I want to dive into a new area today and actually talk about some things I’ve seen in the clinic and in the hospital to try to illustrate some ‘greater truths’. I do this with a little trepidation as I’m a staunch believer in patients’ rights to privacy and talking about an individual patient’s problems is a huge no-no. On the other hand, there’s no doubt that it’s sometimes easier (and more fun) to make points with specific examples. So I’m going to talk about situations more than patients, and if I do mention individuals they are likely to be composites of many persons with lots of made up personal details and ilnesses so as to protect the innocent. It’s the broad points that are important. The details are probably lies.
So onto the subject of this post: mistakes.
I recently saw a very interesting 60 year old woman who’s had rheumatoid arthritis (RA) for about 30 years. The reason she was admitted to hospital is marginally related to this story, but suffice it to say she’d had pain in her ankle. The ankle was aspirated (i.e. fluid was pulled out of the joint using a needle and syringe) in the ER and apparently it looked pretty yucky. The ER folks thought it was infected so she got admitted. I wasn’t convinced.
You see when I saw her, she had lots of joints that were swollen and sore, and infection just doesn’t do that. Septic (infected) joints are almost always single — one at a time, and usually a big joint like a knee or hip, not an ankle. So I figured the ER messed up. But that came later – I hadn’t seen her yet. The ER folks sent the joint fluid for culture to see if it would grow out any bacteria, and admitted this lady to the hospital and started her on antibiotics.
It turns out this lady had had pretty difficult to control RA. She’s been on lots of different meds for it and nothing had completely controlled her pain and joint swelling. But she definitely had RA – there’s a certain joint pattern to RA — it affects the hands symmetrically for instance — and she fit that pattern to a T. She even had characteristic X-rays that showed joint damage typical for RA.
So when I saw her I told the floor team that this was not a septic joint, that they could take her off antibiotics, and that we’d see her in a few weeks in clinic to start her on some other kind of therapy to try to get her disease under control.
And, since the title of this post is ‘Making Mistakes’ there’s clearly more to it than that. I’m clearly going to tell you how the ER made a diagnostic error and somehow generalize it.
Well, that’s partly right. But you see it wasn’t the ER’s mistake; it was mine. OK, the ER got it wrong too, but I got burned worse.
When I went to see this patient with the attending rheumatologist, she agreed with everything I said and did; but she also asked if anyone had ever aspirated one of the patient’s joints before this visit. I had no clue; she was new to our state and our hospital. And so my staff (attending rheumatologist) said ‘well, everyone with a swollen joint deserves at least one tap by a rheumatologist; so since she’s got a swollen knee too, let’s tap it and send the fluid and take a look at it’,’ which we did.
And when I drew back on the stopper, I go this milky white fluid out. Now synovial (joint) fluid should be clear and yellow. This was white, opaque, and viscous. Something was up. This must be what the ER pulled out of her ankle, and this must be why they thought the ankle was infected.
So we sent some of the fluid to the hospital lab, and took a little bit back to our lab to look at under a microscope, and what we saw looked a lot like this …
… except what we saw was about 100X as densely packed. These are monosodium urate crystals. You can tell this because they’re long and needle-shaped, and when you look at them under polarized light with a compensator (as in the above picture) they look yellow when they’re parallel to the compensator axis (the direction of the yellow arrow), and blue when perpendicular. There is absolutely no mistaking them for anything else.
And you don’t get them in RA joints.
You get them with gout.
This lady does not have RA; she has polyarticular gout. Polyarticular (multiple joint) gout is not common. Almost everyone who gets gout gets it in one joint at a time — usually in the big toe first, and then in other joints later — but almost always one joint at a time. In the cases where people do have polyarticular gout, they’ve had gout for years and it’s never been treated, so it gets out of control and ‘spreads’ so that affects lots of joints all at once. But they always tell you that it started in one joint and stayed in one joint for a long time.
Except of course when they don’t. Every now and then a patient will get a bunch of swollen joints just like in RA, but it will really be gout, and the only way to tell is to stick a needle in one of the joints and look at the fluid.
But we don’t always do that with RA patients. You see, the joint distribution in RA is so classic that it’s almost never anything else. And gout presents in one joint, and rules are rules because they work. Except when they don’t.
I was not the first rheumatologist to get burned by this lady. She’s had ‘RA’ for 3 decades and seen I-don’t-know-how-many rheumatologists in that time, and they all thought the same thing I did. She’d been treated with multiple major RA medications which don’t work in gout ; and hence her ‘intractable’ RA.
Oh, and part of the pattern was her X-rays. They look like RA joints. They don’t look like gout joints. Until, of course, you really look at them, and then you start to think that maybe, just maybe, they look a little odd, and just maybe they aren’t RA erosions, but gout erosions.
So, we’re going to treat this lady for gout — we have very good treatment for it — and I fully expect in that in the next few months she’s going to feel a whole lot better.
OK, so it’s time to get to the point, because even though it’s an interesting story in itself, there are some lessons here. Here’s the big one:
Everyone fucks up.
Everyone. Humans exist by pattern recognition. We get through daily life on patterns. Think about driving: do you really make a concious effort to look at the patterns of traffic? Do you try to actually tell how fast the oncoming car is going? Do you try to judge how fast you have to go around the turn to avoid getting hit? Or do you just look, judge, and go, and not even really think about. It’s all patterns.
Bust sometimes the patterns just don’t mean what you think they do. And sometimes that’s tragic. You don’t notice that the truck coming up the hill is coming a lot faster than you think, and he hasn’t signaled that he’s about to change lanes, and you’re just not really thinking about it because you’re working on normal pattern recognition.
This lady has hurt a lot for a long time because she had an ‘RA’ pattern, not a ‘gout’ pattern. Every doc she saw got taken in by it.
I’m sure you’ve all heard about the airliner crash yesterday. This was an error in pattern recognition too, I’m sure. The pilot recognized everything that he was supposed to; he heard the air traffic controller tell him the right things; the chart he was following made sense with what he was seeing. And he was on the wrong runway.
What should we do when this happens? What should we do when humans do things that are human? Because this is, after all, very human behaviour.
I don’t have an answer to that question. Certainly we try to avoid the same mistakes. Next time I see a new RA patient, I’m going to tap a joint and look at the fluid to make sure it’s not a crystal disease like gout, no matter how much I’m sure it’s RA – because it’s simple, easy, safe, causes minimal discomfort, and there’ no good reason not to do it. And the FAA is going to look at why this accident occurred, and try to find ways to prevent accidents like them in the future. And I hope you look twice before you pull out in front of that truck.
And we’ll succeed. And we’ll fail. Because there will always be something else. No matter how careful you are, how redundant you are, how obsessive you are, something will always slip through you’re pattern recognition software, because we are not and can never be perfect and circumstances will always find a way to come together in just the right way to bite us in the ass. All we can do is try to minimize our errors, and minimize the damage that results.
Just remember that as you go through life. I’m not saying don’t try to avoid making mistakes, and not to blame people when they commit egregious errors. But be careful of judging too harshly, because no matter what you do, it’ll happen to you one day. Just pray that the results are minor.