Septic hip in kids: 5 myths of ED evaluation

I had 5 year-old brought into the ED last week, not able to walk on his right leg, or even to move his right hip around much at all. The day before, he had been hopping around to avoid using the leg, but on the day of the ED visit it hurt too much to even do that. Like any active boy, he had had a few recent tumbles, but nothing that he didn't stand up right after. His mother said that he had felt warm at home, and she suspected a fever.
Of course, I was thinking "This either septic arthritis (SA) or transient synovitis (TS) ... how does this again?" So I turned to some trusted EM references, and was reminded of the

THE FOUR BIG RULES

1. The WBC, ESR, and temp can rule out septic arthritis.
2. Send the kid to radiology to get an US of the hip.
3. Send the kid to IR to get the effusion drained.
4. Consult ortho early for help.

At the end of the night, we did the right tests, found some answers, and treated the patient. But looking over the literature the next day, I decided that I would do things differently the next time. Why? Because those FOUR BIG RULES all had problems, and seem especially outdated in this era of emergency physician-performed ultrasound. The RULES aren't really myths, but they are at least only "quasi-truths."

Let's tackle each of them.

"The WBC, ESR, and temp can r/o septic arthritis."
A predictive tool, using the WBC, temperature, ESR, and the ability of the child to walk, was developed back in 1999 by Kocher. While that initial study suggested that using these factors had the potential to rule out SA, a validation study did not show the same utility. Even a new, prospective algorithm, that added a CRP shows limitations, allowing up to a 6% false-negative rate in non-weight-bearing kids.

Part of the problem with these studies is that most of them are retrospective, and/or enrolled very high-risk patients. For example, 1 out of 3 patients in Kocher's validation study had septic arthritis, a very high proportion, while Caird had an even higher rate of SA! On the other hand, Singhal had far more cases of plain ol' TS.

From Singhal

Why does this matter? Because with these radically different populations, it's real hard to derive and validate a decision rule that works in our typical ED patient. Even if you have a kid who can walk, has no fever, and has a normal WBC, ESR, and even a normal CRP...

From Sultan

... you still have a risk of septic arthritis that is somewhere between 0.2% and 17%!

So, if you are going to rely on the labs, temp, and physical exam to avoid doing imaging and arthrocentesis, you have to be willing to ask right now:  

What percent of cases of septic arthritis are you willing to miss? 

I'm betting that your number is closer to 0% than 17%, so you probably should get some imaging!

Speaking of which...

"Send the kid to radiology to get an US of the hip."
Emergency physicians can reliably us ultrasound to find fluid collections in the pericardium, the lungs, and even in the belly. As for joints, we're used to looking for effusions in the elbow, shoulder, knees and ankles. Heck, it doesn't seem like there's a body part that we don't scan for an effusion.

Why should the hip be different? Should we routinely outsource this to radiology, or can we take care of this?

We can probably handle most of these on our own, without having to undergo intensive or prolonged training. As it turns out, even ED physicians with very limited US skills can be quickly trained to be quite accurate with finding hip effusions. 

In a study conducted in a pediatric ED, physicians with only "minimal" prior experience with ultrasound were trained to find effusions in the hip. The training consisted of a 30-minute teaching session, along with 10 practice scans. They only had to find one "positive" scan before being considered adequately trained for the study. How did they do with this crash course - what was their accuracy?

Pretty good!

Of course, if you have a questionable US finding, or you get a result that is discordant with your clinical sense, you should get confirmatory imaging from our friends in radiology, just like we do in other cases!

"Send the kid to IR to get the effusion drained"
Whether it's an abscess, an effusion, a vein, an artery, peritoneal or thoracic fluid, emergency physicians have gotten pretty good at sticking needles into hypoechoic things. So we can probably handle sticking needles into a clear hip effusion!

The literature bears this out. For example, back before he was blogging about ECGs, Stephen Smith reported on a hip he drained back in 1999! Pretty OG... Since then, EPs at NYU and at Boston Medical Center have reported on their experience on draining these themselves (SPOILER: their experiences were good). 

For techniques and tips, I'll refer you to those papers above, as well as an excellent podcast - Check out the Ultrasound Podcast - episode 38, with Mark Goodman.
"Consult ortho early for help."
Just as for every other patient in the ED, we are the folks who are ultimately on the hook for evaluation and management.It's our responsibility if:

• Ortho says "His CRP is low - you don't need to tap the hip."
• Radiology says "We can't US that now - just keep them in the ED until 8 am." 
• IR says "Why don't you just admit them, and we'll tap sometime tomorrow. Also, hold on antibiotics until we get a sample."
• Anesthesia says "We're not going to sedate the kid in the IR suite at this hour - why don't you do it!"

In the end, it's our patient up until admission or operation.

The Bottom Line
It looks like, as we get better at US and diagnosis, we're creating more work for ourselves. Good!