hip arthroscopy.
• The authors are surgeons who receive money from companies that manufacture devices used in hip arthroscopy.
• The studies used in this review are all highly biased in favor of surgery (because the above two factors apply to the authors of those studies as well).

Here’s an example: if a cigarette company pays for a study, and the study concludes that smoking is good for your lungs, we might be suspicious that the authors were influenced by the money from the cigarette company.

They got their money from the cigarette company, and so it would be be more likely for them to come up with a conclusion that speaks positively about the cigarettes.

 

But could that be happening in the world of medicine?

Studies show that it happens regularly in orthopedic research.

So let’s take a close look here.

The first page of the study clearly states:

“The authors declared that they have no conflicts of interest in the authorship and publication of this contribution.”

We can double-check that claim on the official U.S. government website that tracks payments to American doctors.  This website ​​​​sees if anyone is paying the authors and creating hidden conflicts of interest. On that site, we have data from 2013 to 2017 (at the time of writing).

There are four lead authors in this hip impingement study.

One of them has no payment data.  

Three of them have payment data (here, here, and here). And all three receive regular payments from arthroscopic surgery device manufacturers.  

From 2013 to 2017, these three authors were paid a total of $203,078. In 2017, they earned a total of $99,017 in payments, gifts, and grants from arthroscopic surgery device manufacturers.

 

That 2017 number includes a $53,903 grant for research.

Let’s say you know someone who sells cigarettes. He earns several thousand dollars each month selling these cigarettes, and a cigarette company gives him a $50,000 grant to do research on the health benefits of cigarettes.

His study is published and claims that smoking cigarettes is good for your lungs. Would you trust it?

Probably not. He'll make more money if you believe cigarettes are good for you. He makes that money from the cigarette company and from people buying cigarettes directly from him.

He has a conflict of interest. There’s a very high risk of bias.

 

Conflict of Interest?

The lead authors received $203,078 from 2013-2017 from medical device manufacturers. One received over $50,000 as a research grant.

 

The situation is no different with surgeons in this hip arthroscopy review.

Surgeons make money performing surgeries. These surgeons get extra money every year from arthroscopic surgery device manufacturers to publish research that says — unsurprisingly — surgery is good for you.

Is bias in the study likely?

Yes, and later in this article, you’ll see it very clearly.

Finally, this hip impingement study pulled from a pool of low quality research.  Of the 31 studies, 26 are of Level IV evidence. In case you aren’t familiar with the levels, Level IV is one step up from THE BOTTOM. Level I is considered the top.

That means 86% of the studies included are classified as having a high risk of bias (in favor of surgery) and are missing key information.  

So you have biased authors looking at biased studies.

Therefore, the fact that they play with statistics to paint an optimistic picture of surgery for FAI comes as no surprise.

 

What would a high quality study on hip arthroscopy for FAI look like?

A proper study of the efficacy of hip surgery for FAI would have a placebo treatment control group. Historically, orthopedic surgeons strongly resist calls for placebo-controlled studies.

There is a long history of high quality studies debunking other cash-cow orthopedic surgeries. See: knee meniscus surgery, spine surgery, shoulder surgery for impingement.

 

How did they arrive at an 87% return to sport rate for hip arthroscopy for FAI?

This study claims 87.7% of patients returned to sports after hip arthroscopy for hip impingement.

In the abstract (the summary), the authors looked at 31 studies in total and over 1911 patients. This makes the 87.7% seem fantastic. It’s a huge patient population, so the conclusions must be trustworthy!

But the reality is different.

If you read the full paper, you’ll see the authors claim they calculated their 87.7% from only 10 studies and 554 patients. In Figure 3 of the study, they contradict that and actually show 11 studies and 631 patients. It’s a weird inconsistency that we can’t make heads or tails of.

Whether they used 10 or 11 studies isn’t that important. What is important is that the calculation is still not based on 1911 patients. At best, it’s only one-third of the entire sample.

So they did not have 87% of 1911 patients returning to sport. At most, they had 631 patients involved — AND it wasn’t 87% of them returning to sports. They used statistical tricks to make it SOUND that way.

In addition, all the studies they used to calculate their 87.7%  are highly biased Level IV studies. These were all studies published  by the surgeons who performed the surgeries. They had a strong interest in “proving” that their surgeries worked.

There was no oversight on how they collected data, no control groups, no placebo groups — nothing that would make for solid evidence.

 

How did the authors make hip surgery for FAI for athletes look good?

They misused some statistical tools to create rosy numbers. The actual statistical tricks will make your eyes spin in your head, so we’re going to keep our discussion of math to a minimum.

When you study groups of people and try to apply what you find to the general population, you have to think about how representative your groups are of the general population.

Think about it this way: will a pro football player respond to surgery the same way a 70-year-old sedentary woman will? Probably not. This is a difficult thing to measure and quantify, but it’s something to respect when studying humans and medical treatments.  

It's something to pay close attention to when you combine numbers from multiple studies. You have to consider whether big differences in the populations in your studies may have an impact on the trustworthiness of your numbers. 

For example, one study has 153 patients. Their ages range from 50 to 70 years old, and the study doesn’t mention how many of those patients played sports.

Another study has 70 male baseball players. Their ages range from 14 to 35 years old.

The only thing both groups have in common is the hip impingement diagnosis.

If you were studying the effect of ibuprofen on the human digestive system, maybe the age and athletic abilities wouldn't be at the top of your concern list. But when you're talking about the musculoskeletal system and how someone recovers from surgery, there are obvious differences between how a young person and an older person recover. 

The outcomes of an otherwise healthy 20-year-old baseball pitcher may significantly differ from a 60-year-old female who is sedentary.  Their risk of re-operation and their risk of complications may be different, and their likelihood to “return to sports” will obviously be different. 

But you lose that perspective when you combine all these numbers. The differences in your populations are obliterated. That’s why this statistical trick is so sneaky.

Here's an example to make this more clear. We won't use the actual statistical math they used to keep it simple.

Let's say you're doing a super simplified calculation of the chance of winning two gambling games. One game is a coin toss. One game is rolling a six-sided die. A coin has two sides, the die has six.

When you flip the coin, heads is a win. When you roll the die, a six is a win.

For the coin toss, you have a 50% chance of a win on each turn.

For the die, you have a 16.6% chance of a win on each turn.

With a die, you have a 16.6% chance of a win on each turn.

These are clearly different situations. You have a higher chance of winning with the coin than with the die.

But what if we want to know the average chance of winning no matter which game I play? You might come up with up with a 33.3% chance of winning every time you play by averaging out the odds of winning for both games (again, this is not how you would do this statistically, but we’re keeping the math simple).

50 + 16.6 = 66.6.

66.6 / 2 = 33.3%

That lowers the perceived likelihood of winning with the coin and doubles the perceived likelihood of winning with the die.

Of course, this would be incorrect, and it would be silly. You wouldn't combine odds like this because your chance of winning depends entirely on whether you're flipping the coin or rolling the die.

But that’s exactly how statistical tricks are misused in this study. The high-level athletes push the 'return to sport' numbers up (win!), and the bad outcomes (lose) don’t seem as frequent.

This study mixed six different outcome measurements, 18 studies, and 1462 patients. They mixed professional athletes with sedentary people, and they mixed patients with age ranges of 11 to 77 years old.

 

How pro athletes skew the data on return to sport after FAI surgery

This analysis looked at 11 studies on recovery from hip surgery and return to sport. You can take a closer look at all of them in the table below.